2025-01126 SUPPLEMENTAL MATERIAL - Airport South Industrial Annexation (P21-017) (Noticed on 11072025; Published 11072025; Passed for Publication 10212025; Published 10242025) [Updated 11/18/2025 @4:41 PM] - Part 5 — Sacramento City Council, December 2, 2025 (part 1 of 3)

Linking Habitat
Restoration to
Meaningful Units of
Animal Demography
K. Shawn Smallwoodl'2
Abstract
To restore habitat is to restore the functional aspects o{ a place where one or more species are intended to
live. However, habitat restored with the right struc-
tural elements can still fail to support the species due to insufficient space and spatial contiguity (e.9.,
movement corridors) with nearby habitat. The spatial
extent of habitat deterrnines its capacity to support
various demographic units, such as a breeding pair of
individuals, a true population, or a metapopulation.

In scatter-plots oI published density estimates and
their corresponding study area sizes, I found consis-
tent patterns that appeared to represent spatial scale
domains, in which subpopulations, populations, and
multiple populations exist for each species. These
spatial scale domains can help guide habitat restora-
tion by indicating the minimum area needed to sup-
port a population The effectiveness of restored habi-
tat, considered alone or in combination with existing
habitat, should be judged by the habitat's capacity to
sustain a functioning population, which includes the
population's membership in a metapopulation. Addi
tionally, the evidence increasingly demonstrates that
habitat must include sufficient space for the popula-
tion(s) to periodically relocate as resources are de-
pleted locally. The spatial extent o{ proposed habitat
restoration can now be linked to the likely demo-
graphic unit of the species that can be supported
there. I provide two examples of proposed habitat
conservation and their likely effectiveness based on
their spatial areas.
rDepartment of Biological Sciences, California State Univer- sity, Sacramento, CA
2Address correspondence to K. S. Smallwood, 109 Luz Place, Davis, CA 95616, U.S.A., email [email protected].
@ 2001 Society t'or Ecological Restontiotl
Key words: animal demography, habitat capacity,
metapopulation, Natomas Basin Habitat Conservation
Plan, population, restoration.
lntroduction
common target for the conservation of animal
species has been to maintain a minimum viable
population (MVP), with the number of individuals de-
termined by genetics or demographic variables (Boyce
7992; Foley 1997). To do this conservationists have typi- cally used one or more of several approaches. One ap-
proach consists of setting aside land as reserves or in easements (Wright 1998). Another approach consists of
enhancing the quality of existing habitat with the aim of supporting more individuals on the same unit of land.

Yet another approach is to restore habitat where it once existed or establish habitat where it never existed but may exist now or in the future. Even if these approaches are directly tied to attairdng or maintaining a MVP,
they usually target the population as the functionally
significant demographic unit. However, the number of
individuals calculated to compose a MVP, or the habitat area proposed for conserving a population, might be
more representative of some other demographic unit.

Wildlife biologists and ecologists have recognized a
suite of demographic units that have functional signifi- cance, including the (1) individual, (2) mated Pair or re- productive unit (e.9., social group, family group, pack), (3) population, and (4) metapopulation.

Perhaps one reason that demography has been de-
emphasized in conservation planning is that it was al-
ready lost in a key operational term used by ecologists and wildlife biologists to infer relative success of a spe- cies across space and tirne. Density, that is, the number of individuals per unit area, has been one of the most
important operational terms in ecology (Peters 1991). In a single measure density has been used to express both
the distribution and number of organisms on particular
geographical areas (Greig-Smith 1983). It has been used to measure tlre intensity of occurrence of species within ecological communities, ecosystems, or landscapes, or
in vegetation complexes. Density has been a key vari-
able for building theoretical models (Mclntosh 1985; Pe- ters 1991), for measurement and analysis in many field
sfudies, and for management recommendations by
wildlife biologists. However, this seemingly efficient
term informs nothing of demographic organization,
which is lost in the transforrnation of numbers to spatial intensity; that is, if the numbers were associated with demographic units in ihe first place.

Through its use in ecology, animal density has been
recognized to be more complicated than firsi defined
(Elion 1933; Mohr 1940; Odum 1953). Density was de-
Restontion Ecolog! vol. 9 No. 3, pp. 253-261
Hdbitat and Animal Denography
scribed as "crude" when the number of individuals was
divided by the total geographic area encompassing
them. It was termed "ecological" when the number of
individuals was divided by only the habitat area. Addi- tionally ecologists came to realize that density typically changes seasonally and sometimes multi-annually (Keith
1963; Cyr 1997). Density has also been found to change
at habitat "edges" (Yahner 1988) and according to the
presence and width of habitat corridors (La Polla & Bar- rett 1993) and degree of habitat contiguity (Laan & Ver- boom 1990). Density was also found to be scale-depen-
dent (Smallwood & Schonewald 1996), and its influence
on use and availability analysis to identify habitat asso- ciations was found to be scale-deoendent (Smallwood
2001). With these findings, densiiy has turned out to
require considerable caution when it is applied to the- ory and management, and it still provides no informa-
tion regarding demography and its functional signifi-
cance.

Divorced from demographic functionality, density
could mislead conservationists into thinkins some resto- ration project area is successful because dJnsity is reta- tively high. In fact density could be high because the
proiect area is an ecological sink that lures dispersing su- badults faom productive areas while failing to afford
them protection from some exotic predator or on-site
contaminant (van Home 1983). The only way to know
that density indicates success in the restored or con-
served area is to confirm that a functionally significant demographic unit exists there, including ihe appropriate number of individuals within the appropriate space, and including breeding and successful rearing of progeny.

Populations also can be considered either constrained
within a certain area, such as a habitat fragment, or ca- pable of relocating to other sites within a much larger habitat fragment or wiihin large areas of contiguous
habitat. For some species studied across sufficiently
large areas and long time periods, the locations of pop- ulations tend to change periodically, as long as suffi- cient space is available (Taylor & Taylor 1979; den Boer 1981). Taylor and Taylor (1979) proposed three h)?oth-
eses to explain this dynamic spatial distribution of pop- ulations: (1) members of populations move elsewhere
to avoid exhausting their resources; (2) they move in
search of new resource patches after exhausting the re- sources where they were residing; or (3) subadults dis- perse into newly forming populations more frequently
than individuals immigrate into the parent populations.

Anoiher hypothesis for this pattem is that individuals
are attempting to escape their parasite or predator
loads, which intensify around populations. Whatever
the reason, numerous animal species have exhibited a
dlmamic spatial distribution of populations (Taylor &
Taylor 7979; den Boer 1981), which suggests that these
populations may require more space than they occupy
at any given time. Additionally populations are more
likely to persist longer by interacting with other popu- lations of a larger metapopulation (Hanski & Gilpin
1997). Wildlife biologists have provided many estimates of the areas used by individuals and the minimum re-
productive units, but only recently have ecologists esti- mated the areas used by populations (Smallwood 1999)
and metapopulations (Akcakaya & Atwood 1997).

The purpose of this study is to propose an empirical
method for estimating the demographic unit that can be
supported by a given geographical area of habitat. Spe- cifically, i will use a case study to demonstrate how
published estimates of animal numbers can be com-
pared to identify the spatial scale domains of habitat oc- cupied by populations of each species, and how these
spatial scale domains can be used to estimate the effec- tiveness of habitat restoration or the imDact of habitat loss. I will ,rlso suggest a prelimirary means for inter- preting the effectiveness of habitat restoration on meta- populations.

Estimates of areas used to support populations and
metapopulations have not been as forthcoming as esti-
mates of home ranees of individuals and minimum re-
productive units. Sma wood (1999) identified spatial
scale domains in which species' populations likely oc-
cur, averaging 52 adults per population among species
of mammalian Camivora. For each species, I plotted the
number, N, of animals estimated within each specified
study area (hereafter called numerical estlmates) against their corresponding study area sizes. These plots re-
vealed a consistent nonlinear pattern that has been hid- den when log density was regressed on log study area
size (e.g., Blackburn & Gaston 1996; Smallwood &
Schonewald 1996). The numerical estimates increase
proportionally with study area size (scale domain A)
until a threshold area (sensu Smallwood 1999) is reached, at which point numbers vary but no longer increase
with the size of the study area. This numerical plateau continues (scale domain B) until another tfueshold is
reached, at which point numbers double or triple (scale domain C), and so on. Smallwood (1999) interpreted
scale domain B as the range of areas typically occupied by the population. I reasoned that if the numerical pla- teaus in scale dornains C or D represented the popula-
tion, then biologists would have no demographic urrit
to attribute to the numerical plateau of scale domain B, which would be a cluster of individuals more numer-
ous than the minimum reproductive unit and less nu-
merous than the population. Scale domain B and thresh-
old area are empirically derived from published estimates, and may be the first representations of the space needed for the species' population. The following case study il- lustrates how habitat areas restored or removed on two
proiect areas may affect the demographic units of sev-
eral species that can survive on the project areas.

Restoftttion Ecologll SEPTEMBEF 2oo1
Habitat and Animal Demographll
Case Study: Demographic Units ol Threatened Species
in the Natomas Basin
I describe two proposed conservation designs for both
Thamnophis gE s (giant garter snake) and Buteo swain-
so,1l (Swainson's Hawk) in the Natomas Basin, Califor-
nia. The giant garter snake is listed as threatened under the federal Endangered Species Act (FESA), and Swain-
son's Hawk is listed as threatened under the California Endangered Species Aci (CESA). The habitat of both
species will be taken by the proposed land conversions
in the Basin, thus requiring mitigation. One proiect in- volves the conversion of 7,231 ha of field crops and pas- ture to housesi its mitigation plan is described in the Natomas Basin Habitat Conservation Plan (NBHCP:

USFWS 1997). The other project (Northern Territories
Inc., or NTI) involves the conversion of 1,777 ha of field crops and pasture to houses; its proposed mitigation
plan is described in Carrier er al. (1997,1999). These
proiect areas are adjacent to each other, but the proiects are not integrated because so far the U.S. Fish and Wild- life Service has rejected the NTI mitigation plan. These proiect areas harbor one of the last remaining popula-
tions of giant garter snakes (USFWS 1999), which are
found mostly along the irrigation canals and in the rem- nants of a slough known locally as Fisherman's Lake.

Swainson's Hawks forage along the Basin's irrigation
canals and in fields of sugar beets and rice, of which the latter are used by the hawks after the rice harvest
(Smallwood 1995). The spatial extent of conserved habi- tat can be compared to the spatial exient of habitat that would be needed to support meaningful demographic
uruts.

The spatial extent of habitat remaining after either
conservation plan is implemented depends on which
habitat description is invoked, that of the USFWS (1997) or of Carrier et al. (7997). These plans are described in greater detail in Appendix 1. Based on the NBHCP de-
scription of habitat, the NBHCP .would likely reduce
7,23-l ha of contiguous habitat to 165-ha fragments, and the NTI plan would redvce 2,273 ha of contiguous habi-
tat to 496 ha. Based on the Carrier et al. (7997,1999) de- scriotion of habitat. the NBHCP would reduce 211 ha of
habitat to 5 ha, and the NTI plan would increase the
current 66 ha of habitat to 245 ha. Below I use both ver- sions of habitat to compare the habitat areas gained and Iost in the Basin to the geographic areas t)?ically
needed to support various demographic units.

Restoration ecologists can consider five functionally
significant demographic units: the home range of an
adult; the home range of the minimum reproductive
unit, such as the mated pair or the pack; the spatially constrained population; the spatially unconstrained pop- ulation; and the metapopulation. To represent the areas needed to support the smaller of these demographic
units, I collected home-range estimates of adult females and minimum reproductive units. I took averages among
study sites for gophers (Miller 1957; Howard and Childs 1959; Reichman et al. 1982; Bandoli 1987; Cameron et al. 1988), and i used local estimates for giant garter snakes (Wylie 1998) and Swainson's Hawks (Estep 1989). Tho-
momys bottae (pocket gopher) is important prey of the
Swainson's Hawk, and it also constructs burrows that
serve as refugia for the giant garter snake (see Appen- dix 1).

The spatial area of the pocket gophers' minimum re-
oroductive unit was assumed to be the summed aver-
ige home ranges of one female and one male, because
reproductive males and females typically maintain
home ranges adjacent to each other. For the giant Sarter snake I lacked information of what constitutes a mini-
mum reproductive unit and how large an area is re-
quired to support it, so I assumed that the home range
of the average adult female overlaps the ranges of at
least one male. For Swainson's Hawks I used local esti- mates of male home ranges (Estep 1989).

To estimate minimal areas needed to support con-
strained populations, unconstrained populations, and
metapopulations, I started by collecting all the pub-
Iished estimates of seasonal and annual numbers of
pocket gophers (references in Smallwood & Morrison
1999), Swainson's Hawks (references in Smallwood
1995), and garter snakes (Seibert 1950; Carpenter 1952; Asplund 1963; Fitch 1965; Clark 1974; Freedman &
Catling 1978; Blaesing 1979; Gregory et al. 1983; Reich- enbach 1983; Reichenbach & Dalrymple 1986; USFWS
1999). To be included in this exercise, these estimates had to have been made from study area sizes and re-
search methods described in the report, and they could
not consist of extrapolations frorn one site to another or from one site to a larger region. I collected population estimates of 12 species in the genus Thamnophis to rep- resent giant garter snakes, and for six species in the
family Geomyidae to represent the local pocket go-
phers. Without using estimates from these larger taxo-
nomic levels, I lacked the sample size needed to esti-
mate the spatial scale domain within which populations
may function,
For one species at a time, I plotted numerical esti-
mates against their corresponding study areas, both of
which were originally used for estimating densities.

These plots included seasonal and annual estimates of
numbers because I wanted to include all the natural
variation in numbers that was reported. Examining
these plots visually, I identified the largest numerical estimates that included a distinct spatial scale domain, in which numbers of animals no longer increased with
increasing study area size. The plateau of these largest numerical estimates was easiest to see in plots of all es- timates. These numedcal estimates were then removed
sEprEMBEn2ool Restoration Ecolopll
Habitat and Aninal Demograph!
from the next plot, when I also restricted the range of study areas from 0 km2 to the largest area of the remain- ing estimates. The largest and most simllar magnitudes
of the remaining estimates were analyzed in the same
manner as the first group. This routine was repeated
until the smallest magnitude estimates increased pro-
portionally with study area sizes, which had been the
typical pattem observed for numerical estimates ap-
proaching zero (Smallwood 1999). Plots were examined
with decreasing ranges of study area sizes so that I
could identify discontinuities in number, which usually took the form of one or a series of asymptotes, depend- ing on the number of estimates and the range of study
area sizes available.

I assumed that the first numerical asymptote (thresh-
old area), which followed the usual proportional in-
crease in number with study area size (scale domain A), was also the next demographic unit larger than the min- imum reproductive unit. I assumed this organization to
be the population, although the term deze might also be appropriate. Assuming that my interpretation of this
pattern is accurate, habitat areas no larger than the
lower range of scale domain B can support only spa-
tially constrained populations; that is, the populations occupying areas the size of the lower end of scale do-
main B cannot shift locations every generation or so.

These assumptions are discussed further in Smallwood
(1999) and are consistent with theory related to demog- raphy, but ihese assumptions have not been adequately
tested.

To estimate the area needed to support an uncon-
strained population, I searched the literature for esti- mates of the percentage of the available habitat area
across regions. Schaller et al. (1988) found Panthera un- cia (snow leopard) to occupy only 9% of their potential habitat in the Himalayas. Dalsted et al. (1981) and
Knowles (1986) found Cynomys ludoaicianzs (prairie
dog) colonies to occupy only 0.6 to 5.3% of the studied landscapes. Searching throughout California for tracks
five times from 1985 through 1998, Smallwood (1997)
detected mountainlions (Puma concolor) on 77 to 29o/" of the transect, all of which was believed to be in moun-
tain lion habitat at the time of transect establishment.

Akcakaya and Atwood (1997) found 95% of California
Gnatcatchers (Polioptila c. californica) to occur within 24% of 853 km2 of habitat in Orange County, California.

These percentages vary considerably, and they may be
species-specific. For conservation planning purposes, I suggest that it would be reasonably prudent to assume
that other species occupy similar fractions of their avail- able habitat across regions at any given time. For the
purpose of this case study, I assumed that species oc-
cupy about 20% of their available habitats at any given time, so long as the populations of these species are spa- tially unconstrained. Based on this assumption, I multi- plied the threshold area by five to estimate the mini-
mum area needed to support a population of the
species tfuough multiple generations when the species
requires unoccupied habitat to which it can relocate the population for whatever reason. I multiplied the area
supporting this spatially unconstrained population by
three to represent the minimum area needed to support
a metapopulation, although most metapopulatiors prob-
ably consist of more than 3 populations.

Results
I located 38 estimates of garter snake density reported in 10 publications. I found that density was inversely
proportional to the size of the study arca (r2 : 0.97, df - 1,32 RMSE (root mean square error) : 0.36, p < 0.0001):
lo?,Density = 2.428 - |.0llo|Sludy area size r km) t.

This relationship between density and study area size
was similar to those reported for pocket gophers (Small- wood & Morrison 1999) and Swainson's Hawks (Small-
wood 1995). As is obvious from Figure L, very little
variation in log density can be explained by any factors other than study area size. I interpret this pattern as having resulted from garter snake researchers choosing
study sites based on a priori knowledge of high density (or high habitat quality), then varying the sizes of study areas around clusters of garter snakes, thereby includ- ing more or less near-vacant space within the study
area boundaries (Smallwood 1995; Smallwood & Schone-
wald 7996,1.998; Smallwood & Morrison 1999). Figure 1
also indicates that species within the gents Thamnophis exhibit similar patterns of spatial distribution, but the study area sizes used to estimate numbers of T. gigas
were all relatively large.

By converting the density estimates back into numer-
ical estimates and by plotting the raw numerical esti-
mates against their siudy area sizes, a discontinuous
pattern emerged for pocket gophers, garter snakes, and
Swainson's Hawks. Figure 2 illustrates this Pattem for
Swainson's Hawks (for most species, the nonlinear pat-
iem in the plot is difficult to see while viewing the en- tire range of study area sizes). This pattern was similar to that reported for 30 species of mammalian Carnivora
(Smallwood 1999). The numerical estimates made within
study areas of less than the threshold areas were likely aggregations that were spatially constrained within
habitat fragments, based on my assumptions described
earlier. Populations likely were included within the
larger study areas, although the interannual and inter- site variation in number appeared to be great. On areas larger than the threshold areas, and within scale do-
main B, the average and standard deviation of popula-
tion size was 57 :! tM pocket gophers, 315 :! 242 garter snakes (164 a 37 giant garter snakes), and 51 t 31 pairs Resforation Ecology sEprEMBER2ool
Habitat and Animal Denography
Thamnophis giges a Oth6rSpecie. of lhatttoptis tr Swai6on's Hawk lButao swalnsoni) 1,000,000
't 00,000
10,000
t,000
ro0
t0
I
A Scale domrin B
'6
o
125
,t 00
.E
475
€ro
=2s
0 0.0001 0.00'1 0.01 0.'l { 10 100
Area (km2)
Figure 1. Density ol Thamnophis sirtalis, T. brachystoma,T. but- Ieri, T . proximus, T . melanogastor , ar.d T. radi:r regressed on study area size. RMSE stands for root mean square eror in the regression analysis and is proportional to the error in the raw data; b is the estimated slope coefficienq and 12 is the coef- ficient of determination.
of Swainson's Hawks. The threshold areas were 0.0125
kmz for pocket gophers, 0.4 km2 for garter snakes, and
319 km2 for Swainson's Hawks. The smallest area stud-
ied for giant garter snakes was 2.35 km2, which was the area I used as a more conservative estimate of threshold area for this species. At some geographic area larger
than those studied thus far, numerical estimates will
likely increase for each of these species just because
more than one population would be included within
the study area boundary.

All the areas needed to support functionally signifi-
cant demographic units can now be plotted for each
species (Fig. 3). These plots can be useful for assessing the impacts of habitat loss, and they can be used to as- sess the capacity of habitat restoration (Fig. 4). For ex- ample, I can compare the project impacts of the NBHCP
and the NTI plans in terms of each species' demo-
graphic units that are gained or lost.

Using the NBHCP descriptions of habitat, the NBHCP
proiect area started with ihe capacity to support meta- populations of gophers and giart garter snakes but only a mated pair of Swainson's Hawks. At the end of the
project, no individual Swainson's Hawk can be sup-
ported entirely, nor can the remaining area support a
giant garter snake population. Using this same habitat
description, the NTI lands can initially support all de- mographic units of gophers, as well as an unconstrained population o{ giant garter snakes and one pair of Swain- Siz€ ofstudyAroa in km'?
Figure 2. Published estimates of the number of pairs of Swainson's Hawks increase propodionally with increasing study area size as large as threshold area (sources referenced in Smallwood 1995). Estimates made in study areas larger than threshold area vary but no longer increase with study area size (scale domain B). This plot excludes some numerical estimates made at much larger study areas because viewing the entire data range at once hides the nonlinearity observed in this figure.
son's Hawk. After the proiect the remaining land would
be unable to entirely support a pair of Swainson's Hawks, and it would lose the spatial capacity to suPport an un- constrained population of giant garter snakes.
using the carrier et al. (7997, 1999) description of
habitat, the pre-NBHCP area could not support a pair
of Swainson's Hawks, and could barely support a con-
strained population of giant garter snakes. At the com- pletion of the NBHCP, the remaining area would be un-
able to support any giant garter snakes or Swainson's
Hawks, and its capacity to support pocket gophers
would be reduced to a spatially constrained population.

The pre-project NTI area has the spatial capacity to sup- port all demographic units of pocket gophers, but it
cannot support Swainson's Hawks without the area
provided in adjacent lands, and it can support only a
minimum reproductive unit of giant garter snakes.

However, after the project, the habitat would extend far enough to support a constrained population of giant
garter snakes.

Discussion
Caims (193) defined ecological restoration as the rees- tablishment of the structure and function of ecosystems to pre-disturbance conditions. To achieve ecological res- toration, there is no substitute for the ecological space needed to support functionally significani demographic
units of animal species. Animal species require the right habitat structure, as well as the space, for its function- ally significant demographic units to exist. Long peri- sEprEMsEB 2ool Restoration Ecology
Habitat anrl Animal Demography
Figuie 3. The spatial areas of habitat needed to support func- tionally significant demographic units can be plotted for each animal species.
ods of natural selection acting on individuals have
likely established constraints on how many individuals
can be accommodated in a habitat area. Restoration or
enhancement efforts can ensure that species' resources
are ample in the habitat area, though each species ap-
pears to have upper limits on its resource assimilation rate/ resource partitioning in space, and production of progeny. These limits, partly set by mating systems and social behaviors, foraging behaviors, fecundity, and
basal metabolic rate, are collectively known as the car- rying capacity of a species' habitat. There is nothing a restoration ecologist can do to exceed the carrying ca- pacity of a given habitat area.

Besides enhancing habitat to the point that it can
maintain the local carrying capacity of a species, the
spatial extent of habitat is the only resource that can be augmented with a continuous, corresponding benefit to
an animal species. As the habitat area is expanded,
numbers of individuals can be increased along with in-
creasingly greater levels of demographic organization
until multiple, interconnected metapopulations occupy
the region. Ecologists have recognized that spaiially un- constrained populations can be spatially dynamic (Tay-
lor & Taylor 1979) and are interconnected by dispersal
and reproductive exchange (Hanski & Gilpin 1997). To
be minimally viable a population needs more than a
number estimated by examining demographic or ge-
netic factors. It needs adequate space. Without consid- edng the space needed to support spatially uncon-
sirained populations or metapopulations, there is no
operational link between population viability analysis
and restoration ecology. Without considering the spa-
tial capacity of habitat areas to support demographic
units, conservation efforts are poorly linked to the evo- lutionary and ecosystem contexts of the species; this is contrary to sound conservation efforts (Morrison et al. 1998).

Many conservation efforts, as well as supporting re-
search efforts, have targeted species presence as an ir- dicator of the effectiveness of a habitat area. For exam- :' !::
.i-i...
ll '/:,
'.:t
o a..t'.
+
Gophor(o)
Glarlgrrbr Smln.on'3
snare (a) H.wi (E)
Figure 4. The areas needed to suppod functionally significant demographic units of animal species can be compared to
project areas, including aleas to be lost to land conversions and areas to be gained by restoration. The vertical lines cone- spond to habitat areas (km2) that are upland (dashed lines) and both upland and wetland (solid lines) at the start and end of projects, as indicated by the respective origins and direc- tions of the arrows. A$ows pointing left indicate loss of habi- tdt space, dnd arrows pointing right indicate gains.
ple, Walk and Warner (1999) identified minimum areas
needed to maintain a presence of grassland bird species in Illinois. "Presence" connotes a functional signifi-
cance, as the habitat area would appear to support visi- tation or occupancy by at least one individual of the
species. Perhaps it is assumed that these individuals
who are present also participate as members of a popu-
lation, However, presence can also correspond with an
ecological sink (van Horne 1983) or in a fragmented,
minimally functioning populaiion. A species' presence
can indicate successful conservation or restoration as
long as it can be associated with functionally significant demographic units.

The capacity of a habitat area to support a particular
demographic unit also depends on the functioning of
the ecological processes within the habitat area (Rick- lefs et al. 1984), as well as the suitability of the sur- rounding lands and the connectivity afforded by habi-
tat corridors. Beier and Loe (1992) defined a corridor as "a linear habitat whose primary function is to connect
at least two significant habitat areas." The effective size of the conserved habitat area can be exparded when the
surrounding lands can provide the species with oppor-
tunities for travel, refuge, breeding, and foraging. Sur- rounding lands can be classified as optimal, survival,
transition, marginal, secondary, sink, unsuitable, and
inhospitable habitat, among other classifications (Li-
dicker 1995). Similarly, corridors can be classfied by how species perceive and use them or how humans perceive
Restoration Ecolow sEprErrrBEB 2oot
Habitat dnd Aninal Denogrdphy
them (Beier & Loe 7992). For example, corridors can be
distinguished that provide passage or dwelling, that are natural or constructed, or that are of adequate width
and structure. To estimate the spatial capacities of con- served habitat areas, one needs to consider the land-
scape context (Caims 1993) from the perspective of the
species. The actual habitat services provided by corri- dors and larger land uses will determine the degree to
which the demographic unit supported by the restored
habitat area is constrained. These services must be
quantified or classified if they are going to be operation- ally linked to the spatial capacities of habitat areas.

Smallwood (1999) provided the first analytical tool
for using existing population estimates to estimate the spatial scale domain in which a tlpical population may
occur. Although most numerical and density estimates
are attributed to populations, Smallwood (1999) found
few cases in which the investigators provided convinc-
ing evidence that the populations were delineated by a
boundary or completely included within the study area.

Such delineation of population boundaries in contigu-
ous habitat is obviously difficult, as individuals along the edge of the population are mobile and home ranges
there are ephemeral (Dasmarur 1981). The approach
used by Smallwood (1999) makes use of numerical esti-
mates that would otherwise not alone inform of the
area used by the population. However, until wildlife bi- ologists provide many more descriptions of popula-
tions and metapopulations, my case study of the Nato-
mas Basin will remain preliminary and suggestive of
how a tool can be used to assess the effectiveness of a restored or conserved area.

Acknowledgments
I thank Eric Stitt for collecting density estimates of garter snakes, Cameron Doyel and Will Boeger for funding, and
Michael Morrison and Dean Carrier for sharing ideas. I
thank Verna Jigour and an anonymous reviewer for their
helpful comments on a previous draft of this paper.

LITERATURE CITED
Akcakaya, H. R., and J. L. Atwood. 1997. A habitat-based metapo- pulation model o{ the Califomia Gnatcatcher. Conserwation Biology 11:422-434.

Asplund, K. K. 1963. Ecological factors in the distribution of Thannophis brachystomd (Cope). Herpetologica 19:128 132.

Bandoli, J. H. 1987. Activity and plural occuparcy of burrows in Botta's pocket gopher Thomonys bofloe. American Midland Naturalist 118:10-14.

Beier, P., and S. Loe- 1992. A checklist for evaluating impacts to wildlife movement corridors. Wildlife Societv Bulletin 20:
43U,40.

Blackburn T. M., and K. J. Gaston. 1996. Abundance-body size re- lationships: the area you census tells you more. Oikos 75:
303-309.

Blaesing, M. E. 1979. Some aspects of the ecology of the eastern gartet s ake (Thamnophis sbtalis sirtalis) in a semi-disturbed habitat in west-cmtral Illinois. Journal of Herpetology 13:
1n-1A1..

Boyce, M. S. 1992. Population viability analysis. Annual Review of Ecology and Systematics 23:481-506.

Brode, J. M., and C. E. Hansen. 1992. Status and future manage- ment of the giant gatter s ake (Thafinophis gigas) within the southern Anedcan Basin, Sacramento and Sutter counties, California. Idand Fisheries Division, California Department of Fish and Game, Sacramento.

Cairns, J., Jr. 1993. Ecological restoration: replenishing our na- tional and global ecological capital. Pages 193-208 in D. A.

Saunders, R. J. Hobbs, and P. R. Ehrlich, editors. Nature Conservation 3: Reconstruction of fragmented ecosystems.

Surrey Beatty and Sons, Sydney, Australia.

Cameron, S. R., C. N. Spencer, B. D. Eshelman, L. R. Williams, and M. J. Gregory. 1988. Activity and burrow structure of Attwater's pocket gopher (Geonys attuateri). Joumal of
Mammalogy 69:667477.

Carpenter, C. C. 1952. Comparative ecology of the common gar- tet snake (Thaftnophis s. sirtalis\, the ribbon snake (Thamno- phis s. saufitus), a dButler's garter snake (Thamnophis butleri) in mixed populations. Ecological Monographs 22:235-258.

Carier, W. D., K. 5. Smallwood, and M. L. Morrison. 1997. Nato- mas Basin Habitat Conservation Plan: nairow channel
marsh altemative wetland mitigation. Report to Northern Teflito es, Inc., Sacramento, Calitonxa.

Carder, W. D., M. L. Mordson, K. S. Srnallwood, and Vail Engi- neerinS. lsqq. Recommendahions for NBHCP land acqui.i-
tion and enhancernent shategies. Report to Northern Territo- ries, Inc., Sacramento, California.

Clark, D. R., Jr. 1974. The westem ribban snake (Thamnophis proxi zus): ecology of a Texas population. Herpelologica 3A372-379.

Cyr, H. 1997. Does inter-annual vadability in population density increase with tilne? Oikos 79:549-558.

Dalsted, K. J., S. Sather-Blair, B. K. Worcester, and R. Klukas. 1981. Application of remote sensing to prairie dog marlage- ment. Journal of Range Management 34:218-223.

Dasmann, R. F. 1981. Wildlife biology. 2nd edition. John Wiley & Sons, New York.
den Boer, P. J. 1981. On the survival of populations in a heteroge- neous aJrd variable environment. Oecolosia 50:39-53.
fltun C. lq3l. The ecologl of animals. Methuen and Co., Ltd., London, United Kingdom.

Estep, J. A. 1989. Biology, movements, alrd habitat relationships of the Swainson's Hawk in the Central Valley of California, 1986-8a7. California Department of Fish and Game, Non-
game Bird and Mammal Section Report, Sacramento.

Fjtch, H. S. 195,5. An ecological study of the gatter snake (Thamno prrs sirfalis). Museum of Natural History, University of Kan- sas Publications 15:493-564.

Foley, P. 1997. Extinction models for local populations. Pages 215 246 in I. Hanski and M. Gilpin, editors. Metapopulation biology. Academic Press, Nevr' York.

Freedman, W., and P. M. Catling. 1978. Population size and struc- ture of four sympafric species of snakes at Amherstburg, On- tario. Canadian Field-Naturalist 92:167-173.

Gregory, P. T., L. A. Gregory, and f. M. MacArtney. 1983. Color- pattem vadation i^ Thomnophis fielanogastor. Copeia 1983:
530-s34.

Greig-Smith, P. 1983. Quantitative plant ecology. University of Califomia Press, Berkeley. Hansen, G. E. 1986. Status of the giant garter snake Tirmnophis couchi gigas (Fitch) in the south- ern Sacrammto Valley during 1986. Final Report to the Cali- fornia Department of Fish and Game, Contract C-1433.

Hansen, G. E. 1986. Status of the giant garter snake Thamnophis sEprEMBER2O0l ReStO/AtiOn ECOIOIA
Habitat and Animal Demograph!
corrcfti gigds (Fitch) in the southem Sacramento Valley during 1986. Final Report to the Califomia Deparknent of Fish and Came, Contract C-1433.

Hansen, G. E. 1988. Review of the &iarl.t gafier sfiake (Thatnnophis corcrlgigrs) and its suppolting habitat during 1986-47. Final Report to the California Department of Fish and Game, Con- tract C-2060.

Hansen, G. E. 1991- Impacts upon the giant garter snake (T/Mmtlo- pris gigas) of proposed consbuction at the Gray Lodge Wild- Iife Area, Upper Butte Sink, Butte County, California. Pre- pared for the Califomia Department of Fish and Game,
Conhact C-2060.

Hansen, G. E., and J. M. Brode. 1980. Status of the giant garter srlake, Thamnophis cor.ht Itgrs (Fitch). California Department of Fish and Game, Inland Fishedes Endangered SPecies Pro- gram Special Publication 80-5:1-14.

Hanski, L A., and M. E. Gilpin. 1997. Metapopulation biology:
ecology, genetics, and evolution. Academic Press, San Diego, California.

Howard, W. E., and H. E. Childs, Jr. 1959. Ecology of Pocket go- phers with emphasis on ThotflolnVs bottae mewa. Hilgardia 29:277-358.

Keith, L. B. 1963. Wildlife's ten-year cycle. Unive6ity of Wiscon- sin Press, Madison.

Knowles, C. J. 1986. Some relationships of black-tailed Praide dogs to livestock grazing. Great Basin Naturalist 45:198-203.

Laan, R., and B. Verboom. 1990. Effects of pool size and isola- tion on amphibian communities. Biological Conservation 54i 25't162.

I-a Potla, V. N., and G. W. Barett. 1993. Effects of couidor width and presence on the population dynamics of the meadow
vole Microtus pennsyloari.rs. Landscape Ecology 25-37.

Leidy, G. R. 1992. Ecology, status and management of the giant garter snake, Thamnophis gigas. Pacific Environmental Con- sultants, Sacramento, California.

Lidicker, W. 2., Jt. 1995. The landscape concept: something old, something new. Pages 3-19 in W. Z. Lidicker, editor. Land- scape approaches in marnmalian ecology arrd conservation University of Minnesota Press, Minrreapolis.

Mclntosh, R. P. 1985. The background of ecology: concept and theory. Cambddge Unive6ity Press, Cambridge.

Miller, M. A. 1957. Burrows oI the Sacramento Valley pocket go- pher in flood-itigated aifalfa fields. Hilgardra 26:431452.

Mohr, C. O. 1940. Comparative populations of game, fur arrd other mammals. Anedcan Midland Naturalist 241581-584.

Morrison, M. L., B. G. Marcot,.and R. W. Mannan. 1998. Wildlife- habitat relationships: concepts and applications. 2nd edition- University of Wisconsin Press, Madison.

Odum E. P. 1953. Fundamentals of ecology. W. B. Saunders Co., London.

Peters R. H. 1991. A critique for ecology. Cambridge University Press, Cambridge.

Reichenbach, N. G. 1983. AIr aggregation of female garter snakes under corrugated metal sheets. Joumal of Herpetology 17:
41241.3.

Reichenbach, N. G., and G. H. Dalrymple. 1986. Energy use, life histories, and the evaluation of potential competition in two species of garter snake. Journal of Herpetology 20:133-153.

Reichman, O. J., T. G. Whitham, and G. A. Ruffner. 1982. Adap- tive geometry of burrow spacing in two Pocket goPher PoP- ulations. Ecology 63:687-695.

Ricklefs, R. E., Z. Naveh, and R. E. Tumer. 1984. Conservation of ecological processes. The Environmentalist 4, SupPlement 8:
1-16.

Schatter, G. B., R. funrang, and Q. Mingjiang. 1988. Status of the snow leopard Pantheru uncia in Qinghai and Garsu Prov-
inces, China. Biological Conservation 45:179-194.

Seibet, H. C. 1950. Population density of snakes in an area near' Chicago. Copeia 1950:229-230.

Smallwood, K. S. 1995. Scaling Swainson's Hawk population den- sity for assessing habitat-use across aJr agricultulal land- scape.,ournal of Raptor Research 29:172-178.

Smallwood, K. S. 1997. Interpreting puma (Puma concolor) densrty estimates for theory and management. Environmental Con- servation 24:283-289.

Smallwood, K. S. 1999. Scale domains of abundance among sPe- cies oJ mammalian Camivora. Enviroirmental Conservation 26t702-111.

Smallwood, K.S. 2001. Habitat models based on nurnedcal com- parisons. In J. M. Scott, P. J. Heglund, M. Morrison, M.

Raphael, J. Haufler, and B. wall, editors. Predicting species occurences: issues of scale ald accuracy. Island Press, Cov- ello, Califomia. 2001.

Smallwood, K. S., arld M. L. Morrison. 1999. Spatial scaling of pocket gopher (Geomyidae) density. Southwestern Natural- rst 44t73-82.

Smallwood, K. S., and C. Schonewald. 1996. Scaling PoPulation density and spatial pattern for tedestrial, mammalian cami- vores. Oecologia 105:329-335.

Smallwood, K. S, and C. M. Schonewald. 1998. Study design and interpretation for mammalian carnivore density estimates.

Oecolosia 113:474-491.

Smatlwood,k. S., M. L. Morrison, and J. Beyea. 1998. Animal bur- rowing attdbutes affecting hazardous waste management.

Environmental Manaeement 22:831-847.

Taylor, R. A. f., and L. R. Tiylor. 1979. A behavioral model for the evolution o{ spatial dynamics. Pages 1-28 in R. M. Ander- son, B. D. Turner, and L. R. Taylor, edito$. PoPulation dy- namics. Blackwell Scientific Publications, Oxford.

USFWS (U.S. Fish and Wildlife Service). 1997. Natomas Basin Habitat Conservation Plan. U.S. Fish and Wildlife Service Sacramento Field Office, Sacramento, Califomia.

USFWS (U.S. Fish and Wildlife Service). 1999. Dra{t recovery plan for the giant garter snake (Thaflnophis gigas). U.S. Fish and Wildlife SeNice, Portland, Oregon.
van Horne, B. 1983. Density as a misleading indicator of habitat quality. Journal of Wildlife Management 47:893-901 .

Walk, J. W., and R. E. Warner. 1999. Effects of habitat area on the occurence of grassland birds in Illinois. American Midland Naturalist 141:339-344.

Wright J. B. 1998. The role of conservation easement sites in bio- geographic and biological research in the U.S.A. Environ- mental Conservation 25:85-98.

Wylie, G. 1998. Giant garter snake project: 1998 progress report.

U.S. Geological Survey, Biological Resources Division,
Dixon Field SLahon, Di\on, Califomia.

Yahner R. H. 1988. Changes in wildlife communities near edges.

Conservation Biology 21333-339.

Restoration Ecology sEprEMoEn arlt
Habitat and Animal Demography
Appendix. Basis ot habltat proiections. Description of the competing conservation plans for the Natomas Basin.

Those who DreDared the Natomas Basin Habitat Con-
seruation Pl;n GlngCp) regarded rice and its support-
ing irrigation structure as important habitat for both gi- ant garter snakes and Swainson's Hawks (USFWS
1997), even though the prevalence of rice cultivation re- stdcts use of the Basin by Swainson's Hawks (Small-
wood 1995). Thus, spatial comparisons were relatively
straightforward in assessing pre- and post-project sPa- tial capacities for supporting various demographic
units. The NBHCP required 3,616 ha of conserved habi-
tat for the " take" of 7,237 L.a, and required the con- served areas to consist of contiguous habitat of at least 165 ha. One could assess the take of giant garter snakes and Swainson's Hawks due to spatial reductions of
habitat from 7,231 ha to 3,616 ha, and to 165 ha if one as- sumes that each of these minimum habitat areas will be
isolated. No restoration is planned under the NBHCP.

Carrier et al. (1997) defined habiiat more explicitly for these species, with the intent to restore high-quality
habitat as a mitigation strategy. They examined all the reports of giant Barter snake and Swainson's Hawk
habitat associations, based on both anecdotal and scien- tific observations. Various habitat descriptions for the giant garter snake have included terraces overlooking
stagnant or slow-moving water, backed by steep banks
or tall vegetation, along with nearby upland areas
where burrows made by ground squirrels (Spermophilus
beecheyi) and pocket gophers afford daily and seasonal
refugia from floods, predators, and inclement condi-
tions, as well as nearby shallow, warm waters and bare
ground where the snakes can bask with screening cover
of willows and sedges (Hansen & Brode 1980; Hansen
1986, 1988, 199'1.; Leidy 1992; Wylie 1998). Not knowing which attribute of this collective habitat description was more important than another, Carrier et al. (1997) de-
signed a narrow-channel marsh that included most of
the hdbitat attributes ihey had jdentified. Carrier et al. (1997) also designed upland areas and horizontal and
vertical edge to maximize the attractiveness of the area to pocket gophers ald ground squirrels (also see Small- wood et al. 1998). By making the area attractive to bur- rowing animals and by designing upland areas to also
support valley oaks (Quercus lobata) and other large
trees, the narrow channel marsh would also improve
the habitat conditions for Swainson's Hawks bv provid-
inq nest trees and more food.

Northem Territories lnc.'s (NTI) 2,273 ha of land on
the project site currently tncltrdes 27,292 m of ditches and drains (hereafter referred to as canals) within the matrix of fields used for growing rice and sugar beets.

According to Hansen (1986) and Brode and Hansen
(1992) the majodty of the giant garter snakes, if not all of them, have been observed on or within the canal and
its adjacent levees. According to Hansen (1986) giani
sarter snakes have been seen as far from canals as 185
-,
b,rt h" did not mention in which vegetation or land
use this occurred. Without any mention of giant garter
snakes occurring in rice, Carrier et al. (1999) assumed that nearly all the snakes occur within 24 m of habitat spanning the canal and its levees and banks. To be con- servative, Carrier et al. (1999) assumed that all the canal banks are useful as refugia for giant garter snakes dur- ing flooding. Therefore, the Carrier et al. (1999) view of habitat would include 66 ha on NTI's proiect area. As-
suming the NBHCP proiect area includes the same den-
sity of canals as the NTI lands, then Carrier et al. (1999) would estimate 211 ha of giant garter snake and Swain-
son's Hawk habitat occurs there now.

NTI's proposed land conversions would leave 10,370
m of canals within the proposed reserve area, and
76,923 n of canals would be taken (lost). The 66 ha of
current habitat (27,292 x 24 rn) + 10,000 rn2lha : 66 ha, would be reduced to 25 ha. However, NTI proposes to
construct between 50 and 80 canals that will be 3.7 m
wide and O.77 m deep and will extend 250-400 m from
the reserve buffers (peripheries) into the central por- tions of the reserve, each with at least 24 m of upland conditions along at least one side of the canal. This con- struction would add 20,154 m of new canals within the
reserve area, thereby increasing the habitat area to 121 ha, or a 55-ha (83%) net increase in effective habitat for ihe giant garter snake and Swainson's Hawk. Further-
more, Carrier et al. (1999) recommend that these canals be arranged and managed in an experimental marurer,
so that various ideas of what qualifies as high-qualiiy giant garter snake habitat can be replicated and inter- spersed as treatments alon8 these canals for testing
their effectiveness. The upland buffers would further
increase the useful area for giant garter snakes to 245 ha, all of which would lack the destructive canal manage-
ment practices, such as dredging and mechanical re-
moval of bank vegetation, which are the convention in
the Natomas Basin today.
sEprErr€ER 2o0! Re;toration Ecology
Ambient and Emission Trends of Toxic Air Contaminants in California Ralph Propper, * Patrick Wong, Son Bui, Je ﬀ Austin, William Vance, A ́lvaro Alvarado, Bart Croes, and Dongmin Luo *
California Air Resources Board, 1001 “I” Street, P.O. Box 2815, Sacramento, California 95812, United States *S Supporting Information
ABSTRACT: After initiating a toxic air contaminant (TAC) identi ﬁcation and control program in 1984, the California Air Resources Board adopted regulations to reduce TAC emissions from cars, trucks, stationary sources, and consumer products. This study quanti ﬁes ambient concentration and emission trends for the period 1990 −2012 for seven TACs that are responsible for most of the known cancer risk associated with airborne exposure in California. Of these seven, diesel particulate matter (DPM) is the most important; however DPM is not measured directly. Based on a novel surrogate method, DPM concentrations declined 68%, even though the state ’s population increased 31%, diesel vehicle-miles-traveled increased 81%, and the gross state product (GSP) increased 74%. Based on monitoring data, concentrations of benzene, 1,3-butadiene, perchloroethylene, and hexavalent chromium declined 88 −94%. Also, the ambient and emissions trends for each of these four TACs were similar. Furthermore, these declines generally occurred earlier in California than elsewhere. However, formaldehyde and acetaldehyde, which are formed in the air photochemically from volatile organic compounds (VOCs), declined only 20 −21%. The collective cancer risk from exposure to these seven reviewed TACs declined 76%. Signi ﬁcant reduction in cancer risk to California residents from implementation of air toxics controls (especially for DPM) is expected to continue. ■ INTRODUCTION
Background. In response to public concern about exposure to airborne carcinogens, California adopted the Toxic Air Contaminant (TAC) Identi ﬁcation and Control Act in 1984. 1 Since then, California has identi ﬁed 21 TACs for which California’sO ﬃce of Environmental Health Hazard Assessment (OEHHA) has developed inhalation cancer potency factors, 2 and the Air Resources Board (ARB) has implemented
regulations such as airborne toxic control measures (ATCMs) to limit TAC emissions. In 1987, California adopted the Air Toxics “Hot Spots ” Information and Assessment Act, which requires operators of signi ﬁcant-risk facilities to reduce risks from exposure to emitted TACs.
3 The 1990 U.S. Clean Air Act
Amendments 4 listed 187 “hazardous air pollutants ” that California added to its TAC list in 1993, although no
ATCMs have been adopted for added TACs. Since then, the U.S. EPA has developed National Emission Standards for
Hazardous Air Pollutants (NESHAPs)
5 that are delegated to
the states. Local California agencies may adopt more stringent regulations than statewide ATCMs or federal NESHAPs.

Several studies have reviewed ambient TAC trends. Mobile sources of TACs have been studied in California, especially benzene.
6−13 These studies show declining ambient concen-
trations with implementation of emissions control regulations.

In particular, California gasoline reformulation in 1996 led to a large decline in benzene emissions and ambient concen-
trations.
7−11 Because ambient diesel PM (DPM) concen-
trations are not measured directly, two surrogates have been used to estimate their trends: coe ﬃcient of haze (COH) 14 and elemental carbon (EC). 15,16 Some studies show declines in ambient TAC concentrations nationally 17−19 and in speci ﬁc states.20−23 For some TACs, although limited literature was found on statewide or air-basin-wide ambient concentration trends, reductions in ambient TAC concentration have been observed at speci ﬁc locations due to targeted emission controls.
24−26
The study objectives were to (1) quantify ambient and
emission trends for TACs in California and estimate the associated trend in cancer risk, (2) review relationships between declines in ambient concentrations and emissions controls in California, and (3) compare ambient trends in California with trends elsewhere based on data from previous studies.

TACs Studied/Selected. Table 1 shows the seven TACs
selected for review, along with their principal sources, years identiﬁed, and unit risk factors. Cumulatively, they account for most of the known chemically induced cancer risk from
exposure to Californians during the 1990 −2012 study period, with particulate matter from diesel-fueled engines (DPM) the most important in terms of overall cancer risk.

Received: April 21, 2014
Revised: August 31, 2015
Accepted: September 4, 2015
Published: September 4, 2015
Article
pubs.acs.org/est
© 2015 American Chemical Society 11329 DOI: 10.1021/acs.est.5b02766 Environ. Sci. Technol. 2015, 49, 11329 −11339
This is an open access article published under an ACS AuthorChoice License, which permits copying and redistribution of the article or any adaptations for non-commercial purposes.

Downloaded via 71.195.179.26 on April 1, 2025 at 16:51:47 (UTC).

See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

Three of these TACs have been emitted mainly from mobile sources: DPM, benzene, and 1,3-butadiene. Although some stationary sources use diesel-fueled engines, most DPM is emitted from mobile sources mainly trucks, buses, and o ﬀ- road sources such as ships.
27 Benzene is emitted mainly from
on-road vehicles from incomplete fuel combustion and from gasoline evaporation. 27 Its tropospheric lifetime is about 10 days, allowing dispersion throughout an air basin. 28 1,3- Butadiene is emitted mainly from incomplete fuel combustion from mobile sources.
30 Its tropospheric lifetime is typically a
few hours. 31
Two of these TACs have been emitted mainly from
stationary sources: perchloroethylene and hexavalent chromi- um. Perchloroethylene has been used as a solvent, especially for dry cleaning, and in various product formulations.
32 Its
atmospheric lifetime is about four months, su ﬃciently long that global background levels have been reported. 33 Chromium compounds have two oxidation states: trivalent and hexavalent; in the 1980s, 3 −8% of total chromium in California ambient air was hexavalent. 34 Hexavalent chromium has been used for chrome plating. Its half-life was estimated to be about 13 h (∼19 h lifetime).
35 For both of these TACs, most exposure
occurs on a neighborhood scale, with health risks highest near point sources.
27 The remaining two TACs result mainly from photochemical reactions of emitted VOCs: formaldehyde and acetaldehyde.

U.S. ﬁeld studies have shown that primary emissions account for only 8 −30% of ambient formaldehyde concentrations, depending on location.
36−38 Photolysis is the main removal
mechanism for formaldehyde, leading to a lifetime of about 4 h.

Reaction with hydroxyl radical is the main removal mechanism for acetaldehyde, leading to a lifetime of about 9 h.
28 Rain and
fog also remove airborne aldehydes.

Four other TACs were considered for inclusion, 27 but not reviewed. For carbon tetrachl oride, persistent worldwide concentrations represent signi ﬁcant cancer risk; however, emissions have been minimal for decades. For p-dichloroben- zene, California ’s recent ambient data are usually below the limit of detection (LOD). For methylene chloride, California ’s ambient data exhibits such variability that data quality was Table 1. Toxic Air Contaminants Reviewed a
toxic air contaminant principal source
year
identiﬁed
unit risk factor
(μg/m3)−1
diesel particulate matter
(DPM)
mobile sources 1998 8.94 × 10−4
benzene mobile sources 1985 8.12 × 10−5
1,3-butadiene mobile sources 1992 4.87 × 10−4
hexavalent chromium chrome plating 1986 4.54 × 10−1
perchloroethylene dry cleaning 1991 1.71 × 10−5
formaldehyde atmospheric
reactions
1992 1.71 × 10−5
acetaldehyde atmospheric
reactions
1993 8.12 × 10−6
aCalifornia’s Toxic Air Contaminant Identi ﬁcation Reports for each TAC are available at www.arb.ca.gov/toxics/id/summary/summary. htm, which provides links to “health risks ” and “exposure assessment ” documents.

Figure 1. ARB’s toxic air contaminant (TAC) long-term monitoring network. Seventeen sites (circles) provided data for most of 1990 −2012. The two most populous air basins are shaded: South Coast and San Francisco Bay Area. Seven sites (not shown) that operated for shorter periods also provided data: Richmond and Upland (ended 1997); Fontana (ended 1999); Concord, Modesto, and Santa Barbara (ended 2000); and San Pablo (1997−2000).

Environmental Science & Technology Article
DOI: 10.1021/acs.est.5b02766
Environ. Sci. Technol. 2015, 49, 11329 −11339
11330

questionable. For dioxins, the lack of long-term ambient monitoring data precluded a level of review comparable to the seven TACs reviewed.
■ MATERIALS AND METHODS
Data Sources. Emission inventories were obtained from the California Emissions Projection Analysis Model (CEPAM) 39 and the California Emission Inventory Development and
Reporting System (CEIDARS). 40 CEPAM/Norcal Version

1.04 was used to provide a consistent methodology from
which DPM emissions for 1990 −2012 were calculated. ARB ’s emission factor model EMFAC2011
41 was used to provide
input into CEPAM/Norcal Version 1.04 for on-road mobile source emissions. CEIDARS (run-date: October 15, 2014) was used for the other TACs; it provides TAC inventories in two- year increments, for 1996 −2012. CEIDARS incorporates
CEPAM data and data from local air districts for stationary sources. Speciation pro ﬁles were used to estimate TAC
emissions from particulate matter and VOCs. These emission inventories take into account the e ﬀects of growth and emission controls. The data presented here may di ﬀer from some
previously reported values; emission inventory values change due to improved emission pro ﬁles and activity data.

Ambient DPM concentrations were estimated using oxides of nitrogen (NO
X =N O+N O 2) as a surrogate, combined with
scaling factors derived from the ratio of DPM emissions to total NOX emissions as obtained from the ARB inventory. Basin- wide annual average ambient NO X concentrations (1990 − 2012) were calculated using hourly average measurements from over 200 monitoring sites. The following assumptions were made: (1) ambient DPM concentrations are proportional to NO
X concentrations in each air basin; (2) emissions are well- mixed on basin-wide time scales, and (3) background NO X concentrations can be neglected. Section S-IV , Methodology for Estimating Ambient DPM Concentrations, provides details (S = Supporting Information).

The ARB started a TAC monitoring network in 1985 ( Figure 1). At sites selected to represent basin-wide and statewide averages, ambient air samples have been collected and speciated for TACs.42 TAC trends were derived using all monitoring sites in the ARB ’s air quality database. 43 Considering the earliest years for which reliable ambient data (late 1980s) and emission inventory data (1990) were available, 1990 was used as the starting point for this study.

The Bay Area Air Quality Management District (BAAQMD)
also started a TAC monitoring network that same year in the San Francisco Bay Area Air Basin (SFBAAB, Figure S1 ).
44 Some sites were chosen based on regional concerns, such as proximity to oil re ﬁneries;44 others were collocated with ARB sites. Based on correlations between BAAQMD and ARB data sets for perchloroethylene and benzene ( Figure S2 ), data from all ARB and BAAQMD sites that operated for most of the
1990−2012 period were combined to obtain a more robust
data set for trend analyses of these TACs in the Bay Area.

Because only limited BAAQMD data were available for other TACs, no other data sets were combined.

The South Coast Air Quality Management District has
conducted four Multiple Air Toxics Exposure Studies
(MATES) in the South Coast Air Basin (SoCAB). This
paper uses data from MATES-II (1998 −1999),
45 MATES-III
(2004−2006),46 and MATES-IV (2012 −2013).47 Data from
MATES-I (1987) were not rea dily comparable due to
diﬀerences in objectives and sites. Data from national air monitoring networks National Air Toxics Trends Stations (NATTS), Urban Air Toxics Monitoring Program (UATMP),
Photochemical Assessment Monitoring Stations (PAMS) 
were not used, due to limited data in California (NATTS, UATMP), or systematic problems in benzene measurement
(PAMS).
13 However, comparisons were made with U.S. EPA
data that include NATTS and/or UATMP monitoring sites. 48 Determination of Cancer Risk. Ambient DPM concen-
trations were inferred from the scaling factors described above.

For TACs other than DPM, annual mean ambient concen-
trations were determined by aggregating data from all sites, weighed equally. Half of the LOD and 75% completeness
criteria were used for this determination, which is consistent with U.S. EPA methodology.
49 Annual means were calculated
as the average of monthly means.

The statewide cancer risk resulting from exposure to each TAC was de ﬁned as the number of new cases of cancer
projected to occur per million residents. That risk was calculated for each TAC by multiplying its statewide yearly average ambient concentration by its unit risk factor ( Table 1).

These TAC unit risk factors were derived from OEHHA ’s
inhalation cancer potency factors, inhalation exposure rates, Figure 2. Statewide diesel PM ambient concentrations, diesel VMT (million vehicle miles traveled by diesel vehicles), and California Gross State Product (GSP). DPM concentrations are based on NO X concentrations and DPM/NO X emissions ratios. Note: VMT decline 2007 −2009 was due to a national economic downturn (the “Great Recession ”).

Environmental Science & Technology Article
DOI: 10.1021/acs.est.5b02766
Environ. Sci. Technol. 2015, 49, 11329 −11339
11331

individual sensitivity factors, and exposure factors, using the methodology outlined in OEHHA ’s 2012 risk assessment
guidelines.50 These guidelines require the selection of breathing rates; this study used the 95th percentile daily breathing rate (DBR) for children under 2 years old and the 80th percentile DBR for all other ages.
51 ■ RESULTS AND DISCUSSION
Results are presented by source type: mobile source TACs, stationary source TACs, and secondary TACs. Figure 2 shows the statewide ambient DPM trend, and Figure 3 shows annual statewide ambient and emissions trends for the six other TACs reviewed. Figure S3 shows these latter trends normalized, with trend lines added for SFBAAB and SoCAB. Table S1 shows the associated data with standard errors, as well as summarized MATES data. MATES data and ARB ’s SoCAB data show
similar ambient declines.

Mobile Source TACs. Diesel Particulate Matter (DPM).

Figure 2 shows that between 1990 and 2012, statewide ambient DPM concentrations declined 68% (3.1%/year), while vehicle- miles-traveled (VMT) for diesel-fueled vehicles increased 81%,
52 and California ’s Gross State Product (GSP) increased 74% (in constant dollars). 53 This period coincides with ARB ’s adoption of several diesel-related emission control regulations (Table S2 ). The 32% decline for 1990 −2000 is close to the projected 40% emission reduction from ARB ’s regulatory program.
54 In 2000, ARB ’s Diesel Risk Reduction Plan 55
(DRRP) recommended the replacement and retro ﬁt of diesel- fueled engines (e.g., diesel particulate ﬁlters), and the use of ultra-low-sulfur (<15 ppm) diesel fuel. Implementation of the DRRP resulted in a 49% decline for 2000 −2010, and the ARB expects a decline of 71% for 2000 −2020.56 This expectation of a steeper decline is due to state truck regulations that have accelerated ﬂeet turnover, with BC emissions reductions of ≥40% between 2009 and 2011 in Los Angeles. 57
Benzene. Figure 3 a shows that ambient benzene concen-
trations declined 88% from 1990 to 2012. This decrease was achieved despite increased population (31%) and VMT (30%) during that period ( Figure S4). Figure 3a shows similar ambient and emissions trends, indicating an association between regulatory measures and reduced ambient concentrations.

Tailpipe emissions data from ARB ’s vehicle surveillance program show that the 2003 ﬂeet-average benzene rate
decreased by 89% relative to the 1995 ﬂeet.
58 Table S3 lists
major regulations that reduced mobile source emissions and their expected emission reductions, if available.

Two ARB programs signi ﬁcantly reduced mobile source
emissions, including benzene: cleaner fuels and vehicle emission controls. The ARB ’s Low-Emission Vehicles (LEV) regulations
59 resulted in reduced emissions of nonmethane
organic gases (NMOG) from new light- and medium-duty
vehicles. LEV I regulations (1994) led to reduced exhaust and evaporative emissions of hydrocarbons. Since 1996, light-duty vehicles sold in California have been equipped with California ’s second generation On-Board Diagnostic (OBD-II) system.

About half of total car emissions resulted from malfunctions related to emission control devices,
60 thus OBD-II implemen-
tation led to reduced benzene emissions by facilitating repairs.

ARB’s Phase II Reformulated Gasoline (RFG-2) regulation required a reduction in benzene volume content from 1.7% to ∼1%,
30,61 which was con ﬁrmed at several service stations. 6,11 Between 1995 and 1996, vehicular benzene emissions declined 54%; most of this reduction was attributed to lowered levels of benzene and related aromatics in gasoline.
7 The benzene
percentage of NMOC (nonmethane organic compounds) in
tunnel emissions declined from 6% in 1995 to 3.6 −4.2% in later Figure 3. Statewide annual average concentrations and emissions inventory (EI) trends for six toxic air contaminants. EI data were available starting in 1996. (a) Benzene (1990 −2012), (b) 1,3-Butadiene (1990 −2012), (c) Perchloroethylene (1990 −2012), (d) Hexavalent Chromium (1991 − 2012), (e) Formaldehyde (1996 −2012), (f) Acetaldehyde (1996 −2012).

Environmental Science & Technology Article
DOI: 10.1021/acs.est.5b02766
Environ. Sci. Technol. 2015, 49, 11329 −11339
11332

years.6 RFG-2 implementation was projected to cause a 24% decline in ambient concentration, 30 but greater reductions occurred. From 1995 to 1996, the decline was 29%, which was attributed to implementation of RFG-2. 61 In a SFBAAB tunnel, the benzene/VOC weight fraction declined 39% between 1995 and 1996,
11 while SFBAAB ambient concentrations also
declined 39% between 1995 and 1996 ( Table S1 ). In the SoCAB, RFG-2 was implemented over two years (1994 −1996), and benzene concentrations declined 40% ( Table S1) over that period.

Based on these results, a link has been established between RFG-2 implementation and benzene concentration decline.

However, other regulations have allowed multiyear implemen- tation, making it di ﬃcult to link emission controls and TAC concentration declines. Furthermore, mobile sources tend to have overlapping vehicle and fuel regulatory implementations. 1,3-Butadiene. Figure 3 b shows that annual average 1,3- butadiene concentrations declined 85% from 1990 to 2012, and also that ambient and emissions trends are similar. Table S3 lists major regulations that reduced 1,3-butadiene emissions. 1,3-Butadiene emissions depend upon fuel composition, engine type, emission control devices, operating temperature, and the age and condition of the engine,
62 and were expected to decline
due to implementation of motor vehicle regulations. 29 Tailpipe emissions data from ARB ’s vehicle surveillance program show that the 2003 ﬂeet-average 1,3-butadiene rate decreased by 86% relative to the 1995 ﬂeet.58 The adoption of technology-forcing emissions standards led to three-way catalysts to control vehicle exhaust. Vehicles equipped with e ﬀective exhaust catalysts emitted less than a tenth as much 1,3-butadiene as those without functioning catalysts.
63,64 LEV I regulations (1994) led
to reduced exhaust and evaporative emissions of hydrocarbons.

Stationary Source TACs. Perchloroethylene. Figure 3 c
shows that statewide ambient perchloroethylene concentrations declined 92% (1990 −2012), and also shows similar ambient and emissions trends. Table S4 lists regulations that reduced perchloroethylene emissions, and their expected emissions reductions. In 1980, the BAAQMD adopted a rule controlling perchloroethylene emissions from dry cleaners, which led to a 1981 ARB Suggested Control Measure.
65 By 1993, eight local
air districts had rules that required a carbon absorber or refrigerated condenser, and dry-cleaners were transitioning from vented transfer machines to vented dry-to-dry machines and closed-loop machines.
65 In 1993, California adopted a regulation to limit emissions of perchloroethylene from dry cleaning that projected a 78% reduction in emissions from dry cleaners and an overall emissions reduction of 47% (1994 −1998; Table S4 ).
65 Figure
3c shows a lower (33%) ambient decline during that period, perhaps due to lesser emission reductions from other sources, such as degreasing and paint. Due to local air district rules, perchloroethylene concentrations declined steeply before the mid-1990s, when the ARB ’s 1993 dry cleaner regulation was implemented.

Starting in 2003, California provided ﬁnancial incentives for dry cleaners to use other solvents.
32 From 2003 to 2005,
perchloroethylene emissions from dry cleaning decreased by ∼70%.
32 Subsequently, the ARB adopted a regulation for
automotive products that prohibited the use of perchloro- ethylene,66 and regulations banning perchloroethylene use in aerosol coatings and most consumer products. 67 ARB’s dry cleaning regulation was amended in 2007 to require phase-out of perchloroethylene machines by 2023, which would reduce emissions to minimal levels. 32
Hexavalent Chromium. Figure 3 d shows that ambient
concentrations of hexavalent chromium declined 86% (1991 − 2012). Before 1991, the ARB ’s monitoring method allowed conversion of hexavalent to trivalent chromium; 35,68 therefore, reliable monitoring data are only available since 1991. The ambient decline appears steeper prior to 1996; this may re ﬂect a decrease in chrome plating operations, or a switch from hexavalent to trivalent chromium for plating.
27 Figure 3 d also
shows some correspondence between the ambient and
emissions trend.

Table S4 lists ARB-adopted regulations to control hexavalent chromium emissions and their expected emissions reductions.

ARB’s 1988 chrome plating regulation required ≥95% emission reductions, and was expected to reduce emissions by 97% within four years.
69 In 1993, the U.S. EPA adopted a
NESHAP70 with the same emission reduction goal; the ARB regulation was modi ﬁed in 1998 for conformity. 71 In 2005, hexavalent chromium emissions from these facilities were

99.97% less than in 1987, which was more than expected from implementation of these regulations. A 2006 ARB regulation further reduced these emissions, by requiring add-on air pollution control devices and chemical fume suppressants.

Table S4 shows three additional regulations that led to reduced emissions. A 1989 ARB regulation prohibited the use of hexavalent chromium in cooling towers.
72 Others prohibited
the use of hexavalent chromium in motor vehicle and mobile equipment coatings (2001), and in thermal spraying operations (2005).

Signiﬁcant emissions reductions resulted from the 1980s regulations, prior to the ARB ’s 1991 adoption of a reliable hexavalent chromium analytic method for its ambient
monitoring network. Therefore, prior to ARB regulations, hexavalent chromium conce ntrations in the 1980s were
probably much higher than the 1991 statewide average.

Secondary TACs (Formaldehyde and Acetaldehyde).

For the period 1996 −2012, Figures 3 e and 3f show that ambient concentrations of formaldehyde declined 22%, and acetaldehyde declined 21%. (Before 1996, the monitoring method underestimated actual concentrations, but no correc- tion factor is available.
27) During the same period, these ﬁgures
also show that formaldehyde emissions declined 50%, and acetaldehyde emissions declined 14%. The ambient and
emission trends for each aldehyde do not correspond as well as for other TACs, which is likely related to the dominance of their photochemical formation versus direct emissions.

The ARB promulgated a series of increasingly stringent
emission standards for motor vehicle exhaust to limit
hydrocarbon emissions, which also limited aldehyde emissions.

For example, the ARB ’s 1990 Low Emission Vehicle/Clean Fuels (LEV/CF) regulations were expected to reduce form- aldehyde emissions from cars and light-duty trucks over several years.
73 Comparing tailpipe average emission rates between the 1995 and 2003 ﬂeets, formaldehyde emissions declined 11%/ year and acetaldehyde emissions declined 8.6%/year. 58 Based on emissions monitoring in a SFBAAB tunnel (1994 −1999), formaldehyde emissions decreased by 10%/year, and acetalde- hyde by 9.5%/year; attributed to implementation of RFG-2 in 1996.
74 A subsequent SFBAAB tunnel study found that the
formaldehyde emission factor declined 12%/year between 2001 and 2006, and the acetaldehyde emission factor declined 3.8%/ year. These declines were attributed to vehicle ﬂeet turnover Environmental Science & Technology Article
DOI: 10.1021/acs.est.5b02766
Environ. Sci. Technol. 2015, 49, 11329 −11339
11333

and fuel reformulation that occurred during that period; the higher formaldehyde decline was attributed to the contempora- neous switch from MTBE (a formaldehyde precursor) to
ethanol in gasoline.
75 Because incomplete oxidation of ethanol yields acetaldehyde, increased use of ethanol in transportation fuels could lead to an increase in ambient acetaldehyde concentrations. However, other components of California gasoline, such as aromatic compounds and alkenes, were found to be primarily responsible for the formation of acetaldehyde.
76 Figure 3 f shows no
increase in acetaldehyde concentrations between 2002 and 2004, when MTBE in gasoline was replaced with ethanol in California.
75 Although aldehydes are present in the ambient air mainly due to photochemical processes, they are also directly emitted. The major aldehyde direct emission source is combustion, especially from on-road motor vehicles. The largest area-wide source is residential wood burning. Because formaldehyde is used in composite wood products and consumer products, form-
aldehyde also has signi ﬁcant nonmobile emission sources. An ARB regulation adopted in 2008 was designed to reduce
formaldehyde emissions from composite wood products by
57%, or 500 tons per year.
77 Although this represents only 2.4%
of primary emissions, these sources cause disproportionally high exposure due to indoor o ﬀ-gassing.

Comparison with Ambient Trends Obtained from Other
Studies. Table 2 compares ambient TAC declines (%/year) from this study with available reported values, such as from U.S.

EPA’s NATTS monitoring network.
78 These trend compar-
isons are expressed as relative changes over time. This table shows linear reduction rates for the starting and ending years of data for the same period as for each reference provided, except for Burbank and London where natural logarithms (Ln) are compared.
10 Mobile Sources. Because diesel PM is not measured
directly, it is di ﬃcult to compare ambient DPM trends.

However, DPM estimates from MATES-III 46 and -IV47 show a 64% decrease in ambient concentration (2005 −2012) for
SoCAB, compared to a 55% decrease in SoCAB based on this paper ’s methodology. ( Table S5 shows ambient DPM
concentrations for individual air basins.) Table 2 compares the DPM trend from this study with MATES and estimates that result from use of other surrogates, and Figure S5 graphs the comparison with COH
14 and EC (equivalent to BC) 15 trends.

BC was calculated from COH data, 14 but COH monitoring was phased out around 2000. The IMPROVE monitoring net-
work
79 was used to obtain EC measurements, 15 although this network is mainly in rural areas. Using 1990 −2000 data from 11 SFBAAB sites, COH declined about 55%; 14 compared to a contemporaneous 32% decline in statewide DPM ( Figure 2 ).

Possible causes for this disparity are (1) reduced agricultural burning during this period
80 (most COH sites were in or near
agricultural burning areas), and (2) reduced EC/DPM ratios (COH concentrations are proportional to EC concentra-
tions),
14 as observed between MATES-II and -III periods. 46 A
review of 1990 −2012 data from IMPROVE sites found that wildﬁres led to spikes in EC levels in some years. Overall, these data showed a statewide 60% EC decline, similar to the trend in fossil fuel EC emissions (primarily DPM);
15 compared to the
68% decline in statewide DPM in this study. These comparative Table 2. Ambient Toxic Air Contaminant Trends: Comparison with Data from Other Sources this study other sources
period location decline (%/yr) location decline (%/yr) reference Diesel PM
1990−2000 CA (California) 3.2% CA 5.5% (COH) Kirchstetter 14 1990−2012 CA 3.4% CA 3.0% (BC) Bahadur 15
2005−2012 CA South Coast 7.9% CA South Coast 9.1% MATES-IV 47 Benzene
1989−1993 CA 10% Canada 5.0% Dann 1995 83
1994−2000 CA 9.2% National 7.8% Cook 2004 22
1994−2002 CA 7.5% National 9.3% Fortin 2005 13
1994−2009 CA 5.2% National 4.4% EPA 2010 81
2003−2010 CA 6.2% National 4.7% EPA 2014 19
1990−2008 CA 4.8% Camden NJ 3.7% NJ DEP 2011 20
1995−1996 SF Bay Area (CA) 40% SF Bay Area (CA) 39% (tunnel) Kirchstetter 1999 11 42% (ambient) Harley 2006 6
1998−2005 Burbank, CA 11% (Ln) London, UK 26% (Ln) Warneke 2012 10 1,3-Butadiene
1994−2008 CA 5.6% Camden NJ 5.1% NJ DEP 2011 20
1995−2003 CA 6.5% Canada 5.0% Curren 2006 85
2003−2010 CA 6.4% CA 6.4% EPA 2014 82
Perchloroethylene
1991−2006 CA 5.7% eight states 2.6% EPA 2010 81
CA 5.9%
2003−2010 CA 7.8% National 7.7% EPA 2014 82
Hexavalent Chromium
2005−2010 CA 7.8% National 4.9% EPA 2014 82
Formaldehyde
2003−2010 CA 2.6% National 2.4% EPA 2014 82
Acetaldehyde
2003−2010 CA 2.1% National 2.7% EPA 2014 82
Environmental Science & Technology Article
DOI: 10.1021/acs.est.5b02766
Environ. Sci. Technol. 2015, 49, 11329 −11339
11334

trends provide some validation of the scienti ﬁc reliability of this study’s methodology.

Benzene ambient trends from this study were compared with trend data from 22 sites in ten other states for 1994 −2009,81 and from 154 national trend sites for 2003 −2010.19,82 The trend lines appear parallel ( Figure S6 ), but the national trend lags California ’s trend by several years. For 1994 −2009, national concentrations declined 66% compared to 78% in California; for 2003 −2010, the declines were 33% (national) and 43% (California). In Camden NJ, benzene declined 66% (1990−2008)
20 compared to 87% in California. In Canada,
benzene declined 20% (1989 −1993)83 compared to 50% in
California. These data all show steeper declines in benzene concentrations in California, presumably because California initiated regulations earlier than elsewhere. However, in London, benzene declined 91% (1998 −2008),
10 compared to
61% for that period in California. This was attributed to London’s post-1998 adoption of mobile source controls that California had adopted earlier.
10 In New York, the benzene
decline ( ∼50%, 1990 −2003) was found to be consistent with California’s decline; major reasons included the adoption of RFG regulations and California ’s LEV program. 84
For 1,3-butadiene, recent national data (2003 −2010)19,82 show the same decline in ambient concentrations as found in California from this study. However, data for earlier periods show a steeper decline in California compared to other
locations, as follows. Concentrations declined 66% in Camden NJ (1994 −2008),
20 compared to California ’s 78%. Concen-
trations at Canadian urban sites declined 38% (1995 −2003),85 compared to California ’s 61%; consistent with earlier
regulations in California.

The 89% ambient benzene decline and the 88% ambient 1,3- butadiene decline (both 1990 −2012) could be compared with the 65% reduction in ROG emissions reported in the SoCAB (1993−2012).
86 This indicates that regulations targeting
mobile-source TACs have been e ﬀective.

Stationary Sources. For the period 1991 −2006, the U.S.

EPA provided ambient perchloroethylene concentrations data for nine states; the decline in California (88%) was steeper than for the other eight states (39%).
81 For a later period (2003 −
2010), however, the national decline 19,82 was similar to that found in this study. Figure 3 c shows a steep decline in perchloroethylene concentrations until about 2004, consistent with concentrations declining earlier in California than in other states. Hexavalent chromium concentrations declined 33% (2005−2010) nationally,
19,82 compared to 53% for that period
from this study, even though California ’s steepest declines occurred earlier: ARB regulations controlled emissions from plating facilities and cooling towers in the 1980s.

Secondary TACs (Formaldehyde, Acetaldehyde). For
urban areas nationally, mean ambient formaldehyde concen- trations declined 14% from 2003 to 2010, while acetaldehyde declined 17%.
19,82 This study shows similar declines for that
period: 13% for formaldehyde, and 18% for acetaldehyde. The complexity of photochemical formation and removal processes for these TACs increases the di ﬃculty in quantifying their expected ambient trends. Although these aldehyde declines are not as steep as for the other ﬁve TACs reviewed, they are steeper than the 7% decline in ozone (another secondary pollutant) in the western U.S. for 2001 −2010.
87 Continued
ambient declines for these aldehydes is anticipated as the emissions of their VOC precursors continue to decline.

Evaluation of Cancer Risk. Statewide cancer risks were
determined using the calculated unit cancer risk factors ( Table 1) and ambient statewide levels for each TAC, 1990 −2012.

Figure 4 shows a 76% decline in cumulative increased cancer risk per million people, from 1990 to 2012, due to inhalation of ambient air (caption notes interpolation for DPM and
extrapolation for hexavalent chromium and aldehydes). This corresponds to a cumulative risk reduction from 3000 cancer cases per million residents in 1990 to 730 in 2012. This decline has been driven mostly by decreased exposure to DPM,
followed by benzene and 1,3-butadiene. For 1990 −2012, cancer Figure 4. Cumulative increased cancer risk for seven toxic air contaminants, per million California residents. Values are rounded to two signi ﬁcant ﬁgures. Hexavalent chromium values prior to 1991 and aldehyde values prior to 1996 were obtained by linear extrapolation from the three subsequent years. Due to unavailability of 1992 DPM emission inventory, 1992 DPM cancer risk is the average of 1991 and 1992 values. Nonlabeled bars represent cancer risk <100 per million residents.

Environmental Science & Technology Article
DOI: 10.1021/acs.est.5b02766
Environ. Sci. Technol. 2015, 49, 11329 −11339
11335

risk declined 68% for DPM, 89% for benzene, 88% for 1,3- butadiene, 94% for perchloroethylene; 88% for hexavalent chromium, 21% for formaldehyde, and 20% for acetaldehyde.

As a percentage of the combined cancer risk from these seven TACs, the risk from these two aldehydes increased from 3.0% in 1990 to 9.7% in 2012.

Figure 4 shows that from 1990 through 1995, and again from 2003 through 2012, the statewide cancer risk from exposure to benzene was greater than to 1,3-butadiene; but from 1996 through 2002, the opposite was true. In 1996, RFG-2 required a steep reduction in the benzene content of gasoline, which reﬁners generally accomplished by hydrogenating benzene to cyclohexane, a 1,3-butadiene precursor.
88,89 This is consistent
with the increase in cyclohexane in regular gasoline sold in the SFBA from 1.17 weight% in 1995 to 1.84% in 1996. 90
Subsequently, the cyclohexane content of gasoline declined, 91 because re ﬁners found more cost-e ﬀective ways to lower the benzene content. 92
In 1991, the ARB estimated that ambient exposure to
perchloroethylene emissions from eight facilities resulted in up to 14 potential excess cancer cases in 5.5 million people. 93 The ARB’s 2007 dry cleaning regulation is expected to reduce statewide risk to less than one cancer case per million by 2020.
32 Also, ambient monitoring (MATES-III) revealed high
levels of hexavalent chromium that were traced to a cement facility in Riverside. 46 That ﬁnding led to controls that reduced its emissions and resultant exposures. These stationary-source TAC examples indicate that near-source residents may have especially bene ﬁted from TAC emission reductions.

In general, these ambient declines occurred earlier in
California than elsewhere. Emission reductions and lower cancer risk correlate with emissions controls adopted in California, primarily controls on mobile sources and targeted stationary-source controls.

Further reductions in mobile source TAC emissions are
expected from continuing e ﬀorts to improve fuel formulations, reduce vehicle exhaust emissions, and promote less polluting transportation modes. For stationary source TACs, perchloro- ethylene concentrations declined 90% and hexavalent chro- mium declined 86% (1990 −2012); however regulatory controls probably led to even greater emission reductions, considering that California started regulating these TACs in the 1980s.

Implementation of current control measures is expected to lead to further reductions in emissions for both TACs.

Most cancer risk from identi ﬁed TACs is from DPM, and
DPM concentrations represent a larger percentage (71%) of the overall cancer risk in 2012 than in earlier years. This is due to the earlier adoption of control measures for other TACs, and to the relatively long lifetime and slow turnover to cleaner technologies for diesel engines. However, the implementation of ARB ’s recent diesel engine retro ﬁt and replacement requirements
94 has accelerated ﬂeet turnover to cleaner
trucks,26,57 and signi ﬁcant additional reductions are projected statewide.56
The decreased exposure (from 1990 to 2012) to ambient
concentrations of the seven TACs reviewed here resulted in a reduced cancer risk of thousands of fewer cancer cases per million exposed in California. Before 1990, air toxics regulations likely resulted in additional reductions in cancer risk, but these reductions are unquanti ﬁed because they occurred prior to completion of California ’s air toxics monitoring network.

Phased implementation of recent regulations is expected to lead to further reductions in ambient TAC concentrations and cancer risk.
■ ASSOCIATED CONTENT
*S Supporting Information
The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.est.5b02766.

Figures and tables that are described above and a detailed methodology for DPM calculations ( PDF)
■ AUTHOR INFORMATION
Corresponding Authors
*(R.P.) Phone: +1 916-323-1535; e-mail: [email protected]. *(D.L.) Phone: +1 916-324-8496; e-mail: [email protected].

Notes
The authors declare no competing ﬁnancial interest.
■ ACKNOWLEDGMENTS
We acknowledge contributions from four anonymous reviewers and the following ARB sta ﬀ: Jagjeet Arce, Martin Johnson, Robert Krieger, Reza Mahdavi, Hector Maldonado, Nehzat
Motallebi, Thoan Nguyen, Michael Redgrave, Linda Smith, Robert Weller, and Stephen Zelinka. The statements and
opinions expressed in this paper are solely the authors ’ and do not represent the o ﬃcial position of the California Air Resources Board (ARB). The mention of trade names,
products, and organizations does not constitute endorsement or recommendation for use. The ARB is a department of the California Environmental Protection Agency. The ARB ’s
mission is to promote and protect public health, welfare, and ecological resources through e ﬀective reduction of air pollutants while recognizing and considering e ﬀects on the economy. The ARB oversees all air pollution control e ﬀorts in California to meet health-based air quality standards.
■ REFERENCES
(1) Toxic Air Contaminant Identi ﬁcation List ; Air Resources Board:

Sacramento, CA, 2008; www.arb.ca.gov/toxics/id/taclist.htm (ac- cessed April 24, 2015).
(2) Consolidated Table of OEHHA/ARB Approved Risk Assessment Health Values;O ﬃce of Environmental Health Hazard Assessment and Air Resources Board: Sacramento, CA, www.arb.ca.gov/toxics/ healthval/contable.pdf (last updated May 13, 2015).
(3) AB 2588 Air Toxics “Hot Spots ”Program; Air Resources Board:

Sacramento, CA, www.arb.ca.gov/ab2588/ab2588.htm (accessed April 24, 2015).
(4) U.S. Clean Air Act, Section 112 (a) − Deﬁnitions (Hazardous Air Pollutants). www.epa.gov/ttn/atw/112a_def.html (accessed April 24, 2015).
(5) National Emission Standards for Hazardous Air Pollutants ; U.S.

Environmental Protection Agency: Washington, DC, www.epa.gov/ ttn/atw/mactfnlalph.html (accessed April 24, 2015).
(6) Harley, R. A.; Kean, A. J. Chemical Composition of Vehicle-Related Volatile Organic Compound Emissions in Central California , Final Report, Contract 00-14CCOS; San Joaquin Valleywide Air Pollution Study Agency and California Air Resources Board, 2004; www.arb.ca. gov/airways/ccos/docs/II5_0014_Aug04_fr.pdf (accessed April 24, 2015).
(7) Harley, R. A.; Hooper, D. S.; Kean, A. J.; Kirchstetter, T. W.; Hesson, J. M.; Balberan, N. T.; Stevenson, E. D.; Kendall, G. R. Effects of reformulated gasoline and motor vehicle fleet turnover on emissions and ambient concentrations of benzene. Environ. Sci. Technol. 2006, 40, 5084 −5088.

Environmental Science & Technology Article
DOI: 10.1021/acs.est.5b02766
Environ. Sci. Technol. 2015, 49, 11329 −11339
11336

(8) Main, H. H.; Roberts, P. T.; Reiss, R. Analysis of Photochemical Assessment Monitoring Station (PAMS) Data to Evaluate a Reformulated Gasoline (RFG) E ﬀect; STI-997350 −1774-FR2; for U.S. EPA by Sonoma Technology, Inc., 1998; www.epa.gov/ttn/amtic/ ﬁles/ ambient/pams/rfg_oms.pdf (accessed August 17, 2015).
(9) Hammond, D. Ambient trends of benzene in California from 1990 through 1995 ; Technical paper presented at U.S. EPA/AWMA International Symposium on Measurement of Toxic and Related Air Pollutants, Research Triangle Park, North Carolina, 1996; www.arb.ca. gov/aaqm/qmosprog/papers_studies/benzene.htm (accessed October 4, 2012).
(10) Warneke, C.; de Gouw, J. A.; Holloway, J. S.; Peischl, J.; Ryerson, T. B.; Atlas, E. L.; Blake, D. R.; Trainer, M. K.; Parrish, D. D.

Multi-year trends in volatile organic compounds in Los Angeles, California: Five decades of decreasing emissions. J. Geophys. Res. 2012, 117, D00V17.
(11) Kirchstetter, T. W.; Singer, B. C.; Harley, R. A.; Kendall, G. R.; Hesson, J. M. Impact of California reformulated gasoline on motor vehicle emissions. 2. Volatile organic compound speciation and reactivity. Environ. Sci. Technol. 1999, 33, 329 −336.
(12) Fruin, S. A.; St. Denis, M. J.; Winer, A. M.; Colome, S. D.; Lurmann, F. W. Reductions in human benzene exposure in the California South Coast Air Basin. Atmos. Environ. 2001, 35, 1069 − 1077.
(13) Fortin, T. J.; Howard, B. J.; Parrish, D. D.; Goldan, P. D.; Kuster, W. C.; Atlas, E. L.; Harley, R. A. Temporal changes in U.S. benzene emissions inferred from atmospheric measurements. Environ.

Sci. Technol. 2005, 39 (6), 1403 −1408.
(14) Kirchstetter, T. W.; Aguiara, J.; Tonsea, S.; Fairley, D.; Novakova, T. Black carbon concentrations and diesel vehicle emission factors derived from coefficient of haze measurements in California:
1967−2003. Atmos. Environ. 2008, 42, 480 −491.
(15) Bahadur, R.; Feng, Y.; Russell, L. M.; Ramanathan, V. Impact of California’s air pollution laws on black carbon and their implications for direct radiative forcing. Atmos. Environ. 2011, 45, 1162 −1167. (16) Schauer, J. J. Evaluation of elemental carbon as a marker for diesel particulate matter. J. Exposure Anal. Environ. Epidemiol. 2003, 13, 443−453.
(17) McCarthy, M. C.; Hafner, H. R.; Chinkin, L. R.; Charrier, J. G.

Temporal variability of selected air toxics in the United States. Atmos.

Environ. 2007, 41, 7180 −7194.
(18) McCarthy, M. C.; O ’Brien, T. E.; Charrier, J. G.; Hafner, H. R.

Characterization of the chronic risk and hazard of hazardous air pollutants in the United States using ambient monitoring data. Environ.

Health Perspect. 2009, 117, 790 −796.
(19) Second Integrated Urban Air Toxics Report to Congress , EPA-456/ R-14-001; U.S. Environmental Protection Agency: Research Triangle Park, NC, 2014; www2.epa.gov/urban-air-toxics/second-integrated- urban-air-toxics-report-congress (accessed April 24, 2015). (20) New Jersey ’s Environment Trends: Air Toxics , New Jersey Department of Environmental Protection, 2011; www.nj.gov/dep/ dsr/trends/pdfs/air-toxics.pdf (accessed April 24, 2015). (21) McCarthy, M. C.; Hafner, H. R.; Ra ﬀuse, S. M.; Stiefer, P. S.

Temporal Trends in Air Toxics: Supplementary Technical Memorandum , STI-903553 −2582-STM, 2004; www.ladco.org/reports/toxics/sti/ temporal_trends_in_air_toxics_data.pdf (accessed April 24, 2015). (22) Cook, R.; Driver L.; Mullen, M. Trends in Emissions of Air Toxics from Highway Mobile Sources, 1990 to 2002 . 2004. www.epa.gov/ttn/ chief/conference/ei13/toxics/cook.pdf (accessed April 24, 2015). (23) Heirigs, P. L.; Delaney, S. S.; Dulla, R. G. Evaluation of MOBILE Models: MOBILE6.1 (PM), MOBILE6.2 (Toxics), and MOBILE6/ CNG; for American Association of State Highway and Transportation Oﬃcials, 2004; http://onlinepubs.trb. org/onlinepubs/archive/ NotesDocs/25-25%287%29_FR.pdf (accessed April 24, 2015). (24) Cahill, T. A.; Cahill, T. M.; Barnes, D. E.; Spada, N. J.; Miller, R.

Inorganic and organic aerosols downwind of California ’s Roseville Railyard. Aerosol Sci. Technol. 2011, 45 (9), 1049 −1059. (25) Kuwayama, T.; Schwartz, J. R.; Harley, R. A.; Kleeman, M. J.

Particulate matter emissions reductions due to adoption of clean diesel technology at a major shipping port. Aerosol Sci. Technol. 2013, 47 (1), 29−36.
(26) Dallmann, T. R.; Harley, R. A.; Kirchstetter, T. W. Effects of diesel particle filter retrofits and accelerated fleet turnover on drayage truck emissions at the Port of Oakland. Environ. Sci. Technol. 2011, 45, 10773−
10779.
(27) California Almanac of Emissions and Air Quality ; Air Resources Board: Sacramento, CA, 2009; www.arb.ca.gov/aqd/almanac/ almanac09/almanac09.htm (accessed April 24, 2015).
(28) Calvert, J. G.; Atkinson, R.; Becker, K. H.; Kamens, R. M.; Seinfeld, J. H.; Wallington, T. H.; Yarwood, G. The Mechanisms of Atmospheric Oxidation of the Aromatic Hydrocarbons ; Oxford University Press, 2002.
(29) Atkinson, R. Atmospheric chemistry of VOCs and NO X. Atmos.

Environ. 2000, 34 (12−14), 2063 −2101.
(30) California Phase 2 Reformulated Gasoline (CaRFG2) ;A i r Resources Board: Sacramento, CA, 1991; www.arb.ca.gov/fuels/ gasoline/carfg2/carfg2.pdf (accessed April 24, 2015).
(31) Tuazon, E. C.; Alvarado, A.; Aschmann, S. M.; Atkinson, R.; Arey, J. Products of the gas-phase reactions of 1,3-butadiene with OH and NO 3 radicals. Environ. Sci. Technol. 1999, 33 (20), 3586 −3595. (32) Staﬀ Report: Initial Statement of Reasons for the Proposed Amendments to the Control Measure for Perchloroethylene Dry Cleaning Operations and Adoption of Requirements for Manufacturers and Distributors of Perchloroethylene ; Air Resources Board: Sacramento, CA, 2007; www.arb.ca.gov/regact/2007/perc07/isor.pdf (accessed April 24, 2015).
(33) Simpson, I. J.; Meinardi, S.; Blake, N. J.; Rowland, S.; Blake, D.

R. Long-term decrease in the global atmospheric burden of tetrachlorethylene (C 2Cl4). Geophys. Res. Lett. 2004, 31, L08108. (34) Initial Statement of Reasons for Proposed Rulemaking; Public Hearing to Consider the Adoption of a Regulatory Amendment Identifying Hexavalent Chromium as a Toxic Air Contaminant. Sta ﬀ Report; Air Resources Board, Sacramento, CA, 1985; www.arb.ca.gov/toxics/id/ summary/hex.pdf (accessed April 24, 2015).
(35) Grohse, P. M.; Gutknecht, W. F.; Hodson, L.; Wilson, B. M.

Fate of Hexavalent Chromium in the Atmosphere ; Report for Air Resources Board: Sacramento, CA,, 1988; www.arb.ca.gov/research/ apr/past/a6-096-32.pdf (accessed April 24, 2015).
(36) Kawamura, K.; Steinberg, S.; Kaplan, I. R. Homologous series of C1-C10 monocarboxylic acids and C 1-C6 carbonyls in Los Angeles air and motor vehicle exhausts. Atmos. Environ. 2000, 34, 4175 −4191. (37) Lin, Y. C.; Schwab, J. J.; Demerjian, K. L.; Bae, M.-S.; Chen, W.- N.; Sun, Y.; Zhang, Q.; Hung, H.-M.; Perry, J. Summertime formaldehyde observations in New York City: Ambient levels, sources and its contribution to HO X radicals. J. Geophys. Res. 2012, 117, D08305.
(38) Parrish, D. D.; Ryerson, T. B.; Mellqvist, J.; Johansson, J.; Fried, A.; Richter, D.; Walega, J. G.; Washenfelder, R. A.; de Gouw, J. A.; Peischl, J.; Aikin, K. C.; McKeen, S. A.; Frost, G. J.; Fehsenfeld, F. C.; Herndon, S. C. Primary and secondary sources of formaldehyde in urban atmospheres: Houston Texas region. Atmos. Chem. Phys. 2012, 12, 3273 −3288.
(39) Other Search Engines/Databases ;A i rR e s o u r c e sB o a r d :

Sacramento, CA, 2014; www.arb.ca.gov/html/databases.htm (accessed April 24, 2015).
(40) California Emission Inventory Development and Reporting System (CEIDARS) 2.5 Database Structure ; Air Resources Board, Sacramento, CA, www.arb.ca.gov/ei/drei/maintain/dbstruct.htm (accessed Octo- ber 15, 2014).
(41) Mobile Source Emission Inventory Current Methods and Data ; Air Resources Board, Sacramento, CA, 2012; www.arb.ca.gov/msei/ modeling.htm (accessed January, 2013).
(42) Toxic Air Contaminants Monitoring ; Air Resources Board:

Sacramento, CA, 2011; www.arb.ca.gov/aaqm/toxics.htm (accessed April 24, 2015).
(43) iADAM: Air Quality Data Statistics ; Air Resources Board:

Sacramento, CA, 2013; www.arb.ca.gov/adam (accessed April 24, 2015).

Environmental Science & Technology Article
DOI: 10.1021/acs.est.5b02766
Environ. Sci. Technol. 2015, 49, 11329 −11339
11337

(44) 2009 Air Monitoring Network Report ; Bay Area Air Quality Management District: San Francisco, CA, 2010; www.epa.gov/ ttnamti1/ ﬁles/networkplans/CABAAQMDPlan2009.pdf (accessed April 24, 2015).
(45) Multiple Air Toxics Exposure Study in the South Coast Air Basin (MATES-II); South Coast Air Quality Management District: Diamond Bar, CA, 2000; www.aqmd.gov/home/library/air-quality-data-studies/ health-studies/mates-ii (accessed April 24, 2015).
(46) Multiple Air Toxics Exposure Study III (MATES III) ; South Coast Air Quality Management District: Diamond Bar, CA, 2008; www. aqmd.gov/home/library/air-quality-data-studies/health-studies/mates- iii/mates-iii-ﬁnal-report (accessed September 30, 2014). (47) MATES IV Multiple Air Toxics Exposure Study ; South Coast Air Quality Management District: Diamond Bar, CA, www.aqmd.gov/ home/library/air-quality-data-stu dies/health-studies/mates-iv (ac- cessed April 24, 2015).
(48) Air Toxics Data ; U.S. Environmental Protection Agency:

Washington, DC, 2008; www.epa.gov/ttnamti1/toxdat.html #data (accessed August 17, 2015).
(49) Air Toxics Data Analysis Workbook, STI-908304 −3651; U.S.

Environmental Protection Agency: Washington, DC, 2009; www.epa. gov/ttnamti1/ﬁles/ambient/airtox/workbook/T-Workbook_Secs1-8. pdf (accessed August 17, 2015).
(50) Air Toxics Hot Spots Program Risk Assessment Guidelines:

Technical Support Document for Exposure Assessment and Stochastic Analysis; California O ﬃce of Environmental Health Hazard Assess- ment, 2012; http://oehha.ca.gov/air/hot_spots/tsd082712.html (last reviewed April 24, 2015).
(51) Proposed Amended Rule 1401, New Source Review of Toxic Air Contaminants ; South Coast Air Quality Management District:

Diamond Bar, CA, draft sta ﬀ report March, 2015; www.aqmd.gov/ docs/default-source/r ule-book/Proposed-Rul es/212-1401-1401.1- and-1402/draft-staﬀ-report---revised-oehha-guidelines-march-2015.pdf (accessed August 17, 2015).
(52) Tra ﬃc Counts, California Department of Transportation; http://traﬃc-counts.dot.ca.gov (accessed May 26, 2015). (53) Interactive Data ; U.S. Department of CommerceBureau of Economic Analysis; www.bea.gov/iTable/iTable.cfm (accessed August 7, 2015).
(54) Report to the Air Resources Board on the Proposed Identi ﬁcation of Diesel Exhaust as a Toxic Air Contaminant ; Air Resources Board:

Sacramento, CA, 1998; www.arb.ca.gov/toxics/dieseltac/part_a.pdf (accessed April 24, 2015).
(55) Diesel Risk Reduction Plan ; Air Resources Board: Sacramento, CA, 2000; www.arb.ca.gov/diesel/documents/rrpapp.htm (accessed October 4, 2012).
(56) California Almanac of Emissions and Air Quality ; Air Resources Board: Sacramento, CA, 2013; www.arb.ca.gov/aqd/almanac/ almanac13/almanac13.htm (accessed August 13, 2015).
(57) Kozawa, K. H.; Park, S. S.; Mara, S. L.; Herner, J. D. Verifying emission reductions from heavy-duty diesel trucks operating on southern California freeways. Environ. Sci. Technol. 2014, 48, 1475 − 1483.
(58) Pang, Y.; Fuentes, M.; Rieger, P. Trends in the emissions of volatile organic compounds (VOCs) from light-duty gasoline vehicles tested on chassis dynamometers in southern California. Atmos. Environ. 2014, 83, 127 −135.
(59) Low-Emission Vehicle Program ; Air Resources Board: Sacramen- to, CA, 2012; www.arb.ca.gov/msprog/levprog/levprog.htm (accessed April 24, 2015).
(60) On-Board Diagnostic II (OBD II) Systems - Fact Sheet ; Air Resources Board: Sacramento, CA, 2009; www.arb.ca.gov/msprog/ obdprog/obdfaq.htm (accessed April 24, 2015).
(61) California Reformulated Gasoline Program ; Air Resources Board:

Sacramento, CA, 2013; www.arb.ca.gov/fuels/gasoline/gasoline.htm (accessed April 24, 2015).
(62) Priority Substances List Asses sment Report: 1,3-Butadiene ; Environment Canada: Ottawa, Canada, 2000; www.hc-sc.gc.ca/ewh- semt/alt_formats/hecs-sesc/pdf/pubs/contaminants/psl2-lsp2/1_3_ butadiene/1_3_butadiene-eng.pdf (accessed November 20, 2014). (63) Duffy, B. L.; Nelson, P. F.; Ye, Y.; Weeks, I. A.; Galbally, I. E.

Emissions of benzene, toluene, xylenes and 1,3-butadiene from a representative portion of the Australian car fleet - models and measurements. Atmos. Environ. 1998, 32 (14), 2693 −2704. (64) Ye, Y.; Galbally, I. E.; Weeks, I. A. Emission of 1,3-butadiene from petrol-driven motor vehicles. Atmos. Environ. 1997, 31 (8), 1157−1165.
(65) Staﬀ Report, Proposed Airborne Toxic Control Measure for Perchloroethylene Dry Cleaning Operations, and Technical Support Document; Air Resources Board, 1993.
(66) ATCM for Automotive Maintenance and Repair Activities ; Air Resources Board: Sacramento, CA, 2001; www.arb.ca.gov/toxics/amr/ amr.htm (accessed April 15, 2014).
(67) Consumer Products Program - Perchloroethylene Activity ; Air
Resources Board: Sacramento, CA, 2011; www.arb.ca.gov/consprod/ regact/perch/perch.htm (accessed April 25, 2015).
(68) Tirez, K.; Silversmit, G.; Bleux, N.; Adriaensens, E.; Roekens, E.; Servaes, K.; Vanhoof, C.; Vincze, L.; Berghmans, P. Determination of hexavalent chromium in ambient air: A story of method-induced Cr (III) oxidation. Atmos. Environ. 2011, 45, 5332 −5341.
(69) Technical Support Document to Proposed ATCM for Hexavalent Chromium Emissions from Chrome Plating and Chromic Acid Anodizing Operations; Air Resources Board: Sacramento, CA, 1988.
(70) National Emission Standards for Hazardous Air Pollutants; Proposed Standards for Chromium Emissions from Hard and Decorative Chromium Electroplating and Chromium Anodizing Tanks , No. 240, Part II, 58 FR 65768, December 16, 1993; U.S. Environmental Protection Agency: Washington, DC, 1993; Vol. 58, www.epa.gov/ttn/atw/ chrome/fr16de93.pdf (accessed April 25, 2015).
(71) Chrome Plating Operations − Background, Air Resources Board, Sacramento, CA, www.arb.ca.gov/toxics/chrome/background.htm (ac- cessed November 14, 2012).
(72) Cooling Towers Airborne Toxic Control Measure ; Air Resources Board: Sacramento, CA, 1989; www.arb.ca.gov/toxics/atcm/cltwatcm. htm (accessed April 25, 2015).
(73) The California Low-Emission Vehicle Regulations (with Amend- ments E ﬀective December 31, 2012) ; Air Resources Board, Sacramento, CA, 2012; www.arb.ca.gov/msprog/levprog/cleandoc/cleancomplete_ lev-ghg_regs12-12.pdf (accessed April 24, 2015).
(74) Kean, A. J.; Grosjean, E.; Grosjean, D.; Harley, R. A. On-Road Measurement of Carbonyls in California Light-Duty Vehicle Emissions. Environ. Sci. Technol. 2001, 35, 4198 −4204. (75) Ban-Weiss, G. A.; McLaughlin, J. P.; Harley, R. A.; Kean, A. J.; Grosjean, E.; Grosjean, D. Carbonyl and nitrogen dioxide emissions from gasoline- and diesel-powered motor vehicles. Environ. Sci.

Technol. 2008, 42, 3944 −3950.
(76) Air Quality Impacts of the Use of Ethanol in California Reformulated Gasoline, Final Report to the California Environmental Policy Council; Air Resources Board: Sacramento, CA, 1999; www.arb. ca.gov/fuels/gasoline/ethanol/ethfate/Airq/Mainf.pdf (accessed May 12, 2015).
(77) Rulemaking to Consider Adoption of the Proposed Airborne Toxic Control Measure (ATCM) to Reduce Formaldehyde Emissions from Composite Wood Products ; Air Resources Board: Sacramento, CA, 2007; www.arb.ca.gov/regact/2007/compwood07/compwood07.htm (accessed May 4, 2014).
(78) National Air Toxics Trends Stations ; U.S. Environmental Protection Agency: Washington, DC, 2011; www.epa.gov/ttnamti1/ natts.html (accessed April 7, 2014).
(79) Interagency Monitoring of Protected Visual Environments ; Colorado State University, Fort Collins, CO, http://vista.cira. colostate.edu/improve (accessed April 24, 2015).
(80) Rice Straw Management ; Air Resources Board, Sacramento, CA, www.arb.ca.gov/smp/rice/rice.htm (accessed April 24, 2015). (81) Air Quality System ; U.S. Environmental Protection Agency:

Washington, DC, 2010; www.epa.gov/ttn/airs/airsaqs (accessed October 8, 2012).

Environmental Science & Technology Article
DOI: 10.1021/acs.est.5b02766
Environ. Sci. Technol. 2015, 49, 11329 −11339
11338

(82) Ambient Concentrations of Selected Air Toxics ; U.S. Environ- mental Protection Agency: Washington, DC, 2014; http://cfpub.epa. gov/roe/indicator.cfm?i=90 (accessed November 7, 2014). (83) Dann, T. F.; Wang, D. K. Ambient air benzene concentrations in Canada (1989 −1993): Seasonal and day of week variations, trends, and source influences. J. Air Waste Manage. Assoc. 1995, 45, 695−702. (84) Aleksic, N.; Boynton, G.; Sistla, G.; Perry, J. Concentrations and trends of benzene in ambient air over New York State during 1990 − 2003. Atmos. Environ. 2005, 39 (40), 894 −7906.
(85) Curren, K. C.; Dann, T. F.; Wang, D. K. Ambient air 1,3- butadiene concentrations in Canada (1995 −2003): Seasonal, day of week variations, trends, and source influences. Atmos. Environ. 2006, 40, 170 −181.
(86) Lurmann, F.; Avol, E.; Gilliland, F. Emissions reduction policies and recent trends in Southern California ’s ambient air quality. J. Air Waste Manage. Assoc. 2015, 65 (3), 324 −335.
(87) Our Nation ’s Air; Air Quality and Trends ; U.S. Environmental Protection Agency: Washington, DC, www.epa.gov/airtrends/2011/ report/fullreport.pdf (accessed June 4, 2015).
(88) Zhang, H. R.; Eddings, E. G.; Sarofim, A. F. Pollutant emissions from gasoline combustion. 1. Dependence on fuel structural functionalities. Environ. Sci. Technol. 2008, 42, 5615 −5621. (89) Schuetzle, D.; Siegl, W. O.; Jensen, T. E.; Dearth, M. A.; Kaiser, E. W.; Gorse, R.; Kreucher, W.; Kulik, E. The relationship between gasoline composition and vehicle hydrocarbon emissions: a review of current studies and future research needs. Environ. Health Perspect. 1994, 102 (Suppl. 4), 3 −12.
(90) Kirchstetter, T. W.; Singer, B. C.; Harley, R. A.; Kendall, G. R.; Hesson, J. M. Impact of California reformulated gasoline on motor vehicle emissions. 2. Volatile organic compound speciation and reactivity. Environ. Sci. Technol. 1999, 33, 329 −336.
(91) Harley, R. A. On-Road Measurement of Light-Duty Gasoline and Heavy-Duty Diesel Vehicle Emissions ; Final Report for Air Resources Board: Sacramento, CA, 2008; www.arb.ca.gov/research/apr/past/05- 309.pdf (accessed May 18, 2015).
(92) Rock, K.; Judzis, A.; Almering, M. Cost e ﬀective solutions for reducing benzene in gasoline , Presentation for the National Petrochem- ical and Re ﬁners Association, 2008. www.cbi.com/images/uploads/ technical_articles/
CostE ﬀectiveSolutionsforReductio nofBenzeneinGasoline- KerryRockCDTech-AM-08-14.pdf (accessed May 18, 2015).
(93) Initial Statement of Reasons for Rulemaking, Proposed Identi ﬁcation of Perchloroethylene as a Toxic Air Contaminant, Sta ﬀ Report; Air Resources Board: Sacramento, CA, 1991; www.arb.ca.gov/toxics/id/ summary/summary.htm (accessed April 24, 2015).
(94) Truck and Bus Regulation; On-Road Heavy-Duty Diesel Vehicles (In-Use) Regulation ; Air Resources Board: Sacramento, CA, 2012; www.arb.ca.gov/msprog/onrdiesel/onrdiesel.htm (accessed December 30, 2014).

Environmental Science & Technology Article
DOI: 10.1021/acs.est.5b02766
Environ. Sci. Technol. 2015, 49, 11329 −11339
11339

CONTRIBUTORS

DETAILS
All downloadable National Academies titles are free to be used for personal and/or non-commercial academic use. Users may also freely post links to our titles on this website; non-commercial academic users are encouraged to link to the version on this website rather than distribute a downloaded PDF to ensure that all users are accessing the latest authoritative version of the work. All other uses require written permission. (Request Permission)
This PDF is protected by copyright and owned by the National Academy of Sciences; unless otherwise indicated, the National Academy of Sciences retains copyright to all materials in this PDF with all rights reserved.

Visit the National Academies Press at nap.edu and login or register to get:
– Access to free PDF downloads of thousands of publications – 10% off the price of print publications
– Email or social media notiﬁcations of new titles related to your interests – Special offers and discounts

SUGGESTED CITATION

BUY THIS BOOK

FIND RELATED TITLES
This PDF is available athttp://nap.nationalacademies.org/645 Ecological Knowledge and Environmental
Problem-Solving: Concepts and Case
Studies (1986)
400 pages | 6 x 9 | PAPERBACK
ISBN 978-0-309-03645-0 | DOI 10.17226/645
Committee on the Applications of Ecological Theory to Environmental Problems; National Research Council
National Research Council. 1986. Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies. Washington, DC: The National Academies Press. https://doi.org/10.17226/645.

EcologicalEcologicalEcologicalEcological KnowledgeKnowledgeKnowledgeKnowledge andandandand EnvironmentalEnvironmentalEnvironmentalEnvironmental Problem-SolvingProblem-SolvingProblem-SolvingProblem-Solving

CONCEPTS AND CASE STUDIES
Committee on the Applications of Ecological Theory to
Environmental Problems
Commission on Life Sciences
National Research Council

NATIONAL ACADEMY PRESS
Washington, D.C. 1986
i
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved.

National Academy Press 2101 Constitution Ave., NW Washington, DC 20418 NOTICE: The project that is the subject of this report was approved by the Governing Board of the National Research Council, whose members are drawn from the councils of the National Academy of Sciences, the National Academy of Engineering, and the Institute of Medicine. The members of the committee responsible for the report were chosen for their special competences and with regard for appropriate balance.

This report has been reviewed by a group other than the authors according to procedures approved by a Report Review Committee consisting of members of the National Academy of Sci- ences, the National Academy of Engineering, and the Institute of Medicine The Research Council was established by the National Academy of Sciences in 1916 to asso- ciate the broad community of science and technology with the Academy's purposes of furthering knowledge and of advising the federal government. The Research Council operates in accordance with general policies determined by the Academy under the authority of its congressional charter of 1863, which establishes the Academy as a private, nonprofit, self-governing membership corpora- tion. The Research Council has become the principal operating agency of both the National Academy of Sciences and the National Academy of Engineering in the conduct of their services to the government, the public, and the scientific and engineering communities. It is administered jointly by both Academies and the Institute of Medicine. The National Academy of Engineering and the Institute of Medicine were established in 1964 and 1970, respectively, under the charter of the National Academy of Sciences.

This study has been supported by the National Research Council Fund, a pool of private, discre- tionary, nonfederal funds that is used to support a program of Academy-initiated studies of national issues in which science and technology figure significantly. The NRC Fund consists of contributions from a consortium of private foundations, including the Carnegie Corporation of New York, the Charles E. Culpeper Foundation, the William and Flora Hewlett Foundation, the John D. and Cather- ine T. MacArthur Foundation, the Andrew W. Mellon Foundation, the Rockefeller Foundation, and the Alfred P. Sloan Foundation; the Academy Industry Program, which seeks annual contributions from companies that are concerned with the health of U.S. science and technology and with public policy issues with technological content; and the National Academy of Sciences and the National Academy of Engineering endowments.

The study also was supported in part by the U.S. Environmental Protection Agency under Assistance Agreement No. CX-811943-01-0 to the National Academy of Sciences; however, the information in this document may not necessarily reflect the views of the agency and no official endorsement should be inferred.

Library of Congress Cataloging-in-Publication Data
Ecological knowledge and environmental problem-solving: Concepts and case studies Bibliography: p.

1. Nature conservation—Case studies. 2. Environmental protection—Case studies. 3.

Conservation of natural resources—Case studies. 4. Ecology. I. National Research Coun- cil (U.S.). Commission on Life Sciences. Committee on Applications of Ecological The- ory to Environmental Problems.

QH75.U86 1986 363.7 86-817
ISBN 0-309-03645-3
Printed in the United States of America
First Printing, March 1986
Second Printing, October 1987
Third Printing, October 1988
Fourth Printing, October 1989
Fifth Printing, February 1991
Sixth Printing, February 1995
ii
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved.

Committee on the Applications of Ecological Theory to
Environmental Problems
GORDON H. ORIANS, Chairman, Institute for Environmental Studies, University of Washington, Seattle, Washington
JOHN BUCKLEY, P.O. Box 263, Whitney Point, New York
WILLIAM CLARK, International Institute for Applied Systems Analysis, Laxenburg, Austria
MICHAEL E. GILPIN, Department of Biology, University of California at San Diego, La Jolla, California
CARL F. JORDAN, Institute of Ecology, University of Georgia, Athens, Georgia JOHN T. LEHMAN, Division of Biological Science, University of Michigan, Ann Arbor, Michigan
ROBERT M. MAY Department of Biology, Princeton University, Princeton, New Jersey
GORDON A. ROBILLIARD Entrix Inc., Concord, California
DANIEL S. SIMBERLOFF Department of Biological Science, The Florida State University, Tallahassee, Florida
Consultant
W. JAMES ERCKMANN, Institute for Environmental Studies, University of Washington, Seattle, Washington
National Research Council Staff
DAVID POLICANSKY, Staff Officer
NORMAN GROSSBLATT, Editor
AGNES GASKIN, Secretary
iii
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. iv
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved.

Preface
This report is the product of the recognition by the Board on Basic Biology of the National Research Council's Commission on Life Sciences that it should be concerned with the basic biology of ecosystems. The first aspect of ecology to receive attention by the Board was the current status of ecological theory and concepts and their applicability to specific environmental problems. A workshop on this topic had been held September 22-23, 1979, at the National Academy of Sciences in Washington, D.C. The 15 participants represented diverse approaches to ecology and had extensive experiences in marine, freshwater, and terrestrial environments. Discussion during the workshop made it clear that various approaches to ecology could profitably be evaluated from the point of view of the data needed to test them, the difficulty of obtaining those data, the domain over which models can be presumed to be applicable, and special precautions in the use of results. The group also recognized that further exploration of models and concepts would be more productive if it were done in the context of potential applications to environmental problem-solving.

The deliberations of this workshop laid the groundwork for the establishment of the Committee on Applications of Ecological Theory to Environmental Problems, which consists of persons with varied backgrounds in basic ecology, environmental management, and problem-solving. The chief basis of the activities of the committee was the perception that, whereas much about the functioning of ecological systems remains poorly PREFACE v
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. understood, we commonly fail to use even available information when attempting to solve environmental problems. This failure has many causes, among which are the difficulty of determining which components of ecological knowledge might be most usefully applied to particular environmental problems and the difficulty of finding the relevant information. Those difficulties in turn are partly a result of the failure of academic ecologists to attempt to determine where and how their knowledge might be applied and of a lack of communication among generators and users of new knowledge.

The committee recognized that it could not attempt to cover all aspects of ecology without producing a report of unacceptable length. Therefore, we concentrated our attention on ecological knowledge — broadly conceived to include theories, models, data, and concepts — and left methods and techniques of data analysis to be considered in detail elsewhere. This concentration should not be construed as a belief that methods and techniques are unimportant. The committee fully recognizes that such methods as satellite imaging and the use of powerful computers permit analyses of ecological systems on scales of space and time that were previously impossible. These methods are more than just "tools." Nonetheless, the most powerful analytical systems are not substitutes for biological insights or imaginative questioning and hypothesizing. It is to these issues that the committee directed the bulk of its efforts.

The committee also recognized that a report dominated by abstract considerations of ecological knowledge would fail to convey vividly and convincingly how ecological knowledge might improve environmental problem- solving. Therefore, we decided to develop a set of case histories that would illustrate in concrete ways how ecological information has been used and how it has influenced the ways in which problems have been conceived and solutions approached. Because the manifold problems with the quality of assessments of environmental effects of projects have been well noted and thoroughly discussed by others, the committee decided to use examples of the creative use of ecological information, believing that a good example is more instructive than a bad one.

The committee carried out its tasks by means of a series of meetings in which the major format of the report was decided on. Once the plan of the report was determined, subgroups were formed to draft the various chapters of Part I. Another subgroup had the specific task of finding a series of acceptable cases for study, choosing the best ones from among a long list, and selecting persons to write the case studies. The cases were selected to illustrate the use of ecological information in dealing with a wide variety of environmental problems, from management of single spe
PREFACE vi
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. cies to issues involving complete ecosystems over broad areas. Some of the case studies were written by persons actively involved in the projects themselves. Others were written by members of the committee who had familiarity with the examples and who could therefore write about them with some depth of understanding. To make the case studies more informative, the committee prepared an evaluation of each one, pointing out how the use of ecological information had been particularly effective and how even better use of ecological knowledge could have been made. We hope that those comments will help to improve future studies of similar types.

The work of the committee was supported largely by the National Research Council Fund. Additional support was received from the Office of Research and Development of the Environmental Protection Agency for preparing material on biological monitoring ( Chapter 7). The Office of Federal Activities of the Environmental Protection Agency and the Office of Environmental Analysis of the Department of Energy provided support for a joint U.S.-Canadian workshop on cumulative environmental effects. The workshop provided much of the basis of the chapter on cumulative environmental effects (Chapter 9).

A project of this magnitude depends on the cooperation and efforts of many persons in addition to those serving on the committee. In our attempts to find good case studies, we contacted many ecologists in North America and Europe. We give special thanks to the authors of the case studies, who labored so hard to provide us with documents that followed the basic format we had established for them and who were also able to convince us that it was sometimes inappropriate to follow those guidelines slavishly.

The report has been vastly improved by the devoted efforts of the National Research Council staff under the guidance of David Policansky. Dr. Policansky not only provided a valuable interface between the committee and the Research Council, but also contributed extensively to the conceptual development of the project and to the writing of the report, including a case study. He was the person who consistently and insistently reminded the committee members of their duties and responsibilities, thereby preventing the report from appearing even later than it has. Similarly, W. James Erckmann, who assisted in production of the report, played a key role in the writing of the chapters of Part I and in the development of the project and the case studies. We are deeply indebted to Norman Grossblatt of the Commission on Life Sciences for his thoughtful, careful editing; he has immeasurably improved the clarity and style of the report. The report has also benefited from the patient and insightful clerical support of Agnes Gaskin of the Board on Basic Biology and PREFACE vii
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved.

Jeanette Pederson of the Institute for Environmental Studies, as well as from the comments of several reviewers. Without support of such quality and quantity, our report could not have achieved the quality we now believe it possesses.

GORDON H. ORIANS

CHAIRMAN

COMMITTEE ON APPLICATIONS OF ECOLOGICAL THEORY TO

ENVIRONMENTAL PROBLEMS
PREFACE viii
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved.

Contents
Introduction 1
I Kinds of Ecological Knowledge and Their Applications

1. Individuals and Single Populations 23

2. Population Interactions 38

3. Community Ecology 47

4. Materials and Energy 61

5. Scales in Space and Time 68

6. Analog, Generic, and Pilot Studies and Treatment of a Project as an Experiment
75 7. Indicator Species and Biological Monitoring 81

8. Dealing with Uncertainty 88
CONTENTS ix
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved.

9. The Special Problem of Cumulative Effects 93

10. A Scientific Framework for Environmental Problem-Solving 104

11. References 116
II Selected Case Studies

12. North Pacific Halibut Fishery Management
Case Study by David Policansky
137

13. Vampire Bat Control in Latin America
Case Study by G. Clay Mitchell
151

14. Biological Control of California Red Scale
Case Study by Robert F. Luck
165

15. Experimental Control of Malaria in West Africa
Case Study by Robert M. May
190

16. Protecting Caribou During Hydroelectric Development in Newfoundland
Case Study by David J. Kiell, Edward L. Hill, and Shane P. Mahoney
205

17. Conserving a Regional Spotted Owl Population
Case Study by Hal Salwasser
227

18. Restoring Derelict Lands in Great Britain
Case Study by Peter Wathern
248

19. Optimizing Timber Yields in New Brunswick Forests
Case Study by Thom A. Erdle and Gordon L. Baskerville
275
CONTENTS x
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved.

20. Control of Eutrophication in Lake Washington
Case Study by John T. Lehman
301

21. Raising the Level of a Subarctic Lake
Case Study by John T. Lehman
317

22. Ecological Effects of Nuclear Radiation
Case Study by Carl F. Jordan
331

23. Ecological Effects of Forest Clearcutting
Case Study by Carl F. Jordan
345

24. Environmental Effects of DDT
Case Study by John Buckley
358
Index 375
CONTENTS xi
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved.

CONTENTS xii
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved.

Introduction
Concern for the quality of the environment has been evident in the United States for over 100 years. One result of this concern has been the passage of laws, regulations, and treaties that require actions to preserve environmental quality or that inhibit actions detrimental to the environment. These legislative acts have caused the investment of large amounts of time, effort, and money in analyses of the possible environmental impacts of specific development projects or management decisions. For two reasons, the effectiveness of these analyses and the magnitude of resources devoted to them are controversial (Goodman, 1975; Schindler, 1976; Suter, 1981): available scientific information is often not used effectively in their preparation, and research conducted to provide additional information is often inadequate. It is also commonly perceived that massive studies yield predictions that are little better than those produced with much less effort, cost, and time. This report, by the Committee on Applications of Ecological Theory to Environmental Problems, in the Board on Basic Biology of the National Research Council's Commission on Life Sciences, deals with both problems. It explores how the scientific tools of ecology can be used more effectively in dealing with a variety of environmental problems. ''Ecological theory," as described in standard textbooks on ecology, is seldom applied directly to environmental problems. But ecological "knowledge"— including not only theory, but also facts, observations, research results, syntheses, models, and methods of investigation — has been extremely important in developing approaches to a wide range of
INTRODUCTION 1
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. environmental problems. This report discusses knowledge that has proved useful and how it has been used. It deals with a wide range of environmental issues, including prediction and management of environmental impacts, management of renewable resources, protection and restoration of species and ecosystems, control of agricultural and silvicultural pests, and use of generic ecological studies to promote understanding of classes of environmental problems. Rather than focusing on what is wrong with the way ecology is often applied, we use successful applications to show how ecological knowledge can be valuable when used appropriately.

Environmental assessment and management are based on societal goals, which are not always clearly articulated. The social effects of environmental changes and the way goals are developed by society are important aspects of environmental problems. Nevertheless, we treat only the biological side of environmental problems, to avoid taking on an unmanageably large task.

The report does not attempt to cover all aspects of ecology. To have done so would have made our task unacceptably difficult and the report unacceptably long. We have focused on different kinds of ecological knowledge and appropriate arenas for their use, but have restricted the coverage of subjects and have not provided details of methods, except as examples or in the case studies.

The subjects covered were determined by the experiences of the committee members. Ecological subjects that have interfaces with chemistry, geology, and hydrology have been slighted, except in some of the case studies (e.g., Chapters 20, 21, 23, and 24). Thus, detailed discussions of such subjects as CO 2 buildup,
acid deposition, ecotoxicology, and the effects of complex chemical mixtures in the environment are lacking. For treatment of these important topics, the reader is referred to several recent reviews (e.g., Harshbarger and Black, in press; Maki et al., 1980; National Research Council, 1976, 1981, 1982, 1983a,b,c, 1984, in press a,b; Reif, 1984).

Among the most important methods — some of which have changed the scope and scale of ecological research — that do not receive explicit treatment in this report are equations for estimating environmental partitioning of chemicals and related phenomena, such as transformations and transport; hazard evaluation protocols and risk assessment; validation; decision analysis; remote sensing; interpretation of photographs and satellite photography; and bioassays.

Some methods— such as modeling, statistical analysis, and use of computers — are not discussed specifically, but appear in many of the chapters of Part I and the case studies.

INTRODUCTION 2
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved.

STRUCTURE AND AUDIENCE OF THE REPORT
Structure
The report is divided into "What we know" ( Part I ) and "How our knowledge is applied" ( Part II ). Chapters 1-5 treat a number of categories of ecological knowledge and Chapters 6-10 discuss procedures and kinds of knowledge that are common to most environmental problems. Part II presents studies of 13 specific cases of environmental problem-solving.

The first three chapters deal with populations of single species, interactions between two species, and community ecology. They focus on the kinds of information that are available and how that information might be useful in solving environmental problems. Chapter 4 discusses the flow of energy and materials through ecosystems; Chapter 5 deals with the effects of changes in time and space scales on ecological processes and products. Chapter 6 explores the benefits of treating environmental manipulations as large-scale experiments so that we can learn more from them. Chapters 7 and 8 discuss biological monitoring and the management of ecological systems in the face of the substantial uncertainty associated with the behavior of those systems.

Chapter 9 deals with cumulative environmental changes that result from repeated perturbations. There is increasing recognition that some of our most severe environmental problems involve the cumulative effects of many small local actions — individually insignificant, but collectively creating major regional and even global changes. The discussion points out the major issues associated with cumulative effects, which have only recently begun to receive serious consideration, but which present some of the most difficult challenges to decision-makers.

Chapter 10 deals with procedures for application of ecological knowledge.

It begins with a discussion of how ecology can help in conceptualizing a problem and in devising appropriate approaches to its solution. We focus on the design of ecological studies for predicting and managing the effects of manipulations, emphasizing the importance of monitoring projects and decisions to determine their consequences.

Part II discusses 13 cases chosen as instructive examples of the application of ecological knowledge to various kinds of problems. The cases were selected because the committee members and consultants who were asked for suggestions were familiar with them and believed that they would reveal the challenges of applying ecological knowledge, show the difficulty of predicting accurately the behavior of complex systems, and point out the substantial value of monitoring the results of manipulations. The
INTRODUCTION 3
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. cases were chosen to exemplify scientifically sound uses of ecological knowledge. No attempt was made to achieve comprehensive coverage; of necessity, the cases are a small sample of a large universe of potentially informative examples.

Audience
This report is intended for a broad audience, including:
Those who prepare, receive, and use environmental evaluations and management plans.
Legislators and regulators concerned with environmental issues. Teachers and students in the natural and environmental sciences.

TYPES OF ECOLOGICAL KNOWLEDGE
The chapters in Part I discuss the ways in which ecological knowledge has been applied in a variety of environmental problems, from the management of renewable resources to the prediction of environmental impacts. A few common themes run through these chapters. The themes are related to the basic characteristics of ecological systems that are of fundamental importance to understanding their responses to perturbations.

Complex Linkages
Because of the complex linkages between species in ecosystems, the effects of changes are often indirect. Obvious and direct influences on the objects of study are sometimes not as important as less obvious indirect influences. Problem-solvers can find it helpful to develop diagrams that show both direct and indirect influences and show the pathways through which indirect agents exert their influences (Andrewartha and Birch, 1984). Such diagrams are accounting schemes that help to identify important environmental factors. The case studies in Part II contain many examples of indirect effects, such as the effect of DDT on birds far from sites of its application ( Chapter 24) and the mobilization of mercury that results from changing the level of a lake (Chapter 21). These phenomena could not have been predicted by analyses that concentrated on direct effects. In addition, many ecological interactions are characterized by strong nonlinear effects brought about by slight changes in key factors.

INTRODUCTION 4
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved.

Density Dependence
Because environments are finite and resources might not be renewed as quickly as they are harvested, the success of individuals often depends on the number of individuals of the same and other species in their environments.

Often, but not always, conditions deteriorate for individuals as the numbers of individuals of the same species, predators, and parasites increase. Conversely, conditions can improve when the numbers of individuals of those classes decrease. It is therefore often possible to extract much higher sustainable yields from populations than would be expected on the basis of an examination of birth, growth, and death rates in an unharvested population existing close to the limit set by environmental resources. Alternatively, high population densities in most areas could prevent individuals displaced by a perturbation from moving into other areas. Very low densities might influence the probability of finding mates, erode genetic diversity of a population, and increase the probability of its extinction. The existence of density-dependent effects does not necessarily imply that populations exist at steady densities. These effects can cause populations to fluctuate regularly or irregularly (May 1973), as did the blowflies studied by Nicholson and Bailey (1935) and the bean weevil Callosubrochus chinensis studied by Utida (1957). Moreover, density-independent effects can often mask density-dependent ones, because they are often much larger.

The Uniqueness of Individuals
Each individual in a sexually reproducing population is unique, as opposed to many nonliving entities (e.g., all CO
2 molecules can be treated as identical).

A species is composed of different age and sex classes, and genetic variation occurs within each class. Plants with open growth systems often have substantial intraindividual variability that is important for consumers of the plants (Whitham et al., 1984). Because of genetic variability, management practices can create selective pressures that cause genetic and thus phenotypic changes in a population that sometimes subvert the goals of the management practices (Chapter 1).

Keystone Species
Keystone species are those which exert influences over other members of their ecological communities out of proportion to their abundances. Keystone species can have various roles in ecological communities. For example, dominant plants are the major photosynthesizers in communities, INTRODUCTION 5
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. and they also form the physical structure in which many interactions in the system take place. Keystone predators preferentially eat prey that would be competitive dominants in the absence of predation. As a result, keystone predators create conditions favorable for the existence of species that are crowded out of systems with less predation. Keystone predators have been most commonly identified among predators on sessile prey — such as plants and sessile marine invertebrates of rocky shores — that compete for space (Harper, 1969; Hurlbert and Mulla, 1981; Lubchenco, 1978; Morin, 1983; Paine, 1974; Paine and Vadas, 1969). Keystone mutualists are essential for survival of other members of their communities, even though the amount of resources they consume is sometimes small. Well-known examples are pollinators and frugivores, ants that patrol and protect plants, and microorganisms associated with vascular plants (Alexander, 1971; Gilbert, 1975; Howe, 1981; Janzen, 1966). Because changes in the populations of keystone species can influence community dynamics in major ways, astute environmental problem-solvers are alert to keystone relationships in the communities being altered (Chapter 3).

Biological Magnification
Organisms are united through food chains, so materials taken up by prey can accumulate and become concentrated in their predators. Passage of substances through several trophic levels in a community can result in concentrations hundreds of times those initially present in the environment. The greatly magnified indirect effects of pesticides like DDT, which are metabolically stable and fat-soluble, constitute a striking case of biological magnification (Chapter 24).

Population Fragmentation
Human activities often change large patches of, say, forest or prairie into small patches surrounded by land devoted to other purposes (such as agriculture, forest plantations, highways, and cities). Fragmentation of habitats sets in motion processes that can cause unexpected changes in the system.

Among these processes is the loss of species unable to exist in small patches (Chapter 5 ). An important management implication is the need to provide corridors of suitable habitat through which individuals can travel between more extensive patches of appropriate habitat. Ignoring the larger, heterogeneous environment in which local populations of interest are set can lead to poor management of those populations (Chapters 17 and 19).

INTRODUCTION 6
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved.

Stability Boundaries
"Stability," the tendency of ecological systems to remain in a relatively constant state, can result from several processes (Orians, 1975). Of concern to the manager and environmental problem-solver are the resistance of a system or a component of it to change, the speed with which it returns to its previous state when the perturbation ends, and its stability boundaries — the range over which it can be changed without leaving it unable to return to its previous state (Holling, 1973). The crossing of stability boundaries or thresholds by individual organisms can lead to large changes in their functioning. Photoperiods above a threshold might be necessary to induce a plant to flower or an animal to come into breeding condition (e.g., Farner, 1964; Salisbury and Ross, 1978; Sundararaj and Vasali, 1976). A critical minimal density of prey might be necessary for a particular kind of predator to survive in an area. A specific patch size of suitable habitat might be necessary for a particular species to survive (Chapter 17). An organism might be able to withstand temperatures down to or up to a critical point, but not beyond.

Stability boundaries are also important for more complex ecological systems. For example, environmental change and overharvesting can cause persistent changes in the composition of fish stocks (Daan, 1980; Gulland and Garcia, 1984; Chapters 1 and 8). Overgrazed plant communities might be resistant to invasion by new species, even if grazing ceases. These "alternate stable states" are usually less desirable from the human perspective, but some are seen as beneficial, as when a shrub stage resistant to invasion by trees is established as a low-maintenance community along roads and powerline rights of way (Niering and Egler, 1955). Managing systems to achieve greater constancy, a common human objective, can reduce their ability to withstand perturbations. For example, the suppression of spruce budworm populations in eastern Canada through the use of insecticides achieved the intended preservation of the pulp and paper industry in the short run, but left the forests and the economy more vulnerable to outbreaks of a size and intensity that would be impossible in undisturbed forests (Holling, 1973).

It is especially important for the environmental problem-solver to recognize that these thresholds are sometimes not apparent until after they have been crossed. An important role of ecological knowledge is to suggest what such thresholds are likely to be.

Aggregate Variables
Ecologists and other scientists use aggregate variables in several ways.

Typically, entities are grouped on the basis of shared characteristics. These INTRODUCTION 7
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. might be traits shared because of common ancestry, such as the dorsal, hollow nervous system of all vertebrates; they might result from independent adaptations, as in the aggregate entity "herbivores." Ecological guilds — groups of species that use common resources in similar ways — are examples of aggregates consisting of units connected by strong interactions. Aggregates can also be based on some composite process or product, such as productivity or biomass, that is deemed to be of theoretical or applied interest.

Analysis of complex systems requires the use of aggregate variables.

Aggregate variables should be designed to capture the essence of traits or processes. It is the nature of aggregate variables that they lose substantial amounts of information; but essential information should not be sacrificed. A major challenge in using ecological information to help in solving environmental problems is to delineate aggregate variables, their properties, and what will be gained and lost if they are adopted.

Complexity and Uncertainty
Ecological theory is most fully developed for relatively simple systems, which have relatively small numbers of interacting units. Ecology's greatest predictive success occurs in cases that involve only one or two species. That is why, for example, management of game and fish populations through regulation of hunting and fishing is often successful. Predicting the outcome of interactions among many species is much more difficult. Therefore, most environmental problem-solving entails considerable uncertainty. Wise problem- solvers learn to expect the unexpected and to develop approaches that allow them to learn as they proceed, altering their responses as they learn more about the system and avoiding irrevocable actions (Chapter 8).

Scales in Space and Time
Processes and products in ecological systems are strongly influenced by differences in scale (Addicott, 1978; Andrewartha and Birch, 1954, 1984; Levin, 1974; Wiens, 1976; Chapter 5). Large lakes are not simply smaller lakes made bigger, and a large expanse of forest is not the same as a collection of small woodlots. Patchy environments have more edges in proportion to their area than do homogeneous environments; they create dispersal problems for organisms inhabiting them, and they lead to high local extinction rates.

Speciation, particularly of large organisms, requires large expanses of suitable habitat (Soulé and Wilcox, 1980)— larger, in fact, than most parks and reserves.

Success in achieving some local goal
INTRODUCTION 8
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. can set in motion processes that cause regional results opposite to those desired (Chapter 19 ). Populations of some species are maintained by high rates of reproduction in a small fraction of the area that they occupy; populations in the rest of the area depend on a steady supply of immigrants from sites with high reproductive rates. Evolutionary changes can cause both interacting organisms and the results of their interactions to change, often rapidly enough to be important for the environmental problem-solver (Chapter 1).

APPLICATIONS OF ECOLOGICAL KNOWLEDGE
The chapters of Part I are organized in part by their ecological content.

Chapters 1-5 deal with a variety of kinds of ecological knowledge, many having specific applicability to different kinds of problem-solving. Chapters 6-10 deal with knowledge and procedures that are more general in their applicability. We briefly summarize here the parts of the report that deal with particular kinds of application (see Table 1 ) so that users can more easily tap the sources of knowledge most important to them.

Renewable-Resource Management
Renewable-resource management refers to commercial and recreational harvesting of animals and plants. The important topics of population growth, life histories, habitat selection and other behaviors, genetics of populations, and evolution in response to management are dealt with in Chapter 1 . Equally important for managers are interactions between populations ( Chapter 2 ) and energetics— i.e., productivity and nutrient requirements ( Chapter 4 ). Examples of renewable-resource management are discussed
TABLE 1 Kinds of Applications of Ecological Knowledge and Where They Are Discussed
Application Chapter
Renewable-resource management 1, 2, 4, 12, 19
Conservation of species 1-3, 5, 8, 17
Control of pests and diseases 1-3, 13-15, 24
Impact assessment and prediction of effects 6-9, 16, 21-24 Preservation of communities 3-5, 8, 18, 20
Preservation of habitat 7-9, 17, 18, 20, 23
Contaminants and toxic substances 4, 7, 9, 20-22, 24
Mitigation of effects of construction 7, 10, 16, 18, 21 Restoration 3-5, 8, 18
General applications 6-10
INTRODUCTION 9
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. in detail in Chapter 12 (managing a fishery) and Chapter 19 (managing a forest).

Conservation of Species
Closely related to the commercial and recreational harvest of plants and animals is the protection of species that lack commercial value, but that have aesthetic value or that play important roles in ecosystems valued for other reasons. Species conservation requires knowledge of the ecology of the species (Chapter 1), its interactions with other species ( Chapter 2), and its relationships to its ecological community ( Chapter 3 ). The size and shape of the available habitat and the population size of the species needed to ensure its survival for acceptably long periods are central to management decisions ( Chapter 5), as is knowledge of environmental and demographic variability ( Chapter 8 ).

Application of such knowledge to the conservation of populations of spotted owls is discussed in Chapter 17.

Control of Pests and Diseases
The control of pests and diseases usually requires knowledge of the natural history and population characteristics of the pests and their victims ( Chapter 1) and understanding of their interactions ( Chapter 2 ) and of the ecological communities in which they occur ( Chapter 3 ). Attempts to control pests and diseases often result in the evolution of resistance, which tends to reduce the effectiveness of the programs ( Chapter 1 ). Cases of control are discussed in Chapter 13 (vampire bats), Chapter 14 (biological control of citrus pests), Chapter 15 (malaria vectors), and Chapter 24 (results of attempts to control insects with DDT).

Impact Assessment and Prediction of Effects
Predicting effects and assessing potential impacts of activities require knowledge of present conditions ( Chapter 7 ) and of the variability inherent in nature and the uncertainties of science ( Chapter 8 ); without such knowledge, effects cannot be reliably detected and the causes of change cannot be understood. Because the effects of a project can combine with effects of other projects elsewhere, and the effects of distant projects can add to the effects of the project under consideration, potential cumulative effects might require modifications in project design ( Chapter 9 ). It is often useful to compare the situation under consideration with similar ones elsewhere and to carry out pilot studies (Chapter 6). Cases of impact assessment are
INTRODUCTION 10
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. discussed in Chapter 16 (predicting the effects of construction on a herd of caribou), Chapter 21 (predicting the effects of freshwater impoundment), Chapter 22 (studies of the effects of ionizing radiation), Chapter 23 (studies of the effects of clearcutting), and Chapter 24 (predicting and learning the effects of the use of DDT and application of that knowledge to other cases).

Preservation of Communities
When the management goal is the preservation of entire communities of organisms, an understanding of community ecology ( Chapter 3), of productivity and nutrient cycling ( Chapter 4 ), and of the required space and time scales (Chapter 5) is important. Natural variability in climate and community patterns will also affect preservation efforts ( Chapter 8 ). Chapter 18 (restoration of communities of plants on derelict lands) and Chapter 20 (protecting a lake ecosystem from the effects of eutrophication) discuss examples of such applications.

Preservation of Habitat
The conservation of communities and species requires preservation of the habitat in which those organisms live. Monitoring both before and after efforts are initiated is necessary to determine the success of the efforts ( Chapter 7). The variability of natural systems and the cumulative effects of various activities, often far from the site of concern, can strongly influence the outcome of preservation attempts (Chapters 8 and 9). Some aspects of habitat preservation are treated in Chapters 17, 18, 20, and 23.

Contaminants and Toxic Substances
The pervasive and crucial issues of pollution and exposure to toxic chemicals are relatively neglected in this report, but the transport of chemicals is discussed in Chapter 4, indicator species and biological monitoring to detect pollutants and toxic chemicals are discussed in Chapter 7 , and the problem of cumulative effects (e.g., several sources of pollutants) is discussed in Chapter 9.

Specific examples are presented in Chapter 20 (excess nutrients), Chapter 21 (chemistry and hydrology related to heavy metals), Chapter 22 (ionizing radiation), and Chapter 24 (persistent pesticides).

INTRODUCTION 11
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved.

Mitigation of Effects of Construction
Mitigation of undesired side effects of projects is often economically, aesthetically, or legally motivated. If done after the project is completed, it might take the form of restoration (see below); but the best approach to mitigation is to make it unnecessary by appropriate design of the project. This requires careful planning and scoping ( Chapter 10 ), as well as careful monitoring ( Chapter 7 ). Chapter 16 (protecting caribou during hydroelectric development) and Chapter 21 (raising the water level of a lake) provide examples of including mitigation efforts in the design of a project; Chapter 18 (restoration of derelict land) describes mitigation of effects of projects completed many years earlier.

Restoration
To restore an ecosystem, one must have a goal to work toward. This requires some knowledge of previous conditions. Success in reaching the goal depends on knowledge of community structure ( Chapter 3), of productivity and nutrient requirements ( Chapter 4 ), of the space and time scales that will be needed for success ( Chapter 5 ), and of natural variation in biotic and abiotic factors (Chapter 8 ). Chapter 18 describes the restoration of plant communities on derelict lands in Britain.

General Applications
Some issues, procedures, and kinds of knowledge common to most environmental problems are treated in Chapters 6-10. There is always uncertainty, both in the abiotic environment and in organisms' responses to perturbations ( Chapter 8 ). And there is always a need for rigorous scientific procedures and careful scoping, even when time and money are short (Chapter 10). It is always helpful to compare the case at issue with similar ones; it is often useful to conduct pilot experiments, if feasible ( Chapter 6). Potential cumulative effects are often a concern, and monitoring is almost always necessary to ascertain current conditions, to detect variability, and to measure the effects of actions (Chapter 7).

THE CASE STUDIES
The objective of Part II is to show how ecological knowledge has been used in planning and carrying out problem-solving efforts, including how the knowledge has been limited and how it has had to be adapted to the INTRODUCTION 12
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. details of specific problems. The case studies are referred to repeatedly in Part I.

The existence and application of scientific knowledge do not guarantee correct predictions. Ecological systems are complex, situations are often unique, and predictive ecological theories are sometimes inadequate. Nonetheless, when scientific predictions turn out to be inaccurate, the results can reveal important aspects of the limitations of even the best available scientific approaches and show the need to increase the knowledge base on which decision-makers depend.

We selected examples to cover a wide array of types of environmental problem-solving ( Table 2 ). In some of our cases — such as those of the North Pacific halibut, the spotted owl, and the Newfoundland caribou — protection of a single species or achievement of a sustained harvest was the goal. In the malaria, red scale, and vampire bat cases, the goal was to eliminate an organism or at least to reduce it enough to render the damage it caused acceptable.

Problem-solving oriented toward manipulating a single species can draw on general ecological knowledge, such as population dynamics and predator-prey theory, and on the results of previous efforts. As the case studies show, proper targeting of management usually requires a great deal of knowledge about the natural history of the organisms in question.

Other case studies deal with problem-solving at the community level. The objective might have been the achievement of a sustained steady yield of useful products from a combination of species, as in the New Brunswick forest case, or the creation of a functioning ecological community similar to one that existed before the influence of human disturbance, as in the Lake Washington and derelict lands cases. In some cases, the purpose was to anticipate, and thereby to reduce, the undesirable impacts of a future project. For example, in the Southern Indian Lake case, the lake was altered for hydroelectric power generation, in the hope that it would not impair other uses of the lake.

Still other case studies deal with projects in which the objective was to determine the impacts of practices that take place in many locations over a broad area. Some of these practices are subject to legal regulation (use of pesticides, forestry practices, and uses of nuclear energy), and case studies can help to determine what kinds of legal restraints might be properly applied to reduce unwanted consequences in natural environments. As examples we have chosen the use of DDT in the United States, forest clearcutting, and the effects of nuclear radiation on ecological communities. In the latter two, a major objective was scientific.

To ensure consistent emphasis of particular points, we adopted a general format for the presentation of case studies. The preparer of each case INTRODUCTION 13
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. study, whether a committee member or another person intimately involved with the project, was asked to treat in sequence, if possible, the history of the study project, the major environmental problems that it posed, the types of ecological knowledge used in the project, and the role of the knowledge in determining how the problems were perceived, approached, and dealt with. Each case study is followed by a committee comment on the project's strengths and weaknesses and on the lessons that can be drawn from it for other studies of a similar nature. Our concentration on
TABLE 2 Synopsis of Case Studies Presented in Part II
Short Title Type of Case Subject
Halibut (Chapter 12) Renewable-resource
management
Managing an
international marine
fishery
Vampire Bat
(Chapter 13)
Pest control Chemical control of a
parasite on cattle in
Central America
Red Scale (Chapter 14) Pest control Biological control of a pest on citrus crops in
California
Malaria (Chapter 15) Disease control Experimental study of chemical control of a
malaria vector
Caribou (Chapter 16) Impact assessment Protecting a caribou herd during hydroelectric
development
Spotted Owl (Chapter 17) Species conservation Managing a regional population of a rare
species
Derelict Lands
(Chapter 18)
Ecosystem restoration Restoring seminatural
plant communities in
mined areas
New Brunswick Forest
(Chapter 19)
Renewable-resource
management
Developing a model for
long-term regional
forestry management
Lake Washington
(Chapter 20)
Ecosystem protection Predicting and controlling
cumulative impacts of
sewage effluent in an
urban lake
Southern Indian Lake
(Chapter 21)
Impact assessment Predicting impacts of
raising the level of a
subarctic lake
Nuclear Radiation
(Chapter 22)
Generic ecological studies Ecological effects of
nuclear radiation
Forest Clearcutting
(Chapter 23)
Generic ecological studies Ecological effects of
clearcutting forests
DDT (Chapter 24)
Pesticide use Ecological effects of DDT
INTRODUCTION 14
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. ecological knowledge is not meant to suggest that other sources of knowledge were not used or were necessarily less important in dealing with the environmental problems posed; it merely reflects our specific objective of showing how ecological knowledge can be used creatively in helping to solve environmental problems.

RECOMMENDATIONS
This report does not attempt to provide a ''cookbook" for solving environmental problems; that would be impractical. Rather, it is a guide to the process of "cooking," indicating where useful approaches to particular problems can be found in the literature and suggesting how appropriate knowledge and skills might be integrated for dealing with complex environmental problems.

The following recommendations constitute general advice to field workers, managers, and regulators and are intended to form the basis of scientifically sound but flexible environmental problem-solving. The overall approach we recommend is described in Chapter 10 . Here we highlight some major recommendations that apply to most environmental problems. If followed, they should increase the probability that solutions chosen for dealing with environmental problems are imaginative and appropriate. More details can be found in the chapters referred to, including the case studies, which constitute recommendations by example.
Involve scientists from the beginning. Scientists should be involved from the beginning of a project in setting goals, in identifying valued ecosystem components, and in scoping the problem ( Chapter 10 ).

Scientists do not determine the values attached by society to ecosystem components, but they might know which organisms have important roles in the ecosystem that are not understood or appreciated by the general public. Scientists can help to assemble information about a project site and about similar sites and projects elsewhere. They are also helpful in defining goals, because of their knowledge of potential outcomes of manipulations that might be considered. They can advise on the implications of trying to achieve particular goals, on the measurement of values, and on why seemingly compatible values might conflict. The involvement of scientists does not guarantee success, but it should increase the probability that project plans are appropriate.
Treat projects as experiments. Many projects are carried out on a larger scale than scientific experiments can be. To learn the scientific and management lessons that these projects have to offer, we need to treat INTRODUCTION 15
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. them as experiments ( Chapter 6 ). Treating projects as environmental experiments increases the understanding of basic ecological science and allows the testing of predictions made in impact statements.

Careful monitoring is essential to anyone who would understand the effects of projects ( Chapter 7 ), test the predictions made in impact statements ( Chapter 6 ), detect changes in baseline conditions (Chapter 8 ), and detect cumulative effects ( Chapter 9 ). To be most successful, monitoring should be incorporated into the project and study plans (Chapters 7 and 10).
Publish information . Many people associated with projects learn a great deal from them, but without making their knowledge available to others through publication. If the knowledge were made widely available, similar projects could be managed better and could serve as controls. The cases described in Chapters 12, 14, 15, 18-20, 22, and 23 are good examples.
Set proper boundaries on projects. Setting appropriate boundaries (Chapters 5 and 9) involves matching the scale of management to the scales of ecosystems. Environmental problems that cross political boundaries are especially difficult to deal with (e.g., Chapters 12 and 23), so the appropriate jurisdiction for management should be chosen carefully (Chapter 9).
Use natural-history information. Carefully collected natural-history information can help in solving environmental problems ( Chapter 1 ).

The success of the cases described in Chapters 12-21 depended on such information; for example, the control of such pests as vampire bats (Chapter 13) and red scale ( Chapter 15) was based on identifying key natural-history features of the focal organisms.
Be aware of interactions. Interactions among populations and among species in communities are complex (Chapters 2-4), and they can influence the effects of perturbations. It is important to remember that the most obvious connections are not necessarily the most important. Be alert for possible cumulative effects. Small perturbations, even if individually trivial, can have important cumulative effects when they are repeated in space and time. Therefore, many projects need to be evaluated not only in isolation, but also in the context of the overall frequency of their occurrence. Similarly, when a particular environment is being studied, investigators need to be alert for possible effects of activities and perturbations elsewhere (Chapter 9). Plan for heterogeneity in space and time. Environmental patchiness, variation in time, and variation in the species composition of ecological communities conspire to complicate environmental management ( Chapter 5 ). These considerations are particularly important in harvesting populations (for example, the maximal sustainable yield changes as the envi
INTRODUCTION 16
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. ronment changes) and in conservation. Interpopulation variation, both genetic and behavioral, can also complicate management, because of rapid evolutionary changes (e.g., Chapters 1 and 15).
Prepare for uncertainty and think probabilistically. A goal of management is usually to minimize uncertainty. But uncertainty is unavoidable (Chapter 8), and managers are advised to expect it and to plan for it. In addition, pretending that uncertainty is absent leads to bad planning and inflexibility. Therefore, both scientists and managers must be willing to think in terms of probabilities and to deal with them, as weather forecasters and farmers, sailors, fliers, and the general public do every day.

REFERENCES
Alexander, M. 1971. Microbial Ecology. John Wiley & Sons, New York.

Addicott, J. F. 1978. The population dynamics of aphids on fireweed: A comparison of local populations and metapopulations. Can. J. Zool. 56:2554-2564.

Andrewartha, H. G., and L. C. Birch. 1954. The Distribution and Abundance of Animals. University of Chicago Press, Chicago.

Andrewartha, H. G., and L. C. Birch. 1984. The Ecological Web: More on the Distribution and Abundance of Animals. University of Chicago Press, Chicago.

Daan, N. 1980. A review of replacement of depleted stocks by other species and the mechanisms underlying such replacement. Rapp. P-v. Reun. Cons. Int. Expl. Mer 177:405-421 .

Farner, D. S. 1964. The photoperiodic control of reproductive cycles in birds. Am. Sci. 52:137-156 .

Gilbert, L. E. 1975. Ecological consequences of a coevolved mutualism between butterflies and plants. Pp. 210-240 in L. E. Gilbert and P. H. Raven, eds. Coevolution of Animals and Plants. University of Texas Press, Austin.

Goodman, D. 1975. The theory of diversity-stability relationships in ecology. Q. Rev. Biol. 50:237-266 .

Gulland, J. A., and S. Garcia. 1984. Observed patterns in multi-species fisheries. Pp. 155190 in R.

M. May, ed. Exploitation of Marine Communities. Dahlem Konferenzen. Springer, Berlin.

Harper, J. L. 1969. The role of predation in vegetational diversity. Pp. 48-62 in Diversity and Stability in Ecological Systems. Brookhaven Symposia in Biology 22. Brookhaven National Laboratory, Upton, N.Y.

Harshberger, J. C., and J. J. Black. In press. A strategy for using fish bioassays and surveys to identify and eliminate point source environmental carcinogens. In Towards a Transboundary Monitoring Network: Proceedings of a Workshop. International Joint Commission, Washington, D.C., and Ottawa, Ont.

Holling, C. S. 1973. Resilience and stability of ecological systems. Annu. Rev. Ecol. Syst. 4:1-23.

Howe, H. F. 1981. Removal of wild nutmeg (Virola suranimensis) crops by birds. Ecology 62:1093-1106.

Hurlbert, S. H., and M. S. Mulla. 1981. Impacts of mosquitofish (Gambusia affinis) predation on plankton communities. Hydrobiologia 83:125-151 .

INTRODUCTION 17
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved.

Janzen, D. H. 1966. Coevolution of mutualism between ants and acacias in Central America.

Evolution 20:249-275.

Levin, S. A. 1974. Dispersion and population interactions. Am. Nat. 108:207-228 .

Lubchenco, J. 1978. Plant species diversity in a marine intertidal community: Importance of hervibore food preference and algal competition abilities . Am. Nat. 112:23-39.

Maki, A. W., K. L. Dickson, and J. R. Cairns, Jr., eds. 1980. Biotransformation and Fate of Chemicals in the Aquatic Environment. American Society for Microbiology, Washington, D.C.

May, R. M. 1973. Stability and Complexity in Model Ecosystems. Princeton University Press, Princeton, N.J.

Morin. P. J. 1983. Predation, competition, and the composition of larval anuran guilds. Ecol.

Monogr. 53:119-138.

National Research Council. 1976. Vapor-Phase Organic Pollutants: Volatile Hydrocarbons and Oxidation Products. National Academy of Sciences, Washington, D.C.

National Research Council, 1981. Testing for Effects of Chemicals on Ecosystems. National Academy Press, Washington, D.C.

National Research Council 1982. Assessment of Multichemical Contamination. Proceedings of an International Workshop, Milan, April 28-30, 1981. National Academy Press, Washington, D.C.

National Research Council. 1983a. Acid Deposition in Eastern North America. National Academy Press, Washington, D.C.

National Research Council. 1983b. Effects of atmospheric transformations of polycyclic aromatic hydrocarbons . Pp. 3-1 to 3-48 in Polycyclic Aromatic Hydrocarbons: Evaluation of Sources and Effects. National Academy Press, Washington, D.C.

National Research Council. 1983c. Changing Climate. Report of the Carbon Dioxide Assessment Committee. National Academy Press, Washington, D.C.

National Research Council. 1984. Groundwater Contamination. National Academy Press, Washington, D.C.

National Research Council. In press, a. Monitoring and Assessment of Trends in Acid Deposition.

National Academy Press, Washington, D.C.

National Research Council. In press, b. Methods for Assessing the Effects of Mixtures of Chemicals on Humans and Non-Human Biota. National Academy Press, Washington, D.C.

Nicholson, A. J., and V. A. Bailey. 1935. The balance of animal populations. Proc. Zool. Soc. Lond. 3:551-598.

Niering, W. A., and F. E. Egler. 1955. A shrub community of Viburnum lentagon, stable for twenty- five years. Ecology 36:356-360 .

Orians, G. H. 1975. Diversity, stability and maturity in natural ecosystems. Pp. 139-150 in W. H.

Van Dobben and R. H. Lowe-McConnell, eds. Unifying Concepts in Ecology. W. Junk, The Hague.

Paine, R. T. 1974. Intertidal community structure: Experimental studies on the relationship between a dominant competitor and its principal predator. Oecologia 15:93-120.

Paine, R. T., and R. L. Vadas. 1969. The effects of grazing by sea urchins, Strongylocentrotus, on benthic algal populations. Limnol. Oceanogr. 14:710-719 .

Reif, A. E. 1984. Synergism in carcinogenesis. J. Natl. Cancer Inst. 73:25-39 .

Salisbury, F. B., and C. W. Ross. 1978. Plant Physiology. 2nd ed. Wadsworth, Belmont, Calif.

Schindler, D. W. 1976. The impact statement boondoggle. Science 192:509.

Soulé, M. E., and B. A. Wilcox, eds. 1980. Conservation Biology. Sinauer Associates, Sunderland, Mass.

INTRODUCTION 18
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved.

Sundararaj, B. I., and S. Vasali. 1976. Photoperiod and temperature control in the regulation of reproduction in the female catfish. J. Fish. Res. Bd. Can. 33:959-971 .

Suter, G. W., II. 1981. Commentary: Ecosystem theory and NEPA assessment. Bull. Ecol. Soc. Am. 62:186-192 .

Utida, S. 1957. Population fluctuation: An experimental and theoretical approach. Cold Spring Harbor Symp. Quant. Biol. 22:139-151.

Whitham, T. G., A. G. Williams, and A. M. Robinson. 1984. The variation principle: Individual plants as temporal and spatial mosaics of resistance to rapidly evolving pests. Pp. 15-52 in P. W. Price, C. N. Slobodchikoff, and W. S. Gaud, eds. A New Ecology. Wiley- Interscience, New York.

Wiens, J. A. 1976. Population responses to patchy environments. Annu. Rev. Ecol. Syst. 7:81-120.

INTRODUCTION 19
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved.

INTRODUCTION 20
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved.

I
Kinds of Ecological Knowledge and Their
Applications
21 Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. 22 Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. 1 Individuals and Single Populations
Environmental concern commonly focuses on populations of organisms, whether the goal is protection of valued species, harvest of economically profitable species, or control of economically destructive species. We are interested in predicting and controlling changes in size and structure of populations that occur in response to environmental change, whether anthropogenic or not. The ecology of individuals and populations is of particular relevance to this interest. Population biology, the subject of this chapter, encompasses many kinds of research, from the study of the details of life history and behavior to the construction of mathematical models of the dynamics and genetics of multiple populations of a species over a large area.

A life history encompasses an individual's interactions with its physical and biological environment throughout its lifetime. Research into life histories has taken several paths, all of them valuable in the management of populations:
detailed studies of the ecology of individual species (e.g., Chapters 12-17), comparative studies of groups of species (e.g., Chapters 13 and 15), and theoretical treatments of the evolution of life-history patterns (e.g., Stearms, 1976; Chapters 15 and 17). Studies of particular species yield the detailed information needed for management with respect to nutrient and habitat requirements, important interactions with other species, reproductive requirements, and significant behavioral idiosyncrasies. Moreover, research on organisms with complex life cycles has shown the importance of choosing the appropriate stages in the life cycle
INDIVIDUALS AND SINGLE POPULATIONS 23
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. for management intervention. Comparative studies reveal general patterns that help focus individual studies and provide managers with guidance in the absence of detailed information. Theoretical research focuses attention on the elements of life history important in solving long-term management problems (e.g., Beddington, 1974; Lewontin, 1965; May, 1980). Thorough knowledge of species' life histories has a broad range of applicability to problems of population management, including captive propagation programs for endangered species (Frankel and Soul é, 1981), pest and disease control (Chapters 13-15), species protection ( Chapter 17 ), harvesting ( Chapter 12 ), predicting environmental impacts ( Chapter 16 ), and restoring plant communities (Chapter 18).

As valuable as life-history information is for the prediction and control of population behavior, it provides only partial insight into the causes and consequences of changes in population numbers and composition. To determine how a population will respond to an increase in mortality due to harvesting or stress or how effective a given procedure might be for improving reproductive output, we need some understanding of population dynamics. Research in population dynamics has ranged from field studies designed to determine what factors affect population sizes in an area to theoretical studies of how such factors can act together to "regulate" population size over long periods. Some simple models use an "accounting" formulation to calculate future population size on the basis of current size and rates of growth, death, and birth. More complex models deal with such phenomena as dispersal, breeding structure, interchange between populations, environmental variability, and the effects of intraspecific interactions on population behavior.

Traditional models of population dynamics are based on the assumption that organisms do not change genetically during a period of management.

However, genetic changes do occur when populations are exposed to repeated manipulations, and evolution can take place with startling rapidity — e.g., the evolution of insect resistance to pesticides and of bacterial resistance to antibiotics. Both insects and bacteria have the short life spans and high reproductive rates that speed evolution (May and Dobson, in press), but long- lived organisms can also evolve quickly if management results in large differences in mortality or reproduction among individuals. Size-selective harvesting of fish can lead to a reduction in the average age and size at maturity (Ricker, 1981), possibly as a result of changes in competition among age classes and in the frequencies of particular genes. When populations are small, harvesting practices can lead also to loss of genetic variability and can increase the deleterious effects of inbreeding. Preserving a species that is distributed into many small and isolated populations — as many endangered species are — involves an understanding of
INDIVIDUALS AND SINGLE POPULATIONS 24
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. both population dynamics and population genetics (Frankel and Soulé, 1981).

Dynamic and evolutionary effects of manipulations are not always separable. Changing such population characteristics as total numbers, distributions of ages and sizes, and sex ratio not only changes the dynamics of a population, but also establishes a potential for evolutionary change in traits that exhibit genetic variation. Population management that is based only on dynamic considerations can in the long run produce results opposite to those intended.

DDT works miracles on untreated populations of mosquitoes that transmit malaria, but within 5-50 generations the evolution of resistance might largely negate the effectiveness of the chemical; at the same time, human resistance to malaria can decrease in the absence of the disease ( Chapter 15 ). Using large- mesh nets to ensure the harvest of male but not female migrating salmon can lead to an increase in the proportion of early-maturing small male salmon (jacks) (Gross, 1984, 1985). An increase in the proportion of jacks that escape the nets can lead to an increase in the proportion that breed and thus in the proportion that are hatched in the next generation.

IDENTIFYING KEY FACTORS
Organisms are influenced by many components, or factors, of their physical and biological environments. Those factors are not equally important; often, a few dominate the dynamics of populations and need careful study for successful management. Of particular interest are factors that exert especially powerful effects on population size (usually by influencing birth and death rates).

Key-factor analysis is a method for identifying and understanding the stages of an organism's life history in which critical controlling processes occur.

The method has been used for designing pest-control strategies (Clark et al., 1967). The basic data used in key-factor analysis are the number of individuals in each developmental stage or age group, survival from one stage to the next, and fecundity rates — i.e., the standard elements of life tables (Ricklefs, 1979).

When the variance in these elements is partitioned among environmental causes, the key factors are the ones that cause high mortality (Harcourt and Leroux, 1967). Key-factor analysis is less applicable in heterogeneous than in uniform environments (Hassell, 1985).

In organisms with complex life histories (e.g., salmon), different stages use resources differently and sometimes live in different habitats. The behavior of the different stages might therefore be controlled by largely independent events.

Too few individuals of such a species might survive
INDIVIDUALS AND SINGLE POPULATIONS 25
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. one stage to saturate the habitat of the next. In addition, the more independent the habitats are, the less likely it is that their carrying capacities will vary concurrently. Therefore, populations of species with complex life histories often vary more widely in density than they would if all resources were gathered in one habitat (Istock, 1967), and study of a population in only a single stage is unlikely to reveal the causes of fluctuations.

BEHAVIOR
Biological processes of animals are tied to the behavior of individuals as they acquire nutrients, select habitats and mates, avoid predators, and interact socially. Some behavior can be predicted from knowledge about general environmental conditions and the types of species under consideration, but in many cases understanding behavior requires detailed knowledge about particular species and how they make behavioral "decisions." We discuss below decisions made by individual organisms that are important in dealing with environmental problems.

Habitat Selection
By natural selection, species evolve to prefer environments in which survival and reproduction are greatest. Responses to habitats are influenced not only by characteristics of the environment, but also by the presence or absence of individuals of the same species. Other individuals provide information about choices made by previous settlers, and they modify the environment for later settlers (Orians, 1980; Partridge, 1978).

Individuals might not settle in an otherwise suitable area if no other individuals are there. Young individuals might need to learn the locations of suitable sites from older ones, and slight changes in a habitat can lead to rejection, even if it is otherwise suitable for the species. For example, a management plan for the spotted owl ( Chapter 17 ) avoids such rejection by recognizing that the owls require undisturbed old-growth forests for hunting and are reluctant to cross open areas. The collapse of the whitefish fishery in Southern Indian Lake was due not to a decline in the fish population, but to changes in selection of habitats as a result of raising the level of the lake (Chapter 21). The control of malaria was less successful than expected, because of a behavioral polymorphism in habitat selection (Chapter 15).

Attempts to find suitable natural predators to control agricultural and forest pests usually involve searches in areas with climates similar to those of the areas into which the control agent is to be introduced (Chapter 14).

INDIVIDUALS AND SINGLE POPULATIONS 26
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved.

Particular attention is paid to climate, rather than biological interactions, because the required food source — the pest to be controlled — is known to exist in the areas of introduction and because experience has shown that predators and parasites of a single host species typically change over the range of the host. It is rare for a single predator to be an important control agent over the entire range of its host. It is difficult, in practice, to find an agent that can control over even part of a pest's range, because many introduced predators fail to establish viable populations or fail to become common enough to achieve effective control (e.g., Chapter 14 ). Attempts at biological control therefore often involve importing a variety of potential control agents, in the hope that some will prove effective (Chapter 14).

Mating Systems
Mating systems vary widely. In the simplest, sex cells are shed into the surrounding medium, usually water, where fertilization takes place. In more complex animal systems, reproductively mature adults gather, choose mates from the available pool, and associate with one another for various periods after fertilization to care for each other or their offspring. In many species, mate selection is combined with habitat selection for breeding, especially in species in which one adult holds space that contains resources for the reproductive cycle. Much research into the evolution of mate choice is directed at determining the relative roles of mate quality and habitat quality and the criteria used for selection of each (O'Donald, 1967; Searcy, 1982).

A successful biological control program that has capitalized on the use of laboratory-reared sterile males is the control of screwworm populations in the southern United States (Perkins, 1982; Scruggs, 1975). The program depends both on the ability to produce enough sterile males to ensure that most males in the field are sterile and on the mating behavior of the flies. No long-term pair bonds are formed, but the first male to copulate with a female blocks her reproductive tract with a plug that prevents other males from copulating with her. If multiple matings were the rule, a much higher ratio of sterile to fertile males would be required to achieve the same degree of population control. A similar mating system is the basis of strategies to control Mediterranean fruit flies (Dacus).

Social Interactions
Social grooming is a component of social behavior that sometimes accompanies mate selection, but also occurs in other contexts. The observation that vampire bats groom each other extensively suggested a INDIVIDUALS AND SINGLE POPULATIONS 27
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. control strategy that facilitated transfer of a control agent applied to the fur of only a small number of bats ( Chapter 13). The control agent was highly specific and had no effects on other species living in the same caves. The vampire bat case study is an excellent example of the value of detailed natural-history observation for population management.

POPULATION DYNAMICS
Population Regulation
Environmental changes can influence a population in two basic ways, each having important management implications. Fires, bad weather, and other events can reduce a population by removing a fraction of it that is independent of its density. If a population — or a part of it, such as an age group — is limited primarily by such density-independent factors, it can be reduced with little effect on the remaining individuals.

Many environmental changes, however, influence individual growth, birth, and death rates in a manner that depends on population density. Even though populations grow rapidly at low densities, growth often slows as density- dependent effects appear through individual interactions (e.g., competition for food or nest sites) or through the actions of predators, disease, and other factors that increase deaths or reduce births. When individuals are removed from a population, density-dependent compensation can occur — for example, an increase in the birth rate as population size decreases. The operation of density- dependent factors can stabilize population densities and thus cause displaced populations to tend to return to equilibrium, in which the rate of loss of individuals of a species equals the replacement rate. The greatest number of individuals of a species that the environment can support is called the carrying capacity (K); it too can vary. May (1973) has discussed mathematical studies that showed how density dependence can lead to cyclic or even chaotic population change.

Species with a very high reproductive potential (r) in the absence of competition are potentially able to recover quickly from population reduction, but they are more likely than species with a low r to show extreme population fluctuations. Many pest species, such as weeds, are poor competitors, but have good dispersal abilities and high r (Baker, 1974). Species with low r, such as whales and spotted owls ( Chapter 17), often have stable populations consisting of long-lived individuals. Overharvesting of low- r species can easily lead to species extinction. Conversely, low-r species can more easily be controlled with measures used only periodically (e.g., vampire bats, as discussed in Chapter 13).

Commercial or recreational harvesting reduces populations to some INDIVIDUALS AND SINGLE POPULATIONS 28
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. value below the carrying capacity. Managers often try to maintain populations near the size that permits the greatest yield of individuals or biomass. Because yield is a result of both the reproductive rate and the number of individuals reproducing, yield is highest when the population is between low density (few individuals but high individual reproductive rate) and the high density of natural equilibrium (strong density-dependent reduction of average individual reproductive output or survival).

In theory, if environments are constant, every population has a ''maximal sustainable yield" (MSY); but in practice, the MSY is very difficult to identify, let alone achieve. Accurate estimates of abundances are required, especially when an unstable equilibrium exists near the MSY (see below). Both yields and population sizes fluctuate more as the MSY is approached, the effect being most pronounced in large mammals, such as whales, in which density- dependent effects are strongest near K and r is low (May, 1980).

Reducing the risk of errors resulting from population management requires at least estimation of relative changes in population size and composition; good estimates of absolute population size are more desirable. But accurate estimates of the age distributions and sizes of populations are often difficult to obtain (Eberhardt, 1976; Seber, 1982). Although density and age of forest stands can be estimated accurately, errors in density and age estimates in bird, mammal, and fish populations are often large or unknown. Most marine fish populations are managed without knowledge of population sizes (e.g., Pacific halibut, as discussed in Chapter 12 ). Shepherd (1984) discusses the advantages and shortcomings of techniques for managing fish populations in the face of limited information; some difficulties in obtaining the information have been set forth clearly by Larkin (1978).

Populations of many species are often managed indirectly by managing the habitat. Habitat quality is commonly judged by population density, on the assumption that high density indicates high quality of habitat. This assumption is often valid, but data on survival and reproductive success are generally needed to understand differences in population densities. High densities can occur in suboptimal habitats composed mostly of subdominant individuals that have been forced out of other habitats (Van Home, 1983).

Population Stability
A population with a zero growth rate is in equilibrium. The equilibrium is stable if a population displaced above or below the equilibrium point tends to return to it. However, a population can have more than one stable INDIVIDUALS AND SINGLE POPULATIONS 29
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. equilibrium point, and some equilibria are unstable (Berryman, 1981). Unstable equilibria often occur at low population densities of large, long-lived, slowly maturing species with low r (Southwood et al., 1974), as well as in a variety of heavily harvested fish populations (Beverton, 1984). Increased difficulty in finding mates, reduced effectiveness of cooperative defense against predators, and other phenomena can result in a lower threshold of density below which the population collapses unless there is immigration. Density-dependent compensatory mechanisms might also break down at very low densities (Beverton, 1984). That possibility is of particular concern for rare and endangered species with few, isolated populations that are already small.

Environmental fluctuations, disease, and other factors can easily reduce such a population to below its stability threshold. Some fish stocks have collapsed to extremely small numbers after overexploitation, and a few have not recovered even after the release of fishing pressure (e.g., Beverton, 1984; Peterman, 1978). In other species, the maximal yield could be very close to the harvest magnitude that would cause the population to collapse (Ricker, 1963). In any case, environmental variation appears to interact with fishing pressure to complicate the fluctuations in stock size (Chapter 8).

It is difficult to estimate minimal safe population sizes. M. L. Shaffer (unpublished manuscript) has modeled minimal population sizes and areas for grizzly bears in Yellowstone National Park by using successive computer simulations based on estimated population parameters with random variation.

Using several scenarios, he estimated that 35-70 bears were needed to have a 95% probability of population survival for 100 years in that ecosystem. Studies of this type can be helpful in focusing attention on the need to choose the time scales at which to judge the acceptability of a given probability of extinction, but the underlying assumptions are not subject to rigorous testing.

Long-term management plans can accommodate the possible presence of low-density thresholds by providing a substantial safety margin, particularly for isolated populations with little immigration. It might be difficult to predict these thresholds with the available data, and guesses based on experience with similar species could be necessary (Chapter 17; Soulé and Wilcox, 1980).

Many defoliating insects appear to have multiple stable equilibria. Such species appear to be controlled at low densities in a density-dependent fashion by predators, parasites, or pathogens. These agents cannot "control" the pest populations above particular sizes, because other factors prevent the populations of the controlling agent from increasing enough (e.g., Campbell and Sloan, 1977; McNamee et al., 1981). Traditional
INDIVIDUALS AND SINGLE POPULATIONS 30
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. approaches to controlling these species have relied on the massive application of insecticides during outbreaks. However, control might be more effectively applied to subpopulations threatening to escape from the low-density control (Campbell and Sloan, 1978; MacLeod, 1977). Environmentally induced reductions in the vigor of trees can trigger outbreaks of some forest pests, and control techniques can be applied to areas where such conditions appear imminent (Berryman, 1981).

Such techniques use a system of "risk classification" of the habitats of the pests in determining when and where to apply controls. Another successful approach is the use of a biological control agent that has a shorter generation time than the pest and that can respond to outbreaks by increasing its own population. An example is the partial control of gypsy moths in the northeastern United States with bacteria and viruses (Leonard, 1974; Massachusetts Department of Environmental Management, 1981). McNamee et al. (1981) have provided a useful framework for classifying defoliating forest insects so that appropriate control strategies can be chosen.

Dispersion and Population Movements
Species vary in their patterns of dispersion in space (clumped, random, or even), density, dispersal, and migratory behavior. Most terrestrial species occupy habitat patches of various sizes and degrees of isolation. These variations have implications for management. Management of a migratory species must take into account the relationships between populations in breeding and nonbreeding habitats. For example, if the population of salmon at sea is near K, then increasing hatching success in the rivers might have little benefit (Peterman, 1978, 1984).

Fretwell (1972) has shown theoretically and by examples that populations in strongly seasonal environments are ultimately controlled in one "bottleneck" season. For example, winter resources might limit the number of individuals that survive to breed, so an increase in the preceding summer reproduction could cause little change in the number of individuals breeding in the following summer. This idea is particularly important in managing bird species that winter in tropical habitats undergoing extensive deforestation. Recognizing that the protection and enhancement of populations might not be sufficient, in itself, to prevent extinction, legislators of many countries have cooperated in formulating treaties to protect migratory bird species. Migratory patterns can also make species susceptible to local disturbances (e.g., Chapter 16 ). Salmon can be affected by a single dam in a migration of thousands of miles, and some migratory birds
INDIVIDUALS AND SINGLE POPULATIONS 31
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. (such as geese, cranes, and swans) depend on a very small number of resting and feeding areas along their migration routes.

Animals that organize into dense aggregations can easily be harvested to very low numbers, because the profitable unit for harvest is the aggregation.

Fish stocks that form aggregations (e.g., many clupeoids) apparently are subject to catastrophic collapse when fishing effort is continuously increased (May, 1980, 1984; Murphy, 1977). The situation with some very large whales is analogous, in that it can be profitable to hunt and harvest a single animal. May (1980) has discussed various problems encountered in managing the harvest of fish that form schools.

Isolated populations have less potential for recovering from temporary reductions when immigration rates are low, as happens when an occupied habitat becomes an island surrounded by unsuitable habitat. Maintaining travel corridors of appropriate habitat to connect populations, particularly of species at low density and with strict habitat requirements and poor dispersal ability, might be a key component of management (MacClintock et al., 1977; Willis, 1974). The spotted owl ( Chapter 17), a strict obligate of old-growth coniferous forest, is an example of such a species (Forsman et al., 1984). However, if control is the objective, disruption of corridors can be effective in preventing dispersal.

The size of the habitat patch necessary to support even a single breeding female or pair depends on the size, diet, and behavior of the species (Galli et al., 1976; Schonewald-Cox et al., 1983). Carnivores generally require more area than herbivores, and area requirements increase with body size (McNab, 1963; Schoener, 1968). Required areas for managing populations of large social carnivores, such as wolves and bears, can be huge (M. L. Shaffer, unpublished manuscript; Soulé, 1980).

The extinction of bird species on Barro Colorado Island constitutes a particularly revealing example of spatial isolation. The island became isolated from the Central American mainland during the construction of the Panama Canal. One of the best predictors of local extinction was susceptibility to ground predators, such as snakes (Karr, 1982); the populations of these predators were increasing, probably because large predators were decreasing as a result of human activities.

Growth Rates, Age, and Size
The operation of forestry, fisheries, and agriculture depends on an understanding of how growth rates are influenced by interactions with the physical environment and with other organisms. Size is often used as an indicator of age in a continuously growing species, because it is easy to measure and is usually reliable. When age and size are poorly correlated INDIVIDUALS AND SINGLE POPULATIONS 32
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. (e.g., Caswell and Werner, 1978; Policansky, 1983), the use of size as an indicator of age can lead to inappropriate management decisions, as it did in a population of roe deer ( Capreolus capreolus ) in Scotland. The ages of culled deer in that population were estimated on the basis of body size and weight, as is common for deer (Ratcliffe, 1984), but the calves were growing so fast that the technique underestimated their number and overestimated the number of fertile females. The result was overharvesting.

If resources, such as food and light, are limiting to animals or plants, removing large individuals usually increases the growth rate of those remaining.

These interactions are well known to wildlife and fishery managers, foresters, and agriculturalists. In the New Brunswick forest-management case (Chapter 19), the inverse relationship between planting density and growth rates of individual trees was an important part of the management plan. An understanding of the factors affecting growth rates among competing species allowed the successful regeneration of land in the derelict-lands case (Chapter 18).

Environmental manipulations affecting fecundity or survivorship have the greatest effect on population dynamics if the perturbations are introduced at early reproductive stages (Beddington, 1974; Emlen, 1970; Lewontin, 1965).

Expected reproductive output is smaller in late reproductive stages than in early stages, so late stages have a smaller per capita effect on population production.

Similarly, many individuals in prereproductive stages will not survive to breed, so their removal affects population reproduction less than the removal of early breeders. This is the basis for fishing and hunting regulations that specify minimal sizes or ages.

Age Structure
Variations in life expectancy or age structure in a population can influence its responses to management. Length of life and age structure are associated with many variables — such as spatial distribution, reproductive potential, and feeding habits — that are hard to estimate, so generalizations as to the implications of age structure for management can be unreliable. Nonetheless, a few points should be mentioned.

Beverton (1984) classified the expected response of fish species to fishing pressure on the basis of their ecological characteristics (including life history).

Fish populations comprising long-lived individuals in many reproductive age classes are generally more stable and easier to manage than those having only few age classes comprising short-lived individuals. There is also a tendency for species that are densely aggregated when young to be more stable in the face of fishing pressure. Beverton classified
INDIVIDUALS AND SINGLE POPULATIONS 33
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. halibut as "reliable, steady, and robust," and these life-history characteristics probably have been partly responsible for success in halibut management (Chapter 12).

It is important to think about population stability in terms of the life spans of individuals. A population of long-lived organisms might appear to be maintaining itself when in fact it is in danger of serious decline. If habitat degradation by pollution or other factors prevents young organisms from surviving, the long life span of adults can mask the loss of those age classes.

The June sucker (Chamistes liorus ) in Utah has maintained populations without recruitment for 15 years (U.S. Fish and Wildlife Service, 1984) and saguaro cactuses ( Carnegiea gigantea ) in the Sonoran desert remain common in areas that have seen no recruitment for even longer periods (Turner et al., 1969). The other side of this coin is that such populations are able to survive long periods of adverse conditions, even if conditions favorable for reproduction occur only infrequently.

Sex Ratios and Sex Biases
Either intentionally or unintentionally, human activities often produce mortality differences between the sexes and thus biased sex ratios. The effects on population reproduction and dynamics depend on the form of the mating system. The degree to which differences in mortality can be controlled depends on the difficulty of distinguishing the sexes in the field.

In many animal species with polygynous mating systems, one male can fertilize the eggs of many females, so a reduction in the number of males has a relatively small effect on reproductive rates. Management of polygynous species in which the sexes are easily distinguished (e.g., deer, crabs, and pheasants) often involves permitting the harvest only of males. In the sockeye salmon (Oncorhynchus nerka), males are larger than females, so gillnets, which select for larger fish, take disproportionately many males. In spite of this skewing of sex ratio, Mathisen (1962) has shown that a 15:1 ratio of females to males led to egg hatching only 5% lower than that with a 1:1 ratio. In contrast, differences in mortality can have large effects in species whose reproduction is limited by whichever sex is the rarer, such as species in which both parents (e.g., geese) or only the males (e.g., rheas and seahorses) care for the young.

Sex ratios can be manipulated to increase productivity, including food production, as in the case of dairy cattle. Only the female structures of many commercially important crop plants are eaten (seeds and fruits). In plants with separate sexes, fruit yields can be increased by manipulating the sex ratio. In cosexual species— those with both sexual functions in
INDIVIDUALS AND SINGLE POPULATIONS 34
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. the same individual — manipulations are directed toward altering the allocation of investment in female versus male function. Such schemes work when allocation patterns can be genetically selected for or when they are influenced by environmental conditions that can be manipulated. In the Cucurbitaceae (cucumbers, melons, squash, etc.), some genes shift the allocation of resources from male to female function (Kubicki, 1969a,b,c; Robinson et al., 1976), and cultivars that favor female function have been selected (Velich and Satyko, 1974).

GENETIC AND EVOLUTIONARY CONCERNS
Artificial selection has been successful in producing desirable traits in animals and plants, and the genetic resources of domestic species can be maintained by intentional cross-breeding. Manipulations of the population dynamics of nondomesticated organisms, however, can produce the evolution of undesirable genetic consequences, often within a few generations.

Conspicuous examples include the evolution of resistance to pesticides and antibiotics (e.g., Anderson and May, 1982; Peters, 1984, in press) and perhaps increased pathogen virulence (Ewald, 1983), decrease in size (Ricker, 1981), and increased inbreeding depression (see below).

The Evolution of Resistance to Pesticides
One of the best-documented cases of undesired evolutionary change is the development of resistance to DDT ( Chapter 24 ). In retrospect, it should have been obvious that it would happen; in fact, the use of poisons to produce a class of resistant organisms for experimental purposes has been a major technique of microbiology for more than 50 years. The short generation times, high reproductive rates, and large populations of most pests favor rapid evolution (May and Dobson, 1985). The race between chemists and the evolutionary capacities of organisms parallels the race between the evolution of new antiherbivore devices by plants and the evolution of resistance to them by insects (Rhoades, 1983).

A number of approaches can be taken to reduce the rate of evolution of resistance. First, agents that interfere with fundamental and invariant processes, such as eating and oxygen uptake, can be used. Changes in fundamental processes are unlikely to evolve over a short period. However, such agents can affect nontarget species, because the processes in question are shared by many of them. The use of anticoagulants against vampire bats ( Chapter 13) depended on the spread of the materials by intraspecific social grooming, a process that kept the materials within the target species.

Second, two or more chemicals can be used at the same time or in INDIVIDUALS AND SINGLE POPULATIONS 35
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. close alternation (e.g., Peters, 1984, in press; Smith, 1982). This reduces the number of survivors, because survival would require the evolution of two or more types of resistance simultaneously, which is unlikely in view of the small probability that several mutations will occur at once.

Third, agents can be used intermittently. This allows the frequency of evolved resistant types to decrease through competition with nonresistant types (e.g., Levin and Lenski, 1983). However, Plasmodium resistance to chloroquinone actually seems to be associated with a general biological advantage over nonresistant strains (Doberstyn, 1984), so this approach might not be universally applicable.

Genetic Consequences of Differential Harvesting by Sex and Size
Harvesting according to age or sex is a feature of many management schemes. Animals of one sex might be singled out for harvest (e.g., in deer and crabs) or the capture of animals might be sex-biased for other reasons, such as differences in size or activity (consider, for example, the evidence of size and sex selectivity of fishing gear). Nonrandom harvests could have produced such changes— perhaps genetic— as early maturity at smaller size in fish populations, e.g., salmon (Mathisen, 1962; Ricker, 1981), Atlantic cod (Borisov, 1978), an African cichlid (Silliman, 1975), and lake whitefish (Handford et al., 1977); Moav et al. (1978) have suggested that this is a general phenomenon.

Many large male salmon are harvested with size-selective nets before they have a chance to spawn; this harvesting favors the survival of early-maturing, small males (jacks). Increased proportions of jacks, which have low commercial and sport value, are a matter of growing concern to the managers of exploited salmon populations (Ricker, 1981). Theoretical studies (Gross, 1984, 1985) have suggested that jacks are not abnormal, as previously thought, and that they and larger males exist in an equilibrium that depends on both differences in mortality and conditions in spawning areas. The theory suggests ways of reducing the proportion of jacks despite heavy harvesting of larger males. For example, spawning sites could be manipulated to reduce the cover in which the jacks seek protection from larger males, thus lowering their reproductive success and the proportion of jacks in the next generation.

The jack story shows how management can lead to undesirable evolution.

For example, trophy hunting, which removes the finest specimens from the breeding population, leads to the evolution of deer genotypes with smaller antlers. Hunting in general is known to make animals furtive, and genetic changes in dispersal, foraging, and social behavior are possible. In short, the evolutionary effects of management must be considered
INDIVIDUALS AND SINGLE POPULATIONS 36
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. if it can lead to differences in reproductive success of different genotypes (Law, 1979).

Genetic Consequences of Small Population Size
Genetic deterioration has often led to extinctions in captive (including laboratory) populations, and small natural populations might also be in such danger, because of inbreeding depression and loss of genetic variability (Frankel and Soul é, 1981; Schonewald-Cox et al., 1983; Soul é and Wilcox, 1980)— the former reduces fitness, and the latter reduces adaptive potential.

Efforts to protect species threatened by the genetic effects of small population size involve estimating a minimal "effective" population size (minimal N e) that
is based on estimates of tolerable degrees of inbreeding, which in turn are based on experience with domestic and laboratory animals. (Ne is the number of potential breeding individuals in an "ideal" population — one with random mating, a sex ratio of 1:1, discrete generations, constant size, and a Poisson distribution of family size — that retains the same amount of selectively neutral genetic variability as the population under consideration. If the population departs from "ideal" conditions, as occurs commonly in nature, N e is lower than
actual N.)
Selecting tolerable rates of loss of genetic variability is much more difficult than choosing tolerable degrees of inbreeding, because the long-term consequences of reduced variability are not known and could depend on the species involved. Soul é (1980) crudely estimated a minimal N e of 50 on the
basis of inbreeding alone, according to an expectation of surviving extinction for 1.5 N
e, or 75, generations. Franklin (1980) suggested that, for long-term adaptive potential, Ne should be 10 times as high, i.e., about 500. The consequences of inbreeding might be less serious in species that normally engage in much inbreeding or that are already strongly homozygous (e.g., many self-fertilizing plants).

Sex-biased harvesting can also lower genetic variability, perhaps enough to have serious consequences. Ryman et al. (1981) have shown how various moose- and deer-hunting policies can lead to loss of genetic variability, with N e
perhaps as low as 5% of N.

INDIVIDUALS AND SINGLE POPULATIONS 37
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. 2 Population Interactions
Every species in an ecological community is connected to many others.

Each is both a predator (if plants can be regarded as predators on photons) and a prey, each can compete with other species that use the same resources, and each can engage in mutualistic interactions with other species. In a particular environmental problem, usually one of the roles that a species can play is of prime concern, but neglect of other interactions might cause management efforts to fail or produce unwanted side effects.

The processes in which species are involved are measured in different units. Rates of photosynthesis are measured in units of mass change; feeding rates are measured in energetic units. Competitive interactions might decrease reproductive rates or cause species to be absent from some areas. Not all these effects can be expressed in energetic terms. If disruption of mutualisms lowers seed set or dispersal, the energy content of the lost seeds is only a small component of the significance of those changes. The overall unit for expressing these effects is fitness (relative reproductive success), but total fitness rarely can be measured in the field, and even partial measures can be difficult to obtain. As a result, various units are used by environmental problem-solvers, the most appropriate one depending on the problem and the objectives of manipulation.

Human uses of ecological systems commonly involve altering the nature and extent of interactions among populations, whether or not this is the prime objective of the management programs. We commonly eliminate large carnivores, because they are dangerous and because they prey on wild or domesticated species that we wish to exploit. Herbivorous insects POPULATION INTERACTIONS 38
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. compete with us for the tissues of crop plants, and weeds compete with those plants for light, water, and nutrients. Our efforts to remove these species or reduce their abundances are based on the assumption that doing so will increase the yields of valued products from the species we wish to protect. The assumption is probably true in most cases, but there are remarkably few quantitative data on the effects of weeds on yields of crop plants (Mortimer, 1984). There are many more estimates of losses of crops to herbivores. Damage is often severe enough not only to justify control, but to cause economic hardship to agriculturalists (Barrons, 1981; May, 1977; Pimentel et al., 1980). It is clear, however, that control measures are often initiated when populations of pests are so small that economically significant damage is unlikely to occur (Pimentel et al., 1980). Restricting the use of control measures to times and places when they are cost-effective is important, because many of the materials used are toxic and exert other, usually detrimental, effects on ecological communities (Barrons, 1981; Bull, 1982; Carson, 1962; Dunlap, 1981) and because resistance to toxic chemicals evolves as a function of the frequency with which they are used (Chapter 1).

Green plants account for about 97% of all the carbon fixed in terrestrial and aquatic ecosystems, and their photosynthetic activity supports almost all other components of those systems. Plants carry out photosynthesis by only three different mechanisms, which differ in a variety of ways, including optimal temperature and amount of carbon that can be fixed per unit of water lost (Berry, 1975; Bjorkman and Berry, 1973). Nearly all woody plants at all latitudes and herbaceous plants at high and middle latitudes use the same pathway. Because of the relative uniformity of mechanisms of photosynthesis, plants are generally replaceable in their role as photosynthesizers. That is the basis of single-crop agriculture and the reason why forest productivity is largely independent of the number of species of trees growing on the site. Plants differ greatly, however, in physical structure, chemistry, depth of soil from which they extract water and minerals, and kinds and numbers of mutualistic interactions.

Those differences make it possible for particular species of plants to play different roles.

All animals require high-energy organic compounds as food, so much of their diversity is related to what they eat and how they find it. Animals are involved in many kinds of competitive interactions, and they participate in many coevolved mutualisms with plants (e.g., pollination, fruit dispersal, and plant protection), other animals (e.g., mimicry and protection), and microorganisms (e.g., digestion and bioluminescence). Animals exert most of their effects through consumption of prey. Because energy-rich molecules are usually found at low to moderate densities in ecological POPULATION INTERACTIONS 39
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. communities, most animals are motile and move around to find food. In special environments, however, such as coral reefs and the rocky intertidal zone, each wave brings a new supply of food that cannot be depleted by the foraging activities of the attached animals, and animals compete for space instead of food and actually form the dominant structural elements of the ecological communities.

Like animals, microorganisms are diverse in their modes of obtaining energy. They use various substrates for their synthetic abilities, use various photosynthetic pigments, and derive energy by oxidizing many simple substrates, such as gaseous hydrogen, inorganic nitrogen, sulfur, and iron.

Microorganisms live in the bodies of all species of larger organisms. They are able to synthesize compounds required by plants and animals (e.g., vitamins and nitrates), and they are the only organisms that can break down many biologically important molecules (e.g., cellulose, waxes, and lignins).

PREDATOR-PREY INTERACTIONS
It is useful to think about predator-prey interactions from the perspective of the relative sizes of predators and prey. ''Typical" predators are larger than their prey, kill and consume all or most of each prey item they capture, and must find many different prey items daily or over their life spans; in the extreme, predators like baleen whales feeding on krill are so much larger than their prey that the prey are handled en masse. However, predators can be smaller than their prey and consume only parts of them. If they live externally on their prey, these predators are called herbivores or external parasites. If they live internally, they are usually called parasites or parasitoids (Price, 1980; Thompson, 1982).

These differences are important from a management perspective. Predators that eat most types of prey captured are unlikely to be sensitive to the loss of any particular prey species. Predators that are specialists, eating only a single species or a few closely related species of prey, are desirable for many biological control purposes ( Chapter 14), because they are unlikely to feed on nontarget prey species (which would cause unexpected and unwanted side effects). Knowledge of size relations can be helpful in directing research efforts, because specialist predators are often smaller than their prey. Predator species being considered for biological control must, of course, be screened individually to determine their diets.

Narrow diets can evolve in the absence of size differences between predators and prey. For example, dietary specialists should be more prevalent in stable habitats, where the availability of specific resources fluctuates over rather narrow limits and particular prey types are therefore
POPULATION INTERACTIONS 40
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. reliably available throughout the year (Charnov and Orians, 1973; Cowie, 1977; Pyke et al., 1977; Schoener, 1971; Tinbergen, 1981; Werner and Hall, 1974).

Monophagy is prevalent among predators with special adaptations for capturing and eating prey with unusual defenses, such as armor, spines, and toxic chemicals. Predators can be dietary specialists because their simple nervous systems leave them unable to use other than simple criteria for recognizing prey (Levins and MacArthur, 1969). Monophagy can be so narrow that predators eat tissue of only some types found in their prey; herbivores of large, woody plants are predators of this kind (Crawley, 1983; Strong et al., 1984).

The consequences for an ecological system of adding or removing predators depend on whether the system is predator-controlled or "donor- controlled" (Pimm, 1979, 1980). In a donor-controlled system, the supply of prey is determined mainly by factors other than predation; therefore, the predators have little influence on their food supplies. Familiar examples of donor-controlled systems are communities of sessile, planktivorous animals of rocky intertidal shores, where a fresh supply of food is delivered with each successive wave; consumers of dead plants and animals; and consumers of fruits and seeds. In predator-controlled systems, the predators, by their feeding, reduce the supply of prey and their reproductive ability (i.e., the future supply).

The distinction is important, because removal of predators in donor-controlled systems has little effect on population dynamics and interactions among their prey species, whereas removal of predators in predator-controlled systems often results in outbreaks of the prey and large changes in the relative abundances of the prey species (Estes et al., 1982; Simenstad et al., 1978). The ability to predict such changes is obviously important for the design of management schemes.

An important and rapidly growing application of predator-prey relationships is the use of the natural defenses of plants against their predators as a substitute for synthetic toxic chemicals. A major advantage of using natural defenses is that effects on nontarget organisms are minimized and the plant itself synthesizes and distributes the defenses. Breeding of pest-resistant crop plants has a long tradition, but improved knowledge of the chemical bases of resistance now enables managers to search for specific kinds of chemical and physical defenses in plants to deal with particular herbivores, rather than relying simply on randomized field trials to see what happens to work (Maxwell and Jennings, 1980).

Defensive chemicals are in two major classes — acute toxins and digestibility-reducing substances. Acute toxins are present in plant tissues in small amounts and exert their effects in the bodies of the herbivores by interfering with a basic metabolic process, such as nerve transmission, protein synthesis, or hormone balance. They are effective because animals POPULATION INTERACTIONS 41
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. have many tissue types not present in plants — tissue types that are targets for chemicals not toxic to the plants producing them. Digestibility-reducing substances act in the guts of the herbivores (technically outside their bodies, in as much as no cell membrane has to be crossed), by combining with proteins in the food to reduce their availability to the herbivore. Tannins and resins are common substances of this type. Acute toxins are most effective against generalized herbivores, because specialized herbivores rapidly evolve the ability to detoxify chemicals in their host plants. It is more difficult to evolve counteradaptations to digestibility-reducing substances, so they are often effective against specialized herbivores (Feeny, 1976; Rhoades, 1979; Rhoades and Cates, 1976). Herbivores can, however, evolve resistance to hydrolyzable tannins (Fox and Macauley, 1977). The use of such information in screening for strains of crop plants to use in different situations is still largely a task of the future, but the necessary ecological knowledge is being accumulated.

An important recent discovery is that plants can respond to herbivore attacks by turning on inducible defenses. These induced defenses can cause marked reductions in growth rates, survival, and pupal weights among insects feeding on foliage of previously attacked branches or trees (Haukioja and Hakala, 1975; Haukioja and Neimelä, 1979; Ryan and Green, 1974; Schultz and Baldwin, 1982); Rhoades (1979) has reviewed the evidence. The full significance of this discovery is not yet clear, but once the mechanisms underlying such responses are identified, they might be manipulated to induce defenses in advance of any attacks. Moreover, damaged plants might release volatile materials that induce neighboring plants to increase their defenses (Baldwin and Schultz, 1983; Rhoades, 1982); if so, that opens up still other possibilities to manipulate plant defenses against herbivores to reduce reliance on toxic substances.

COMPETITIVE INTERACTIONS
Competition occurs when a number of individuals use common resources for which the demand exceeds the supply or when resources are not scarce and organisms harm one another in the process of seeking those resources (Birch, 1957). Competition can occur among individuals of a single species (intraspecific competition) or among individuals of different species (interspecific competition). Interactions can be especially strong between two species or spread out over a larger number of species, each of which contributes a small part of the total (diffuse competition). Understanding competition can be helpful in solving environmental problems.

The most direct method of assessing the importance of competition is POPULATION INTERACTIONS 42
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. to remove individuals of one species and measure the responses of the remaining species. Often such a manipulation cannot be made, however, and even if it can, it reveals only the short-term responses to the reduction of competition. Longer-term responses, which might involve genetic changes in the component species, cannot be measured this way, except in the case of very short-lived species, such as bacteria, protists, algae, and some insects.

Competition is much more difficult to study than predation. Unlike predation, a more or less continuous process (all predators must eat regularly), competition can be intermittent, with periods of competition alternating with long periods in which competition is weak or absent. Nonetheless, patterns due in part to competition might persist through periods when competition is absent.

Indeed, if species evolve traits that are adaptations to environments with particular sets of competitors, their behavior might reflect competition that occurred long ago — a phenomenon referred to as the "ghost of competition past" (Connell, 1980).

Most studies of competition have not involved measuring competition directly, but have attempted to predict and find the patterns that should be found in nature if competition were occurring. Necessary though this procedure might be in many cases, it increases the probability of mistakenly assuming either the presence or absence of competition because predictions about resulting patterns have been incorrect. The literature is biased, owing to underrepresentation of negative results, and, not surprisingly, there is much controversy among ecologists about the importance of competition in nature and about the extent to which patterns observed in ecological communities can be attributed to it (Cody, 1974; Diamond and Gilpin, 1982; Gilpin and Diamond, 1982; Strong et al., 1979, 1984; Tilman, 1982). All observers do agree, however, that competition occurs and that it must be understood better if we are to understand ecological communities.

Except for competition for space in marine communities, where animals and plants compete vigorously (Paine, 1969, 1980), competition generally occurs among organisms at the same trophic level in a community. Some authors have suggested that competition is stronger at some trophic levels than at others. For example, Hairston, Smith, and Slobodkin (1960), observing that the world is green, suggested that competition was intense among plants, rare among herbivores, common among carnivores, and common among detritivores (coal and oil do not appear to be forming at appreciable rates today). They stressed that their conclusions were trophic-level generalizations and did not imply that all carnivores or no herbivores compete. Menge and Sutherland (1976), working in rocky intertidal environments, suggested that the importance of competition should rise with
POPULATION INTERACTIONS 43
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. trophic level and that competition should be especially noticeable in trophically simple communities. They differed from Hairston et al. in predicting low competition among plants.

These and other predictions about competition have recently been examined by Connell (1983) and Schoener (1983). Connell's criteria for inclusion of studies in his sample were more stringent than Schoener's, so the two reached different conclusions. Schoener found strong support for the Hairston et al. hypothesis in both terrestrial and freshwater communities, but only weak support in marine communities. An important exception was that some groups of terrestrial carnivores, such as spiders and predatory insects, often do not compete strongly. The reason might lie in the fact that these small vulnerable carnivores are preyed on by many potential competitors for their prey (Schoener, 1983). Connell found little support for the Hairston et al. hypothesis.

Thus, the general importance of competition in nature is not settled; but there is general agreement about some patterns in competitive interactions.

First, small animals are more vulnerable than large ones to vagaries in the physical environment and predators and, as a result, are less likely to compete.

Second, most competitive interactions are highly asymmetrical in their effects (Connell, 1983; Lawton and Hassell, 1981; Schoener, 1983), i.e., the removal of one species has a strong effect on the other, but the reverse is not true. Indeed, many cases are so asymmetrical that the effects on one of the species are almost undetectable. Such a situation was predicted by theoretical considerations: a slight difference in competitive abilities between two species can lead to the elimination of one of them in simple environments (Gause, 1934; Park, 1948); therefore, in nature, even a slightly subordinate species might be much less common and found in many fewer habitat types than if the dominant species were not present. Indeed, the experimental removal of the larger of two competitors usually has a much greater effect than the removal of the smaller — this was true in 27 of 32 cases in Schoener's sample.

Knowledge that removal of the larger of two competing species is likely to result in substantial population increases among the smaller species and that the larger, despite its competitive dominance, is more likely to be eliminated by human-caused perturbations might suggest likely consequences to which responses should be prepared, especially in cases of biological control where two or more predators are introduced and they differ in size and hence in behavioral dominance. Human activities often introduce into ecological communities new species that, through their competitive and predatory activities, strongly affect indigenous species. Indeed, some of the most important biological problems have been caused by the introduction of species, whether inadvertent or deliberate. The
POPULATION INTERACTIONS 44
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. general response to such problems has been the enactment of strict legislation governing importation of species. However, because importation is certain to continue, better understanding of its likely consequences will be of great value to the environmental problem-solver.

MUTUALISTIC INTERACTIONS
Mutualism occurs when two species benefit one another. As in competition, the reciprocal effects are rarely of equal strength. Mutualisms, which are found among all major groups of living organisms, range from obligate (as in the association of some algae and fungi to form lichens) to facultative. Most mutualisms are based on the transfer of energy and materials between the partners, and many mutualisms are believed to have evolved from predator-prey interactions (Thompson, 1982). Mutualistic interactions are of particular importance for the environmental problem-solver, because loss or severe reduction of a mutualist can have a major impact on a target species that would be unexpected if the nature of the mutualistic interaction were not known.

Most frugivores eat many different kinds of fruits over the course of the year, and many include animals in their diets (Wheelwright, 1983). However, many plants have brief fruiting seasons and are visited by only a few species of frugivores. Therefore, plants might be more seriously affected by the loss of one species of frugivore than would a frugivore by the loss of one of the species of plants whose fruits it eats. In contrast, most plants are visited by many species of potential pollinators, whereas many species of pollinators, especially bees, restrict their foraging to one or a few species of plants. (Agriculture, however, depends heavily on the honey bee, Apis mellifera, which has a catholic diet and ready access to the flowers of most cultivated plants. The loss of bees, as sometimes occurs when pesticides are used extensively in a region, can cause serious losses of fruits and seeds, even when plant growth remains normal and the harvesting activity of the bees is a rather minor component of the total energy flux in the plant community.)
Interactions among plants, their pollinators, and disseminators of their fruits and seeds are generally rather obvious, and managers are usually sensitive to the need to maintain these mutualistic relationships. Relationships between plants or animals and microorganisms, although less apparent, are equally important. Fixation of nitrogen in terrestrial ecosystems depends on the presence of a few species of bacteria and cyanobacteria associated with the roots of some species of plants, especially legumes (Alexander, 1971; Bond, 1967). Digestion of cellulose by most animals depends on the presence of specific microorganisms in their guts.

POPULATION INTERACTIONS 45
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved.

Generally, however, environmental problem-solvers need not be seriously concerned about the preservation of these mutualisms, because the microorganisms appear to be nearly universally distributed and the probability that a plant or animal will fail to establish its required mutualistic relationships is very low, even in the presence of severe environmental perturbations.

INDIRECT EFFECTS
We have considered direct effects of interactions, but species often influence other species with which they have no direct contact. For example, adult Heliconius butterflies in the American tropics depend for food primarily on the flowers of vines in the genera Anguria and Gurania, both in the cucumber family (Cucurbitaceae). The ability of the butterflies to lay eggs and hence their rate of attack on the larval food plant, passionflower ( Passiflora spp.) vines, can depend on the abundance of cucurbits, even though the two groups of vines do not interact directly. Starfish ( Pisaster), by preying on competitively dominant mussels in rocky intertidal environments, make possible the presence of many species of animals that they do not eat or otherwise directly affect (Paine, 1980). Such indirect effects are less well documented than are direct effects of interactions among species, because more careful experimentation would be needed; and such effects are likely to be missed, simply because investigators do not think to monitor the relevant species in the system likely to be affected by indirect interaction.

CONCLUSIONS
A component of many environmental problems is alteration in the interactions among species. Such a perturbation is a likely candidate for unexpected side effects of a project, most of which are undesirable from a human perspective. Assessing which species are likely to be lost from systems as a result of planned perturbations and which species are likely to be affected (whether beneficially or adversely) by those losses is an important part of preproject analysis and one in which ecological knowledge is central. The knowledge needed to avoid serious problems is often easy to obtain, but no pertinent data are likely to be gathered unless the effects of interactions are considered.

POPULATION INTERACTIONS 46
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. 3 Community Ecology

INTRODUCTION
Every species population is part of an assemblage of species — plants, animals, and microorganisms — that share space and interact. We speak of this group of interacting organisms as an "ecological community." It is difficult to define that term precisely, for two reasons. First, whether communities are discrete entities with higher-order "emergent" properties (not readily derivable from an analysis of constituent species) or simply groups of species from the available pool is controversial (Krebs, 1985; Simberloff, in press). Second, whether sympatric species (occupying the same area) should be considered parts of a community if they do not interact with many of the other species present is debatable. The debate, in part, is over the strengths of interactions and how they contribute to community structure. Many environmental problems arise because the alteration of some of these interactions leads to new arrangements of species populations that, from a human perspective, are less desirable than the former ones.

Because the population dynamics of species depend on the kind and intensity of their interactions with other species — species that prey on them and compete with them and on which they prey — a knowledge of these interactions is often valuable in managing individual species. Chapter 2 discusses population interactions as though they are generally independent of each other.

Yet we know that each species interacts with many others and that the interactions are often indirect. For example, grazers
COMMUNITY ECOLOGY 47
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. depend for food on the productivity of grasses, whose growth depends on the activity of earthworms in the soil, which can be affected by the addition of toxic materials to the soil. Human manipulations of the environment can therefore affect individual species, not only directly, but also in indirect ways that can be understood only in the context of the functional structure of the whole community (Davidson et al., 1984).

Ecological communities have numerous properties that transcend those of their constituent species and that require study themselves — trophic structure, rates of energy and nutrient flow, growth form and physical structure, number of species and their numerical distribution, stability characteristics, and ecotones (zones of transition between two habitat types), to name a few. The composition of a community changes in space and time as a result of physical and biological processes. This variability makes it difficult to detect changes caused by human intervention, even if long-term data are available.

Communities are usually named for the commonest or "most important" kinds of organisms found in them. Thus, we speak of "chaparral communities" and "blue mussel communities.'' The organisms chosen to identify communities are usually the ones that provide physical structure for them, such as plants in terrestrial environments, or that are the bases of food chains in systems lacking fixed structural elements, such as plankton in the ocean.

In addition to its value in managing individual species, a knowledge of community ecology is essential when the object is to manage communities themselves. For example, we often wish to maintain a diversity of organisms in an area, as mandated by law in the U.S. national forests ( Chapter 15 ), or to maintain a particular set of species together, as in parks or agricultural areas.

Restoring degraded habitats requires not only a knowledge of the physical conditions that favor the growth of individual species, but also an understanding of how the species interact under different conditions ( Chapter 18). We might need to know how the introduction of a species can alter community makeup or how well a new species can substitute for another in maintaining community stability. Or we might wish to know how much disturbance a community can tolerate before undergoing important compositional change.

SPECIES COMPOSITION
The species composition of a community — the number and kinds of species present— is one of its most obvious features. A knowledgeable ecologist can deduce a great deal about environmental conditions in an area simply by looking at a site and inspecting a list of species present COMMUNITY ECOLOGY 48
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. (Dayton, in press). Systems under the influence of strong perturbations typically show reductions in the number of species that are numerically dominant. For example, continued Eutrophication of lakes usually leads to an excess of some nutrients, a deterioration in water clarity due to algal proliferation, and a reduction in available oxygen due to increased rates of decay. Species that depend on clear water for foraging or on high oxygen content for respiration disappear from the lakes and are replaced by species that can use the increase in nutrients under relatively anoxic conditions. In Lake Washington ( Chapter 20), continued addition of sewage led to the appearance of the blue-green alga, Oscillatoria rubescens, now known to be characteristic of severe eutrophication; when the sewage was reduced, Oscillatoria disappeared and later the zooplankton community changed.

The presence or absence of an indicator species ( Chapter 7 ) is in itself, however, not always a reliable sign of conditions resulting from human-induced perturbations. Species "typical" of some communities are not invariably present.

Local populations can die out for various reasons, including disease, high predation rates, and unusually harsh weather. A species might have failed to colonize an area because of its isolation. Consequently, reliable use of species as indicators of perturbations requires knowledge of their distribution under "normal" conditions and knowledge of past conditions at the site. In addition, knowledge of the composition of an entire community can substantially improve the usefulness of the indicator approach. Patrick et al. (1967), for example, related a variety of stream pollutants to changes in the relative abundance of groups of algal species that had different tolerances to the pollutants.

The number of species in a community (species richness) often changes in response to disturbance, and species richness has been used as an indicator of disturbance. Some types of stream pollution simplify the stream environment and reduce the number of available niches; others kill off many species outright (Patrick et al., 1967). But moderate disturbance can produce less clear-cut effects, and some perturbations can alter the relative abundance of species without changing the number of species present (Dickman, 1968).

More sophisticated measures and indexes of community composition weight the number of species with the relative abundance or biomass of each (Krebs, 1985). A diversity index increases both as the number of species increases and as the numerical distribution of species becomes more even. A large environmental change often leads to local extinction of many sensitive species and to the predominance of a few "disturbance-tolerant" organisms or organisms capable of using the new conditions for increased growth, so diversity indexes have been used as measures of
COMMUNITY ECOLOGY 49
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. disturbance in a community. Such uses of diversity indexes have been controversial, because they have been applied with little regard for the functional changes that occur in disturbed ecosystems.

Many ecologists hoped that diversity measures would capture key changes in fundamental community processes and thus obviate more detailed analyses of species composition. These hopes have been largely unrealized, and there is increasing recognition that the most important information is often discarded in calculating these indexes (May, 1985). A diversity index should be used with caution, for several reasons:
Many factors other than the disturbance of concern can cause a change in the index. This problem is particularly acute when communities in different areas are compared only once. Site differences in physical and biological factors — nutrient availability, presence or absence of key species, climatic differences, etc. — can cause differences in diversity.
The species present in a community can change substantially without any significant change in diversity indexes.
Some disturbances can increase diversity if they increase habitat heterogeneity, reduce the influence of competitively dominant species, or create opportunities for new species to invade (discussed below).

Most of the complexities of the processes that change diversity are not captured in diversity indexes, which are appropriately used only when we are confident that they reflect the behavior of the system being measured. However, because comparison of long species lists from several communities is cumbersome and trends can be difficult to communicate with such lists, changes in diversity indexes within a community can be used to capture the most salient features of that community. The complete lists of species are still needed, but they need not always be presented in full.

FACTORS AFFECTING SPECIES DIVERSITY
Several processes that contribute to change in the number of species in a community can be of concern when the goal of management is to preserve or increase diversity. A primary determinant of species diversity is environmental heterogeneity. On a large scale, species diversity increases as habitat types are added to the environment ( Chapter 5). More niches are also added as structural complexity increases within a habitat. For example, vertical complexity in the form of an increase in the number of foliage layers is associated with an increase in bird species diversity, because birds often partition a habitat by occupying different horizontal strata (MacArthur, 1965). Plantings that change the number of foliage
COMMUNITY ECOLOGY 50
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. layers can change the number of bird species that a park can support (Gavereski, 1976).

Bird species diversity is correlated much less with plant species diversity than with the structural heterogeneity of vegetation (Karr and Roth, 1971; MacArthur and MacArthur, 1961; Orians, 1969; Recher, 1969). Birds forage widely every day and visit many plants in their search for food, responding differently to plant species primarily when those species are structurally very different, as are coniferous and broad-leaved trees (MacArthur and MacArthur, 1961) or spiny and nonspiny desert shrubs (Orians and Solbrig, 1977). In contrast, many herbivorous insects spend their entire foraging lives on one plant and choose their host plants on the basis of chemical characteristics, which often vary substantially with species (Caswell et al., 1973; Ehrlich and Raven, 1965; Fox, 1981; Rhoades, 1979; Westoby, 1978).

Ground-dwelling vertebrates often partition habitats horizontally, and species diversity increases with increase in horizontal heterogeneity (Pianka, 1966). Horizontal heterogeneity also contributes to bird species diversity:
patchier habitats support more species (Roth, 1976). Structural heterogeneity is positively associated with insect diversity, with many species supported on a single plant, each specialized for foraging or hiding on a different substrate, such as upper leaf surface, twig, and trunk (Heinrich, 1979; Ricklefs and O'Rourke, 1975; Schultz, 1983a). In addition, chemical variations within a plant cause insects to move more often (Schultz, 1983b). Manipulation of physical and chemical structures of crop plants and mixtures of plants can thus be an effective way of combating pests (Crawley, 1983; Hare, 1983; Whitham et al., 1984).

Repeated perturbations in a community change its makeup as species undergo local extinction and reinvade. The number of species is usually relatively small in highly disturbed communities, because few populations are able to re-establish themselves before they are reduced by later disturbances. In contrast, a low rate of disturbance provides few opportunities for pioneering species and might allow competitively dominant species to usurp limiting resources, particularly in space-limited systems, such as rocky intertidal zones or some terrestrial plant communities. Therefore, the number of species in a community is often greater at intermediate rates of disturbance (Connell, 1978; Huston, 1979) than at either low or high rates.

Some climax plant communities seem to require periodic disturbance for long-term maintenance. For example, some California chaparral communities (Biswell, 1974) and some grasslands (Wells, 1965) appear to be maintained by periodic fires. When fire is controlled, these communities are replaced by others.

COMMUNITY ECOLOGY 51
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved.

Predation
Predation and periodic disturbance of other types influence species diversity in similar ways. By removing competitively dominant species, predators can increase species diversity. For example, experimental removal of starfish from a rocky intertidal zone allows competitively dominant mussels to usurp space from other species; when present, starfish open up space for other species by selectively removing mussels (Paine, 1966, 1974). However, although mussels in the absence of starfish predation can take over space and reduce the diversity of macroinvertebrates in the rocky intertidal zone, the mussel beds provide vertical structure that actually increases total diversity when microinvertebrates are also considered; this shows how the effects of spatial heterogeneity and predation can interact (Suchanek, 1979).

Herbivores can exert similarly powerful effects on community structure in terrestrial grazing systems. The effects of grazing on the diversity of plants depends on whether herbivores selectively graze on the competitively dominant species, which increases diversity, or on poorer competitors, which decreases diversity (Harper, 1969). The selective grazing of herbivores can lead to the replacement of naturally dominant but palatable species with species that are spiny or toxic. The influence of predators on species diversity seems to be most powerful in space-limited systems.

Freshwater predators often select their prey by size, and that can result in large changes in the makeup of plankton communities (Zaret, 1980). In some cases, the introduction of planktivorous fish into fish-free lakes can reduce the numbers of larger, competitively dominant zooplankton and lead to increases in smaller species (Brooks and Dodson, 1965), although the changes can be complex and difficult to predict (DeMott and Kerfoot, 1982).

Competition
Competition can influence community structure by causing the elimination of some species from local regions or habitats and by reducing the abundances of species in the habitats in which they occur. There are reasons for expecting competition to be strongest among closely related species (Darwin, 1859; Lack, 1954), and many such cases of competition have been documented. However, competition has been especially looked for among closely related species, and the frequency of competition among more distantly related organisms could be much higher than currently believed. Competition among distantly related species is especially prevalent when space is the limiting resource. Plants of all taxonomic groups
COMMUNITY ECOLOGY 52
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. compete strongly with one another and, in the rocky intertidal zone, animals of different phyla compete with one another and with algae (Connell, 1975; Paine, 1966; Underwood and Denley, 1984).

Territorial exclusion, a form of competition for space, can also occur between species. Interspecific territoriality is most common among closely related organisms, but does occur among more distantly related ones as well, especially among fish (Ebersole, 1977; Myerberg and Thresher, 1974). Some cases of interspecific territoriality among birds also involve distantly related species (Cody, 1969; Moore, 1978; Orians and Willson, 1964).

Competition occurs among distantly related grazing mammals, such as moose and hares (Belovsky, 1984). In desert ecosystems, ants and rodents compete strongly for seeds (Davidson et al., 1980, 1984; Kodric-Brown and Brown, 1979). Many more cases of competition among distantly related species are likely to be uncovered as ecologists devote more effort to the study of such competition.

Competition seems to be rare among herbivorous arthropods (Strong, 1984). This suggests that management practices that exert their effects precisely on the target insect species are unlikely to result in unintended side effects on many other species in the community. The use of toxic substances that adversely affect many species has led to greatly magnified influence on population dynamics of other species ( Chapter 24 ). Careful targeting of management toward the focal species decreases the likelihood of side effects that undermine the goals of management.

Productivity
Field studies of more or less natural ecosystems have shown a positive relationship between the number of species in an ecosystem and its productivity (Connell and Orias, 1964). The most common interpretation is that in productive ecosystems more resources are above the minimal abundance required to support users than in unproductive ecosystems (Connell and Orias, 1964; MacArthur, 1972) and that animals in productive ecosystems can therefore specialize on resources that in less productive environments can be used only by generalists. Also, if productivity affects the number of species of plants, then the richness of species of animals that depend on the plants automatically increases as a consequence.

In contrast, it is commonly observed that eutrophication of lakes reduces, rather than increases, species richness (Rosenzweig, 1972). There is no generally accepted explanation of this response. One model assumes that increasing productivity makes predators more effective in eliminating some of their prey or in inducing wide oscillations in their abundances, COMMUNITY ECOLOGY 53
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. which are likely to lead to extinction from a number of causes (Rosenzweig, 1972). Another suggests that competitive exclusion is more important in enriched environments; enriched environments favor "weedy" species that dominate typical members of less productive environments (Huston, 1979).

Managers must be alert to the possibility that changes in ecosystem productivity, a common goal or by-product of human intervention, will lead to unexpected changes in abundances and distributions of many species in a system. Among the affected species are likely to be some of aesthetic or commercial value.

Spatial Factors
Habitat patch size and isolation, two factors that can have strong effects on diversity, are discussed in detail in Chapter 5. Species diversity increases as the area of an "island" of habitat increases, and species diversity in a given habitat patch generally decreases as the patch becomes more isolated either by distance or by the unsuitability of intervening habitat. How these factors affect design and management of ecological reserves and biological control programs is important, and there is much discussion over how to apply understanding of them to the long-term preservation of species and communities (Frankel and Soulé, 1981; Simberloff and Abele, 1976). For example, to conserve some communities of species, it is necessary to maintain a mosaic of various successional stages (Pickett and Thompson, 1980). The resulting habitat patchiness allows species adapted to each stage to find suitable areas. A mosaic of stages can also afford temporary refuge to prey or host species; they will eventually be eliminated in any particular patch by predators or parasites, but by then they will have colonized other patches (Dodd, 1959).

The spatial configuration of habitat patches also determines the extent and nature of ecotones. Ecotones support many species that would not be present in pure communities. Some forest management plans attempt to maximize diversity by creating configurations of habitat patches with much ecotone while retaining patch sizes and proximities that can support species that rely on single community types (Thomas, 1979). What constitutes the most appropriate configuration of patch size, shape, and spacing depends on the requirements of the species to be maintained.

Summary
The exact roles of the factors that influence biogeographic patterns of species are controversial (Brown, 1984), but the factors reviewed above are all known to affect diversity. An understanding of these factors can COMMUNITY ECOLOGY 54
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. provide an environmental manager with a powerful set of tools for manipulating the environment to bring about or limit changes in diversity. Because management is usually targeted to particular species or groups of species and not toward diversity itself, however, additional knowledge of the natural history and population dynamics of the species of interest is required.

Competition, predation, and mutualistic interactions combine with components of the environment to influence species richness in various ways.

Plant species richness is often high in the presence of low soil fertility and periodic disturbance, both of which interact to slow down the takeover of a site by competitively dominant species (Huston, 1979; Tilman, 1982; Chapter 18).

A similar phenomenon occurs in rocky intertidal habitats, where animals are the dominant competitors for space (Menge and Sutherland, 1976; Paine, 1966).

COMMUNITY ORGANIZATION
Species richness is only one property of a community that influences its structure and dynamics. Abiotic factors, such as moisture and temperature (Holdridge, 1967), and the biological processes of competition (Strong et al., 1984), predation (Paine, 1980), evolution (Orians, 1975), and trophic structure (May, 1983; Paine, 1980; Pimm, 1982, 1984) act to influence the structure of a community by determining its makeup and the constraints under which its constituent species live. Because the linkages between species can be at once complex, indirect, and strong (Paine, 1984), investigating the effects of perturbations by studying single species can be misleading (Kimball and Levin, 1985).

Nonetheless, a small number of factors often dominate the organization of a given community and determine its response to particular stresses. For example, soil nutrients in many tropical forests are primarily tied up in the vegetation and superficial soil layers. When all the vegetation of these forests is removed for cultivation and the soils are unprotected from nutrient leaching, the soil can lose its capacity for regeneration for a long period (Gomez-Pompa et al., 1972). Particular species often dominate the visual appearance and structure of communities, providing physical structure for the existence of many other species. For example, coral reef communities depend critically on the reef- building activities of living corals. In terrestrial communities, vascular plants provide the dominant substrate on which most biological interactions are carried out.

A single species can be critical to the maintenance of a community in its "normal" state, such as starfish in some rocky intertidal communities. Those "keystone predators" exert an influence on community makeup COMMUNITY ECOLOGY 55
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. out of proportion to their numbers or biomass. The idea of keystone species was first applied to predators that have sessile prey, especially when the preferred prey is a dominant competitor for space in the absence of predation. For example, the elimination of the sea otter along the Pacific coast of North America contributed to a large increase in the number of sea urchins, a major food item of otters; and the proliferation of urchins resulted in the decline of kelp beds through excessive urchin grazing (Duggins, 1980; Estes et al., 1982).

Lobsters can function similarly as keystone species by preying on urchins off the Atlantic coast (Mann and Breen, 1972). The African elephant exerts a large effect on the landscape by destroying shrubs and trees; that results in the proliferation of grasses, an increase in the frequency of fires, and the conversion of woodland to grassland (Krebs, 1985).

STABILITY AND RESILIENCE OF ECOLOGICAL

COMMUNITIES
Ecologists hold diverse opinions about the relationships between the numbers of species in an ecosystem and the complexity of their interactions and about the system's responses to perturbations. Some have asserted that simple ecosystems are much less stable than complicated ecosystems (Elton, 1958; Hutchinson, 1959; MacArthur, 1955; Watt, 1964), and others have asserted the opposite (Gilpin, 1975; Goodman, 1975; Horn, 1974; May, 1973; Pimm, 1979).

This diversity of opinion reflects inadequacies in information and the use of different definitions of stability and different types of communities and perturbations. "Stability" has been used to refer to lack of fluctuations (constancy), resistance to being changed by external perturbations (inertia), speed of recovery from perturbations (resilience), and other ideas (Goodman, 1975; Holling, 1973; Orians, 1975). Frank (1968) has pointed out that a community of long-lived species can appear to have some aspects of stability merely because the component species live a long time.

A general relationship between stability in any general sense and species richness is unlikely. Many natural ecosystems are species-poor, but nonetheless stable by some definition mentioned above (e.g., Arctic tundra). And some species-rich systems are sensitive to disturbance, because of the intricacies of the connections among their component species (e.g., tropical rain forests).

Moreover, human-induced perturbations not only change species richness, but also create new patterns of interactions (e.g., Cairns, 1980). Until the species have adjusted through evolution to those new patterns, the systems might behave in ways that reflect not simply
COMMUNITY ECOLOGY 56
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. their altered richness, but the evolutionary novelty of the interactions (May, 1973).

For management purposes, it is important that the meaning of "stability" most appropriate for the problem at hand be clearly specified. In some cases, such as preservation of valued species, it could be most important to prevent the system from being changed very much by the planned actions ( Chapter 16). In other cases, such as control of erosion, it could be more important to quicken the return to a former condition of the community, because the problem depends primarily on the duration of a disturbance. Ecological knowledge probably will never be able to provide answers that are general and yet precise enough to replace the need for understanding specific systems and perturbations. Such knowledge can be expected, however, to help in focusing research more narrowly on the most important interactions.

INVADABILITY
An early survey of invasion by plants and animals was carried out by Elton (1958). He concluded that invaders were more likely to establish populations in cultivated and otherwise disturbed environments than in pristine environments, and he noted that islands were more susceptible to invasion than mainland areas. This general perspective has been supported by recent research, although the reasons for the relationships are not much clearer than they were 30 years ago. Determining whether a species might invade new areas requires knowledge about its life history, relationships with other species, and responses to various agents that perturb ecosystems.

Herbaceous plants have been among the most successful invaders of new environments. The flora of California now contains nearly 1,000 exotic plants, and much of the intermountain west is dominated by European and Asian annuals (Mack, in press; Mack and Thompson, 1982). Communities of freshwater fish also appear to be unusually susceptible to invasion by exotic species (Courtenay and Stauffer, 1984). Birds do not invade new areas as easily.

Only three natural invasions of North America by birds have occurred during the last century: those of the cattle egret and two gulls. All three exploit food resources that have greatly expanded in recent decades (Orians, in press).

Deliberately introduced species, such as starlings and house sparrows, primarily exploit human-modified environments. Many of the escaped captive birds that have established feral populations in North America also exploit new food resources, particularly those provided by extensive plantings of ornamental trees and shrubs in southern cities.

COMMUNITY ECOLOGY 57
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved.

Invasions by herbivorous insects are complex, but most species feed on plants that are closely related to the species on which they feed in their native range (Furniss and Carolin, 1980). Many insects colonize introduced plants in all parts of the world, but their natural food plants are generally unknown (Strong et al., 1984).

SUBSTITUTABILITY
When a species is removed from an ecological community, its roles are sometimes taken up entirely or in part by other species. The degree to which this occurs is referred to as substitutability. Because all species are involved in many different interactions, substitutability probably varies with the particular role being considered. For example, the blight-caused loss of the American chestnut in Appalachian forests resulted in only a temporary reduction in rates of photosynthesis in those forests, because other trees replaced chestnuts in the canopy. However, species that are specialists on the tissues of chestnut trees (folivores, frugivores) must have suffered major losses that will continue as long as chestnuts are rare.

The likely effects of species losses on community dynamics depend on the details of current interactions, so an important part of project planning is a survey of competitive, predator-prey, and mutualistic interactions of an obligate and specialized nature. Such information can help in predicting which species losses are most likely to affect other species in the system. The significance of the potential effects can be evaluated, and steps to reduce the likelihood of their occurrence can be included in the project plan.

ECOLOGICAL SUCCESSION
As long as physical conditions do not change greatly, more or less distinct communities tend to replace others after disturbance in a predictable way.

Although ecological succession was originally thought of as a community process, examination of particular successions has shown that abrupt, wholesale extinction of the constituent species of one community with concurrent colonization by the species of another is rare (Drury and Nisbet, 1973). The fates of some pairs or groups of species are inextricably intertwined, as are the fates of some mutualists, but these linkages are in a minority. Typically, the times of appearance and disappearance of most species in a succession are generally independent of those of others, and some species that seem late are present early, but in an inconspicuous form.

COMMUNITY ECOLOGY 58
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved.

The nature of the interactions among species that determine their turnover during a succession and the relative stability of the climax stage are poorly understood for many successions, partly because no community is exactly like any other. Recent research has suggested that processes and patterns of succession differ among communities and depend on which species are present at the start and are available to colonize later and on their life histories (Horn, 1976). Some early species modify the environment to facilitate growth and recruitment of other species, as colonizers of sand dunes stabilize the soil and so allow others to become established (Olson, 1958). Many pioneering plant species are so intolerant of shade that the shade they create inhibits growth of their own seedlings. Some species inhibit others chemically (Rice, 1974). Some late successional species persist because they are more tolerant of potential sources of mortality, such as fire or grazing (Harper, 1969; Sousa, 1984).

Although successions are highly variable in detail, most have some characteristics in common. Odum (1969) listed many patterns of change in energy flow, biomass, and physical structure that are predictable. Some of these patterns, such as the ratio of gross production to respiration ( PG/R) for the community as a whole, can indicate the stage of a succession and how long a given stage is likely to persist without intervention.

Early successional stages typically have relatively high PG/R, whereas later stages have ratios approaching 1:1. Part of the reason is that early species are usually herbaceous, with most of each plant's resources devoted to growth and reproduction; many later species, which persist longer, support more woody tissue and devote more resources to competition than to reproduction. Thus, early stages do not lose in respiration most of the matter produced by photosynthesis, as do later stages, and usable (net) production is relatively high.

The high net production of early succession is harvestable for human use, and this is taken advantage of in agriculture and forestry.

Human societies usually try to maintain early successional stages precisely because they are more productive, but maintaining them in the face of the natural tendency for change requires large expenditures of energy, effort, and materials. Prolonging normally short-lived early successional stages by calculated disturbances (such as plowing and weeding) or the use of chemicals (such as herbicides and pesticides) entails environmental and health problems (e.g., see Chapters 14, 23, and 24). Simply harvesting in the same site for a long period can result in slow degradation of the soil by erosion and leaching of nutrients. As the soil loses its capacity for production, the economic and environmental costs of maintenance grow with the use of fertilizers.

The challenge for ecologists is to help to identify ways of minimizing COMMUNITY ECOLOGY 59
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. the large expenditures and environmental costs of maintaining early successional communities. Integrated pest-management programs aim at reducing the role of pesticides by integrating the use of pesticides with other modes of control (e.g., crop rotation and biological control) on the basis of detailed studies of pest life history and ecology. Soil erosion can be reduced by such techniques as reducing tillage and selecting optional contours (Greenland and Lal, 1977). Herbicide applications to powerline rights of way can be reduced by planting shrubs that impede succession (Niering and Egler, 1955).

CONCLUSIONS
All human-induced environmental disturbances alter interactions among species in some way that leads to direct and indirect affects on the composition of ecological communities and their dynamics. Generally, the direct effects on species of concern are more readily identified and anticipated than are the indirect effects, especially the effects that influence community properties that are the summation of activities of many species. The major question is sometimes how long a community will remain in an altered state. At other times, the main question is how seriously a community is changed. Major changes might be intolerable, even if the community eventually returns to its predisturbance state. The problems described in this chapter are among the most difficult to deal with and are accordingly those for which careful planning and monitoring of a project are especially important, if unexpected and undesired ecological changes are to be avoided or reduced.

COMMUNITY ECOLOGY 60
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. 4 Materials and Energy
Organisms depend on the input of energy (in the form of sunlight or high- energy molecules), water, and mineral nutrients for metabolism and growth.

Human societies depend on harvesting organisms or parts of them, so many of our perturbations of ecosystems are intended to maintain or increase production of organisms. Methods of increasing production of organisms useful to people include domestication, increasing the proportion of production that comes from species of economic value, removing predators, reducing competition, and increasing supplies of the resources that support production (e.g., by fertilization, irrigation, and modification of microclimates). Some valued ecosystem traits, such as processing of wastes and generation of aesthetically attractive landscapes, depend on the total production of species in the system, but often only a portion of total ecosystem production is of immediate value.

Environmental problem-solving is often aimed at increasing allocation of production to particular forms, such as wood production in forests or seed production in agriculture.

The overall determinants of biological production in terrestrial ecosystems are generally well known. The rate of photosynthesis depends on solar radiation, the availability of water in the soil, abundance of mineral nutrients, and temperature. All these are commonly manipulated to increase productivity of agroecosystems.

Productivity is tied to fluxes of energy and matter. Sometimes it is measured in units of mass per unit time, as when one assesses the growth of vegetation or livestock. Sometimes it is measured in units of energy MATERIALS AND ENERGY 61
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. per unit time, especially when the objective is to estimate the efficiency with which one form of biomass (say, a plant) is converted to another (an animal).

Viewed as energy transformation, productivity conforms to the laws of thermodynamics; viewed as material flux, it exhibits conservation of mass.

Hence, the measurement of nutrients and nutrient fluxes is a logical companion to productivity studies.

Production, as opposed to productivity, is the material produced and hence is measured in units of biomass. It can be differentiated into gross production and net production. Not all gross production ( P
G) is available for harvest,
because materials and energy are expended metabolically by living organisms through respiration ( R) to maintain themselves. Net production ( PN) is the portion of gross production that exceeds maintenance ( P G = PN + R). In most
ecosystems, NP corresponds to the amount of material that can be harvested or processed by detritivores.

Measuring the production of entire ecosystems is difficult, and only incomplete data are available on even the most intensively studied ecosystems.

Nonetheless, enough is known to enable us to see general patterns in the allocation of the products of photosynthesis to compartments in the systems.

For example, interesting patterns are revealed by examining ratios of energy stored in long-lived tissues to energy stored in short-lived tissues of plants. If the ratio is high, much energy is tied up in tissues that remain intact for most of the life of the plant, such as trunks, branches, and large roots of trees and shrubs and rhizomes of some perennial herbs and grasses. The patterns revealed by such a comparison are as follows (Jordan, 1971):
The ratio of production of wood (long-lived) to production of litter (short-lived) in forests generally increases as solar energy available during the growing season decreases — that suggests the increasing importance of energy storage in low-energy environments. The ratio of wood production to litter production decreases as precipitation decreases — that suggests that increases in size are less valuable than the production of photosynthetically active tissue in dry environments.

An important implication of these patterns is that a higher proportion of energy tied up in long-lived tissues means that less energy is available for use by other organisms in the community, because long-lived tissues are not heavily used as food.

Consideration of ratios directs attention to trade-offs between different patterns of allocation. A plant, for example, can allocate a quantity of energy to wood, leaves, flowers, pollen, nectar, fruits, or defensive chemicals. An increase in the allocation to one comes at the expense of allocation to at least one of the others. Domestication involves major changes in allocation patterns to increase the energy devoted to products of value to
MATERIALS AND ENERGY 62
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. people (Snaydon, 1984). The trade-off often expresses itself as a reduction of resistance to harsh climatic conditions, predators, parasites, and pathogens.

Protection from these agents must be provided by people — e.g., through weeding, application of pesticides, and killing of predators.

PERTURBATIONS AND PRODUCTIVITY
Natural perturbations — such as floods, droughts, windstorms, and fire — can cause extensive and profound changes in an environment, sometimes even the elimination of living organisms. Perturbations caused by people can be just as extensive geographically, and their cumulative effects can be profound, e.g., conversion of forest to croplands or cropland to residential areas.

Productivity often increases as a result of perturbation, but decreases as perturbation becomes more severe (Odum et al., 1979). Sulfur dioxide and nitrous oxide at low concentrations in the air over a cornfield can increase production, because of the fertilizing effect of the sulfur and nitrogen. But, as their concentrations increase, productivity of the cornfield decreases, eventually to far below the preperturbation level.

Common physical perturbations are those imposed deliberately in land management, e.g., logging or selective harvesting of forests (Chapters 19 and 23), channeling or damming to alter water flow ( Chapter 21 ), and mining (Chapter 18). The disturbance might be simply the harvesting of the species of interest ( Chapter 12 ). Development projects can also inadvertently affect species (Chapter 16).

Effects of perturbations caused by chemicals can be more subtle than those caused by physical alterations. Pollutants that have been taken up by photosynthesizers can be passed on to grazers or detritivores, stored in those organisms, and concentrated (as in the case of some radionuclides and fat- soluble substances). When the grazers are eaten by predators or scavengers, the pollutant burden is passed along. Other predators or scavengers might eat those and thus continue the process of concentration. Eventually, pollutant concentration can be high enough to cause significant biological effects (Chapter 22).

CHEMICAL PATHWAYS AND BIOLOGICAL

CONCENTRATION
Biological concentration of organic chemicals, especially hydrophobic, lipid-soluble ones, can be extreme. For example, polychlorinated biphenyls (PCBs) can accumulate to more than 100,000 times the concentration in water.

Eating 0.5 kg of Lake Erie fish can cause as much PCB intake as drinking 1.5 × 10 6 L of Lake Erie water. Such pollutants cycle both
MATERIALS AND ENERGY 63
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. through the inorganic parts of the environment and through the biota, and their presence can affect primary and secondary productivity directly or indirectly.

The DDT case study ( Chapter 24) shows how a pollutant can affect populations by altering reproductive performance.

The pollutants that follow a decomposition pathway might appear, in some form, in the sediments of a water body or in the soils. Some metals released into the environment are transformed by bacterial action and become either more or less toxic to higher organisms. For example, relatively insoluble inorganic mercury, which is not highly toxic, can be methylated and become biologically mobile and highly toxic.

Effects of chemical perturbations become evident when death or illness results from excessive concentrations of toxic materials. Chemical perturbations can also affect ecosystem processes and patterns. For example, increased input of phosphorus often results in eutrophication of lakes, with substantial increases in primary productivity and changes in species composition (Chapter 20).

NUTRIENT FLUXES
The major elements of organisms — e.g., carbon, nitrogen, oxygen, sulfur, and phosphorus— are involved in massive global cycles that involve both living and nonliving components of the biosphere. Except for phosphorus, these elements can be present as gases, and the magnitudes of their movement are being substantially modified by human activity. For example, the anthropogenic fluxes of sulfur today are approximately equal to those of the natural sulfur cycle (Andreae and Raemdonck, 1983; Franey et al., 1983). Changes in global biogeochemical cycles affect environmental problem-solving at specific sites and are beyond the control of managers. But they are the most important of the cumulative effects of individual projects, because the perturbations are largely the sum of the inputs of specific projects. One of the most challenging problems in maintaining high environmental quality is to find ways of reducing undesirable effects of individual projects on biogeochemical cycles. This requires dealing with the cumulative impacts of many projects.

Patterns and rates of nutrient fluxes have traditionally been studied along with biological productivity, especially in agricultural systems where relations between nutrients and productivity were first demonstrated. Construction of ''nutrient budgets" for ecosystems by tabulating the incomes and outputs of various elements is an application of the first law of thermodynamics, i.e., net energy changes in a system are determined by the inputs and outputs, whatever the internal pathways or mechanisms. Nutrient conditions in both Lake Washington (Chapter 20) and Southern
MATERIALS AND ENERGY 64
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved.

Indian Lake ( Chapter 21 ) were predicted essentially in this way. Attaching biological relevance to nutrient values, however, requires an understanding of physiological processes. The importance of phosphorus as a fundamental and controllable limiting nutrient was central to arguments about Lake Washington, for instance. Among the elements that constitute major proportions of cellular mass (carbon, nitrogen, oxygen, and phosphorus), phosphorus is unique in not having a gaseous atmospheric phase. For this reason, plant production in lakes is often limited by rates of supply of phosphorus and can be manipulated over long periods. Because the potential for plant production rises and falls with the availability of phosphorus in the water, growth decreases if phosphorus is removed from the water and its input is lessened. (The rate of decrease depends on the size and shape of the lake and the flushing rate of water.) Although early work on phosphorus in lake waters discounted its importance as a regulating nutrient (Juday and Birge, 1931; Juday et al., 1928), experience with eutrophic systems showed its pivotal role by the middle of the twentieth century (Hasler, 1947). Domestic sewage is especially rich in phosphorus from animal wastes and the polyphosphate complexes that are commonly used as builders or surfactants in laundry detergents.

Relationships between production and light are sometimes more important than those between production and nutrients. For example, Southern Indian Lake remained thoroughly mixed after impoundment, and suspended solids from an eroding shoreline increased abiotic turbidity and diminished light penetration (Chapter 21). Lake Washington was moving to a somewhat similar situation of low water transparency during 1963 and 1964, the peak years of enrichment (Chapter 20).

In Southern Indian Lake, the element that received most attention after impoundment was not the one that had been targeted by the impact assessments.

Impact studies had considered phosphorus because it is often a limiting nutrient in freshwater lakes, but mercury emerged as more important. The environmental problem became chemically much less like that of Lake Washington than like the problem with DDT ( Chapter 24). In Southern Indian Lake, the deleterious material was not manufactured and added artificially to the system, but was released as a consequence of the manipulation.

INTERACTIONS AMONG PRODUCTIVITY, BIOMASS, AND

NUTRIENTS
Development of our understanding of productivity, biomass, and their relations to nutrients was spurred in part by practical concerns. During the 1960s, a debate arose about the causes of lake eutrophication and the MATERIALS AND ENERGY 65
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. most potent management options (Edmondson, 1974; Likens, 1972; National Research Council, 1969). That phosphorus and phosphorus loading were basic for controlling productivity in eutrophic basins was shown by convincing logic and careful analyses of cases in which loading was altered by design (e.g., Edmondson, 1969, 1972; Schindler, 1974).

Current models are based on findings that are embodied in quantitative relations among hydrological conditions, basin structure, and nutrient and chlorophyll concentrations in lakes (Chapra and Reckhow, 1983; Dillon and Rigler, 1974; Vollenweider, 1969, 1976). These models help to relate physical and chemical circumstances to productivity and trophic state. They are relevant both to applied issues and to studies of the constraints on energy transfer among trophic levels and of the magnitude of the production base. A key feature of the models, however, is not so much what they explain, but what they do not. They cannot predict algal biomass, chlorophyll, or related entities even to within an order of magnitude when a set of heterogeneous basins are compared.

Substantial residual variability in lake productivity cannot be explained by nutrient loading (Carpenter and Kitchell, 1984; Harris, 1980). For single lake basins or lake districts, the fits are usually far better, particularly when variation in externally derived nutrient loading is large (Edmondson and Lehman, 1981; Schindler et al., 1978), but even in such basins, changes in chlorophyll and water transparency can be extreme, despite relatively constant nutrient loads (Edmondson and Litt, 1982).

INDEXES OF ECOSYSTEM FUNCTIONING
Productivity is used as an index of perturbation in a number of ways. In aquatic ecosystems, following the ideas of Eppley and Peterson (1979) and Harrison (1980), primary production can be considered as a sum of production based on "new" nutrients and production based on "regenerated" nutrients. The "new'' portion of production is the primary carbon fixation that is supported by externally supplied nutrients, e.g., from upwelling, chemical weathering, and stream or overland runoff. The "regenerated" portion of production is the availability of nutrients in situ from excretion and decomposition of cells.

Production that is based on regeneration cannot be harvested without compromising the production base itself. Measuring "new" and "regenerated" nutrients is not easy, but the data needed are not much different from those needed to construct quantitative nutrient loading models (Vollenweider, 1969, 1975, 1976). The techniques for doing so have become very sophisticated (Reckhow, 1979).

The cycling index (Finn, 1976, 1978) is a useful measure of the effects MATERIALS AND ENERGY 66
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. of perturbations on nutrient fluxes in ecosystems, especially terrestrial ones.

The cycling index is the ratio of the amount of nutrients recycled in an ecosystem per unit time to the amount of nutrients moving through the system.

For most undisturbed terrestrial ecosystems, the index is between 0.6 and 0.8.

The 20-40% of the nutrients lost must be made up either from the atmosphere or from weathering of bedrock, if the ecosystem is to remain in a steady state. In agricultural or other disturbed systems, the index is often much lower, and productivity is maintained by fertilization.

CONCLUSIONS
Our efforts to extract materials from ecosystems, to use them as waste processors, and to manage them for productivity or maintenance of species richness often conflict. Managing the conflict requires understanding of the determinants of biological productivity and of how those determinants are influenced by human-induced perturbations. The most difficult changes to predict are those involving dynamics of important nutrients, and most surprises result from incorrect prediction of the effects of projects on the behavior of nutrient elements. The same problem is central in understanding the influence of human activities on global air circulation patterns and climate — issues addressed only peripherally here. Changes in behavior of nutrients seldom attract attention, because they are often not directly perceivable by the unaided senses, although they do become obvious, for example, when they impair visibility in scenic areas.

Other conflicts arise over decisions to divert energy flow patterns in ecosystems toward species of indirect value to people and away from species with no current value. Indeed, much of the difficulty in preserving species richness is related not to overexploitation of species, but to conversion of the ecosystems to uses that are incompatible with preservation of the community of species originally present.

MATERIALS AND ENERGY 67
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. 5 Scales in Space and Time
When we alter the size, shape, and spatial distribution of patches of particular ecological communities, we alter the population dynamics of the species that live in them. The ability of a population to withstand environmental fluctuations, for example, depends not only on the life history of the species, but also on population size and the availability of immigrants from other populations. Analogously, the temporal characteristics of environmental manipulations can influence the kind and strength of their effects. Perturbations of longer duration or greater frequency might exceed the capacity of a community or species to absorb or recover from them, whereas short or single disturbances might not. Focusing on the consequences of single perturbations can lead to a failure to perceive the patterns and cumulative effects of those perturbations over time and space. If a population or community is repeatedly disturbed for long enough, changes qualitatively different from and more serious than the effects of single perturbations often occur. An appropriate choice of scale for thinking about, analyzing, and manipulating these processes is crucial.

PATCHINESS AND COMMUNITY COMPOSITION
Species-Area Relationship
Large areas tend to have more species than small areas of similar habitat type (Cain, 1938; Connor and McCoy, 1979; Gleason, 1922; MacArthur SCALES IN SPACE AND TIME 68
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. and Wilson, 1967; Preston, 1960), partly because larger areas typically have more habitat variation within the general habitat type. Each variant contains species adapted specially to it. However, even when habitats in a region are uniform, there is a relationship between area and number of species. A second reason for this relationship, which is especially pronounced in small areas, is that each species has a minimal viable population size for a given probability of extinction (Shaffer, 1981). As the area of a habitat decreases, local populations get smaller and extinction becomes more likely. In addition, the species-area relationship is partly a result of sampling (Connor and McCoy, 1979), i.e., large areas receive more immigrants than small ones and therefore obtain larger samples from the species pool. These three reasons for the species-area relationship are not mutually exclusive and can be difficult to distinguish (Connor and McCoy, 1979).

Extinction of Small Populations
At least five forces increase the probability of extinction of small populations:
Demographic stochasticity. Random fluctuations of demographic events (birth, death, and determination of sex) endanger small populations. For example, the probability that all individuals in a generation will be male is much greater in a small than in a large population.
Genetic stochasticity, consisting of inbreeding depression and production of homozygotes for lethal or severely deleterious recessives. Inbreeding depression is the general decrease in traits that contribute to fitness, such as fertility. It has been documented in many plant and animal species and appears to be associated with the increased homozygosity (in which the two copies of a particular gene are identical) that results when near relatives mate. Because mating with relatives is more frequent in small populations, inbreeding depression is greater. In addition, greater homozygosity for the population as a whole might be associated with reduced genetic variability and lead to reduced ability of the population to adapt to environmental change (Soul é, 1980). Mating with near relatives also increases the likelihood of producing individuals homozygous for recessive traits that are lethal or severely deleterious. Environmental stochasticity. Random variation in the physical or biotic environment of a species affects demographic values, whether the population is large or small. Such variation, even if not severe, can threaten the very existence of a small population, however, because the probability
SCALES IN SPACE AND TIME 69
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. that all individuals will be killed is much greater for a small than for a large population.
Disasters and catastrophes. The once-in-a-century flood or fire, for instance, can destroy a local population.
Social behavior. Some animal species have stylized forms of social behavior (e.g., predation, defense, thermoregulation, and mating displays) that break down if there are too few individuals. A breakdown can lead to breeding failure and endanger the population.

Minimal viable population size varies widely among species, for a number of reasons. In general, the minimal number of individuals necessary to support a population for a long period increases as average population density decreases.

The average area necessary to support an individual animal is greater for predators than herbivores, and in general it increases with body size within groups of similar species (McNab, 1963; Schoener, 1968). Species differ in their ability to tolerate the increase in inbreeding that occurs in small populations. In general, plant populations appear to be able to survive longer on smaller sites than animal populations. It is easier to conserve plants than animals by artificial means (e.g., cold storage and seed banks).

Patch Geometry and Edge Effects
The shapes of habitat patches cause effects similar to those due to patch size. As patches deviate from circular to linear, the proportion of their area close to an edge increases, as it does when they decrease in size. Species adapted to conditions found at the interfaces between patches of different types can exploit an increasing fraction of the areas of small patches. They can compete with species adapted to the interiors of patches, parasitize them (Brittingham and Temple, 1983), or function as predators against which interior species are not well adapted (Wilcove, 1985). Human modifications of environments often create patches that are much longer than they are wide (Godron and Forman, 1983), thereby exacerbating the effects of patchiness itself.

The shape and orientation of patches can have important ecological consequences. Cutting of forests into strips causes less erosion if the strips follow the contours of the terrain, rather than being oriented at right angles to them (Hornbeck et al., 1975). At high latitudes, direct sunlight might penetrate to the ground only at dawn and dusk in narrow clearcuts oriented in an east- west direction, whereas in north-south patches direct sunlight is present at ground level at midday— the time of most intense solar radiation.

SCALES IN SPACE AND TIME 70
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved.

DISTRIBUTION OF PATCHES IN SPACE AND TIME
Spatial or temporal patchiness is sometimes obvious, but more often difficult to detect. For example, the distribution of herb species in a field might appear random when actually it is determined by the microspatial heterogeneity of soil nutrient conditions (Tilman, 1982). Patchy distribution of organisms can result from variability in the physical environment (such as soil types), physical disturbance, and patchiness in biological interactions (see also Chapter 3 ).

Studies of intertidal and subtidal communities have shown the importance of local heterogeneity generated by physical and biological disturbances in both temperate areas (Dayton, 1971; Menge and Sutherland, 1976; Paine, 1966; Paine and Levin, 1981) and tropical areas (Connell, 1978; Porter et al., 1981).

The spatial and temporal variability of lake and deep-sea benthos is well known (Berg, 1938; Brinkhurst, 1974; Grassle et al., 1975; Jonasson, 1972).

Because marine systems are dominated by species that do not derive nutrients from the substrate (animals and nonvascular plants), substrate-related variability is due primarily to the physical environment and stability of the substrate and the type of anchorage it provides (e.g., Dayton, 1984, 1985). In terrestrial systems, however, soils differ primarily in their ability to supply nutrients and water, so variability in distribution of plants, burrowing animals, and species that depend on plants is related to these properties. Soil scientists could be included in terrestrial research teams more often than they usually are.

Spatial Considerations
Population dynamics in patchy environments are determined primarily by the rates of individual movements between patches and rates of local population extinction. The properties that enable populations to maintain themselves in patchy environments include high dispersal rates, tendencies to cross unsuitable habitats, high growth rates, early reproduction, and high reproductive rates (Baker and Stebbins, 1965; MacArthur and Wilson, 1967). These traits increase the probability that isolated patches will be found, that reproduction will occur before the patch becomes unsuitable, and that new colonists will be generated.

Species with the traits increase in abundance in environments that are heavily modified by people.

Patches are not static, but change with time, primarily as a result of the growth of and interactions among colonizing organisms. The succession of organisms over time is driven by several processes that are not mutually exclusive, but that differ in their relative importance under SCALES IN SPACE AND TIME 71
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. different conditions (Horn, 1974, 1976; Shugart and West, 1980). Species are adapted to conditions at different stages before populations become extinct as a result of within-patch changes. Early successional stages typically are much shorter than later successional ones. Therefore, even if rates of disturbance are low, populations of early successional species are more likely to become extinct. However, current rates of human-caused disturbance are so high in most areas of the world that species requiring late stages of succession, such as old-growth forests, are in the most precarious positions ( Chapter 17 ). In addition, many organisms have complex life histories in which different stages require distinct habitats. Not only must individuals find all the necessary habitat types at the correct time, but populations are affected by fluctuations in the availability of habitats for each stage; these fluctuations can be independent of each other (Istock, 1967). The supply of appropriate habitat must be adequate at the correct time during the year; its abundance at other times can be irrelevant.

Even within an apparently uniform patch, interactions among subunits can be complex. This complexity was not a factor in the case of Lake Washington (Chapter 20), because the high inflow of water during winter and early spring, when the lake is isothermal, causes free circulation throughout the lake.

Southern Indian Lake ( Chapter 21 ), however, contains several subbasins, and the major outlet has been close to the inlet since the diversion of the Churchill River. The lake is thus not a well-mixed body, and its properties differ between regions, whether or not they are affected by the flow. The high quality of the whitefish fishery before impoundment was maintained by confining fishing to areas where the stocks were virtually free from cestode infestation. After impoundment and diversion of the normal river flow, the stocks became redistributed, and the fish were concentrated in areas with high infestation rates.

Local populations occasionally become extinct because of predation, disease, or physical disturbance. The rate of recolonization is inversely related to the distance from other occupied sites that are sources of immigrants.

Increasing isolation of patches increases the probability that locally extinct populations will not be replaced, and creation of communities that would normally develop without help might need to be managed because of a lack of immigrants (Chapter 18). In general, species characteristic of later successional stages are poorer dispersers than "weedy" species of earlier stages. In plants, late stages show a striking increase in the average size of seeds (Harper et al., 1970; Salisbury, 1942); in animals, the later stages show less tendency to disperse and greater reluctance to cross stretches of unsuitable habitat.

SCALES IN SPACE AND TIME 72
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved.

Temporal Considerations
Community dynamics are strongly affected by interactions among perturbation characteristics (intensity, duration, and frequency), succession, and the rate at which a community recovers. Very different outcomes are possible if relative rates of disturbance and recovery are altered. For example, tropical slash-burn agriculture is compatible with long-term soil fertility if plots are small, are farmed for only a few years, and are allowed to remain fallow for several decades (Gomez-Pompa et al., 1972). However, if plots are made larger and are recut after shorter fallow periods, soil fertility cannot recover within the period of a single cycle and rapidly declines (Myers, 1984).

If pesticides are used infrequently, time might be available between applications for susceptible genotypes of pests to replace resistant types that are at a disadvantage in the absence of the pesticides. If pesticides are used more often, this process does not go to completion, and many resistant genotypes are still present when the pesticide is used again. As a result, the pesticide becomes progressively less effective, applications are increased in frequency and magnitude, and the evolution of resistance among the pests is accelerated (Chapters 1 and 24).

Repeated or continuous perturbations can lead to qualitative changes in community structure, because the ability of the system to remove or recover from the disturbance is exceeded. Striking changes took place in the plankton communities of Lake Washington when the addition of sewage at multiple points exceeded the flushing rate of the lake (Chapter 20).

In addition to the deleterious genetic effects that often occur in small populations (Franklin, 1980; Selander, 1983), there are long-term evolutionary consequences of patch size and distribution. Evolution in response to spatial or temporal change in the environment is retarded by small population size, because genetic variability is reduced in small populations (see Endler, 1977, for a discussion of the effect of migration on evolution). For example, dividing agricultural plots into different sections and applying a different pesticide to each should retard the evolution of resistance.

CONCLUSIONS
The great importance of size and spatial relationships in the working of ecological processes points to the importance of dealing explicitly with scales in space and time in all efforts to solve environmental problems. The major changes in processes and products that accompany changes in spatial and temporal scales can escape attention, if efforts are not directed SCALES IN SPACE AND TIME 73
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. specifically at them in all phases of environmental problem-solving. Well intentioned efforts can be undermined if they are planned for too short a term or for areas that are too small (Soul é and Wilcox, 1980). However, key ecological processes might be obscured if inappropriately large temporal and spatial scales are used. Averaging over large areas can mask the importance of local patchiness for the survival of particular species. Individual trees could be especially susceptible to attack by herbivores, and the maintenance of large populations of trees could depend critically on patches. Similarly, patches of high concentrations of nutrients resulting from defecation of zooplankton might be essential to survival of algae, and patches of high concentrations of plankton might be important to the survival of marine fish larvae (Sissenwine, 1984).

SCALES IN SPACE AND TIME 74
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. 6 Analog, Generic, and Pilot Studies and
Treatment of a Project as an Experiment

INTRODUCTION
The previous five chapters emphasized the application of ecological knowledge of many general types, but they did not deal directly with knowledge specifically applicable to particular problems. However, ecological systems often respond to perturbations in a site-specific fashion, and large or novel perturbations can produce effects not predictable from basic ecological principles, so specific knowledge is often necessary for the understanding and prediction of many environmental effects. This chapter briefly discusses the value and limitations of using specific information and of treating projects and other environmental actions as experiments when potentially serious effects cannot be adequately predicted from information acquired in advance.

Observation of the actual effects of an action can sometimes be the only source of information for management decisions.

ANALOG STUDIES
An analog study concerns an action similar to one of immediate interest.

The most useful analog study would be conducted in the same area, on the same species or ecosystem, on a similar scale, and under similar conditions and would be thoroughly analyzed. Even if the analogs are not perfect, study of them can be useful in scoping a problem, identifying major potential effects, and designing appropriate studies.

ANALOG, GENERIC, AND PILOT STUDIES AND TREATMENT OF A PROJECT AS

AN EXPERIMENT
75 Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved.

When a project is similar to many others, analog studies typically accumulate, especially if they are legally required. For example, in the legally mandated Maryland Power Plant Siting Program for managing cumulative impacts, study of the environmental effects of each power plant should increase the ability to predict the effects of future power plants (e.g., Maryland Department of Natural Resources, 1984). But the use of data on one power plant to predict effects of another also requires knowledge of the differences between the receiving ecological systems and the differences between power plants. Variations in physical and ecological conditions between areas can lead to serious errors in extrapolation. For example, in the Southern Indian Lake project (Chapter 21), no specific analog studies dealt with changing of the level of a shallow lake in a permafrost zone. Because the best analogs — deep lakes in Siberia— differed in important respects from Southern Indian Lake, investigators failed to predict the magnitude of bank erosion that occurred and the consequent changes in water turbidity and mercury release.

In the Lake Washington case ( Chapter 20 ), valuable analogs were available from experience with nutrient enrichment in temperate lakes in Europe. In the derelict lands case ( Chapter 18 ), no studies of methods for diversifying plant communities on previously mined land were available, but attempts to diversify roadside vegetation informed and motivated the research studies. Analog studies were of value in the caribou case ( Chapter 16 ) in identifying the major issues, developing field techniques, formulating a study strategy, and locating important resource people who were familiar with the problem.

Some cases not only relied on analog studies for their design, but were designed so that they became valuable analogs themselves. The Lake Washington experience now serves as a model for many lake eutrophication studies and has contributed importantly to our understanding of nutrient loading in lakes. The comprehensive review and analysis of the Southern Indian Lake project increased its instructional value. Analog studies are most valuable if their results are analyzed and summarized in an easily accessible form.

GENERIC STUDIES
A generic study is designed to increase knowledge of the physical and biological phenomena common to a group of environmental problems. It might be designed to reveal the processes that underlie ecological responses to manipulations in a particular environment— e.g., a study of the effects of marine oil pollution along coastal beaches or of oil pollution in aquatic

ANALOG, GENERIC, AND PILOT STUDIES AND TREATMENT OF A PROJECT AS

AN EXPERIMENT
76 Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. environments in general. It might deal with a restricted subset of environmental perturbations, such as the uses of DDT, or try to provide a model of the effects of a large class of perturbations, such as pollution by organochlorides in general. Because ecological effects vary with the environmental manipulation or insult, the location, the biological communities present, the scale, and other factors, a major question concerning the usefulness of generic (and analog) studies is the extent to which their results can be applied to a particular site and problem. The most useful generic studies identify, characterize, and explain the major ecological effects of a particular environmental insult or manipulation and provide a framework for understanding the effects of variations in location and other factors.

The distinction between analog and generic studies is not always clear. So much was learned from the Lake Washington experience, for example, that it might serve as a generic model. A major practical difference between analog and generic studies is that generic studies are usually well known to biologists, whereas analog studies are often buried in government or industry files. Also, generic studies are usually designed to address a specific question, which means that the postproject monitoring is often good enough to provide comparative data.

When government and industry cooperate in addressing problems of mutual concern, studies can be coordinated and research resources used efficiently. For example, before issuing a permit for exploratory drilling in state- owned offshore lease areas, California requires extensive studies designed by a committee of representatives of academe, government, and industry.

Coordination of studies among a number of applicants for leases in the Santa Barbara Channel allows each to conduct studies on specific problems common to all drilling programs, such as the effects of discharge fluids on hard-bottom organisms; results from all the studies can be used to develop permit stipulations for other applicants. The advantage of such coordinated endeavors is that duplication is minimized, so more research is possible.

Environmental studies, for practical reasons, cover limited areas, relatively short periods, and conditions dictated by the time and place chosen. Moreover, many generic studies involve simulation of the environmental change of interest, often on a smaller scale and with a narrower range of variables and variation than encountered in the actual situation. For example, the aquatic effects of thermal effluents from a nuclear power plant might be studied by injection of hot water into a river (if the appropriate permits required by the Clean Water Act were obtained first). But such an injection would be made over a much shorter time and a much smaller

ANALOG, GENERIC, AND PILOT STUDIES AND TREATMENT OF A PROJECT AS

AN EXPERIMENT
77 Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. area than the usual discharge of effluent, and the injected water might lack some contaminants present in the normal effluent.

Some other limitations of generic studies are discussed in the review of clearcutting studies ( Chapter 23). The Hubbard Brook clearcutting experiments of the 1960s (Bormann et al., 1968) were for a long time accepted as general models of the effects of clearcutting on stream eutrophication, soil nutrient loss, and the regeneration potential of land. However, herbicides were applied after the clearcut in the Hubbard Brook experiments, whereas rapid revegetation is encouraged in many forestry operations. Errors in extrapolation from a generic model to a specific situation can be reduced if generic studies focus on a particular application and location.

Generic reviews can yield ecological knowledge of great potential value.

They might focus on various topics of potential use to environmental managers:
a group of related problems; natural history and ecology of particular species, areas, or communities; general approaches to problems; and broadly useful methods or techniques. A list of useful generic reviews and related literature can be found in the annotated bibliography on ecological impact assessment prepared by Duinker and Beanlands (1983).

PILOT-SCALE EXPERIMENTS
Pilot-scale experiments are designed to investigate the effects of specific kinds of environmental perturbations on time and space scales much smaller than those of a planned project or action. They are appropriate when there is little confidence in the hypotheses underlying the prediction of effects and when realistic simulation of the perturbation is possible.

The major uncertainty in connection with pilot-scale experiments is whether their results can be generalized to similar large-scale phenomena (Hilborn and Walters, 1981). Although the overall effects of large developments, such as hydroelectric dams, cannot be realistically simulated on a small scale, some individual effects of large developments can be studied experimentally with the traditional scientific procedure of dividing a problem into its components for analysis.

Experiments were used to advantage in the derelict lands case (Chapter 18), to investigate the performance of groups of plants under different soil conditions; there is no obvious reason why the results of these studies cannot be generalized to larger-scale applications. Pilot-scale experiments were conducted to study the potential effects of the Alaska pipeline on caribou movements (Cameron et al., 1979). A variety of small-scale experiments were conducted on the effects of DDT (Chapter 24). Labo

ANALOG, GENERIC, AND PILOT STUDIES AND TREATMENT OF A PROJECT AS

AN EXPERIMENT
78 Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. ratory and field experiments were used to investigate toxicity in several species and to determine residence time of DDT under different environmental conditions. Field and laboratory experiments confirmed the occurrence of bioaccumulation. The Atomic Energy Commission program for studying the effects of nuclear radiation ( Chapter 22) is a good example of the combination of a variety of laboratory and field experiments to improve understanding of a complex problem.

TREATING A PROJECT OR ACTION AS AN EXPERIMENT
If reliable prediction of the ecological effects of a project or action is not feasible, even with extensive field investigations, the limits of predictive ability should be openly recognized, so that planners, decision-makers, and managers can take into account the uncertainty of the ecological outcomes of particular actions (Paine, 1981). Viewing a project or action as an experiment can aid in designing a program to monitor effects. Such monitoring can have two major advantages: the detection of unexpected effects can be used as a basis for altering procedures (Holling, 1978), and the monitoring information can be used in the planning and design of similar projects or actions.

The idea of a project as an experiment can be applied to a variety of environmental problems, particularly in impact assessment and resource management. As Beanlands and Duinker (1983) pointed out, however, most impact assessments concentrate on the prediction of impact almost to the exclusion of followup studies to determine actual impact. Even in postproject studies, little effort is spent in determining whether predictions were incorrect and why (Larkin, 1984). An attempt by a study group to audit the results of a large number of impact assessments in the United Kingdom revealed several major impediments to learning from such experiences (Anonymous, undated):
predictions were absent or vague, monitoring schemes were inadequate, documentation was poor, and relevant parties were unwilling to cooperate.

Chapter 10 offers some general suggestions for designing monitoring studies to make them more useful for planning projects or actions. Among the projects described in case studies, the best examples of the idea of a project as an experiment are the Garki malaria project ( Chapter 15), the project to raise the level of Southern Indian Lake ( Chapter 21), the eutrophication and cleanup of Lake Washington ( Chapter 20 ), and the efforts to protect caribou during development of a hydroelectric dam in Newfoundland (Chapter 16).

ANALOG, GENERIC, AND PILOT STUDIES AND TREATMENT OF A PROJECT AS

AN EXPERIMENT
79 Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved.

Walters and Hilborn (1976) and Walters (1984) have suggested that natural resources can be managed experimentally, not only by following up the consequences of management actions, but by basing management decisions on the need for increased knowledge. In that paradigm, managers must weigh the trade-offs between the value of potential immediate yield and the value of information. The value of information is greatest when understanding of the system is poor and the consequences of incorrect decisions most severe.

ANALOG, GENERIC, AND PILOT STUDIES AND TREATMENT OF A PROJECT AS

AN EXPERIMENT
80 Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. 7 Indicator Species and Biological Monitoring
Living organisms can be used to monitor movements, accumulations, and modifications of materials in their environments and to monitor the biological effects of those materials. They can also be used to indicate the effects of habitat alterations and fragmentation and the effectiveness of management schemes designed to preserve or change individual species or community-level patterns.

Toxic substances are used by people because they confer economic or other benefits that are believed to outweigh the dangers they pose. The regulations devised for their use are guided by knowledge of their biological effects. Chemical and physical monitoring can tell the quantities of materials entering or already in the environment and sometimes the fraction that is anthropogenic. But only biological monitoring can tell us what those materials are doing to organisms. Living organisms not only are essential for determining the biological effects of pollutants, but have several advantages over physical and chemical monitoring:
Living organisms can function as continuous monitors. They accumulate records of past conditions in their tissues. These records can sometimes be read and monitored more economically than can records obtained by establishing stations that continuously monitor environmental conditions directly. But living organisms can also distort or obscure the records by altering chemicals in their bodies; therefore, detailed information is needed to determine which organisms are best suited to provide continuous records of various contaminants.

INDICATOR SPECIES AND BIOLOGICAL MONITORING 81
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. Living organisms can increase the sensitivity of monitoring. Many species can concentrate materials in their tissues and so amplify weak environmental signals. Lipid-soluble pesticides are concentrated in body fat and further concentrated when one organism eats another.

Foraging bees bring pollen and nectar to their hives from many nearby sources. Sampling of pollen from bodies of bees has provided a much more complete picture of the distribution of several pollutants in the Puget Sound region than was obtained by more expensive chemical monitoring (Bromenshenk et al., 1985).
Living organisms can provide information about complex mixtures of materials. Laboratory studies, for reasons of expense and ability to control concentrations, typically expose organisms to potential toxicants one at a time. In nature, however, organisms often encounter toxicants in complex mixtures. By observing changes in such functions as locomotion, growth, and reproduction in living organisms, we can determine the effects of those mixtures, even when their exact nature is unknown. (Fortunately, appropriate control measures can often be instituted without knowledge of all the details of a mixture.) Living organisms can provide information about how pollutants affect performance under natural conditions. Test organisms in a laboratory are typically kept in ideal conditions and are seldom subjected to inclement weather, disease, predators, crowding, or food shortages. In nature, however, these insults can occur simultaneously with the presence of the toxic materials of concern. Under such conditions, concentrations of toxicants that would have no effects in the laboratory might be detrimental. Because our concern is to protect organisms under natural conditions, biological monitoring must be an integral part of monitoring schemes, although it must be kept in mind that the variability of natural systems often makes it hard to be certain that an apparent effect is a genuine one.
The use of communities of living organisms for monitoring can provide information about how toxic materials influence patterns of interactions among organisms, community patterns, and processes of concern (see Chapter 3).

CHOICE OF ORGANISMS TO USE FOR BIOLOGICAL

MONITORING
Groups of organisms can be chosen for particular monitoring purposes on the basis of the required speed of response (Cairns et al., 1973) or on the basis of sensitivity to temperature changes (Cairns, 1977), for instance.

INDICATOR SPECIES AND BIOLOGICAL MONITORING 82
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved.

In all cases, the monitoring systems should be chosen with a clear understanding of the goals to be achieved, the status of current knowledge, and the ability and commitment to act on the results obtained.

Life-history characteristics of organisms strongly influence their utility for various types of biological monitoring. For example, short-lived organisms respond quickly to environmental changes, whereas long-lived ones might integrate stresses over years, decades, or even centuries. Species with high metabolic rates and, hence, usually high growth rates are often more sensitive to contaminants than are species with low metabolic rates. Sessile organisms are exposed to all the contaminants that enter their immediate environments, whereas mobile organisms can escape many of them by leaving the area. Many organisms stop reproducing under stressful conditions, so changes in fecundity rates can be important signals of environmental change; but special care is required here, because stress stimulates reproduction in some organisms, particularly plants.

Even within a category, some organisms are better indicators of environmental change than others. Vascular plants are effective detectors of air pollution, because particular species are especially sensitive to particular pollutants. For example, the toxic components of photochemical smog in California were unknown until they were revealed by plant assays. Lesions on beans, spinach, and grape leaves indicated the presence of ozone (Middleton, 1956); lesions on annual bluegrass leaves indicated the presence of peroxyacetylnitrate (Bobrov, 1955). The ways in which different plants respond to major air pollutants are documented by Weinstein and McCune (1970).

Simpler terrestrial plants, such as lichens and mosses, are often more sensitive to airborne pollutants than are vascular plants, because they absorb water and nutrients directly from air and rainwater. As a result, they concentrate pollutants and exhibit toxic effects more quickly than vascular plants, even though they are not generally more sensitive (Hawksworth, 1971; Lawrey and Hale, 1979). A measure of the magnitude of pollution is the lichen species diversity index, which combines the number of species present and their relative abundances (the abundances are based on extent of coverage, growth form, or degree of luxuriance).

In aquatic environments, algae and cyanobacteria are useful indicators of various changes. Increases in nutrients (eutrophication) can be assessed according to increases in biomass of algae ( Spirogyra, Oedogonium, Stigeoclonium, and Cladophora) and several genera of cyanobacteria (such as Oscillatoria in Lake Washington, as described in Chapter 20 ). Diatoms and some other algae concentrate heavy metals and radioactive materials by a factor of several thousand.

Mollusks have been used as monitors in aquatic environments with a INDICATOR SPECIES AND BIOLOGICAL MONITORING 83
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. monitoring network known as Mussel Watch (Goldberg et al., 1978). Species that have been used include some in the genera Mytilus, Crassostrea, and Ostrea in outer estuaries and coastal marshes; in Geukensia in tidal marshes; and in Anodonta and related genera in freshwater lakes, rivers, and streams (Goldberg et al., 1978; Patrick and Kiry, 1976). Much is known about how pollutants are deposited in mollusk shells and preserved for long periods.

Metals, pesticides, and hydrocarbons are also deposited in soft tissue; they are often rapidly flushed from some soft tissue and can be accumulated there, too.

Much more work needs to be done before we will know whether concentrations of most materials in soft tissue can be used effectively to indicate current environmental changes or past environmental contamination.

The utility of organisms as biological monitors of environmental changes is greatest when the functional relationships between perturbation and response are understood. Nonetheless, organisms can be useful as indicators even if causal relationships are obscure. If a valued ecosystem component disappears or is reduced, attention is directed to a change that might be important. Caution is always necessary, because natural environments are highly variable. Species that are ''typical" of some types of environment can be absent from specific locations for reasons having nothing to do with human disturbance, and jumping to a premature conclusion might retard progress toward the goal of understanding the causes of environmental changes and how they affect organisms.

Some general guidelines in selecting species most useful for particular monitoring purposes are nonetheless possible. For example, top carnivores (the organisms that develop the highest concentrations of persistent, fat-soluble pesticides) are often of special value in detecting environmental changes. They have large home ranges and they typically forage over wide areas, where they can be exposed to materials with a patchy distribution that are difficult to detect with fixed monitoring stations. Sea birds are valuable for monitoring, because species that nest in a single location might forage at widely differing distances from the breeding colony; therefore, knowing which species are encountering which chemicals (as revealed by analysis of regurgitated food) helps in locating sources of and areas contaminated by toxic materials (Boersma, in press; Gilman et al., 1979). Sea birds are useful also because the relative ease of measuring their distributions and abundances makes them good monitors of changes in populations of fish and aquatic invertebrates, whose populations are much more difficult to assess. The herring gull ( Larus argentatus) was chosen to monitor pollutants in the Great Lakes of Canada and the United States, because it accumulates pollutants only from the lakes, rather than from land, and because it is a year-round resident (pollutant concentrations INDICATOR SPECIES AND BIOLOGICAL MONITORING 84
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. in its tissues and eggs are not influenced by pollution in distant places). Several environmentally important compounds were first identified in herring gull tissues, and the gull has helped to monitor the fate of persistent organochlorides after control measures were established (Mineau et al., 1984).

Butterflies and moths are well enough known for changes in their distributions and abundances to be detected, and they can evolve rapidly enough for genetic changes induced by environmental changes to appear within a few decades. The familiar cases of industrial melanism among lepidoptera in western Europe and northeastern North America are among the best- documented cases of evolution by natural selection in which environmental pollution was the primary factor in changing the survival rates of some genotypes (Kettlewell, 1973; May and Dobson, in press).

Many animals are used by veterinarians, toxicologists, and physicians to monitor human toxicants in the environment (Buck, 1979; Harshbarger and Black, in press; Schwabe, 1984). The expertise of those specialists could usefully be combined with that of ecologists to improve the choice and use of animals as environmental monitors.

MONITORING AND ENVIRONMENTAL SPECIMEN

BANKING
Living organisms can preserve records of environmental materials in their tissues, but these records might change as the organisms age. Even more important, the death and decay of an organism eliminates such records. It is therefore useful for a program of biological monitoring to include specimen banking. The first indication of the utility of specimens as records of former environments was the documented thinning of eggshells as a result of the widespread use of DDT (Cooke, 1973; Peakall, 1975; Chapter 24). This benefit was, of course, fortuitous — the eggs were collected for other purposes. The potential value of preservation and methods for preserving and deciding what to preserve are treated in detail by Lewis, Stein, and Lewis (1984), who point out that appropriate specimens can provide records of trends of pollution over long periods. Moreover, specimens remain available as analytical techniques improve, so we can use new methods retrospectively. Similarly, as new chemicals become of concern, banked specimens can be examined for them, even though no attention was paid to them when the specimens were collected.

Banked specimens can also serve as records to determine the effectiveness of pollution control programs. Nonbiological materials, such as lake and marine sediments, also have some of these advantages and should therefore be included in a global banking scheme; but they cannot provide the full INDICATOR SPECIES AND BIOLOGICAL MONITORING 85
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. range of information obtainable from specimens of living organisms collected at specified times and places.

The results of monitoring are difficult to evaluate unless the organisms collected can be identified. Faulty identifications could lead to erroneous conclusions about environmental changes and their causes. For this reason, in spite of the cost, it would be useful for monitoring programs to be associated with regional taxonomic centers that could care for the collections and identify the materials. The lack of adequate taxonomic collections and trained persons able to care for them is a serious scientific problem in the United States, as elsewhere, and it adversely affects our ability to follow and interpret human- caused environmental changes.

MONITORING OF BIOLOGICAL RESOURCES
We have discussed the role of biological monitoring in the detection of pollutants and their effects. Biological monitoring is also needed to provide inventories of biological resources. Human modifications of habitats are profoundly influencing distributions and abundances of species. Species that thrive in disturbed habitats — such as croplands, pastures, early successional habitats, and urban environments — are increasing at the expense of species that require old-growth forests, riparian environments, wetlands, estuaries, and flowing waters. Indeed, many experts believe that current rates of deforestation in the tropics could cause the extinction of as many as a million species within the next half-century, many of them before they have even been named and described (Ehrlich and Ehrlich, 1981; Lovejoy, 1979; Myers, 1979, 1980). An important role of biological monitoring is to determine which species are increasing and which are decreasing in abundance, where losses of species are most serious, and hence where conservation efforts should be directed. Much of this information is compiled in the "Redbooks" on rare and endangered vertebrates and invertebrates published by the International Union for the Conservation of Nature.

Biological monitoring is especially valuable in helping to identify the effects of habitat fragmentation, a common form of alteration of terrestrial environments that results from human activity. Monitoring can help to determine the minimal sizes of patches required by species and how the rate of occupancy of suitable sites declines as distance between habitats increases. For example, the selective loss of neotropical migrant birds with decreasing size of forest patches in eastern North America has been documented by monitoring.

Birds are good organisms for biological monitoring, because there are relatively few species, they are nearly all easily recognized, and there are INDICATOR SPECIES AND BIOLOGICAL MONITORING 86
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. large numbers of amateurs who know them well and observe them regularly.

Christmas bird counts and breeding-bird censuses have been carried out for many decades and are standardized. They provide excellent long-term records of patterns of distribution and abundances. The northward spread of a number of species of nonmigratory birds as a result of extensive winter feeding of birds in eastern North America has been well documented by Christmas bird counts.

The striking spread of the house finch (Carpodacus mexicanus) in the northwestern United States during the last 20 years can be seen clearly in the breeding-bird censuses, including changes in abundances of species with which the house finch might compete (Mindinger and Hope, 1982). These changes probably indicate as yet unknown environmental changes.

MONITORING AND THE IDEA OF A PROJECT AS AN

EXPERIMENT
A major theme of this report is that many projects intended to produce something of value to human society take place on spatial and temporal scales much greater than can be duplicated in experiments designed purely for scientific purposes. Treating these projects as scientific experiments is a key component of effective environmental decision-making ( Chapter 10 ).

Monitoring can be a vital part of the use of such projects as scientific experiments. The scales of projects are often large, so monitoring often needs to be extensive. In addition, the results of monitoring are more valuable if they are reported and analyzed in forms readily available to scientists and managers. As in other experiments, specimens collected as part of such monitoring should be retained in repositories, where they will be available to future researchers.

INDICATOR SPECIES AND BIOLOGICAL MONITORING 87
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. 8 Dealing With Uncertainty
This chapter concerns the management of ecological systems in the face of uncertainty. Some uncertainty is unavoidable in ecologists' predictions about ecological systems, but decisions that might lead to unexpected environmental changes have to be made. To deal with uncertainty in ecological prediction, we must first identify its sources and consequences — this information is often the most useful that an ecologist can give to an environmental manager.

SOURCES OF UNCERTAINTY
Many ecological systems — among them those most affected by human activities— are poorly understood. We often lack the detailed information necessary for making accurate predictions, particularly under the unusual conditions that we impose on ecological systems. Several broad categories of uncertainty make the precise prediction of ecological changes difficult.

Complexity
Because the relationships among species are complex, changes in one can lead to unexpected changes in another. Indirect effects — those propagated through a number of species links — are particularly difficult to predict (Brown, in press). For example, when DDT was introduced into terrestrial environments, no one imagined that it would eventually be found in marine fish ( Chapter 24).

Equally hard to predict are the effects of
DEALING WITH UNCERTAINTY 88
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. multiple perturbations, such as the introduction of different kinds of pollutants from multiple sources into a river (Chapter 9).

The response of a population or community can become nonlinear at some degree of perturbation, perhaps with a threshold beyond which the response changes qualitatively. Fish populations can undergo rapid collapse when overharvested (Murphy, 1977; Ricker, 1963), and lake communities can undergo distinct changes in composition when nutrient concentrations change (Chapter 20). Continued reduction and fragmentation of habitat can also lead to the extinction of local populations and whole species ( Chapter 17 ). In some cases, nonlinearity of population dynamics can lead to chaotic behavior even in environments that are homogeneous in space and time (May, 1981; Chapter 2).

Natural Variability
Populations and communities vary in space and time because of both intrinsic processes and changes in their physical and biological environments (see also Chapter 5). In most species, reproduction is seasonal, and mortality is neither constant throughout the year nor equally distributed with respect to sex and age. An estimate of the makeup of a population or community at any particular time is like a snapshot — often one with poor resolution — of one state among many possible states. The responses of populations and communities are all too often influenced by their site. The presence or absence of a particular species, for example, can change a community's response to environmental change. Ecological systems commonly respond in a site-specific or situation- specific fashion, so our predictive capability is likely to be poor when appropriate analog studies for nearby or similar sites are unavailable.

Temporal variability in ecological systems can have several implications.

Some systems are highly variable because of periodic natural disturbances or intrinsic cyclic behavior. For example, the marine intertidal communities on the western coast of North America are subject to severe storms that can temporarily destroy them. Many organisms in such systems are adapted to those disturbances and have characteristics that facilitate recovery (Sousa, 1984).

Changes in the systems must be interpreted through an understanding of the natural history of the organisms present. Similarly, changes in plankton communities that undergo seasonal turnover must be interpreted in the light of population changes in natural conditions.

Changes in climate can also complicate assessment of human-caused environmental changes. For example, it is often unclear whether changes in fish populations are due to fishing pressure or to changes in ocean conditions (e.g., Chapter 12; Parrish and MacCall, 1978; Steele, 1984;
DEALING WITH UNCERTAINTY 89
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved.

Wooster, 1983). Environmental changes can interact with biological phenomena to produce large and complex changes in population numbers and perhaps alternative stable states (Steele and Henderson, 1984; Wooster, 1983). Changes in climate can be severe but brief, such as El Ni ños (e.g., Barber and Chavez, 1983), or they can take place over hundreds of years with large effects on the animals and plants distributed over wide areas (Bryson and Murray, 1977).

Natural variability makes it difficult to establish baseline conditions. The possibility that changes in populations or communities are due to natural changes in the abiotic environment should always be kept in mind in making predictions about ecological systems.

Random Variation
Because many forces acting on populations are more or less random, populations behave in probabilistic rather than deterministic ways, and their responses to perturbations can never be predicted precisely. Population numbers can fluctuate randomly and on a periodic or cyclic basis — seasonally or over longer or shorter periods. If the scale of observation does not match the scale of natural fluctuations, even cyclic (e.g., seasonal) fluctuations can be perceived as random noise that makes determination of the state of a system difficult.

Because variation is sometimes random and sometimes systematic, we need to know about the variability of important population or community measures — mean values are not enough — if we are to manage them adequately and predict the effects of changes.

Errors of Estimation
In addition to errors stemming from random variation, measurement error is unavoidable, and it can be large. For example, independent observers counting migrating animals can disagree substantially if the animals pass them rapidly and in great numbers. Estimates of population size must often be made indirectly, as when fish catch statistics are used to estimate stock size (e.g., Chapter 12 ). Whether estimation is direct or indirect, inevitable errors of estimation add to the uncertainty of management projections.

Errors of estimation commonly combine with natural variability to make the detection of ecological effects challenging. Sampling programs necessary to distinguish an effect from background noise become more elaborate, more time- consuming, and more expensive as the accuracy of estimation required and the extent of natural variability increase. In view
DEALING WITH UNCERTAINTY 90
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. of the practical difficulty of distinguishing among effects, it is surprising that we traditionally worry about concluding that an effect has occurred when it has not (type 1 error), but rarely worry about concluding that an effect has not occurred when it has (type 2 error) (Zar, 1976). If multiple type 2 errors are made for a given class of environmental change, serious cumulative ecological effects can occur (Chapter 9).

Lack of Knowledge
We often lack experience with a particular type of environmental perturbation. When neither appropriate studies nor reliable theories or models are available, accurate prediction is not possible, and it might be necessary to conduct extensive ecological studies as an aid to management decisions.

However, if even the best possible ecological studies are inadequate to support prediction of acceptable accuracy, a planned project or action must be viewed as an experiment in itself (Chapters 6 and 21).

MANAGING IN THE FACE OF UNCERTAINTY
The point of discussing the many obstacles to making accurate predictions is not to argue the futility of trying, but to show that the process of prediction must be viewed as complex and probabilistic. An appropriate approach to managing ecological systems recognizes the random component of population dynamics by dealing with measures of system variability, the possibility of nonlinearity, and the consequences of errors (e.g., Beddington 1984a,b; Walters, 1984). For example, attempting to achieve maximal sustainable yields in many fisheries can result in more variable yields, greater population fluctuations, and a greater likelihood of population collapse (May, 1980).

Regulating a fishery at higher stock size can stabilize yield and decrease the likelihood of catastrophic failure of the fishery. When populations are very variable and exceeding a critical threshold would have severe consequences, the best strategy could be to hedge one's bets — to reduce risk by trying not to exceed the threshold (Beddington, 1984a; Jewell and Holt, 1981).

When we have little understanding of either the major potential effects of a perturbation or the dynamics of the receiving system, our predictive capability is likely to be extremely poor. Developing a project or initiating a management plan can then be viewed as conducting a large-scale experiment. Information from well-designed monitoring studies can allow management to adapt to the occurrence of unexpected effects (Holling, 1978) and can be valuable in the planning of similar actions (Chapter 12).

DEALING WITH UNCERTAINTY 91
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved.

In some cases, particularly those involving the management of renewable biological resources, information can be gathered by managing experimentally — purposefully manipulating the system to learn about the effects (Bar-Shalom, 1976; Beddington, 1984a; Ludwig and Hilborn, 1983; Walters, 1984). Models can aid in conceptualizing a problem and in developing hypotheses that can be used to design research and monitoring studies (Holling, 1978). The use of multiple working hypotheses permits one to compare the potential effects of several possible perturbations.

A number of methods are available for dealing with randomness in ecological systems. Error analysis is a formal method of dealing with random effects when models are formulated explicitly (Meyer, 1975). Random error is inserted into a model to simulate variation, and the model is explored by Monte Carlo simulation or simple first-order analytical techniques (Walters, in press).

This method of modeling has been used to determine the minimal viable population for endangered species (Shaffer, 1981). Survival is a probabilistic affair, and one who would determine the minimal tolerable population size must decide for how many generations a species is to be protected and the acceptable probability of extinction. Such modeling efforts are usually limited to analyzing the effects of random error, but it is also possible to model systematic errors, in an attempt to identify potential causes of extinction.

Sensitivity analysis is a method for assigning different values to the parameters of a model to explore the consequences of errors in choosing and measuring them (Meyer, 1975). The analysis helps to identify the variables to which the system is most sensitive. It permits exploration of alternative management plans and can be used to decide which parameters need to be estimated most accurately.

The obstacles to making accurate quantitative predictions of the behavior of populations and communities are formidable. The natural variability and complexity of ecological systems will always limit our ability to make precise predictions. In the long run, a better understanding of natural variability and of the factors that make some systems more responsive to change than others will improve our predictive ability. But as long as we continue to alter our environment in new ways, uncertainty will always be associated with the effects of our manipulations. Environmental manipulations will always be experimental to some extent, and our most promising course is to structure each one so that we can learn as much as possible from it.

DEALING WITH UNCERTAINTY 92
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. 9 The Special Problem of Cumulative Effects

THE NATURE OF THE PROBLEM
The continuing degradation of ecological systems due to repeated perturbations is a difficult environmental and scientific problem. Because current science and policy focus primarily on the environmental effects of single projects or actions, little progress has been made in dealing with cumulative effects. Recent awareness of such serious cumulative effects as acid rain, the rapid loss of tropical forests, and the threatened extinction of many species has brought increasing political and scientific attention to cumulative environmental effects.

Many cumulative effects result from the movement of materials through the environment. An analysis of the propensity for natural processes to concentrate or disperse materials can help to determine where cumulative effects will be most severe. Several processes lead to the concentration of materials. The most important is movement of the medium itself, such as atmospheric mixing and stratification, water currents, and downward movement of water in soil. Other processes are relatively independent of the medium itself, as when chemical interactions cause flocculation and settling of suspended particles entering estuaries and when particles settle because of gravity.

Living organisms are responsible for considerable concentration of materials, and their patterns of movement and dispersal can result in increased exposure to environmental contaminants. Some organisms migrate great distances, and many species aggregate for feeding or
THE SPECIAL PROBLEM OF CUMULATIVE EFFECTS 93
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. breeding. Passage of material through food chains can lead to great concentration; top carnivores are exposed to much higher concentrations of contaminants than are plants at the bottom of the food chain (Chapter 24).

Processes that lead to dispersal of materials are sometimes similar to those concentrating them. Water currents, wind, the migration of organisms, and the dispersal of their propagules can all serve either to disperse or to concentrate materials. Dispersal can either increase or decrease the ecological effects of materials, depending on its rate in relation to the biological activity of the materials. Dispersal of highly toxic substances that remain active for a long time might simply increase the area of effects. An inadequate understanding of how such processes work in specific cases can lead to undesirable results.

Recent discoveries of areas of toxic waste buildup in Puget Sound, Washington, indicated that the processes of dilution originally counted on to disperse toxic materials and reduce them to harmless concentrations were not functioning as expected. The balance of processes that disperse and concentrate materials in different environments can result in the manifestation of effects over widely differing scales.

The effects of environmental perturbations accumulate when the frequency of individual perturbations is so high that one comes before the system has recovered from the previous one or when the ecological effects of perturbations in adjacent areas overlap. Thus, cumulative effects result from spatial and temporal crowding of environmental perturbations. The combined effects of repeated perturbations can be more severe than and sometimes qualitatively different from the sum of the effects of individual events.

Comparing the problem of cumulative effects with the ''tyranny of small decisions" described by the economist Alfred Kahn, Odum (1982) pointed out that, when numerous small decisions on related environmental issues are made more or less independently, the combined consequences of the decisions are not addressed; therefore, no provision is made for analyzing the patterns of the perturbations or their effects over large areas or long periods.

A step toward identifying the major issues involved in the assessment of cumulative impact was made in a recent international workshop held in Toronto, Ontario, under the joint sponsorship of the present committee and the Canadian Environmental Assessment Research Council. Some of this chapter draws on the proceedings of the Toronto workshop (Beanlands et al., in press).

THE SPECIAL PROBLEM OF CUMULATIVE EFFECTS 94
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved.

KINDS OF CUMULATIVE EFFECTS
Several types of perturbation can produce cumulative effects. Many involve the addition of materials to the environment, for example, the addition of sewage effluent to Lake Washington from multiple sources over many years (Chapter 20). In that case, a major cumulative effect was the rapid deterioration of water quality with attendant changes in species composition. The materials added can also be toxic (Chapter 24).

A second major type of cumulative effect involves the repeated removal of materials or organisms from the environment. In forestry, for example, both trees and minerals are removed (Geppert et al., 1984). The effect of harvest rate on population dynamics is seldom a linear function of population density, and the response of a population to an incremental increase in harvesting rate can vary substantially with conditions. Increasing the intensity of harvesting can lead to population collapse if a critical threshold is passed, especially if the increase in harvesting is combined with natural environmental changes (Beverton, 1983; Murphy, 1977; Parrish and MacCall, 1978; Peterman, 1978).

A third kind of cumulative effect can occur when management decisions result in environmental changes over large areas and long periods. In the New Brunswick forest (Chapter 19), forestry was based on single stands, instead of a large area of forest. The cumulative effect was a decrease in average yields with an increase in annual variability. Harvesting practices that seem appropriate from the standpoint of population dynamics often result in cumulative genetic effects, such as loss of genetic variability or inadvertent selection for undesired traits ( Chapter 1 ). A classic example of the cumulative genetic effects of repeated interventions is the evolution of pesticide resistance (Chapter 1).

A fourth, more complicated situation arises when stresses of different types combine to produce a single effect or suite of effects. For example, the construction of too many wells, developments, and drainage canals lowered the water table in the Florida Everglades substantially. Similarly, the construction of roads in national forests can trigger logging, recreational activities, and poaching, which might combine with acid rain and other influences to threaten sensitive populations of forest species.

Complex cumulative effects also occur when many individual areas in a region are repeatedly altered. The result can be dramatic changes in the mix, arrangement, and internal characteristics of the habitats of species. Large habitats can become fragmented into patches separated by areas that are inhospitable to many organisms. As habitat patches become smaller and more isolated, species that depend on them become less able to find them and to maintain populations in them (Chapter 17). The conservation THE SPECIAL PROBLEM OF CUMULATIVE EFFECTS 95
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. of species in the face of severe habitat fragmentation could be our greatest terrestrial environmental problem.

Cumulative effects need not be adverse, but might be considered by many as improvements in the environment. Sometimes, repeated attempts to "improve" the environment have the greatest cumulative effects of all. The conversion of deserts to arable land by massive irrigation projects destroys desert habitats, but creates economic benefits and new habitats that favor species adapted to exploit irrigated farmland. Even the attempt to keep some environments in a constant state through repeated interventions can have unexpected consequences. Fire is often controlled to protect habitats, but some communities require periodic fires to maintain their typical species composition (Kozlowski and Ahlgren, 1974). Moreover, reducing the frequency of fires can increase the severity of fires that do occur by fostering the accumulation of flammable material.

DEFINITION OF CUMULATIVE ENVIRONMENTAL EFFECTS
Time-crowded perturbations. Cumulative effects can occur because perturbations are so close in time that the effects of one are not dissipated before the next one occurs. An example is repeated harvesting of agricultural crops or forests that removes some nutrients faster than they are regenerated between harvests (Geppert et al., 1984; Krebs, 1985, Ch. 28). Similarly, the evolution of resistance to pesticides occurs because the susceptible genotypes are repeatedly reduced in number each time the pesticide is applied (Georghiou et al., 1983; May and Dobson, in press).
Space-crowded perturbations. Cumulative effects can occur when perturbations are so close in space that their effects overlap. An example is power plants close enough that the heat plumes of their cooling water overlap (e.g., Slawson and Marcy, 1976).
Synergisms. Different types of perturbations occurring in the same area can interact to produce qualitatively and quantitatively different responses by the receiving ecological communities. For example, several pollutants might interact to produce toxic mixtures (e.g., National Research Council, 1982, 1983b; for examples of mixtures toxic to humans, see Reif, 1984).
Indirect effects. Cumulative effects can be produced after or away from the initial perturbation or by a complex pathway. For example, when the level of Southern Indian Lake in Manitoba was raised, the inundation of the lake shorelines resulted in the release of mercury into the lake (Bodaly et al., 1984) and increased the turbidity of the water (Hecky, 1984). Neither of these consequences was predicted by knowledgeable limnologists (Hecky et al., 1984).

THE SPECIAL PROBLEM OF CUMULATIVE EFFECTS 96
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. Nibbling. Incremental and decremental effects are often involved in each of the above categories, but not always. However, this "nibbling" of environments is so important that it should be given its own category. The numerous examples include time and space crowding (the addition of power plants to a river one at a time or the introduction of several pollutant sources into a lake), as well as removal of habitat piece by piece, as in the degradation of Chesapeake Bay (Flemer et al., 1983).
Other types of impact have sometimes been equated with cumulative effects, such as threshold developments that stimulate additional activity in a region (e.g., new energy developments in northern Canada) or projects whose environmental effects are delayed (time lags) or are felt over large distances (space lags). These types of effects can be cumulative if their impacts overlap in time or space or are synergistic with those of other developments.

DIFFICULTIES IN PREDICTING AND CONTROLLING

CUMULATIVE EFFECTS
A fundamental problem in predicting and controlling cumulative effects is the frequently large mismatch between the scales or jurisdictional boundaries of management authority and the scales of the ecological phenomena involved or their effects. Affected environments — such as river basins, airsheds, and estuaries— usually cross local and state, and sometimes even national, boundaries. Conversely, an environmental insult that originates in several jurisdictions might exert its effects largely or wholly in one area. The history of the Delaware River Basin Commission project is a striking example of the complexity of managing such a situation scientifically and equitably (Ackerman et al., 1974).

As difficult as management of cumulative effects can be for regional authorities, management by local authorities is even more so. Local authorities are ill equipped to deal with regional trends in environmental deterioration.

They often lack the legislated responsibility, motivation, expertise, and resources to perform the necessary analyses on an appropriate scale. When effects appear far from their sources, the local jurisdictions with the sources might not even become involved.

A major impediment to the control of cumulative effects is the perception that the effects are minor. How often are projects stopped because they appear to be contributing (however slightly) to a trend in environmental deterioration? What happens if scientists recognize a trend, but cannot detect or distinguish the effects of individual actions? The problem of multiple "insignificant" effects is at the core of our failure to deal with the continued nibbling of coastal and other habitats. It is extremely difficult
THE SPECIAL PROBLEM OF CUMULATIVE EFFECTS 97
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. to predict the point of attrition that constitutes a critical threshold. Below what population size will a species become unable to recover ( Chapter 14)? Beyond what degree of fragmentation will an ecosystem begin to lose species rapidly or change markedly in its functioning? At what concentration will an environmental contaminant cause qualitative changes in community organization (Chapter 20)?
Predicting the effects of perturbations is particularly difficult when many sources of different types act in concert. Accurate information on the nature of each source is difficult to obtain, and their interactions can be complex. The current controversy over the sources of acid rain and its effects in particular areas is a case in point (National Research Council, 1983a, in press; U.S.

Environmental Protection Agency, 1980). The "proof" required in this socially and politically charged case is prohibitively expensive to acquire. Control of eutrophication in lakes has been the greatest success in solving a multiple- source problem, possibly because of the limited areas involved. When effects are global, as with the long-term buildup of atmospheric CO 2, assigning
causality and regulating the sources are extremely complex, both scientifically and politically (National Research Council, 1983c).

It is clear that continued loss of communities and ecosystems will, in the long term, constitute a "cost" to society, but there is little agreement on how to express this cost. We lack an accepted framework for determining how the cost of habitat deterioration, with resulting fragmentation, is apportioned among nonresponsible parties, because the observable effects of each incremental loss are often small and local. Assigning a value to an increment of loss can be very difficult, but at some point these small and primarily local effects might be expressed in different forms on a regional scale. Knowing the nature and values of ecological thresholds would give us a benchmark for associating costs with changes, but would not solve the problem of valuing changes that do not threaten to exceed a threshold. Research that will help us to identify the threshold of rapid, observable deterioration is badly needed.

SCALE AND THE RATES OF CRITICAL PROCESSES
Choosing appropriate scales on which to analyze cumulative effects requires an understanding of the relative balance of important concentrating and dispersing phenomena and of the rates of other processes that influence the removal of or recovery from individual stresses in the environment being studied (see also Chapter 5). Because the rate of atmospheric mixing is so high and because the atmosphere is relatively uncompartmentalized, many materials that have low rates of removal from
THE SPECIAL PROBLEM OF CUMULATIVE EFFECTS 98
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. the atmosphere (by decomposition, chemical transformation, settling, etc.) are spread on a global scale. In the atmosphere, dispersing processes typically predominate over concentrating processes, with notable exceptions, as when inversions and scavenging of particles by precipitation lead to local concentration. Atmospheric pollution has been modeled largely on a regional and global scale appropriate to the scale of effects.

Mixing in the oceans is much more compartmentalized and less complete.

Concentrating and dispersing processes are more evenly balanced, and the dominating process varies with local conditions. In many freshwater systems, such as lakes, concentrating processes predominate. Individual lakes often behave independently of one another, so the appropriate scale for analyzing cumulative effects is likely to be an individual lake and its associated watershed.

Terrestrial environments are particularly complex spatially and temporally.

Spatial heterogeneity is marked, and the existence of many persistent and relatively impermeable structural components results in a high degree of compartmentalization. Because the basic substrate is solid, movement of materials is restricted, and many impacts remain local. Processes that affect the concentration or dispersal of materials are complex in terrestrial systems.

Chemical reactions in the soil result in leaching of materials through the substrate and into groundwater or, if precipitation is less, in their concentration in different layers of the soil.

Spatial scales in terrestrial systems can be particularly important when communities or habitats are increasingly fragmented by nibbling and alteration.

The sizes of habitat patches and the distance between them can profoundly influence the survival of species that depend on them. The probability of local extinction increases as population size decreases, and the chance that a given habitat patch will be recolonized depends on the dispersibility of the species, the proximity of sources of potential colonizers, and the ability of a population to be established by a small number of colonizing individuals (Chapters 1 and 5). Seminatural plant communities on derelict lands in Great Britain often fail to re-establish themselves naturally, because the sources of seeds of desired plants in undisturbed habitats are too widely scattered ( Chapter 18 ). Adjacent areas provide mostly seeds of introduced weedy species. To maintain natural communities, managers must artificially increase the supply of propagules.

The common approach of treating cumulative effects in distinct environmental compartments is artificial and often inadequate. Atmospheric phenomena, such as wind and precipitation, are intimately involved in the transport of materials into and out of terrestrial systems. Contaminants in terrestrial systems eventually find their way into aquatic systems, and THE SPECIAL PROBLEM OF CUMULATIVE EFFECTS 99
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. contaminants in aquatic systems often find their way into terrestrial organisms through food chains. The DDT case study ( Chapter 24) is a classic example of the complex interchanges that occur between environments. DDT sprayed on crops drifts on the wind and is transported by runoff into aquatic systems, where it is biologically concentrated through food chains. Eggshell thinning problems have followed in terrestrial raptors that feed on contaminated fish or in other birds that eat contaminated aquatic organisms.

Whether an ecological system is adversely affected when subjected to repeated perturbations depends not only on the rate and intensity of perturbations, but also on the rate at which their effects are absorbed by or removed from the system. If toxic chemicals are rapidly degraded biologically to a harmless form, a receiving system will be capable of absorbing a relatively high rate of toxicant input. Ecological systems that are characterized by very low rates of growth and reproduction, such as the arctic tundra, are extremely vulnerable to repeated disturbance.

MANAGING CUMULATIVE EFFECTS: BEYOND A CASE-BY-

CASE APPROACH
Improvement in our efforts to predict and to limit or reduce cumulative effects will require a restructuring of both scientific and institutional approaches and a considerable improvement in the interchange of information between scientists and managers. If one accepted the dictum that "everything in ecology is connected to everything else," the already formidable challenge of dealing with cumulative effects would become overwhelming. Fortunately for both scientists and managers, connections are not equally strong, and a first step in trying to predict cumulative effects is to determine which interactions are most important (to reduce the problem to as simple a set of relationships as possible) and which sources of perturbation are likely to affect the ecosystem components of concern (Chapter 10).

As part of the Sustainable Development of the Biosphere project being carried out at the International Institute for Applied Systems Analysis in Austria, scientists have been developing a promising matrix-based framework for scoping potential cumulative-effects problems in the atmosphere (Clark, in press). The approach is designed to identify the many kinds of perturbations that affect particular valued atmospheric components — and the multiple effects of particular perturbations — through an analysis of the mechanisms of interaction that underlie direct and indirect pathways of cause and effect. This approach can be applied to environments other than the atmosphere.

THE SPECIAL PROBLEM OF CUMULATIVE EFFECTS 100
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved.

By indicating the relative strengths of the effects of all potential perturbations of a system, the framework can help an environmental manager to choose control variables that offer the most promise and to avoid attempts to control a system by manipulating variables with little potential influence.

Because it also identifies the multiple effects of a given perturbation, the framework might also help to focus attention on actions that have the greatest general impact on the environment.

When major effects and their pathways have been identified, the primary tasks of scientists are to track the state variables of ecological systems, to identify trends in environmental quality, and to predict critical ecological thresholds. Considerable attention has been given to tracking environmental trends ( Chapter 7 ), and some refreshingly novel low-cost approaches to biological monitoring have been developed (Bromenshenk et al., 1985). But methods for predicting critical thresholds are sorely lacking. Our understanding of the factors involved in determining population thresholds has improved (Chapters 2, 5, and 17), but much more basic research is needed.

Dealing more adequately with cumulative effects will require changes in institutional approaches of both scientists and managers of cumulative effects.

Assessment involves many scientific disciplines and requires both interdisciplinary and disciplinary scientists who are comfortable with and adept at working with experts in a variety of fields. Unfortunately, the reward systems for both faculty and graduate students at most universities discourage such activities.

The Toronto workshop mentioned above brought together representatives of applied and basic science, environmental management, business, and resource agencies. The most critical immediate need in management practices identified in the workshop was the need to improve the match in scales between the ecological phenomena involved in cumulative effects and the management jurisdiction. A good start in this direction would be to improve communication between scientists and managers, and the nomogram used by Erdle and Baskerville (Chapter 19) to indicate the long-term benefits and costs of a variety of forestry options is a useful tool for this purpose.

Comprehensive generic reviews of the scientific issues germane to particular cumulative-effects problems, including a discussion of scale, would be valuable to a manager. Local decision-makers, who often lack access to the expertise and resources necessary to deal with cumulative effects, also need syntheses of regional ecological knowledge. For example, Cooper and Zedler (1980) have proposed a system for land-use planning that allows development projects to be analyzed on the basis of regional ecological knowledge of ecosystem sensitivity to stress.

THE SPECIAL PROBLEM OF CUMULATIVE EFFECTS 101
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved.

The Toronto workshop made it clear that the project-specific nature of environmental impact assessment is inadequate for the management of cumulative effects. The management of cumulative effects must in some way match the scale of the ecological phenomena involved.

When important cumulative effects have been identified in aquatic environments, regional management authority has often been proposed or implemented. When valued ecosystem components have been readily identified and have been recognized and generally agreed on by the public, this approach has sometimes been successful. For example, the formation of a metropolitan regulatory authority to control sewage pollution in Lake Washington was very effective, because the public readily comprehended the issues and agreed on the value of a clean lake ( Chapter 20 ). The effort to establish regional control of multiple sources of pollution in the Delaware River Basin was much more complex (Ackerman et al., 1974): many issues of conflict were involved, there was less public agreement on what was most valued, the scientific questions were more difficult, the regional authority was far less successful, and attempts to expand the scope of the metropolitan authority (like that formed to protect Lake Washington) to deal with other and less clear issues, such as transportation, failed to win public support.

Regional authority might need some ground swell of public support if it is to succeed. In a number of recent cases, ad hoc nongovernment organizations, such as "save the bay" groups, have helped greatly in bringing particular cumulative-effects issues to public attention and have created and supported the public debate that eventually provided the impetus and pressure for government agencies to act. Perhaps more important, such groups can serve as intermediaries between scientists and regulators, thereby fostering exchange of information and ideas on controversial topics.

RECOMMENDATIONS
Some of the recommendations produced at the Toronto workshop for improving the scientific and institutional approaches to managing cumulative effects follow; they are adapted from the proceedings of that workshop (Beanlands et al., in press). Some of them are addressed to scientific research, and some to improving institutional handling of cumulative effects. Cumulative environmental effects should be placed in readily seen time and space scales, as described for atmospheric effects (Clark, in press). This should help to identify different susceptibilities of different THE SPECIAL PROBLEM OF CUMULATIVE EFFECTS 102
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. environments and ecosystems to cumulative effects caused by many kinds of perturbations.
A better match between scales of management and scales of environmental effects is needed. To this end, cases in which cumulative environmental problems have been dealt with — both successfully and unsuccessfully — should be reviewed to understand how the setting of management and ecological boundaries influenced their success or lack of it.
Research should be conducted to determine what rates of addition of materials to environments and harvesting of resources from them are consistent with protection of valued components of various environmental systems, as well as the systems themselves. Studies of response rates and recovery times should be included in this research. Research should be conducted to determine the types of indicators, thresholds, and environments most likely to be useful for assessing and managing different kinds of cumulative effects.
Monitoring should be built into the design of projects, and cumulative effects taken into account where appropriate. This requires an understanding of the appropriate boundaries. In addition, monitoring should be frequent enough and carried out for long enough to detect cumulative effects.
Agreements between decision-makers, managers, and scientists with respect to the appropriate time and space boundaries for dealing with cumulative effects should be documented. This will force clear thinking about important issues and will provide a record so that procedures can be improved.

THE SPECIAL PROBLEM OF CUMULATIVE EFFECTS 103
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. 10 A Scientific Framework For
Environmental Problem-Solving
To manage the effects of environmental manipulations, we must be able to predict them. However, knowledge is seldom sufficient to allow accurate prediction, so studies are necessary to provide the information needed to make decisions. Such studies must be carefully planned, because they are expensive in time, money, and effort. This chapter presents a general framework for identifying, scoping, and planning studies of environmental problems. The framework is, in essence, an admonition to think before acting and to use established scientific principles. Table 1 makes it clear that deficiencies in environmental impact assessment are due not only to scientific difficulties — the ones with which this chapter is primarily concerned — but also to political, administrative, and economic difficulties.

Despite their bewildering variety, environmental problems share some basic features, including actions that result in environmental changes, public and scientific concern about those changes, a need for methods for predicting environmental responses to human actions, and limited resources for the acquisition and analysis of relevant ecological information. We draw heavily on a number of recent efforts to make environmental assessment and management scientifically more credible (Andrews et al. 1977; Anonymous, 1980; Council on Environmental Quality, 1978; Fritz et al., 1980; Holling, 1978; Larkin, 1984; Rosenberg et al., 1981; Sanders et al., 1980; Sharma et al., 1976; States et al., 1978; Walters, in press; Ward, 1978) and in particular on a recent Canadian review (Beanlands and Duinker, 1983).

A SCIENTIFIC FRAMEWORK FOR ENVIRONMENTAL PROBLEM-SOLVING 104 Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved.

TABLE 1 Some Recent Criticisms of Ecological Impact Assessment, Based Primarily on Beanlands and Duinker (1983), Carpenter (1976), Rosenberg et al. (1981), and Skutch and Flowerdeu (1976)
Guidelines too elaborate and requirements too diverse
Time and money constraints not recognized
Unreasonable expectations of decision-makers
Tendency to start gathering baseline data immediately, at the expense of careful planning
Failure to formulate objectives clearly and to develop a study strategy Unwarranted belief that ecological principles used in managed systems are as appropriate to unmanaged systems
Failure to recognize the value of early input from those who might later be involved in review, leading to an adversarial process
Failure to define project boundaries
Failure to consider cumulative effects
Failure to state the bases of value judgments
Lack of scientific standards for impact assessment
Lack of respect in academe for impact assessment
Vague and unverifiable predictions
Lack of a rigorous, quantitative approach, especially in monitoring Lack of continuity in studies conducted during planning, developmental, and operational phases of a project
Failure to follow actions with adequate monitoring studies Use of impact assessment for disclosure, rather than for learning Failure to recognize the scientific value of experimentation and monitoring Failure to consider the recovery potential of species and ecosystems Poorly written reports in which major points are buried in enormous amounts of information
Inordinate expenditure of effort on descriptive studies with little potential for predictive value
Inaccessibility of reports and results of studies, making them difficult to evaluate and learn from

DEFINING ENVIRONMENTAL GOALS AND SCIENTIFIC

QUESTIONS
In spite of the difficulties and controversies associated with identifying environmental goals, a clear statement of goals early on can help to focus research and can increase the chance of protecting components of the environment likely to be identified as valuable to society. The first step in defining such goals is to identify the components of the environment perceived as valuable, such as salmon in rivers of the northwestern and northeastern United States, a ''natural-looking" community of plants on reclaimed land (Chapter 18 ), clean water ( Chapter 20 ), clean air, forest productivity (Chapter 19), and fishery productivity (Chapter 12).

The second step is to determine the desired degree of protection, exploitation, or control. This decision usually involves choosing a state in which to maintain the ecological system in question and a length of time A SCIENTIFIC FRAMEWORK FOR ENVIRONMENTAL PROBLEM-SOLVING 105 Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. for which to maintain it. For example, we might wish to reduce the population of a pest or the amount of damage it causes, increase the yield of a harvested species, or maintain the species composition of a valued habitat. The period of management might be months, decades, or centuries.

The third step is to learn the environmental and financial costs of managing the system. Maintaining an ecological system in other than the "natural" condition usually requires some expenditure and might produce unwanted side effects. Harvesting a population for maximal yield can increase the variability of the yield and the likelihood of overharvesting ( Chapter 1 ).

Increasing production of an agricultural crop or forest often involves the use of hazardous pesticides ( Chapter 24 ), which can have several deleterious cumulative side effects (Chapters 1, 3, and 4).

Identifying environmental goals is complex and requires input from the public and from scientists. The public is concerned primarily with the choice of environmental goals. Scientists can help to identify Overharvesting goals and can indicate the environmental and economic costs involved. Scientists are also needed to translate environmental goals into scientific objectives, which show what information is needed to answer the major questions and hence help in the planning of studies. As in any research plan, scientific objectives are based not only on the need for particular information, but also on how easily that information can be obtained.

The issues on which environmental goals are based are specified in part by law and in part by public concern (see Table 2 ). The National Environmental Policy Act requires early public and professional input in identifying those issues (Council on Environmental Quality, 1978). The goal of protecting the Southern Indian Lake whitefish fishery was economically motivated (Chapter 21 ). Provincial wildlife biologists recognized caribou migration as a major public concern in the Newfoundland hydroelectric development case (Chapter 16). In the case of Lake Washington ( Chapter 20), interested scientists and the public cooperated to define goals and develop an appropriate response; the DDT case ( Chapter 24 ) shows how such interactions can lead to new understanding and to legislation.

Attempts to achieve a goal sometimes have unexpected results. The New Brunswick forest case study shows how attempting to maximize forest timber production on the basis of individual stands might not only fail to maximize yield over the whole forest, but fail to provide consistency in yields over a long period. In fisheries, managing for maximal sustainable yield often produces large variations in both yields and stock abundance (May 1980), making overexploitation and population collapse more likely
A SCIENTIFIC FRAMEWORK FOR ENVIRONMENTAL PROBLEM-SOLVING 106 Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. than more conservative management would (e.g., Murphy, 1977). Harvesting or managing populations over long periods can also produce undesired cumulative genetic changes (Chapter 1).

TABLE 2 Some Common Criteria for Identifying Important Issues and Valued Ecosystem Components in Impact Assessment
Legal requirements
Air and water quality standards
Public health
Rare, threatened, and endangered species
Protected areas or habitats
Aesthetic values
Landscape appeal
Attractive communities
Appealing species (e.g., large ungulates, colorful birds, cacti) Species at higher trophic levels (e.g., eagles and tigers) Clear air and water
Economic concerns
Species or habitats of recreational or commercial interest Ecosystem components
Environmental values and concerns
Ecosystem rarity or uniqueness
Sensitivity of species or ecosystems to stress
Ecosystem "naturalness"
Genetic resources
Ecosystem services
Recovery potential of ecosystems
"Keystone" species

SCOPING THE PROBLEM
Scoping involves bringing together all interested parties — public, business, government, and scientific — so that they can interact and express their views before major actions or studies are initiated (Council on Environmental Quality, 1978). Early scoping can help to identify the important issues and potential environmental effects associated with planned actions. It can help to define scientific objectives and guide the design of ecological studies. Scoping can also be useful during a program or project to ensure that the most important issues are being addressed, that studies are producing useful results, and that important new issues are noted (Fritz et al., 1980; Sanders et al., 1980).

Once the valued ecosystem components, significant issues, and major potential effects have been identified, ecologists can establish scientific objectives. When prediction of environmental effects is a major purpose A SCIENTIFIC FRAMEWORK FOR ENVIRONMENTAL PROBLEM-SOLVING 107 Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. of studies, four questions are usefully posed (Beanlands and Duinker, 1983):
Are valued ecosystem components expected to be affected either directly or indirectly by the project or action?
Can the valued ecosystem components be studied directly? (In the absence of adequate guidance from experience or the literature, pilot investigations might be needed to indicate the feasibility of particular studies.)
Is it possible to study valued ecosystem components indirectly? For example, because large carnivores are often difficult to study directly, the effects of an action on their prey base or their habitat could be studied instead. Studies of indirect effects are most appropriate when they are reliably associated with effects on the valued ecosystem components.
Would the use of indicators of impact be helpful? (See Chapter 7.) Formulating a conceptual model of the relationships between the proposed action and the receiving environment can help to identify pertinent questions and potential environmental effects. The purpose of such models is to identify the physical and biological pathways by which an action can produce ecological effects. By focusing on relationships important to the manifestation of effects, they help to develop specific, testable hypotheses to explain why particular effects should or should not occur. Conceptual models can also help to identify logical errors, to highlight factors that require special study, to synthesize ideas and knowledge, and to communicate information (Beanlands and Duinker, 1983); guidance for developing conceptual models can be found in Holling (1978), Ward (1978), Fritz et al. (1980), and Beanlands and Duinker (1983).

Multidisciplinary workshops can be used to articulate a problem and plan studies (Holling, 1978) and have been used to advantage in this way (ESSA, 1982).

Given adequate time and resources, sophisticated modeling should be considered (Munn, 1979). Basic guidance in development and use of such models can be found in Frenkiel and Goodall (1978), Holling (1978), and Ward (1978). Quantitative modeling can help by forcing assumptions to be made explicit, by making their consequences clear, and by revealing the sensitivity of outcomes to details of various assumptions. Simulation models are necessarily based on numerous unverified assumptions and cannot predict quantitative changes very accurately (Hilborn, 1979; Walters, 1975). But they can be useful in identifying potential qualitative effects and exploring the consequences of alternative management plans. Sensitivity analysis allows an exploration of the consequences of altering
A SCIENTIFIC FRAMEWORK FOR ENVIRONMENTAL PROBLEM-SOLVING 108 Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. the assumptions of a model. Simulation models are often used in connection with freshwater systems ( Chapter 21 ), in which the driving physical and chemical processes are fairly well understood; fishery and wildlife management (Chapter 12); epidemiology (Chapter 15); and forest management (Chapter 19).

ESTABLISHING STUDY BOUNDARIES
One of the first and most important tasks in the design of research on environmental effects is to establish a set of boundaries — temporal, spatial, and ecological. When might effects appear, and how long might they last? How long must studies last to allow reasonable predictions and reliable diagnosis of effects? Over what area will effects occur? Are there any natural barriers to the transmission of effects? Are any physical or biological processes likely to spread effects to other areas? What ecosystem components will be affected? At what levels of biological organization will effects appear? What species or ecosystem processes need to be studied and over what area? Boundaries in open systems, such as the ocean or atmosphere, are the most difficult to define.

Variations in ecosystem components of interest can strongly influence the time required for biological effects to appear (Holling, 1973).

Making judgments about boundaries is difficult, and many surprises have occurred. For example, large water impoundments can influence local climate or induce earthquakes (Baxter and Glaude, 1980). When DDT was first used as a pesticide, no one expected it to appear in animals in the ocean ( Chapter 24).

The DDT story and similar cases (e.g., that of acid rain) have shown that environmental effects can spread by subtle pathways. Assumptions implicit in management decisions might result in setting boundaries that omit critical processes. Attempts to increase stocks of anadromous fish by increasing reproduction in rivers might fail if survival in the ocean is already limited by food supply (Peterman, 1984). The cumulative effects of multiple actions have taught us that specific projects and actions must be viewed in the context of related actions (Odum, 1982). Recent efforts to protect and conserve species have shown how management of a population requires consideration of its relationship with other populations (Frankel and Soul é, 1981; Schonewald-Cox, 1983; Soulé and Wilcox, 1980). And only recently has it been recognized that systematic management procedures — e.g., sex-biased or size-biased harvesting — can lead to undesirable genetic changes over remarkably short periods (Chapter 1).

A SCIENTIFIC FRAMEWORK FOR ENVIRONMENTAL PROBLEM-SOLVING 109 Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved.

The establishment of boundaries is constrained by administrative, project- related, technical, and ecological factors (Beanlands and Duinker, 1983).

Administrative constraints include jurisdictional limits, insufficient time or funding, and political factors. Spatial boundaries are often obvious, unless long- range transport phenomena are involved. Temporal boundaries might be less well defined, because of political and other uncertainties; as we go from short- term, local effects at the population level to long-term, regional effects at the ecosystem level, we are less able to predict them (Christensen et al., 1976).

Other technical constraints are imposed by environmental variability, project location, and logistical problems.

Setting appropriate temporal and spatial boundaries is important in the management of species populations, whether for protection, control, or harvest.

When populations become small, patterns and rates of interchange of individuals and genes between populations become critical. The sizes of populations needed for the long-term maintenance of the spotted owl in the Pacific Northwest depend on whether the Columbia River is a dispersal barrier (Chapter 17 ). When timber is managed on a forest-wide basis, rather than by stands, yields are higher and more consistent ( Chapter 19 ). Physical and chemical processes can be critical in defining boundaries in aquatic systems, particularly when the spread and accumulation of pollutants are involved. The control of eutrophication in Lake Washington depended on knowledge of flow rates through the lake and the low turnover of phosphate in lake sediments (Chapter 20).

DEVELOPING AND TESTING HYPOTHESES
Statements about relationships between proposed actions and ecosystem components or processes are, in effect, hypotheses that can be tested. Studies designed to test them can increase our ability to predict environmental effects.

In addition, the explicit statement of hypotheses helps us to identify important assumptions and formulate specific objectives for ecological studies. However, despite the acknowledged value of testing hypotheses in solving environmental problems, many studies are not designed and conducted to do so. Many studies in wildlife management, for example, involve elaborate collection of field data with only after-the-fact attempts at explanation (Romesburg, 1981). What happened as a result of a project is rarely studied (Beanlands and Duinker, 1983; Larkin, 1984).

In practice, most general hypotheses are evaluated by testing specific predictions that arose from them. In environmental impact assessment, the predictions themselves are a major product of preproject research. It is often helpful to develop several hypotheses about possible effects and their causes, so that studies can be designed to distinguish among them.

A SCIENTIFIC FRAMEWORK FOR ENVIRONMENTAL PROBLEM-SOLVING 110 Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved.

A hypothesis can be tested by studies before a project and by treating the project itself as a test. Methods for testing ecological hypotheses in preproject studies include the use of microcosms (Crow and Taub, 1979; Heath, 1979), field and laboratory experiments (Giddings, 1980; Suter, 1982; Ward, 1978), and computer simulations (Frenkiel and Goodall, 1978). Despite the difficulty of assigning causality in field experiments (Sharp et al., 1979) and of extrapolating from small studies to large problems (Hilborn and Walters, 1981), pilot-scale perturbation studies could be the most productive research tool for impact assessment, although underused (Beanlands and Duinker, 1983; Ward, 1978).

If projects are to be treated as large-scale experiments, baseline data must be collected before the project begins (Beanlands and Duinker, 1983; Hilborn and Walters, 1981; Larkin, 1984). The baseline can best be viewed as a description of the mean values and natural variability in the system (Hirsch, 1980). Judgments of how much information is needed are often difficult, because of periodic cycles, random events, and spatial heterogeneity and because many variables can change systematically during the baseline study period (e.g., owing to succession).

Statistical guidance is available for designing baseline and monitoring programs once the variables of interest have been identified (Cowell, 1978; Eberhardt, 1976, 1978; Green, 1979; Kumar, 1980; Lucas, 1976; Sharp et al., 1979; Ward, 1978; Zar, 1976). Two common problems that make it difficult to design projects as experiments properly are the lack of adequate controls (Cowell, 1978) and the lack of true replicates (Eberhardt, 1976). Eberhardt (1976) suggested a "pseudodesign" with baseline data on a control area and the project site. They can be compared with data collected when the project is complete, with replicates in time substituting for spatial replicates.

Baseline and monitoring studies are most effective if they are statistically designed to detect changes of the magnitude expected (Zar, 1976). This expectation in turn determines the extent of sampling required (Hartzbank and McCusker, 1979). In highly variable systems, adequate sampling might be too expensive, and resources might be better used in carrying out less direct studies.

Baseline information can sometimes be derived after impacts have already occurred (e.g., Cowell and Syratt, 1979).

The Lake Washington case ( Chapter 20) is an excellent example of testing hypotheses concerning the effect of lake fertilization changes on the makeup of plankton communities. A great deal was learned from this case, because monitoring continued throughout the development and treatment of the problem. Similarly, scientists at Southern Indian Lake ( Chapter 21) were able to test hypotheses derived from the results of other lake studies and current limnological models. Carefully designed studies before
A SCIENTIFIC FRAMEWORK FOR ENVIRONMENTAL PROBLEM-SOLVING 111 Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. and after project development showed that some of the predictions were wrong because the models used were not based on knowledge of lakes in a permafrost zone.

Field and laboratory experiments can be used to test hypotheses. The Garki malaria project (Chapter 15) was itself a large-scale experiment to investigate a model for controlling malaria through a combination of drugs and mosquito control. Careful monitoring studies before, during, and after applications allowed the model to be tested, and an important phenomenon — exophily— was discovered.

SPECIFYING PREDICTIONS AND DETERMINING THE

SIGNIFICANCE OF EFFECTS
A major purpose of developing general ecological hypotheses is to generate a set of specific predictions of ecological change that can be used in decision-making. The predictions should be as clearly and precisely stated as possible. The period over which a change is expected to occur, the bases of the prediction, and the degree of uncertainty should be specified.

Determining the significance of predicted or observed ecological changes is often very difficult, because ecological systems are not fully understood. A clear distinction, if it can be made, between the magnitude of a change and its biological importance is useful. The rates of change and recovery are often important components of ecological effects (e.g., Cairns and Dickson, 1980).

The overall significance of an effect is tied closely to the definition of environmental goals. The best course for scientists is to predict or describe changes precisely. Whether or not a change is "significant" is a judgment that transcends science and is best made by all interested parties.

Several of the cases studied were organized around tests of specific predictions. In the derelict lands restoration case ( Chapter 18), predictions were derived from empirical results of other restoration efforts and basic plant ecological theory. The bases of these predictions were clearly stated, and tests of them produced results of value to other restoration projects. Predictions in the Atomic Energy Commission radiation studies were derived from knowledge of food-chain dynamics and laboratory studies, and hypotheses were continually revised as predictions were tested experimentally (Chapter 22).

In the Lake Washington case, scientists predicted not only specific changes in water quality, but also the periods over which deterioration and recovery would occur ( Chapter 20 ). In the Southern Indian Lake studies ( Chapter 21 ), predictions were based on analogs and limnological
A SCIENTIFIC FRAMEWORK FOR ENVIRONMENTAL PROBLEM-SOLVING 112 Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. principles, but were mostly qualitative. In both cases, careful monitoring was incorporated into comprehensive postproject analyses to improve understanding.

MONITORING
Biological monitoring is used in ecological studies in two basic ways.

Studies conducted during or after an action or project are designed to learn what ecological changes resulted. Anticipatory monitoring is designed to track the effects of activities that might be cumulative or pose hazards to human health (Baker, 1976). Properly done, monitoring provides continuous indexes of environmental quality that can signal environmental degradation or improvement (Chapter 7).

In the event of unexpected environmental changes, monitoring can facilitate adaptive changes in management and in the design of ecological studies (Hilborn et al., 1980; Holling, 1978; Walters and Hilborn, 1976). From a broader perspective, followup monitoring and retrospective analysis are ways to learn from experience and improve the prediction of ecological effects.

Monitoring is most effective when it is designed to test ecological hypotheses and when preproject studies have provided baseline information (see Beanlands and Duinker, 1983). Postproject studies of the accuracy of predictions are useful, but are not as useful as followup monitoring that coordinates preproject and postproject sampling and that tests relevant hypotheses.

Periodic analysis of results can help to detect unexpected changes and evaluate sampling programs, allowing them to be changed in a timely way.

Thus, an iterative approach to monitoring — with results fed into study design — is often effective, particularly when methods have not been well tested and when effects are uncertain. Any changes in sampling, however, must be made carefully, to ensure that new data are statistically comparable with those already collected.

Baseline monitoring of characteristics with substantial variation has a low probability of helping to detect changes due to a project. Measurements of baseline variability can help to identify the characteristics that it will be useful to measure in followup studies (Green, 1979) and can be used, with estimates of the duration of effects, to determine how long followup should continue.

Even if all the above criteria are met, followup studies of ecological effects can help in planning only if they are made available in an easily digestible form, ideally as published summaries and as complete postproject analyses (Hilborn and Walters, 1981).

In several case studies, followup monitoring was shown to be part of A SCIENTIFIC FRAMEWORK FOR ENVIRONMENTAL PROBLEM-SOLVING 113 Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. an overall study design to test hypotheses (Chapters 15, 22, and 23). The hydroelectric development in Southern Indian Lake (Chapter 21) was treated as a large-scale experiment, and monitoring provided increased understanding of artificial lakes. Because the results were analyzed and published, they can be applied to other cases.

In the caribou case ( Chapter 16 ), monitoring of caribou movements and herd productivity began before development and continued during construction.

Daily information on caribou movement was incorporated into constraints on construction activities. Annual monitoring of catch and fishing effort is used by the International Pacific Halibut Commission to set fishing quotas. Monitoring of conditions in Lake Washington allowed scientists to detect changes, predict trends in eutrophication, and predict and document recovery of the lake after action was taken; because the work was published, its lessons are readily available to managers of similar projects. Monitoring for DDT in the environment first identified the spread of another important group of toxic chemicals, PCBs (Chapter 24).

SUMMARY: DEVELOPING A STUDY STRATEGY
A study strategy is a plan for conducting ecological studies to help to predict and manage ecological effects. It is motivated by the environmental goals identified in scoping and is organized around the scientific objectives defined on the basis of the goals. Scoping identifies what information is required, and the study strategy specifies how to acquire it. A problem must be carefully thought through before studies aimed at solving it begin (Beanlands and Duinker, 1983).

Potential study objectives should be evaluated, so that efforts can be devoted to studies with some chance of producing useful results. What information is needed? Why is it needed? Is it possible to acquire adequate information? How will the information be used to satisfy the ecological goals? How will it be used in decision-making? Highly accurate characterization of a variable is of little use if decisions are made on the basis of other considerations. Decision analysis helps to ensure that modeling and research remain focused on the objectives.

A basic first step in designing studies is a review of what is already known about the problem. Larkin (1984) believes that literature review can provide more than 50% of the information needed in most initial impact assessments, and as much as 75% when coupled with brief reconnaissance surveys.

To summarize, a broad ecological study strategy:

A SCIENTIFIC FRAMEWORK FOR ENVIRONMENTAL PROBLEM-SOLVING 114 Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. Is based on thought-out environmental goals.
Is organized around clearly defined scientific objectives designed to satisfy the environmental goals.
Includes a description of the boundaries established for the problem, with demonstration of their appropriateness.
Is designed to evaluate hypotheses about how the ecological system functions and will be affected by perturbations.
Specifies predictions that will be tested, with the basis of the predictions and a statement of confidence in their accuracy. Defines the basis for choosing environmental goals and evaluating their significance.
Explains clearly how each part of the study fits into the overall design. Provides for baseline and followup monitoring to determine the effects of the project or perturbation.
Allows the results of the study to be used to evaluate the plan and to modify it if necessary.

A SCIENTIFIC FRAMEWORK FOR ENVIRONMENTAL PROBLEM-SOLVING 115 Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. 11 References
Ackerman, B. A., S. Rose-Ackerman, J. W. Sawyer, and D. W. Henderson. 1974. The Uncertain Search for Environmental Quality. Free Press (Macmillan), London.

Alexander, M. 1971. Microbial Ecology. John Wiley & Sons, New York.

Anderson, R. M., and R. M. May, eds. 1982. Population Biology of Infectious Diseases. Dahlem Konferenzen. Springer, Berlin.

Andreae, M. O., and H. Raemdonck. 1983. Dimethyl sulfide in surface oceans and marine atmosphere: A global view. Science 221:744-747 .

Andrews, R. N. L., P. Cromwell, G. A. Enk, E. G. Farworth, J. R. Hibbs, and V. L. Sharp. 1977.

Substantive Guidance for Environmental Impact Assessment: An Exploratory Study. The Institute of Ecology, Washington, D.C.

Anonymous. 1980. Biological Evaluation of Environmental Impacts. FWS/OBS-80/26 . U.S.

Council on Environmental Quality and Fish and Wildlife Service, U.S. Department of the Interior, Washington, D.C.

Anonymous. Undated. Post-development Audits to Test the Effectiveness of Environmental Impact Prediction Methods and Techniques. Final Report. Project Appraisal for Development Control, Environmental Impact Assessment and Planning Unit, Department of Geography, University of Aberdeen, Aberdeen, Scotland.

Baker, H. G. 1974. The evolution of weeds. Annu. Rev. Ecol. Syst. 5:1-24.

Baker, H. G., and G. L. Stebbins. 1965. The Genetics of Colonizing Species. Academic Press, New York.

Baker, J. M. 1976. Biological monitoring — Principles, methods and difficulties. Pp. 4153 in J. M.

Baker, ed. Marine Ecology and Oil Pollution. John Wiley & Sons, New York.

Baldwin, I. T., and J. C. Schultz. 1983. Rapid changes in tree chemistry induced by damage:

Evidence for communication between plants. Science 221:227-280 .

Barber, R. T., and F. P. Chavez. 1983. Biological consequences of El Niño . Science 222:1203-1210.

Barrons, K. C. 1981. Are Pesticides Really Necessary? Regnery Gateway, Chicago.

REFERENCES 116
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved.

Bar-Shalom, Y. 1976. Caution, probing, and the value of information in the control of uncertain systems. Ann. Econ. Soc. Meas. 5:323-337.

Baxter, R. M., and P. Glaude. 1980. Environmental Effects of Dams and Impoundments in Canada:

Experience and Prospects. Can. Bull. Fish. Aquat. Sci. 205.

Beanlands, G. E., and P. N. Duinker. 1983. An Ecological Framework for Environmental Impact Assessment in Canada. Institute for Resource and Environmental Studies, Dalhousie University, Halifax, Nova Scotia, and Federal Environmental Assessment Review Office, Ottawa, Ont.

Beanlands, G. E., W. J. Erckmann, G. H. Orians, J. O'Riordan, D. Policansky, M. H. Sadar, and B.

Sadler, eds. In press. Cumulative Environmental Effects: A Binational Perspective.

Canadian Environmental Assessment Research Council, Ottawa, Ont., and National Research Council, Washington, D.C.

Beddington, J. R. 1974. Age structure, sex ratio, and population density in the harvesting of natural populations. J. Appl. Ecol. 11:915-924 .

Beddington, J. R., rapporteur. 1984a. Management under uncertainty. Group report. Pp. 227-244 in R. M. May, ed. Exploitation of Marine Communities. Dahlem Konferenzen. Springer, Berlin.

Beddington, J. R. 1984b. The response of multispecies systems to perturbations. Pp. 209-225 in R.

M. May, ed. Exploitation of Marine Communities. Dahlem Konferenzen. Springer, Berlin.

Belovsky, G. E. 1984. Moose and snowshoe hare competition and a mechanistic explanation from foraging theory. Oecologia 61:150-159 .

Berg, K. 1938. Studies on the bottom animals of Esrom Lake. K. Danske Vidensk. Selsk. Skr. Nat.

Mat. Afd. 9(8):1-255.

Berry, J. A. 1975. Adaptation of photosynthetic processes to stress. Science 188:644-650 .

Berryman, A. A. 1981. Population Systems: A General Introduction. Plenum, New York.

Beverton, R. J. H. 1983. Science and decision-making in fisheries regulation. Pp. 919-938 in G. D.

Sharpe and J. Csirke, eds. Proceedings of the Expert Consultation to Examine Changes in Abundance and Species Composition of Neritic Fish Resources . FAO Fishery Report 291.

Food and Agriculture Organization, Rome.

Beverton, R. J. H., rapporteur. 1984. Dynamics of single species. Group report. Pp. 1358 in R. M.

May, ed. Exploitation of Marine Communities. Dahlem Konferenzen. Springer, Berlin.

Birch, L. C. 1957. The meanings of competition. Am. Nat. 91:5-18.

Biswell, H. H. 1974. Effects of fire on chaparral. Pp. 321-364 in T. T. Kozlowski and C. E. Ahlgren, eds. Fire and Ecosystems. Academic Press, New York.

Bjorkman, O., and J. Berry. 1973. High-efficiency photosynthesis. Sci. Am. 229(4):80-93 .

Bobrov, R. A. 1955. Use of plants as biological indicators of smog in the air of Los Angeles County. Science 121:510-511 .

Bodaly, R. A., R. E. Hecky, and R. J. P. Fudge. 1984. Increases in fish mercury levels in lakes flooded by the Churchill River diversion, northern Manitoba. Can. J. Fish. Aquat. Sci. 41:682-691 .

Boersma, P. D. In press. Seabirds reflect petroleum pollution. Science.

Bond, G. 1967. Fixation of nitrogen by higher plants other than legumes. Annu. Rev. Plant Physiol. 18:107-126 .

Borisov, V. M. 1978. The selective effect of fishing on the population structure of species with a long life cycle. J. Ichthyol. 18:896-904 .

Bormann, F. H., G. E. Likens, D. W. Fisher, and R. S. Pierce. 1968. Nutrient loss accelerated by clear-cutting of a forest ecosystem. Science 159:882-884.

REFERENCES 117
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved.

Brinkhurst, R. O. 1974. The Benthos of Lakes. St. Martin's Press, New York.

Brittingham, M. C., and S. A. Temple. 1983. Have cowbirds caused forest songbirds to decline? BioScience 33:31-35 .

Bromenshenk, J. J., S. R. Carlson, J. C. Simpson, and J. M. Thomas. 1985. Pollution monitoring of Puget Sound with honey bees. Science 227:632-634.

Brooks, J. L., and S. I. Dodson. 1965. Predation, body size, and composition of the plankton.

Science 150:28-35.

Brown, J. H. 1984. On the relationship between abundance and distribution of species. Am. Nat. 124:255-279 .

Brown, J. H. In press. Discussion: Cumulative impacts in terrestrial communities. In G. E.

Beanlands, W. J. Erckmann, G. H. Orians, J. O'Riordan, D. Policansky, M. H. Sadar, and B. Sadler, eds. Cumulative Environmental Effects: A Binational Perspective. Canadian Environmental Assessment Research Council, Ottawa, Ont., and National Research Council, Washington, D.C.

Bryson, R. A., and T. J. Murray. 1977. Climates of Hunger. University of Wisconsin Press, Madison.

Buck, W. B. 1979. Animals as monitors of environmental quality. Vet. Human Toxicol. 21:277-284.

Bull, D. 1982. A Growing Problem: Pesticides and the Third World Poor. Oxfam, London.

Cain, S. A. 1938. The species-area curve. Am. Midl. Nat. 20:573-581 .

Cairns, J., Jr. 1977. Effects of temperature changes and chlorination upon the community structure of aquatic organisms. Pp. 129-144 in M. Marois, ed. Towards a Plan of Actions for Mankind. Vol. 3. Biological Balance and Thermal Modifications. Pergamon Press, Oxford, Eng.

Cairns, J., Jr. 1980. The Recovery Process in Damaged Ecosystems. Ann Arbor Sciences, Ann Arbor, Mich.

Cairns, J., Jr., and K. L. Dickson. 1980. Risk analysis for aquatic ecosystems. Pp. 73-83 in Biological Evaluation of Environmental Impacts. FWS/OBS-80/26. U.S. Council on Environmental Quality and Fish and Wildlife Service, U.S. Department of the Interior, Washington, D.C.

Cairns, J., Jr., G. R. Larra, R. E. Sparks, and W. T. Waller. 1973. Developing biological information systems for water quality management. Water Res. Bull. 9:81-99.

Cameron, R. D., K. R. Whitten, W. T. Smith, and D. D. Roby. 1979. Caribou distribution and group composition associated with construction of the trans-Alaska pipeline. Can. Field Nat. 93:155-162.

Campbell, R. W., and R. J. Sloan. 1977. Natural regulation of innocuous gypsy moth populations.

Environ. Entomol. 6:315-322 .

Campbell, R. W., and R. J. Sloan. 1978. Natural maintenance and decline of gypsy moth outbreaks.

Environ. Entomol. 7:389-395 .

Carpenter, R. A. 1976. The scientific basis of NEPA — Is it adequate? Environ. Law Reporter 6:50014-50019 .

Carpenter, S. R., and J. F. Kitchell. 1984. Plankton community structure and limnetic primary production. Am. Nat. 124:159-172.

Carson, R. 1962. Silent Spring. Houghton Mifflin, Boston.

Caswell, H., and P. A. Werner. 1978. Transient behavior and life history of teasel (Dipsacus svlvestris Huds.). Ecology 59:53-66.

Caswell, H., F. Reed, S. N. Stephenson, and P. A. Werner. 1973. Photosynthetic pathways and selective herbivory, a hypothesis. Am. Nat. 107:465-480.

Chapra, S. C., and K. H. Reckhow. 1983. Engineering Approaches for Lake Management. Vol. 2.

Ann Arbor Sciences, Ann Arbor, Mich.

REFERENCES 118
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved.

Charnov, E. L., and G. H. Orians. 1973. Optimal Foraging: Some Theoretical Explorations.

University of Utah Press, Salt Lake City.

Christensen, S. W., W. Van Winkle, and J. S. Mattice. 1976. Defining and determining the significance of impacts: Concepts and methods. Pp. 191-219 in R. K. Sharma, J. D.

Buffington, and J. T. McFadden, eds. Proceedings of a Workshop on the Biological Significance of Environmental Impacts. NR-CONF-002. U.S. Nuclear Regulatory Commission, Washington, D.C.

Clark, L. R., P. W. Geier, R. D. Hughes, and R. F. Morris. 1967. The Ecology of Insect Populations in Theory and Practice. Methuen, London.

Clark, W. C. In press. The cumulative impacts of human activities on the atmosphere. In G. E.

Cody, M. L. 1969. Convergent characteristics in sympatric species: A possible relation to interspecific competition and aggression. Condor 71:223-239.

Cody, M. L. 1974. Competition and the Structure of Bird Communities. Princeton University Press, Princeton, N.J.

Connell, J. H. 1975. Some mechanisms producing structure in natural communities: A model and evidence from field experiments. Pp. 460-490 in M. L. Cody and J. M. Diamond, eds.

Ecology and Evolution of Communities. Harvard University Press, Cambridge, Mass.

Connell, J. H. 1978. Diversity in tropical rain forests and coral reefs. Science 199:1302-1310.

Connell, J. H. 1980. Diversity and the coevolution of competitors, or the ghost of competition past.

Oikos 35:131-138.

Connell, J. H. 1983. On the prevalence and relative importance of interspecific competition:

Evidence from field experiments. Am. Nat. 122:661-696.

Connell, J. H., and E. Orias. 1964. The ecological regulation of species diversity. Am. Nat. 98:399-414.

Connister, R. Z. 1964. The mechanism of prey digestion in the spotted crab, Cancer ocellatus. J.

Am. Syst. Assoc. 42:36-54 .

Connor, E. F., and E. McCoy. 1979. The statistics and biology of the species-area relationship. Am.

Nat. 113:791-833 .

Cooke, A. S. 1973. Shell-thinning in avian eggs by environmental pollutants. Environ. Pollut. 4:85-157.

Cooper, C. F., and P. H. Zedler. 1980. Ecological assessment for regional development. J. Environ.

Manage. 10:285-296.

Council on Environmental Quality. 1978. Regulations for Implementing the Procedural Provisions of the National Environmental Policy Act. Reprint 43 FR 55978-56007, November 29, 1978, 40CFR Parts 1500-1508. Washington, D.C.

Courtenay, W. R., Jr., and J. R. Stauffer, Jr., eds. 1984. Distribution, Biology, and Management of Exotic Fishes. Johns Hopkins University Press, Baltimore, Md.

Cowell, E. B. 1978. Ecological monitoring as a management tool in industry. Ocean Manage. 4:273-285 .

Cowell, E. B., and W. J. Syratt. 1979. A technique for assessing ecological damage to the intertidal zone of rocky shores for which no previous baseline data is available. Pp. 29-39 in Proceedings of the Ecological Damage Assessment Conference. Society of Petroleum Industry Biologists, Los Angeles.

Cowie, R. J. 1977. Optimal foraging in great tits (Parus major). Nature 268:137-139 .

REFERENCES 119
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved.

Crawley, M. J. 1983. Herbivory: The Dynamics of Plant-Animal Interactions. Blackwell Scientific, Oxford, Eng.

Crow, M. E., and F. B. Taub. 1979. Designing a microcosm bioassay to detect ecosystem level effects. Int. J. Environ. Stud. 13:141-147.

Darwin, C. 1859. On the Origin of Species by Means of Natural Selection, or the Preservation of Favoured Races in the Struggle for Existence. John Murray, London.

Davidson, D. W., J. H. Brown, and R. S. Inouye. 1980. Competition and the structure of granivore communities. BioScience 30:233-238.

Davidson, D. W., R. S. Inouye, and J. H. Brown. 1984. Granivory in a desert ecosystem:

Experimental evidence for indirect facilitation of ants by rodents. Ecology 65:1780-1786 .

Dayton, P. K. 1971. Competition, disturbance, and community organization: The provision and subsequent utilization of space in a rocky intertidal community. Ecol. Monogr. 41:351-389 .

Dayton, P. K. 1984. Processes structuring some marine communities: Are they general? Pp. 181-197 in D. R. Strong, Jr., D. Simberloff, L. Abele, and A. Thistle, eds. Ecological Communities:

Conceptual Issues and the Evidence. Princeton University Press, Princeton, N.J.

Dayton, P. K. 1985. Ecology of kelp communities. Annu. Rev. Ecol. Syst. 16:215-245 .

Dayton, P. K. In press. Cumulative impacts in the marine realm. In G. E. Beanlands, W. J.

Erckmann, G. H. Orians, J. O'Riordan, D. Policansky, M. H. Sadar, and B. Sadler, eds.

Cumulative Environmental Effects: A Binational Perspective. Canadian Environmental Assessment Research Council, Ottawa, Ont., and National Research Council, Washington, D.C.

DeMott, W. R., and W. C. Kerfoot. 1982. Competition among cladocerans: Coexistence of Bosmina and Daphnia. Ecology 63:1949-1966.

Diamond, J. M., and M. E. Gilpin. 1982. Examination of the ''null" model of Connor and Simberloff for species co-occurrences on islands. Oecologia 52:64-74.

Dickman, M. 1968. The relation of freshwater plankton productivity to species composition during induced successions . Ph.D. thesis, University of British Columbia, Vancouver.

Dillon, P. J., and F. H. Rigler. 1974. The phosphorus-chlorophyll relationship in lakes. Limnol.

Oceanogr. 19:767-773.

Doberstyn, E. B. 1984. Resistance of Plasmodiumfalciparum. Experientia 40:1311-1317.

Dodd, A. P. 1959. The biological control of prickly pear in Australia. Pp. 565-577 in A. Keast, R. L.

Crocker, and C. S. Christian, eds. Biogeography and Ecology in Australia. Monographiae Biologicae. Vol. 8. W. Junk, The Hague.

Drury, W. B., and I. C. T. Nisbet. 1973. Succession. J. Arnold Arbor. 54:331-368.

Duggins, D. O. 1980. Kelp beds and sea otters: An experimental approach. Ecology 61:447-453.

Duinker, P. N., and G. E. Beanlands. 1983. Ecology and Environmental Assessment: An Annotated Bibliography. Institute for Resource and Environmental Studies, Dalhousie University, Halifax, Nova Scotia, and Federal Environmental Assessment Review Office, Ottawa, Ont.

Dunlap, T. R. 1981. DDT: Scientists, Citizens, and Public Policy. Princeton University Press, Princeton, N.J.

Eberhardt, L. L. 1976. Quantitative ecology and impact assessment. J. Environ. Manage. 4:27-70.

Eberhardt, L. L. 1978. Appraising variability in population studies. J. Wildl. Manage. 42:207-238.

Ebersole, J. P. 1977. The adaptive significance of territoriality in the reef fish Eupomacentrus leucostictus. Ecology 58:914-920.

REFERENCES 120
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved.

Edmondson, W. T. 1969. Eutrophication in North America. Pp. 124-149 in Eutrophication: Causes, Consequences, Correctives. National Academy of Sciences, Washington, D. C.

Edmondson, W. T. 1972. Nutrients and phytoplankton in Lake Washington. Am. Soc. Limnol.

Oceanogr. Spec. Symp. 1:172-193 .

Edmondson, W. T. 1974. Environmental Phosphorus Handbook. Limnol. Oceanogr. 19:369-374 . (book review)
Edmondson, W. T., and J. T. Lehman. 1981. The effect of changes in the nutrient income on the condition of Lake Washington. Limnol. Oceanogr. 26:1-29 .

Edmondson, W. T., and A. H. Litt. 1982. Daphnia in Lake Washington. Limnol. Oceanogr. 27:272-293.

Ehrlich, P. R., and A. Ehrlich. 1981. Extinction: The Causes and Consequences of the Disappearance of Species. Random House, New York.

Ehrlich, P. R., and P. H. Raven. 1965. Butterflies and plants: A study in coevolution. Evolution 18:586-608.

Elton, C. S. 1958. The Ecology of Invasions by Animals and Plants. Methuen, London.

Emlen, J. M. 1970. Age specificity and ecological theory. Ecology 51:588-601 .

Endler, J. A. 1977. Geographic Variation, Speciation, and Clines. Princeton University Press, Princeton, N.J.

Eppley, R. W., and B. Peterson. 1979. Particulate organic matter flux and planktonic new production in the deep ocean. Nature 282:677-680.

ESSA (Environmental and Social Systems Analysts Ltd.). 1982. Review and Evaluation of Adaptive Environmental Assessment and Management. Environment Canada, Vancouver, B.C.

Estes, J. A., R. J. Jameson, and E. B. Rhode. 1982. Activity and prey selection in the sea otter:

Influence of population status on community structure. Am. Nat. 120:242-258.

Ewald, P. W. 1983. Host-parasite relations, vectors, and the evolution of disease severity. Annu.

Rev. Ecol. Syst. 14:465-485 .

Feeny, P. 1976. Plant apparency and chemical defense. Rec. Adv. Phytochem. 10:1-40.

Finn, J. T. 1976. Measures of ecosystem structure and function derived from analysis of flows. J.

Theoret. Biol. 56:363-380.

Finn, J. T. 1978. Cycling index: A general definition for cycling in compartment models. Pp. 138-164 in D. C. Adriano and I. L. Brisbin, eds. Environmental Chemistry and Cycling Processes. CONF-760429. Technical Information Center, U.S. Department of Energy, Washington, D.C.

Flemer, D. A., G. B. Mackiernan, W. Nehlsen, R. B. Biggs, D. Blaylock, N. H. Burger, L. C.

Davidson, D. Haberman, K. S. Price, and J. L Taft. 1983. Chesapeake Bay: A Profile of Environmental Change. Chesapeake Bay Program, U.S. Environmental Protection Agency, Annapolis, Md.

Forsman, E. D., E. C. Meslow, and H. M. Wight. 1984. Distribution and biology of the spotted owl in Oregon. Wildl. Monogr. 87:1-64.

Fox, L. R. 1981. Defense and dynamics in plant-herbivore systems. Am. Zool. 21:853-864.

Fox, L. R., and B. J. Macauley. 1977. Insect grazing on Eucalyptus in response to variation in leaf tanning and nitrogen. Oecologia 29:145-162.

Franey, J. R., M. V. Ivanov, and H. Rhodhe. 1983. The sulphur cycle. Ch. 2.5 in B. Bolin and R. B.

Cook, eds. The Major Biogeochemical Cycles and Their Interactions. Scientific Committee on Problems of the Environment Publ. 21. John Wiley & Sons, Chichester, Eng.

Frank, P. W. 1968. Life histories and community stability. Ecology 49:355-356 .

REFERENCES 121
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved.

Frankel, O. H., and M. E. Soul. 1981. Conservation and Evolution. Cambridge University Press, New York.

Franklin, I. R. 1980. Evolutionary change in small populations. Pp. 135-149 in M. Soul é and B.

Wilcox, eds. Conservation Biology: An Evolutionary-Ecological Perspective. Sinauer Associates, Sunderland, Mass.

Frenkiel, F. N., and D. W. Goodall, eds. 1978. Simulation Modeling of Environmental Problems.

SCOPE Report 9. International Council of Scientific Unions-Scientific Committee on Problems of the Environment. John Wiley & Sons, Chichester, Eng.

Fretwell, S. D. 1972. Populations in a Seasonal Environment. Princeton University Press, Princeton, N.J.

Fritz, E. S., P. J. Rago, and 1. D. Murarka. 1980. Strategy for Assessing Impacts of Power Plants on Fish and Shellfish Populations. FWS/OBS-80/34. National Power Plant Team, Fish and Wildlife Service, U.S. Department of the Interior, Ann Arbor, Mich.

Furniss, R. L., and V. M. Carolin. 1980. Western Forest Insects. Miscellaneous Publication 1339.

Forest Service, U.S. Department of Agriculture, Washington, D.C.

Galli, A. E., C. F. Leck, and R. T. T. Forman. 1976. Avian distribution patterns in forest islands of different sizes in central New Jersey. Auk 93:356-364 .

Gause, G. F. 1934. The Struggle for Existence. Williams and Wilkins, Baltimore, Md.

Gavereski, C. A. 1976. Relation of park size and vegetation to urban bird populations in Seattle, Washington. Condor 78:375-382.

Georghiou, G. P., A. Lagunes, and J. D. Baker. 1983. Effect of insecticide rotations on the evolution of insecticide resistance. Pp. 183-189 in J. Miyamoto, ed. Pesticide Chemistry: Human Welfare and the Environment. Pergamon Press, New York.

Geppert, R. R., C. W. Lorenz, and A. G. Larson. 1984. Cumulative Effects of Forest Practices on the Environment: A State of the Knowledge. Ecosystems, Inc., Olympia, Wash.

Giddings, J. M. 1980. Field experiments. Pp. 315-331 in F. S. Sanders, S. M. Adams, L. W.

Barnthouse, J. M. Giddings, E. E. Huber, K. D. Kumar, D. Lee, B. Murphy, G. W. Suter, and W. Van Winkle, eds. Strategies for Ecological Effects Assessment at DOE Energy Activity Sites. Environmental Sciences Division Publ. 1639. ORNL/TM-6783. Oak Ridge National Laboratory, Oak Ridge, Tenn.

Gilman, A. P., D. B. Peakall, D. J. Hallett, G. A. Fox, and R. J. Norstrom. 1979. Animals as Monitors of Environmental Pollutants. National Academy of Sciences, Washington, D.C.

Gilpin, M. E. 1975. Limit cycles in competition communities. Am. Nat. 109:51-60.

Gilpin, M. E., and J. M. Diamond. 1982. Factors contributing to non-randomness in species co- occurrences on islands. Oecologia 52:75-84 .

Gleason, H. A. 1922. On the relation between species and area. Ecology 3:158-162 .

Godron, M., and R. T. T. Forman. 1983. Landscape modification and changing ecological characteristics. Pp. 12-28 in H. A. Mooney and M. Godron, eds. Disturbance and Ecosystems. Springer, New York.

Goldberg, E. D., V. T. Bowen, J. W. Farrington, G. Harvey, J. H. Martin, P. L. Parker, R. W.

Riseborough, W. Robertson, E. Schneider, and E. Gamble. 1978. The Mussel Watch.

Environ. Conserv. 5:101-125.

Gomez-Pompa, A. C., C. Vasquez-Yaues, and S. Guerarra. 1972. The tropical rainforest: A nonrenewable resource. Science 177:762-765.

Goodman, D. 1975. The theory of diversity-stability relationships in ecology. Q. Rev. Biol. 50:237-266.

Grassle, J. F., H. L. Sanders, R. R. Hessler, G. T. Rowe, and T. McLellan. 1975. Pattern REFERENCES 122
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. and zonation: A study of the bathyal megafauna using the research submersible Alvin.

Deep-Sea Res. 22:457-481.

Green, R. H. 1979. Sampling Design and Statistical Methods for Environmental Biologists. John Wiley & Sons, Toronto.

Greenland, D. J., and R. Lal. 1977. Soil Conservation and Management in the Humid Tropics. John Wiley & Sons, Chichester, Eng.

Gross, M. R. 1984. Sunfish, salmon, and the evolution of alternative reproductive strategies and tactics in fishes. Pp. 55-75 in G. W. Potts and R. J. Wootton, eds. Fish Reproduction.

Academic Press, New York.

Gross, M. R. 1985. Disruptive selection for alternative life histories in salmon. Nature 313:47-48.

Hairston, N. G., F. E. Smith, and L. B. Slobodkin. 1960. Community structure, population control and competition. Am. Nat. 94:421-425 .

Handford, P., G. Bell, and T. Reimchen. 1977. A gillnet fishery considered as an experiment in artificial selection. J. Fish. Res. Bd. Can. 34:954-961.

Harcourt, D. G., and E. J. Leroux. 1967. Population regulation in insects and man. Am. Sci. 55:400-415.

Hare, J. D. 1983. Manipulation of host suitability for herbivore pest management. Pp. 655-680 in R.

F. Denno and M. S. McClure, eds. Variable Plants and Herbivores in Natural and Managed Ecosystems. Academic Press, New York.

Harper, J. L., P. H. Lovell, and K. G. Moore. 1970. The shapes and sizes of seeds. Annu. Rev. Ecol.

Syst. 1:327-356 .

Harris, G. P. 1980. Temporal and spatial scales in phytoplankton ecology: Mechanisms, models and management. Can. J. Fish. Aquat. Sci. 37:877-900.

Harrison, W. G. 1980. Nutrient regeneration and primary production in the sea. Pp. 433460 in P. G.

Falkowski, ed. Primary Productivity in the Sea. Plenum, New York.

Harshbarger, J. C., and J. J. Black. In press. A strategy for using fish bioassays and surveys to identify and eliminate point source environmental carcinogens. In Towards a Transboundary Monitoring Network: Proceedings of a Workshop. International Joint Commission, Washington, D.C., and Ottawa, Ont.

Hartzbank, D. J., and A. McCusker. 1979. Establishing criteria for offshore sampling design. Pp. 59-78 in Proceedings of the Ecological Damage Assessment Conference. Society of Petroleum Industry Biologists, Los Angeles.

Hasler, A. D. 1947. Eutrophication of lakes by domestic sewage. Ecology 28:383-395.

Hassell, M. P. 1985. Insect natural enemies as regulating factors. J. Anim. Ecol. 54:323-334.

Haukioja, E., and T. Hakala. 1975. Herbivore cycles and periodic outbreaks. Formulation of a general hypothesis. Rep. Kevo Subarctic Res. Stn. 12:1-9.

Haukioja, E., and P. Niemel ä. 1979. Birch leaves as a resource for herbivores: Seasonal occurrence of increased resistance in foliage after mechanical damage of adjacent leaves. Oecologia 39:151-159.

Hawksworth, D. 1971. Lichens as litmus for air pollution: A historical review. Int. J. Environ. Stud. 1:281-296.

Heath, R. T. 1979. Holistic study of an aquatic microcosm: Theoretical and practical implications.

Int. J. Environ. Stud. 13:87-93.

Hecky, R. E. 1984. Thermal and optical characteristics of Southern Indian Lake before, REFERENCES 123
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. during, and after impoundment and Churchill River diversion. Can. J. Fish. Aquat. Sci. 41:579-590 .

Hecky, R. E., R. W. Newbury, R. A. Bodaly, K. Patalas, and D. M. Rosenberg. 1984.

Environmental impact prediction and assessment: The Southern Indian Lake experience.

Can. J. Fish. Aquat. Sci. 41:720-732.

Heinrich, B. 1979. Resource heterogeneity and problems of movement in foraging bees. Oecologia 40:235-245.

Hilborn, R. 1979. Some failures and successes in applying systems analysis to ecological systems. J.

Appl. Syst. Anal. 6:25-31 .

Hilborn, R., and C. Walters. 1981. Some pitfalls of environmental baseline and process studies.

Report 3. Cooperative Fisheries Research Unit. University of British Columbia, Vancouver.

Hilborn, R., C. S. Holling, and C. J. Walters. 1980. Managing the unknown: Approaches to ecological policy design. Pp. 103-113 in Biological Evaluation of Environmental Impacts.

FWS/OBS-80/26. Council on Environmental Quality and Fish and Wildlife Service, U.S.

Department of the Interior, Washington, D.C.

Hirsch, A. 1980. The baseline study as a tool in environmental impact assessment. Pp. 84-93 in Biological Evaluation of Environmental Impacts. FWS/OBS-80/26. Council on Environmental Quality and Fish and Wildlife Service, U.S. Department of the Interior, Washington, D.C.

Holdridge, L. R. 1967. Life Zone Ecology. Tropical Science Center, San Jose, Costa Rica.

Holling, C. S. 1973. Resilience and stability of ecological systems. Annu. Rev. Ecol. Syst. 4:1-23.

Holling, C. S., ed. 1978. Adaptive Environmental Assessment and Management. Int. Ser. on Applied Systems Analysis 3, International Institute for Applied Systems Analysis. John Wiley & Sons, Toronto.

Horn, H. S. 1974. The ecology of secondary succession. Annu. Rev. Ecol. Syst. 5:25-37 .

Horn, H. 1976. Succession. Pp. 187-204 in R. M. May, ed. Theoretical Ecology: Principles and Applications. Blackwell Scientific, Oxford, Eng.

Hornbeck, J. W., G. E. Likens, R. S. Pierce, and F. H. Bormann. 1975. Strip cutting as a means of protecting site and streamflow quality when clearcutting northern hardwoods. Pp. 209-225 in B. Bernier and C. H. Winget, eds. Forest Soils and Forest Land Management.

Proceedings of the Fourth North American Forest Soils Conference. Les Presses de l'Université Laval, Quebec, Que.

Huston, M. 1979. A general hypothesis of species diversity. Am. Nat. 113:81-101 .

Hutchinson, G. E. 1959. Homage to Santa Rosalia, or why are there so many kinds of animals? Am.

Nat. 93:145-159.

Istock, C. A. 1967. The evolution of complex life cycle phenomena: An ecological perspective.

Evolution 21:592-605.

Jewell, P. A., and S. Holt, eds. 1981. Problems in Management of Locally Abundant Wild Mammals. Academic Press, New York.

Jonasson, P. M. 1972. Ecology and production of the profundal benthos in relation to phytoplankton in Lake Estrom. Oikos 14(Suppl.):1-148.

Jordan, C. F. 1971. A world pattern in plant energetics. Am. Sci. 59:425-433 .

Juday, C., and E. A. Birge. 1931. A second report on the phosphorus content of Wisconsin lake waters. Trans. Wis. Acad. Sci. Arts Lett. 26:353-382 .

Juday, C., E. A. Birge, G. I. Kemmerer, and R. J. Robinson. 1928. Phosphorus content of lake waters of northeastern Wisconsin. Trans. Wis. Acad. Sci. Arts Lett. 23:233-248 .

REFERENCES 124
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved.

Karr, J. K. 1982. Avian extinction on Barro Colorado Island, Panama: A reassessment. Am. Nat. 119:228-239.

Karr, J. R., and R. R. Roth. 1971. Vegetation structure and avian diversity in several New World areas. Am. Nat. 105:423-435 .

Kettlewell, H. B. D. 1973. The Evolution of Melanism. Clarendon Press, Oxford, Eng.

Kimball, K. D., and S. A. Levin. 1985. Limitations of laboratory bioassays: The need for ecosystem- level testing. BioScience 35:165-171 .

Kodric-Brown, A., and J. H. Brown. 1979. Competition between distantly related taxa in the coevolution of plants and pollinators. Am. Zoo. 19:1115-1127 .

Kozlowski, T. T., and C. E. Ahlgren, eds. 1974. Fire and Ecosystems. Academic Press, New York.

Krebs, C. J. 1985. Ecology: The Experimental Analysis of Distribution and Abundance. 3rd ed.

Harper and Row, New York.

Kubicki, B. 1969a. Investigations on sex determination in cucumber ( Cucumis sativus). Parts 3-8.

Genet. Pol. 10:5-144 .

Kubicki, B. 1969b. Sex determination in muskmelon (Cucumis melo). Genet. Pol. 10:145-166 .

Kubicki, B. 1969c. Comparative studies on sex determination in cucumber ( Cucumis sativus) and muskmelon (Cucumis melo). Genet. Pol. 10:167-184.

Kumar, K. D. 1980. Statistical considerations. Pp. 333-348 in F. S. Sanders, S. M. Adams, L. W.

Barnthouse, J. M. Giddings, E. E. Huber, K. D. Kumar, D. Lee, B. Murphy, G. W. Suter, and W. Van Winkle, eds. Strategies for Ecological Effects Assessment at DOE Energy Activity Sites. Environmental Sciences Division Publ. 1639. ORNL/TM6783. Oak Ridge National Laboratory, Oak Ridge, Tenn.

Lack, D. 1954. The Natural Regulation of Animal Numbers. Clarendon Press, Oxford, Eng.

Larkin, P. A. 1978. Fisheries management— An essay for ecologists. Annu. Rev. Ecol. Syst. 9:57-73.

Larkin, P. A. 1984. A commentary on environmental impact assessment for large projects affecting lakes and streams. Can. J. Fish. Aquat. Sci. 41:1121-1127 .

Law, R. 1979. Optimal life histories under age-specific predation. Am. Nat. 114:399-417 .

Lawrey, J. D., and M. E. Hale, Jr. 1979. Lichen growth responses to stress induced by automobile exhaust pollution. Science 204:423-424.

Lawton, J. H., and M. P. Hassell. 1981. Asymmetrical competition in insects. Nature 289:793-795 .

Leonard, D. E. 1974. Recent developments in ecology and control of the gypsy moth. Annu. Rev.

Entomol. 19:197-229.

Levin, B. R., and R. E. Lenski. 1983. Coevolution in bacteria and their viruses and plasmids. Pp. 99-127 in D. J. Futuyma and M. Slatkin, eds. Coevolution. Sinauer Associates, Sunderland, Mass.

Levins, R., and R. H. MacArthur. 1969. An hypothesis to explain the incidence of monophagy.

Ecology 15:910-911 .

Lewis, R. A., N. Stein, and C. W. Lewis, eds. 1984. Environmental Specimen Banking and Monitoring as Related to Banking. Martinus Nijhoff, Boston.

Lewontin, R. C. 1965. Selection for colonizing ability. Pp. 77-94 in H. G. Baker and G. L. Stebbins, eds. The Genetics of Colonizing Species. Academic Press, New York.

Likens, G. E., ed. 1972. Nutrients and eutrophication: The limiting nutrients controversy. Am. Soc.

Limnol. Oceanogr. Spec. Symp. 1:1-328 .

Likens, G. E., and F. H. Bormann. 1972. Nutrient cycling in ecosystems. Pp. 25-67 in REFERENCES 125
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved.

J. A. Wiens, ed. Ecosystem Structure and Function. Oregon State University Press, Corvallis.

Lovejoy, T. E. 1979. A projection of species extinctions. Pp. 328-329 in Council on Environmental Quality and Department of State. The Global 2000 Report to the President: Entering the Twenty-First Century. Council on Environmental Quality, Washington, D.C.

Lucas, H. L. 1976. Some statistical aspects of assessing environmental impact. Pp. 295-306 in R. K.

Sharma, J. D. Buffington, and J. T. McFadden, eds. Proceedings of a Workshop on the Biological Significance of Environmental Impacts. NR-CONF-002. U.S. Nuclear Regulatory Commission, Washington, D.C.

Ludwig, D., and R. Hilborn. 1983. Adaptive probing strategies for age-structured fish stocks. Can. J.

Fish. Aquat. Sci. 40:559-569.

MacArthur, R. H. 1955. Fluctuations of animal populations, and a measure of community stability.

Ecology 36:533-536 .

MacArthur, R. H. 1965. Patterns of species diversity. Biol. Rev. 40:510-533 .

MacArthur, R. H. 1972. Geographical Ecology. Harper and Row, New York.

MacArthur, R. H., and J. W. MacArthur. 1961. On bird species diversity. Ecology 42:594-598.

MacArthur, R. H., and E. O. Wilson. 1967. The Theory of Island Biogeography. Princeton University Press, Princeton, N.J.

MacClintock, L., R. F. Whitcomb, and D. L. Whitcomb. 1977. Island biogeography and "habitat islands" of eastern forest. II. Evidence for the value of corridors and minimization of isolation in preservation of biotic diversity. Am. Birds 32:6-16.

Mack, R. N. In press. Alien plant invasion into the Intermountain West: A case history. In H. A.

Mooney, ed. Ecology of Biological Invasions in North America and Hawaii . Springer, New York.

Mack, R. N., and J. N. Thompson. 1982. Evolution in steppe with few large, hooved mammals. Am.

Nat. 119:757-773.

MacLeod, J. M. 1977. Discontinuous Stability in a Sawfly Life System and Its Relevance to Pest Management. Working Paper W-22, Institute of Animal Resource Ecology, University of British Columbia, Vancouver, B.C.

Mann, K. H., and P. A. Breen. 1972. The relation between lobster abundance, sea urchins, and kelp beds. J. Fish. Res. Bd. Can. 29:603-605.

Maryland Department of Natural Resources. 1984. Power Plant Cumulative Environmental Impact Report. PPSP-CEIR-4. Maryland Department of Natural Resources, Annapolis, Md.

Massachusetts Department of Environmental Management. 1981. Gypsy Moth Control. Final Environmental Impact Report. Commonwealth of Massachusetts, Boston.

Mathisen, O. A. 1962. The effect of altered sex ratios on the spawning of red salmon. Pp. 137-245 in T. Y. Koo, ed. Studies of Alaska Red Salmon. University of Washington Press, Seattle.

Maxwell, F. G., and P. R. Jennings, eds. 1980. Breeding Plant Resistance to Insects . John Wiley & Sons, New York.

May, R. M. 1973. Stability and Complexity in Model Ecosystems. Princeton University Press, Princeton, N.J.

May, R. M. 1977. Food lost to pests. Nature 267:669-670 .

May, R. M. 1980. Mathematical models in whaling and fisheries management. Am. Math. Soc.

Lect. Math. Life Sci. 13:1-64.

May, R. M. 1981. Models for single populations. Pp. 5-29 in R. M. May, ed. Theoretical Ecology.

Sinauer Associates, Sunderland, Mass.

REFERENCES 126
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved.

May, R. M. 1983. The structure of food webs. Nature 301:566-568.

May, R. M., ed. 1984. Exploitation of Marine Communities. Dahlem Konferenzen. Springer, Berlin.

May, R. M. 1985. Evolution of pesticide resistance. Nature 315:12-13.

May, R. M., and A. P. Dobson. In press. Population dynamics and the rate of evolution of pesticide resistance. In National Research Council. Management of Resistance to Pesticides:

Strategies, Tactics, and Research Needs. National Academy Press, Washington, D.C.

McNab, B. 1963. Bioenergetics and the determination of home range size. Am. Nat. 97:133-140.

McNamee, P. J., J. M. McLeod, and C. S. Holling. 1981. The Structure and Behavior of Defoliating Insect/Forest Systems. Research Publication R-25. Institute of Animal Resource Ecology, University of British Columbia, Vancouver.

Menge, B. A., and J. P. Sutherland. 1976. Species diversity gradients: Synthesis of the roles of predation, competition, and temporal heterogeneity. Am. Nat. 110:351-369.

Meyer, S. L. 1975. Data Analysis for Scientists and Engineers. John Wiley & Sons, New York.

Middleton, J. T. 1956. Response of plants to air pollution. J. Air Pollut. Control Assoc. 6:7-9.

Mindinger, P. C., and S. Hope. 1982. Expansion of the range of the house finch. Am. Birds 36:347-353.

Mineau, P. G., G. A. Fox, R. J. Norstrom, D. V. Weseloh, D. J. Hallett, and J. A. Ellenton. 1984.

Using the herring gull to monitor levels and effects of organochlorine contaminants in the Canadian great lakes. Pp. 425-452 in J. O. Nriagu and M. S. Simmons, eds. Contaminants in the Great Lakes. John Wiley & Sons, Toronto.

Moav, R., T. Brody, and G. Hulata. 1978. Genetic improvement of wild fish populations. Science 201:1090-1094.

Moore, F. R. 1978. Interspecific aggression: Toward whom should it be addressed? Behav. Ecol.

Sociobiol. 3:173-176.

Mortimer, A. M. 1984. Population ecology and weed science. Pp. 363-388 in R. Dirzo and J.

Sarukhan, eds. Perspectives on Plant Population Ecology. Sinauer Associates, Sunderland, Mass.

Munn, R. E., ed. 1979. Environmental Impact Assessment: Principles and Procedures. SCOPE 5.

John Wiley & Sons, Chichester, Eng.

Murphy, G. I. 1977. Clupeoids. Pp. 283-308 in J. Gulland, ed. Fish Population Dynamics. John Wiley & Sons, New York.

Myerberg, A. A., Jr., and R. E. Thresher. 1974. Interspecific aggression and its relevance to the concept of territoriality in reef fishes. Am. Zool. 14:81-96.

Myers, N. 1979. The Sinking Ark. Pergamon Press, Oxford, Eng.

Myers, N. 1980. Conversion of Tropical Moist Forests. National Academy of Sciences, Washington, D.C.

Myers, N. 1984. The Primary Source. Norton, New York.

National Research Council. 1969. Eutrophication: Causes, Consequences, Correctives. National Academy of Sciences, Washington, D.C.

National Research Council. 1981. Testing for Effects of Chemicals on Ecosystems. National Academy Press, Washington, D.C.

National Research Council. 1982. Assessment of Multichemical Contamination. Proceedings of an International Workshop, Milan, April 28-30, 1981. National Academy Press, Washington, D.C.

National Research Council. 1983a. Acid Deposition in Eastern North America. National Academy Press, Washington, D.C.

National Research Council. 1983b. Effects of atmospheric transformations of polycyclic aromatic hydrocarbons. Pp. 3-1 to 3-48 in Polycyclic Aromatic Hydrocarbons: Evaluation of Sources and Effects . National Academy Press, Washington, D.C.

National Research Council. 1983c. Changing Climate. Report of the Carbon Dioxide Assessment Committee. National Academy Press, Washington, D.C.

National Research Council. In press. Monitoring and Assessment of Trends in Acid Deposition.

National Academy Press, Washington, D.C.

Niering, W. A., and F. E. Egler. 1955. A shrub community of Viburnum lentagon, stable for twenty- five years. Ecology 36:356-360.

O'Donald, P. 1967. A general model of sexual and natural selection. Heredity 22:499-518.

Odum, E. P. 1969. The strategy of ecosystem development. Science 164:262-270.

Odum, E. P., J. T. Finn, and E. H. Franz. 1979. Perturbation theory and the subsidystress gradient.

BioScience 29:349-352.

Odum, W. E. 1982. Environmental degradation and the tyranny of small decisions. BioScience 32:728-729.

Olson, J. S. 1958. Rates of succession and soil changes on southern Lake Michigan sand dunes .

Bot. Gaz. 119:125-170 .

Orians, G. H. 1969. The number of bird species in some tropical forests. Ecology 50:783-801 .

Orians, G. H. 1975. Diversity, stability and maturity in natural ecosystems. Pp. 139-150 in W. H.

Van Dobben and R. H. Lowe-McConnell, eds. Unifying Concepts in Ecology. W. Junk, The Hague.

Orians, G. H. 1980. Some Adaptations of Marsh-nesting Blackbirds. Princeton University Press, Princeton, N.J.

Orians, G. H. In press. Site characteristics favoring invasions. In H. A. Mooney, ed. Ecology of Biological Invasions in North America and Hawaii. Springer, New York.

Orians, G. H., and O. T. Solbrig, eds. 1977. Convergent Evolution in Warm Deserts. Dowden, Hutchinson & Ross, Stroudsburg, Pa.

Orians, G. H., and M. F. Willson. 1964. Interspecific territories of birds. Ecology 17:736-745 .

Paine, R. T. 1966. Food web complexity and species diversity. Am. Nat. 100:65-75.

Paine, R. T. 1969. The Pisaster-Tegula interaction: Prey patches, predator food preference, and intertidal community structure. Ecology 50:950-961 .

Paine, R. T. 1974. Intertidal community structure: Experimental studies on the relationship between a dominant competitor and its principal predator. Oecologia 15:93-120 .

Paine, R. T. 1980. Food webs: Linkage, interaction strength and community infrastructure. J. Anim.

Ecol. 49:667-685 .

Paine, R. T. 1981. Truth in ecology. Bull. Ecol. Soc. Am. 62:256-258.

Paine, R. T. 1984. Ecological determinism in the competition for space. Ecology 65:1339-1348 .

Paine, R. T., and S. A. Levin. 1981. Intertidal landscapes: Disturbance and the dynamics of pattern.

Ecol. Monogr. 51:145-198 .

Park, T. 1948. Experimental studies of interspecies competition. I. Competition between populations of the flour beetles Tribolium confusum Duval and Tribolium castaneum Herbst. Ecol.

Monogr. 18:265-308 .

Parrish, R. H., and A. D. MacCall. 1978. Climatic Variation and Exploitation in the Pacific Mackerel Fishery . Fish Bull. 167. California Department of Fish and Game, La Jolla.

Partridge, L. 1978. Habitat selection. Pp. 351-376 in J. R. Krebs and N. B. Davies, eds. Behavioral Ecology. Sinauer Associates, Sunderland, Mass.

Patrick, R., and R. R. Kiry. 1976. Estuarine surveys, biomonitoring and bioassays. Department of Limnology of the Academy of Natural Sciences, Philadelphia.

Patrick, R., B. Crum, and J. Coles. 1967. Temperature and manganese as determining factors in the presence of diatom or blue-green algal floras in streams. Proc. Natl. Acad. Sci. U.S.A. 64:472-478.

Peakall, D. B. 1975. Physiological effects of chlorinated hydrocarbons on avian species. Pp. 343-360 in R. Hauge and V. Freed, eds. Environmental Dynamics of Pesticides. Plenum, New York.

Perkins, J. H. 1982. Insects, Experts, and the Insecticide Crisis: The Quest for New Pest Management Strategies. Plenum, New York.

Peterman, R. M. 1978. Testing for density-dependent marine survival in marine salmonids. J. Fish.

Res. Bd. Can. 35:1434-1450.

Peterman, R. M. 1984. Interaction among sockeye salmon in the Gulf of Alaska. Pp. 187-199 in W.

G. Pearcy, ed. The Influence of Ocean Conditions on the Production of Sockeye Salmonids. Oregon State University Sea Grant Communications, Corvallis.

Peters, W. 1984. New answers through chemotherapy? Experientia 40:1351-1357.

Peters, W. In press. Resistance to antiparasitic drugs and its prevention. Parasitology.

Pianka, E. R. 1966 . Latitudinal gradients in species diversity. Am. Nat. 100:33-46 .

Pickett, S. T. A., and J. N. Thompson. 1978. Patch dynamics and the design of nature reserves. Biol.

Conserv. 13:27-37 .

Pimentel, D., D. Andow, R. Dyson-Hudson, D. Gallahan, S. Jacobson, M. Irish, S. Kroop, A. Moss, I. Schreiner, M. Shepard, T. Thompson, and B. Vinzant. 1980. Environmental and social costs of pesticides: A preliminary assessment. Oikos 34:126-140 .

Pimm, S. L. 1979. The structure of food webs. Theor. Pop. Biol. 16:144-158.

Pimm, S. L. 1980. Food web design and the effect of species deletion. Oikos 35:139-147 .

Pimm, S. L. 1982. Food webs. Chapman and Hall, London.

Pimm, S. L. 1984. The complexity and stability of ecosystems. Nature 307:321-326.

Policansky, D. 1983. Size, age and demography of metamorphosis and sexual maturation in fishes.

Am. Zool. 23:57-63 .

Porter, J. W., J. D. Woodley, G. J. Smith, J. E. Neigel, J. F. Battey, and D. G. Dallmeyer. 1981.

Population trends among Jamaican reef corals. Nature 294:240-250 .

Preston, F. W. 1960. Time and space and the variation of species. Ecology 29:254-253.

Price, P. W. 1980. Evolutionary Biology of Parasites. Princeton University Press, Princeton, N.J.

Pyke, G. H., H. R. Pulliam, and E. L. Charnov. 1977. Optimal foraging: A selective review of theory and tests. Q. Rev. Biol. 52:137-154.

Ratcliffe, P. R. 1984. Population dynamics of red deer (Cervus elephas L.) in Scottish commercial forests. Proc. R. Soc. Edinburgh 82B:291-302.

Recher, H. F. 1969. Bird species diversity and habitat diversity in Australia and North America.

Am. Nat. 103:75-80.

Reckhow, K. H. 1979. Empirical lake models for phosphorus: Development, applications, limitations and uncertainty. Pp. 193-221 in D. Scavia and A. Robertson, eds. Perspectives in Lake Ecosystem Modeling. Ann Arbor Sciences, Ann Arbor, Mich.

Reif, A. E. 1984. Synergism in carcinogenesis. J. Natl. Cancer Inst. 73:25-39 .

Rhoades, D. F. 1979. Evolution of plant chemical defense against herbivores. Pp. 3-54 in G. A.

Rosenthal and D. N. Janzen, eds. Herbivores: Their Interaction with Secondary Plant Metabolites. Academic Press, New York.

Rhoades, D. F. 1982. Response of alder and willow to attack by tent caterpillars and fall webworms:

Evidence for pheromonal sensitivity of willows. Pp. 55-68 in P. Hedin, ed. Plant Resistance to Insects. ACS Symposium Series 208. American Chemical Society, Washington, D.C.

Rhoades, D. F. 1983. Offensive-defensive interactions between herbivores and plants: Their relevance in herbivore population dynamics and ecological theory. Am. Nat. 125:205-238 .

Rhoades, D. F., and R. G. Cates. 1976. Toward a general theory of plant antiherbivore chemistry.

Rec. Adv. Phytochem. 10:168-213 .

Rice, E. L. 1974. Allelopathy. Academic Press, New York.

Ricker, W. E. 1963. Big effects from small causes: Two examples from fish population dynamics. J.

Fish. Res. Bd. Can. 20:257-264 .

Ricker, W. E. 1981. Changes in the average size and average age of Pacific salmon. Can. J. Fish.

Aquat. Sci. 38:1636-1656 .

Ricklefs, R. E. 1979. Ecology. 2nd ed. Chiron Press, New York.

Ricklefs, R. E., and K. O'Rourke. 1975. Aspect diversity in moths: A temperate-tropical comparison. Evolution 29:313-324 .

Robinson, R. W, H. M. Munger, T. W. Whitaker, and G. W. Bohn. 1976. Genes of the Cucurbitaceae. Hortoscience 11:554-568 .

Romesburg, H. C. 1981. Wildlife sciences: Gaining reliable knowledge. J. Wildl. Manage. 45:293-313.

Rosenberg, D. M., V. H. Resh, et al. 1981. Recent trends in environmental impact assessment. Can.

J. Fish. Aquat. Sci. 38:591-624.

Rosenzweig, M. L. 1972. Stability of enriched aquatic ecosystems. Science 175:562-565.

Roth, R. R. 1976. Spatial heterogeneity and bird species diversity. Ecology 57:773-782.

Ryan, C. A., and T. R. Green. 1974. Proteinase inhibitors in natural plant protection. Rec. Adv.

Phytochem. 8:123-140.

Ryman, N., R. Baccus, C. Reuterwall, and M. Smith. 1981. Effective population size, genetic interval, and potential loss of genetic variability in game species under various hunting regimes. Oikos 26:257-266.

Salisbury, F. B. 1942. The Reproductive Capacities of Plants. George Bell & Sons, Ltd., London.

Sanders, F. S., S. M. Adams, L. W. Barnthouse, J. M. Giddings, E. E. Huber, K. D. Kumar, D. Lee, B. Murphy, G. W. Suter, and W. Van Winkle, eds. 1980. Strategies for Ecological Effects Assessment at DOE Energy Activity Sites. Environmental Sciences Division Publ. 1639.

ORNL/TM-6783. Oak Ridge National Laboratory, Oak Ridge, Tenn.

Schindler, D. W. 1974. Eutrophication and recovery in experimental lakes: Implications for lake management. Science 184:897-898.

Schindler, D. W., E. J. Fee, and T. Ruszczynski. 1978. Phosphorus input and its consequences for phytoplankton standing crop and production in the Experimental Lakes Area and in similar lakes. J. Fish. Res. Bd. Can. 35:190-196.

Schoener, T. W. 1968. Sizes of feeding territories among birds. Ecology 49:123-141.

Schoener, T. W. 1971. Theory of feeding strategies. Annu. Rev. Ecol. Syst. 2:369-404.

Schoener, T. W. 1983. Field experiments on interspecific competition. Am. Nat. 122:240-285.

Schonewald-Cox, C. 1983. Conclusions: Guidelines to management: A beginning attempt. Ch. 25 in C. Schonewald-Cox, S. M. Chambers, B. MacBryde, and L. Thomas, eds. Genetics and Conservation: A Reference for Managing Wild Animal and Plant Populations. Benjamin/ Cummings, Menlo Park, Calif.

Schonewald-Cox, C., S. M. Chambers, B. MacBryde, and L. Thomas, eds. 1983. Genetics and Conservation: A Reference for Managing Wild Animal and Plant Populations. Benjamin/ Cummings, Menlo Park, Calif.

Schultz, J. C. 1983a. Habitat selection and foraging tactics of caterpillars in heterogeneous trees. Pp. 61-90 in R. F. Denno and M. S. McClure, eds. Variable Plants and Herbivores in Natural and Managed Systems. Academic Press, New York.

Schultz, J. C. 1983b. Impact of variable plant defensive chemistry on susceptibility of insects to natural enemies. Pp. 37-54 in P. A. Hedin, ed. Plant Resistance to Insects. ACS Symposium Series 208. American Chemical Society, Washington, D.C.

Schultz, J. C., and I. T. Baldwin. 1982. Oak leaf quality declines in response to defoliation by gypsy moth larvae . Science 217:149-151.

Schwabe, C. 1984. Animals as monitors of the environment. Pp. 562-578 in Veterinary Medicine and Human Health. 3rd ed. Williams and Wilkins, Baltimore, Md.

Scruggs, C. G. 1975. The Peaceful Atom and the Deadly Fly. Jenkins, Austin, Tex.

Searcy, W. A. 1982. The evolutionary effects of mate selection. Annu. Rev. Ecol. Syst. 13:57-85.

Seber, G. A. F. 1982. The Estimation of Animal Abundance and Related Parameters. Griffin, London.

Selander, R. K. 1983. Evolutionary consequences of inbreeding. Pp. 201-215 in S. C. Schonewald- Cox, S. M. Chambers, B. MacBryde, and L. Thomas, eds. 1983. Genetics and Conservation: A Reference for Managing Wild Animal and Plant Populations. Benjamin/ Cummings, Menlo Park, Calif.

Shaffer, M. L. 1981. Minimum population sizes for species conservation. BioScience 31:131-134.

Sharma, R. K., J. D. Buffington, and J. T. McFadden, eds. 1976. Proceedings of a Workshop on the Biological Significance of Environmental Impacts. NR-CONF-002. U.S. Nuclear Regulatory Commission, Washington, D.C.

Sharp, J. M., S. G. Appan, M. E. Bender, T. L. Linton, D. J. Reisch, and C. H. Ward. 1979. Natural variability of biological community structure as a quantitative basis for ecological impact assessment. Pp. 257-284 in Proceedings of the Ecological Damage Assessment Conference. Society of Petroleum Industry Biologists, Los Angeles.

Shepherd, J. G. 1984. The availability and information content of fisheries data. Pp. 95-109 in R. M.

May, ed. Exploitation of Marine Communities. Dahlem Konferenzen. Springer, Berlin.

Shugart, H. H., and D. C. West. 1980. Forest succession models. BioScience 30:308-313.

Silliman, R. P. 1975. Selective and unselective exploitation of experimental populations of Tilapia mossambica. Fish. Bull. 73:495-507.

Simberloff, D. In press. Ecological communities. In McGraw-Hill Encyclopedia of Science and Technology. 6th ed. McGraw-Hill, New York.

Simberloff, D., and L. G. Abele. 1976. Island biogeography theory and conservation practice.

Science 191:285-286.

Simenstad, C. A., J. A. Estes, and K. W. Kenyon. 1978. Aleuts, sea otters, and alternate stable-state communities. Science 200:403-411 .

Sissenwine, M. P. 1984. Why do fish populations vary? Pp. 59-94 in R. M. May, ed. Exploitation of Marine Communities. Dahlem Konferenzen. Springer, Berlin.

Skutch, M. M., and R. T. N. Flowerdeu. 1976. Measurement techniques in environmental impact assessment. Environ. Conserv. 3:209-217 .

Slawson, G. C., Jr., and B. C. Marcy, Jr. 1976. Evaluation of effects of multiple power plants on a river ecosystem. Ch. 17-3 in International 123 Conference on Environmental REFERENCES 131
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved.

Sensing and Assessment. Vol. 1. Institute of Electrical and Electronics Engineers, New York.

Smith, C. E. G. 1982. Practical problems in the control of infectious diseases. Pp. 177190 in R. M.

Anderson and R. M. May, eds. Population Biology of Infectious Diseases. Dahlem Konferenzen. Springer, Berlin.

Snaydon, R. W. 1984. Plant demography in an agricultural context. Pp. 369-407 in R. Dirzo and J.

Sarukhan, eds. Perspectives on Plant Population Ecology. Sinauer Associates, Sunderland, Mass.

Soulé, M. E. 1980. Thresholds for survival: Maintaining fitness and evolutionary potential. Pp. 151-169 in M. E. Soul é and B. A. Wilcox, eds. Conservation Biology. Sinauer Associates, Sunderland, Mass.

Soulé, M. E., and B. A. Wilcox, eds. 1980. Conservation Biology. Sinauer Associates, Sunderland, Mass.

Sousa, W. P. 1984. The role of disturbance in natural communities. Annu. Rev. Ecol. Syst. 15:353-391.

Southwood, T. R. E., R. M. May, M. P. Hassell, and G. R. Conway. 1974. Ecological strategies and population parameters. Am. Nat. 108:791-804 .

States, J. B., P. T. Haug, T. G. Shoemaker, L. W. Reed, and E. B. Reed. 1978. A Systems Approach to Ecological Baseline Studies. FWS/OBS-78/21. Fish and Wildlife Service, U.S.

Department of the Interior, Fort Collins, Colo.

Steams, S. C. 1976. Life-history tactics: A review of the ideas. Q. Rev. Biol. 51:3-47 .

Steele, J. H. 1984. Kinds of variability and uncertainty affecting fisheries. Pp. 245-262 in R. M.

May, ed. Exploitation of Marine Communities. Dahlem Konferenzen. Springer, Berlin.

Steele, J. H., and E. W. Henderson. 1984. Modeling long-term fluctuations in fish stocks. Science 224:985-987.

Strong, D. R. Jr. 1984. Exorcising the ghost of competition past: Phytophagous insects. Ch. 11 in D.

R. Strong, Jr., D. Simberloff, L. G. Abele, and A. B. Thistle, eds. Ecological Communities:

Conceptual Issues and the Evidence. Princeton University Press, Princeton, N.J.

Strong, D. R., Jr., L. A. Szyska, and D. S. Simberloff. 1979. Tests of community-wide character displacement against null hypotheses. Evolution 33:897-913.

Strong, D. R., Jr., D. Simberloff, L. G. Abele, and A. B. Thistle, eds. 1984. Ecological Communities: Conceptual Issues and the Evidence. Princeton University Press, Princeton, N.J.

Suchanek, T. H. 1979. The Mytilus californianus Community: Studies on the Composition, Structure, Organization, and Dynamics of a Mussel Bed. Ph.D. thesis, University of Washington, Seattle.

Suter, G. W., II. 1981. Commentary: Ecosystem theory and NEPA assessment. Bull. Ecol. Soc. Am. 62:186-192.

Suter, G. W., II. 1982. Terrestrial perturbation experiments for environmental assessment. Environ.

Manage. 6:43-54.

Thomas, J. W., ed. 1979. Wildlife Habitats in Managed Forests: The Blue Mountains of Oregon and Washington. Agriculture Handbook 553. Forest Service, U.S. Department of Agriculture, Washington, D.C.

Thompson, J. N. 1982. Interaction and Coevolution. John Wiley & Sons, New York.

Tilman, D. 1982. Resource Competition and Community Structure. Princeton University Press, Princeton, N.J.

Tinbergen, J. M. 1981. Foraging decisions in starlings (Sturnus vulgaris L.). Ardea 69:1-67.

Turner, R. M., S. M. Alcom, and S. Hope. 1969. Mortality of transplanted saguaro seedlings.

Ecology 50:835-844.

Underwood, A. J., and E. J. Denley. 1984. Paradigms, explanations, and generalizations in models for the structure of intertidal communities on rocky shores . Pp. 151-180 in D. R. Strong, Jr., D. Simberloff, L. G. Abele, and A. B. Thistle, eds. Ecological Communities:

Conceptual Issues and the Evidence. Princeton University Press, Princeton, N.J.

U.S. Environmental Protection Agency. 1980. Acid Rain. U.S. EPA Report 600/ 9-79-036 . Office of Research and Development, U.S. Environmental Protection Agency, Washington, D.C.

U.S. Fish and Wildlife Service. 1984. Endangered classification proposed for four fishes in Southeast and Utah. Endang. Species Tech. Bull. 9(8):1, 10-11.

Van Horne, B. 1983. Density as a misleading indicator of habitat quality. J. Wildl. Manage. 47:893-901 .

Velich, I., and L. Satyko. 1974. Possibilities of increasing earliness in melon by means of the genetics of sex and floral biology. Agrartudomanyi Kozlemenyek 33:459-472 .

Vollenweider, R. A. 1969. M öglichkeiten und Grenzen element ärer Modelle der Stoffbilanz von Seen. Arch. Hydrobiol. 66:1-36 .

Vollenweider, R. A. 1975. Input-output models with special reference to the phosphorus loading concept in limnology. Schweiz. Z. Hydrol. 37:53-84 .

Vollenweider, R. A. 1976. Advances in defining critical loading levels for phosphorus in lake eutrophication. Mem. Ist. Ital. Idrobiol. 33:53-83.

Walters, C. J. 1975. An interdisciplinary approach to development of watershed simulation models.

J. Fish. Res. Bd. Can. 32:177-195.

Walters, C. J. 1984. Managing fisheries under biological uncertainty. Pp. 263-274 in R. M. May, ed.

Exploitation of Marine Communities. Dahlem Konferenzen. Springer, Berlin.

Walters, C. J. In press. Adaptive Management of Renewable Resources. Macmillan, New York.

Walters, C. J., and R. Hilborn. 1976. Adaptive control of fishing systems. J. Fish. Res. Bd. Can. 33:145-159
Ward, D. V. 1978. Biological Environmental Impact Studies: Theory and Methods. Academic Press, New York.

Watt, K. E. F. 1964. Comments on fluctuations of animal populations and measures of community stability. Can. Entomol. 96:1434-1442.

Weinstein, L., and D. McCune. 1970. Field surveys, vegetation monitoring. Pp. G1-4 in J. S.

Jacobson and A. C. Hill, eds. Recognition of Air Pollution Injury to Vegetation: a Pictorial Atlas. Air Pollut. Control Assoc., Pittsburgh.

Wells, P. V. 1965. Scarp woodlands, transplanted grassland soils, and the concept of grassland climate in the Great Plains region. Science 148:246-249 .

Werner, G. E., and D. J. Hall. 1974. Optimal foraging and the size selection of prey by the bluegill sunfish (Lepomis macrochirus). Ecology 55:1042-1052 .

Westoby, M. 1978. What are the biological bases of varied diets? Am. Nat. 112:627-631 .

Wheelwright, N. T. 1983. Fruits and the ecology of resplendent quetzals. Auk 100:286-301.

Wilcove, D. S. 1985. Nest-predation in forest tracts and the decline of songbirds. Ecology 66:1211-1214 .

Willis, E. O. 1974. Populations and local extinctions of birds on Barro Colorado Island, Panama.

Ecol. Monogr. 44:153-169 .

Wooster, W. S., ed. 1983. From Year to Year: Interannual Variability of the Environment and Fisheries in the Gulf of Alaska and the Eastern Bering Sea. University of Washington, Washington Sea Grant, Seattle.

Zar, J. H. 1976. Statistical significance and biological significance of environmental impacts. Pp. 285-293 in R. K. Sharma, J. D. Buffington, and J. T. McFadden, eds. Proceedings of a Workshop on the Biological Significance of Environmental Impacts. NR-CONF-002. U.S.

Nuclear Regulatory Commission, Washington, D.C.

Zaret, T. M. 1980. Predation and Freshwater Communities. Yale University Press, New Haven, Conn.

II
Selected Case Studies
The case studies in this part are presented as examples of how ecological knowledge has been used in planning and carrying out problem-solving efforts, including how the knowledge has been limited and how it has been adapted to specific problems. The case studies presented here are referred to frequently in Part I . Their selection and preparation are described in the Introduction. The Committee is responsible for the introductory statements and comments. 135
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. 136
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. 12 North Pacific Halibut Fishery Management
The goal of management of a commercially important resource is to resolve the conflict between maintenance and exploitation. Unfettered exploitation often leads to disappearance of the resource, as in the cases of whales, passenger pigeons, and buffaloes. But management that is too conservative leads to inefficient use of the resource. One approach to the conflict in fishery management has been to use the idea of some maximal sustainable yield that can be taken from the fishery. Unfortunately, owing largely to unpredictable variations in the environment, maximal yields are usually not sustainable for very long. The management of the Pacific halibut by the International Pacific Halibut Commission (IPHC) is an example of the responsive (adaptive) approach to management. Additionally, in this case there has been strong commitment to an understanding of the biology of the species, and the activities of IPHC have been superbly documented, as has the biological information obtained.

Case Study
DAVID POLICANSKY, National Research Council, Washington, D.C.

INTRODUCTION
The International Pacific Halibut Commission (IPHC) was established by a convention between the United States and Canada in 1923, which was revised in 1930, 1937, and 1953 (Bell, 1969) and again in 1983. (The commission was originally named the International Fisheries Commission and renamed in 1953; I use the abbreviation IPHC throughout.) The purpose of the Commission was to provide a mechanism for joint management of the Pacific halibut (Hippoglossus stenolepis Schmidt), whose abundance had been declining up to 1923. The management objective, originally the maximization of sustainable yield, was changed in 1983 to the optimization of sustainable yield (R. B.

Deriso, personal communication); the idea of ''optimal" sustained yield includes the "quality of the fishery," as well as the weight of the harvest (Roedel, 1975).

The primary ecological problem has been stock assessment; some related issues have also been important.

The Commission originally had four members and now has six, drawn from industry and government. Half the members are Canadian and half are from the United States. The Commission is supported by a scientific staff, headquartered in Seattle, with a full-time director appointed by the commissioners; a Conference Board composed of fishermen and vessel owners, which makes recommendations to the Commission with respect to regulations; and an Advisory Group composed of 14 fishermen, dealers, and fish-processors, half of whom are selected by the Conference Board and half by the Halibut Association of North America. Members of the Board participate in the Commission's meetings as observers.

The history of IPHC can be traced in its many thorough reports and in other sources; the following account relies especially on Thompson and Freeman (1930) and Bell (1981). Until 1888, the Indians had conducted an important halibut fishery, with a catch probably exceeding 3 million pounds a year. The Northern Pacific Railroad was completed in 1888, and the Canadian Pacific Railroad in 1892. These new railroads profoundly affected the development of the halibut fishery by providing ready access to large markets for halibut in the East. Attracted by the new profitability of the Pacific halibut fishery, men and boats from the eastern fishery arrived, and the annual catch increased rapidly. The next 30 years saw depletion of known halibut banks, maintenance of yield by discovery and exploitation of new banks, and technical innovations. By the time of World War I, it was obvious NORTH PACIFIC HALIBUT FISHERY MANAGEMENT 138
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. that, from an economic point of view, the banks were being overexploited. For this reason and because winter fishing was dangerous and expensive, a winter closed season was desirable to the fishermen.

Thus, economic pressure was the impetus for the birth of IPHC. The institution of a closed season required international regulation, which required a treaty. Many attempts to enact fishery treaties between the United States and Canada had failed, because they included both conservation and unrelated economic considerations (such as reciprocal port privileges). However, when the halibut treaty was finally ratified in 1924, it contained only a conservation measure (a closed season) and provision for the establishment of IPHC.

An IPHC report to the two governments in 1928 offered specific recommendations for the development of the fishery and the conservation of the resource. The report detailed the decline in abundance of halibut in all the areas where they were exploited. It recommended establishment of management areas in each of which the total catch of halibut could be reduced until the yield was stable, with the amount of the reduction being responsive to the catch per unit of effort (CPUE); closure of the nursery grounds; prohibition of the use of destructive gear; extension of the closed season; provision for future modifications of the closed season; and licensing of all vessels for statistical and other purposes. These recommendations, based on scientific activities of IPHC under the direction of W. F. Thompson (including tagging experiments, analyses of catch statistics, and hydrographic studies), resulted in the Halibut Convention of 1930, which gave IPHC broad regulatory authority.

BIOLOGICAL BASIS OF MANAGEMENT
The overriding issue here is the ecological problem of stock assessment, i.e., knowing how many fish are in the sea. If that is known and the catch is known, then the effect of fishing on the fish stocks can be determined. Other aspects of the biology of the halibut have also been studied and are discussed below.

Assessment of Stock
Knowledge of stock abundance is desirable if a fishery is to be monitored.

The assessment of fish stocks is perhaps the major fishery problem and is often intractable. IPHC has relied heavily on CPUE as an index of stock abundance (Skud, 1978); but the catch has been measured in biomass (weight), rather than numbers of fish. Number of fish is also important, because declining numbers can be masked by increased growth rate if only information on biomass is used (Schmitt and Skud, 1978). In part for that reason, mark-and-recapture (tagging) experiments have been done.

In addition to the estimates just mentioned, a number of models have been used (they are discussed below). In general, these models have provided estimates of abundance relative to that at some specified time. As long as the relativity coefficients vary by only small amounts, knowledge of absolute abundance is not needed for successful fishery management, although it would be intellectually satisfying.

Relationship Between Stock and Recruitment
The relationship between stock and recruitment is difficult to quantify — the density-related form described by Ricker's equation (Ricker, 1954) is seldom accurate enough to be used in management. The reason for this is probably that recruitment is determined largely by larval survival, and larval survival seems to be very much affected by environmental conditions (for recent reviews, see Beverton, 1984; Hunter, 1983; Sissenwine, 1984; and Wooster, 1983). Although there have been attempts to relate stock and recruitment in halibut (e.g., Schmitt and Skud, 1978), the relationship is not directly used in managing the halibut fishery. It appears as a term in some of the models discussed below, but they usually assume that recruitment is independent of stock size.

Catchability
The catchability of fish with a given type of fishing gear is a measure of the efficiency of that gear. If the catchability coefficient were a constant for a particular gear, then catch would depend on fishing effort and fish abundance only. But catchability varies with fish behavior, as well as with gear type, and must be known if comparisons of CPUEs with different types of gear are to be valid. Such comparisons are required for tracking the history of a fishery over periods in which different types of gear were used; in other words, they are necessary for an understanding of the effect of fishing on stock abundances.

Some of the difficulties associated with catchability estimates are discussed below.

The commonest gear in the halibut fishery is the setline, and the functional unit of this longline gear is the "skate" — an 1,800-foot length of line, gangions (branch lines), and hooks. The hooks, regularly spaced on a given line, are 9-26 feet apart. The catch not only is proportional to the number of hooks, but varies with their spacing; therefore, neither the skate nor a number of hooks is a good unit of fishing effort. Changes in these variables complicate analyses of historical trends in CPUE (Skud, 1978). Hook shape also influences catchability, and a much more efficient hook has recently been introduced into the halibut fishery (Deriso, personal communication). Other variables influence catchability as well (Hoag and
NORTH PACIFIC HALIBUT FISHERY MANAGEMENT 140
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved.

Deriso, 1984; IPHC, 1982; Myhre and Quinn, 1984). All these variables affect estimates of CPUE and hence of stock abundance, and considerable effort has been devoted to the understanding of their effects (Hamley and Skud, 1978; Myhre and Quinn, 1984; Skud, 1975, 1978).

Migration Patterns
The halibut fishery extends over a very large area, from the Oregon coast to the Bering Sea. For intelligent management, it is obviously important to know whether this area is occupied by one large, intermingling stock or by a number of independent stocks. The fishery is divided into management areas; the largest catches come from area 2 (Cape Spencer, in the middle of the Alaska panhandle, south to Oregon) and area 3 (Cape Spencer to the western end of the Aleutian Islands). Tagging experiments before the mid-1950s suggested that adult halibut from the two areas did not mix, and studies of the distributions of eggs and larvae coupled with the use of drift bottles suggested that there was not much interchange of those stages either.

Skud (1977a) re-examined the data in the early life stages and considered more recent data on later stages. He concluded that, although there is little drift of the early stages from area 3 to area 2, there is substantial drift northward from area 2 to area 3, much migration of juvenile stages between the two areas, and seasonal movement of adults.

The longline fishery currently operates only in the summer and preferentially takes fish that return to the same grounds each summer. Because of this, Skud concluded that the areas could reasonably be managed separately, if the fishery continued to operate only in the summer. He did not comment on the implications of the intermingling of the earlier life stages, but noted that a trawl fishery operates year round and preferentially takes juvenile halibut, which migrate extensively. For these reasons, the areas should not be managed independently (Skud, 1977a).

Natural and Fishing Mortality
It is extremely difficult in practice to separate causes of mortality in a fished population. The estimates of natural and fishing mortality are closely related and are important in models of stock abundance (as discussed later).

Growth Rates
Growth rates have been studied by IPHC for a variety of management purposes (e.g., Schmitt and Skud, 1978), most recently for use in mathematical models of stock abundances. Because catch in this fishery is measured NORTH PACIFIC HALIBUT FISHERY MANAGEMENT 141
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. in biomass, rather than numbers of fish, a knowledge of growth rates is essential to an understanding of the population dynamics of the resource.

Growth rates have been estimated by weighing fish and determining their ages by otolith counts. The growth rates of halibut have varied with location and over time, with marked increases since the 1950s (Schmitt and Skud, 1978).

Both natural environmental changes and density-dependent responses to fishing appear to have influenced growth rates (Schmitt and Skud, 1978; Southward, 1967).

GENERAL APPROACH TO MANAGEMENT
The general approach, as indicated earlier, has been that an understanding of the population dynamics of halibut is required for successful management of the fishery. The Pacific halibut was studied before the existence of IPHC, to a large extent by W. F. Thompson, who became its first director. The 1923 convention required the Commission to "make a thorough investigation into the life history of the Pacific halibut." The aim of the Commission "has been to establish beyond doubt the actual condition of the fishery" by "close adherence to the facts and avoidance of unsupported theory" (Babcock et al., 1931).

Thompson was familiar with the already large European literature on fishery biology and was probably guided in his studies by that literature.

The specific objective of IPHC's program, stock assessment, has not yet been reached, although relative abundances are used successfully for management.

Boundaries
The boundaries, originally set by the range of the boats and the demand for fish, represented the range of fishing activities. Fishing is now limited in the south by the abundance of halibut, and in the north and west by deep water and low abundance. The setting of the jurisdictional boundaries by the formation of IPHC required many years of difficult negotiations (Bell, 1981; Thompson and Freeman, 1930).

Monitoring
The nature of this case has made monitoring an essential part of the scientific program from its inception. There is continuous monitoring of many biological parameters, including numbers, biomass, growth rate, fecundity, and catchability. Information from the commercial catch and NORTH PACIFIC HALIBUT FISHERY MANAGEMENT 142
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. results of research sampling of the fish are both used. The information derived from this monitoring is applied in regulating the fishery.

Cumulative Effects
The population has changed in a number of ways since 1930. The fish grow faster, fecundity increased up to about 1960 and then declined, and abundance declined and recently increased. The sex ratio might also have become more female-biased (Schmitt and Skud, 1978). These changes could reflect the results of fishing, which is selective (e.g., Myhre, 1969), and they might therefore have a genetic basis, at least in part. If the changes are the result of evolution under fishing pressure, then fishing has had an important cumulative effect.

SOURCES OF KNOWLEDGE
Ecological Facts
All the ecological facts that IPHC used in managing the fishery were obtained through study of the halibut fishery itself and from persons involved with it. No useful facts were obtained from the published literature, data banks and files, or other stored information. However, general knowledge in fishery biology includes many ecological theories, principles, and models that can be found in textbooks of ecology or fishery biology, and some were tested on the halibut fishery.

A great deal of interesting and new ecological information has been gathered by IPHC, but it appears that only total catch, CPUE, catch-at-age, growth rates, and migration information from tagging were actually used in managing the fishery. Other facts— such as egg size, fecundity, and age and size at metamorphosis— could not be used, because there was no practical basis for incorporating them. Although other available data were not used and there has been at least some success in managing the fishery, additional ecological information would probably improve management. But the additional information would probably come from improvement of the types of information already used, e.g., stock numbers and biomass, recruitment, and catchability.

Theory and General Principles
Of the theoretical aspects of fishery management that are actually used in management of the halibut fishery, the law of accumulated stock is perhaps the most fundamental. The crux of this idea is that yield is some NORTH PACIFIC HALIBUT FISHERY MANAGEMENT 143
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. function of stock size. If the stock is too small to provide adequate yield, and "if a greater yield is desired, the only possible way to obtain it is not by increasing the amount of gear fished but by decreasing it" (Thompson, 1937). Other theories and principles, some of them implicit, are discussed in the next section.

Specific Models of Stock Abundance
Several models have been used in the attempt to assess stock abundances.

The more important include the use of CPUE as a measure of stock abundance; constant or random stock-recruitment models; von Bertalanffy and other growth- rate models; models of mortality rate ranging from constant mortality to variable, age-specific (usually fishing) mortalities; population age-structure models; and such complex models as cohort analysis, delay-difference models, and nonlinear catch models. Some of these models were developed in connection with the halibut fishery, but most were taken from the general fisheries literature or developed with no specific fishery in mind. Some of the important ones are briefly discussed below.
CPUE. This method uses CPUE as an index of biomass (not numbers) each year and assumes that catchability is constant. Because gear can become saturated (there is a limit to the proportion of hooks that will catch fish, however abundant the population is), CPUE can fail to increase even if the population increases. In addition, catchability varies, as discussed earlier, and gear does not catch halibut at random, but selects for different sizes of fish (Myhre, 1969). Thus, the relationship between biomass and numbers changes when the age composition of the population or the growth rate changes. Because increased growth can counteract decreased abundance, the use of biomass instead of numbers can be misleading (Schmitt and Skud, 1978).
Cohort Analysis (Pope, 1972). In this method, very similar to virtual population analysis (Gulland, 1965), one starts with the number of living fish in the oldest year class and follows the year class back through preceding years, estimating abundances by using the natural mortality (M
i) and fishing mortality (Fi) for each year. The information needed includes an estimate of the ages of the fish caught in the preceding years. In general, natural mortality is assumed to be constant at around 0.2, and fishing mortality is estimated from catch information (Hoag and McNaughton, 1978). If natural mortality is constant from year to year, then fluctuations in CPUE are due to changes in abundance.
Delay-Difference Population Model (Deriso, 1980). This model was developed as a general fishery model, although it has been specifically NORTH PACIFIC HALIBUT FISHERY MANAGEMENT 144
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. applied to halibut. It is in effect a simulation model that takes into account starting biomass, a growth coefficient, catch, natural survival (1 - natural mortality), and a spawner-recruit function. The model is run with available information; the measurable variables include starting relative abundance or starting biomass, growth rates, and catch. As in cohort analysis, natural mortality has to be estimated (by tagging or other means) or assumed to be constant. The stock- recruitment relationship has defied elucidation and is assumed to be independent of density. The model is attractive, because it demonstrates the validity of using biomass of sexually mature fish, rather than numbers; the rationale is that fecundity is more closely proportional to the biomass of sexually mature (female) fish than to their numbers. The model also uses parameters whose values can be estimated independently of CPUE data.
Nonlinear Catch-Age (Catch at Age) Analysis. Developed by Pope (1974) and Doubleday (1976), this method estimates abundance by estimating relative year-class strength and cumulative fishing and natural mortalities. It takes into account the differential selectivity of gear for different age classes and the inconstancy of catchability. But it assumes that age selectivity is constant over time; if growth rates change, that assumption is not valid.

A comparison of the four models, which are not completely independent, indicates fairly close agreement among them in their estimates of stock abundance; the most reliable are cohort analysis and the delay-difference population model (Quinn et al., 1984).
Annual Surplus Production (ASP). ASP is defined as the amount of fish that can be taken in a year without changing the biomass of the stock over the course of the year. It is estimated by adding the catch in a given year to the annual change in stock biomass as estimated by one or more of the four methods previously described. Although ASP has been referred to as "equilibrium yield" (Schaefer, 1954), it is not necessary to view it that way, because neither the population size nor the catch need be at equilibrium.

Project as Experiment
The adaptive management approach to the halibut fishery implies an element of experimentalism of the form "let's try X and see if it works," where X is based on the scientific activities of IPHC. In fact, in the 1970s IPHC deliberately increased the catch to test estimates of maximal sustainable yield to demonstrate that the Bering Sea stock was fully utilized and therefore qualified for exemption from Japanese fishing. However,
NORTH PACIFIC HALIBUT FISHERY MANAGEMENT 145
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. there were no controls, and the great number of variables involved makes it impossible to evaluate the results with certainty.

ROLE OF ECOLOGICAL KNOWLEDGE
The main role of ecological knowledge in this case has been in management, and the most important piece of information has been the relative abundance of the stock. IPHC has based its regulations on stock assessments, derived at first only from CPUE and total catch and later from additional estimates. When the estimated stock size declined, the total allowable catch was decreased; when the estimated stock size increased, the total catch was allowed to increase.

To this day, it is uncertain how much of the decline in abundance up to 1923 was due to fishing mortality and how much the increase after 1930 was due to the regulations imposed by the Commission, although fishing and regulations probably had at least some effect (Skud, 1975). The primary uncertainty concerns estimates of stock abundance and CPUE, because the effect of changes in fishing gear is unknown. An alternative hypothesis to fishing mortality is the one of environmentally caused fluctuations advanced by Burkenroad (see Skud, 1975, for review).

Management areas were established in recognition of the differences in the fishery between different places: fish grow at different rates, spawn at different ages, and respond differently to fishing pressure in the different areas. Although more recent work has revealed much migration between the areas (Skud, 1977a), there is some ecological justification for the establishment of the management areas, and certainly a management justification.

Estimating fishing mortality from sources other than the directed longline fishery for halibut posed an assessment problem. At first, IPHC had to deal only with the effects of the directed fishery on the halibut stock. The closed seasons imposed were in winter, when very little other fishing took place. However, trawl fisheries and directed line fisheries for black-cod, Anoplopomafimbria (Pallas), became more common and resulted in incidental catches of halibut. "Incidental" catches were also made by foreign vessels from the late 1950s on.

Detection of unreported catches is essential to any rational fishery management. At first, IPHC reduced the legal catch, to maintain CPUE at an acceptable level without estimating the illegal catch. That is, legal fishing was reduced to counterbalance illegal, unreported fishing. In effect, the resource was being preserved for the outlaws. It thus became necessary for IPHC to estimate unreported catches, and it resorted to persuasion and a number of clever stratagems, including the use of undercover agents
NORTH PACIFIC HALIBUT FISHERY MANAGEMENT 146
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. and monitoring of the sales of halibut livers for comparison with the reported catch. Little biological sophistication is required; halibuts have one liver apiece, and there is a statistically valid ratio of liver weight to halibut weight. Skud (1977b) has described IPHC's regulations fully.

IPHC has used monitoring, biological information, and models in management of the halibut fishery. Regulations are based on biological information and on models using that information, and the population is monitored to ascertain their effects. The results of this monitoring lead to annual, or even more frequent, revisions of the regulations.

IPHC is aware of impediments to management that it is powerless to deal with (Skud, 1976). It has no control of such matters as limiting entry to the fishery, enforcement, price of the catch, and efficiency of harvest. However, these are not biological issues, although they exert strong influences on the biological issues.

ACKNOWLEDGMENTS
The International Pacific Halibut Commission staff, particularly R. B.

Deriso, has been extremely generous with information and advice. R. B. Deriso, R. M. May, and other members of the Committee on Applications of Ecological Theory to Environmental Problems have made many helpful comments on this paper.

REFERENCES
Babcock, J. P., W. A. Found, M. Freeman, and H. O'Malley. 1931. Report of the International Fisheries Commission. International Fisheries Commission, Seattle, Wash.

Bell, F. H. 1969. Agreements, Conventions and Treaties between Canada and the United States of America with Respect to the Pacific Halibut Fishery. Int. Pac. Halibut Comm. Rep. 50.

International Pacific Halibut Commission, Seattle, Wash.

Bell, F. H. 1981. The Pacific Halibut: The Resource and the Fishery. Alaska Northwest Publishing Co., Anchorage, Alaska.

Beverton, R. J. H., rapporteur. 1984. Dynamics of single species. Group report. Pp. 13-58 in R. M.

May, ed. Exploitation of Marine Communities. Dahlem Konferenzen. Springer, Berlin.

Deriso, R. B. 1980. Harvesting strategies and parameter estimation for an age-structured model.

Can. J. Fish. Aquat. Sci. 37:268-282.

Doubleday, W. G. 1976. A least squares approach to analyzing catch at age data. Int. Comm.

Northwest Atl. Fish. Res. Bull. 12:69-81.

Gulland, J. A. 1965. Manual of Methods for Fish Stock Assessment. FAO Manuals in Fisheries Science 4. U.N. Food and Agriculture Organization, Rome.

Hamley, J. M., and B. E. Skud. 1978. Factors affecting longline catch and effort: II. Hook-spacing.

Pp. 16-24 in Int. Pac. Halibut Comm. Sci. Rep. 64. International Pacific Halibut Commission, Seattle, Wash.

Hoag, S. H., and R. B. Deriso. 1984. Recent Changes in Halibut CPUE: Studies on Area NORTH PACIFIC HALIBUT FISHERY MANAGEMENT 147
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved.

Differences in Setline Catchability. Int. Pac. Halibut Comm. Sci. Rep. 71. International Pacific Halibut Commission, Seattle, Wash.

Hoag, S. H., and R. J. McNaughton. 1978. Abundance and Fishing Mortality of Pacific Halibut, Cohort Analysis, 1935-1976. Int. Pac. Halibut Comm. Sci. Rep. 65. International Pacific Halibut Commission, Seattle, Wash.

Hunter, J. R. 1983. Commentary: On the determinants of stock abundance. Pp. 11-16 in W. S.

Wooster, ed. From Year to Year: Interannual Variability of the Environment and Fisheries of the Gulf of Alaska and the Eastern Bering Sea. Washington Sea Grant, University of Washington, Seattle.

IPHC (International Pacific Halibut Commission). 1982. The Pacific Halibut: Biology, Fishery, and Management. Int. Pac. Halibut Comm. Tech. Rep. 16. International Pacific Halibut Commission, Seattle, Wash.

Myhre, R. J. 1969. Gear Selection and Pacific Halibut. Int. Pac. Halibut Comm. Rep. 51.

International Pacific Halibut Commission, Seattle, Wash.

Myhre, R. J., and T. J. Quinn II. 1984. Comparison of Efficiency of Snap Gear to Fixed Hook Setline Gear for Catching Pacific Halibut. Int. Pac. Halibut Comm. Sci. Rep. 69.

International Pacific Halibut Commission, Seattle, Wash.

Pope, J. G. 1972. An investigation of the accuracy of virtual population analysis using cohort analysis. Int. Comm. Northwest Atl. Fish. Res. Bull. 9:65-74.

Pope, J. G. 1974. A Possible Alternative Method to Virtual Population Analysis for the Calculation of Fishing Mortality from Catch at Age Data. Int. Comm. Northwest Atl. Fish. Res. Doc. 74/20. International Commission of the Northwest Atlantic Fisheries, Dartmouth, N.S.

Quinn, T. J., II, R. B. Deriso, S. H. Hoag, and R. J. Myhre. 1984. A summary of methods of estimating annual surplus production for the Pacific halibut fishery. Int. North Pac. Fish.

Comm. Bull. 42:73-81.

Ricker, W. E. 1954. Stock and recruitment. J. Fish. Res. Bd. Can. 11:559-623.

Roedel, P. M., ed. 1975. Optimum Sustained Yield as a Concept in Fisheries Management.

Proceedings of a Symposium, September 1974. Special Pub. 9. American Fisheries Society, Washington, D.C.

Schaefer, M. B. 1954. Some aspects of the dynamics of populations important to the management of commercial marine fisheries. Int. Am. Trop. Tuna Comm. Bull. 1:27-56.

Schmitt, C. C., and B. E. Skud. 1978. Relation of Fecundity to Long-Term Changes in Growth, Abundance and Recruitment. Int. Pac. Halibut Comm. Sci. Rep. 66. International Pacific Halibut Commission, Seattle, Wash.

Sissenwine, M. P. 1984. Why do fish populations vary? Pp. 59-94 in R. M. May, ed. Exploitation of Marine Communities. Dahlem Konferenzen. Springer, Berlin.

Skud, B. E. 1975. Revised Estimates of Halibut Abundance and the Thompson-Burkenroad Debate.

Int. Pac. Halibut Comm. Sci. Rep. 56. International Pacific Halibut Commission, Seattle, Wash.

Skud, B. E. 1976. Jurisdictional and Administrative Limitations Affecting Management of the Halibut Fishery. Int. Pac. Halibut Comm. Sci. Rep. 59. International Pacific Halibut Commission, Seattle, Wash.

Skud, B. E. 1977a. Drift, Migration, and Intermingling of Pacific Halibut Stocks. Int. Pac. Halibut Comm. Sci. Rep. 63. International Pacific Halibut Commission, Seattle, Wash.

Skud, B. E. 1977b. Regulations of the Pacific Halibut Fishery, 1924-1976 . Int. Pac. Halibut Comm.

Tech. Rep. 15. International Pacific Halibut Commission, Seattle, Wash.

Skud, B. E. 1978. Factors affecting longline catch and effort: I. General review. Pp. 5- NORTH PACIFIC HALIBUT FISHERY MANAGEMENT 148
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. 14 in Int. Pac. Halibut Comm. Sci. Rep. 64. International Pacific Halibut Commission, Seattle, Wash.

Southward, G. M. 1967. Growth of Pacific Halibut. Int. Pac. Halibut Comm. Rep. 43. International Pacific Halibut Commission, Seattle, Wash.

Thompson, W. F. 1937. Theory of the Effect of Fishing on the Stock of Halibut. Int. Fish. Comm.

Rep. 12. International Fisheries Commission, Seattle, Wash.

Thompson, W. F., and N. L. Freeman. 1930. History of the Pacific Halibut Fishery. Int. Fish.

Comm. Rep. 5. International Fisheries Commission, Seattle, Wash.

Wooster, W. S. 1983. On the determinants of stock abundance. Pp. 1-10 in W. S. Wooster, ed. From Year to Year: Interannual Variability of the Environment and Fisheries of the Gulf of Alaska and the Eastern Bering Sea. Washington Sea Grant, University of Washington, Seattle.

Committee Comment
The management of marine fisheries has not been distinguished by many great successes. In the United States, striped bass ( Roccus saxatilis ), haddock (Melanogrammus aeglefinus), and Atlantic halibut ( Hippoglossus hippoglossus) were once abundant on the East Coast (Smith, 1833), as were Pacific sardines (Sardinops sagax ) and chub mackerel ( Scomber japonicus ) in California (Murphy, 1966; Parrish and MacCall, 1978). They are now much reduced, although overfishing cannot be given all the blame (see Chapter 8 ). Fifty-five years after establishment of IPHC, and in the face of continuous, intensive harvesting, the Pacific halibut fishery is still viable and valuable.

It is not clear how much the survival of the halibut fishery is due to the activities of the Commission and how much is due to luck, but the intelligent approach to management taken by IPHC seems to have contributed. The idea of adaptive fishery management, which seems at least partly responsible for the success of IPHC, is not peculiar to that organization; it has been recommended specifically (e.g., Parrish and MacCall, 1978) and as a general approach (e.g., Walters and Hilborn, 1976).

From the beginning, the Commission's primary focus has been on the stock (population) size of the fishery. Regulations have been based on relative stock sizes; when the stock seems to have decreased, fishing pressure is reduced, and when the stock seems to have increased, fishing pressure is allowed to increase.

There has been no serious attempt to manage the fishery on any equilibrium or static basis, and that is a major strength of the management program. Because a responsive (adaptive) approach is used, management is not tied to an estimation procedure that could be wrong, and errors are not compounded from year to year. This is strikingly different from the management of striped bass on the East NORTH PACIFIC HALIBUT FISHERY MANAGEMENT 149
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved.

Coast of the United States, where all the regulations, strict as some are, are static and not directly tied to stock sizes.

Beginning with Thompson, the first director of IPHC, there has been a commitment to an understanding of the biology of the halibut. This is based on the premise that better understanding of the biology of the halibut can lead to better management. As a result of this commitment, there is an enormously detailed, long-term, and intelligently collected set of data on halibut biology.

Although it is possible to manage a fishery for short periods with minimal biological information (Shepherd, 1984), the halibut data have been put to good use. They have allowed the development of sophisticated models of the fishery, have made it possible to analyze the long-term effects of previous management practices (i.e., there has been excellent monitoring), and can be used to acquire new biological knowledge (i.e., the project has become a well-designed and well-executed research project). The documentation also allows this case to inform other attempts to manage fisheries.

The case study describes an excellent solution to the problem of managing in more than one jurisdiction, which makes many fish stocks particularly difficult to manage. The Fishery Conservation and Management Act of 1976 resulted in part from a recognition by the United States that the sharing of its marine fisheries by many diverse fishing nations made it very difficult to manage them. One of the Act's major provisions places almost all fisheries within 200 miles of the coast under the sole jurisdiction of the United States, and this appears to have resulted in an increase in the stocks of many fisheries (Finch, 1985). Similar legislation is now common elsewhere. The case also illustrates how difficult it is to achieve agreements and how strong the incentive for agreement must be.

References
Finch, R. 1985. Fishery management under the Magnuson Act. Mar. Policy 9:170-179.

Murphy, G. I. 1966. Population biology of the Pacific sardine ( Sardinops caerulea ). Proc. Calif.

Acad. Sci. 34:1-84.

Parrish, R. H., and A. D. MacCall. 1978. Climatic Variation and Exploitation in the Pacific Mackerel Fishery. Fish. Bull. 167. Calif. Department of Fish and Game, La Jolla, California.

Shepherd, J. 1984. The availability and information content of fisheries data. Pp. 95-109 in R. M.

May, ed. Exploitation of Marine Communities. Dahlem Konferenzen. Springer, Berlin.

Smith, J. V. C. 1833. A Natural History of the Fishes of Massachusetts; Embracing a Practical Essay on Angling. Allen and Ticknor, Boston.

Walters, C. J., and R. Hilborn. 1976. Adaptive control of fishing systems. J. Fish. Res. Bd. Can. 33:145-159.

NORTH PACIFIC HALIBUT FISHERY MANAGEMENT 150
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. 13 Vampire Bat Control in Latin America
To find an effective control agent that affects only the species of concern normally requires knowledge of the life history of the target species. The current control of vampire bats in Latin America rests on the following aspects of the biology of the bats: they are much more susceptible than cattle to the action of anticoagulants, they roost extremely close to each other, they groom each other, their rate of reproduction is low, they do not migrate, and they forage only in the absence of moonlight. These pieces of information were acquired in a search for control methods, and they yielded methods that are effective and that have minimal or no effect on nontarget species. The approach in this case was to find and use weak links in the life history of the animal in question. It also reflects an attempt to save money by comparing the cost and availabilities of various chemicals.

Case Study
G. CLAY MITCHELL, Denver Wildlife Research Center, Denver, Colorado

INTRODUCTION
When the Conquistadors arrived in Mexico in 1527, their horses were immediately attacked by the common vampire bat, Desmodus rotundus (Molina Solis, 1896). Over four centuries later, Bernardo Villa (1969) made the following statement after reviewing the vampire bat problem in Brazil: ''No notable success has been achieved in controlling bats of the family Desmodontidae, particularly the species Desmodus rotundus, in any of the countries of tropical America."
Of the three species of vampire bats, only the common vampire, the subject of this review, is economically damaging. The others, Diphylla ecaudata and Diaemus youngii, have specialized feeding habits, preferring to feed on the blood of birds (Uieda, 1982; Villa, 1966). Vampire bats range from tropical Mexico to northern Argentina and northern Chile, and they cost the Latin American livestock industry $350 million a year: $100 million in direct losses to rabies and $250 million in secondary losses, such as reduction in milk production and secondary infections (Kverno and Mitchell, 1976). In addition, several humans die each year of rabies transmitted by vampire bats (Baer, 1975; Irons et al., 1957; Venters et al., 1954).

Previous attempts to reduce vampire bat populations in Latin America have been ineffective, dangerous, destructive, impractical, too localized, or too expensive. Methods tried included the use of flame throwers in Trinidad (Greenhall, 1970), the dynamiting of several thousand caves in the State of Grande do Sul, Brazil (Villa, 1969), the placement of a strychnine-syrup mixture at old bite sites (Greenhall, 1963), the gassing of thousands of caves in Latin America (Arteche, 1969), the use of Japanese mist nets (Dalquest, 1954; Greenhall, 1963), and the placement of traps at the entrances of caves (Constantine, 1969). Bats have also been killed with clubs or firearms in stables and dwellings where they attack their prey (Constantine, 1970).

In 1967, under the authority of the Foreign Assistance Act, the U.S.

Agency for International Development (AID) asked the U.S. Department of the Interior's Fish and Wildlife Service to conduct research on vertebrate pests (rodents, pest birds, and vampire bats) in developing countries. In VAMPIRE BAT CONTROL IN LATIN AMERICA 152
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved.

June 1968, the Service's Denver Wildlife Research Center (DWRC) initiated a program of research on the vampire bat problem under the auspices of AID. The program consisted of two phases: (1) research conducted by a laboratory team at DWRC and a field team at the Instituto Nacional de Investigaciones Pecuarias in Palo Alto, Mexico; and (2) a utilization and training phase, developed after species-specific control methods were available.

THE ENVIRONMENTAL PROBLEMS
The main objective of the program was to increase livestock production in Latin America by reducing vampire bat populations. The main environmental problems were how to accomplish this objective without destroying habitat, contaminating the environment with pesticides, destroying desirable species, and adversely affecting an endangered species, the white-winged vampire (Diaemus youngii ). It was most important to develop species-specific control methods that were effective, inexpensive, and safe and required little training to apply.

APPROACHES TO CONTROLLING VAMPIRE BATS
When this project began, the morphology, physiology, and distribution of vampires (primarily Desmodus) were reasonably well known, but little was known of their social structure, behavior, movement, population dynamics, and interspecific relationships, and no techniques were available for determining their relative or absolute abundance (Linhart, 1975). Hence, the proposed initial research was aimed at acquiring some knowledge of the behavior and ecology of vampires.

The following studies were undertaken during the research phase from 1969 to 1973:
Pharmacological evaluation of chemical agents, such as toxicants and substances that alter physiology or behavior, for controlling vampire bats.
Laboratory evaluation of techniques for applying control agents, including feeding and grooming behavior of captive vampires, carrier compounds or liquid vehicles for application of control agents to livestock, and the application of control agents directly to livestock. Field evaluation of techniques for applying control, including devices for applying carrier agents and their persistence on livestock, feeding behavior of free-ranging vampires, and treatment of cattle with control agents.

VAMPIRE BAT CONTROL IN LATIN AMERICA 153
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. Estimation of actual numbers and repopulation or recovery rates of natural populations of vampire bats.

Additional studies focused on vampire bat movements, sensory mechanisms associated with locating prey, and pharmacological hazards associated with the application of control agents to livestock.

Ecology of Vampire Bats
Research identified two potential weak links in vampires. First, Desmodus produce only one or possibly two young each year (Burns, 1970), although each has a potential longevity of 13-14 years (Linhart, 1973; Trapido, 1946).

Therefore, population recovery after reduction by control is slow. Second, vampire populations are concentrated near herds of livestock to which they fly nightly (Constantine, 1970).

Vampires do not migrate, although they can move locally during the breeding season (Burns and Flores Crespo, 1975). They forage every night, except when there is a full moon, but only during the darkest part of the night, either before the moon rises or after it sets (Flores Crespo et al., 1972). Finally, vampire bats form mobile communities that use multiple roosts; the number of vampires in a colony might remain the same, but there can be nightly exchange of individuals among roosts (Mitchell et al., 1973). The most practical time to capture bats in mist nets set around corralled cattle is during the week after a full moon, when the period of complete darkness is still short.

Development of Control Methods
Two species-specific control methods for reducing vampire bat populations were developed in this project: the systemic method, in which cattle are treated with a chemical; and the topical method, in which captured vampire bats are treated with the same chemical.

Systemic Method. The idea of treating cattle with systemic toxicants originated at a Communicable Disease Center (CDC) laboratory working on bats and rabies in Las Cruces, New Mexico. Denny G. Constantine, a CDC veterinarian, was told that a Bayer systemic insecticide, Neguvon, would kill vampire bats. He treated cattle with Neguvon, but no vampires that fed on the treated cattle died. Although Neguvon did not kill vampires in Constantine's experiments, other systemic insecticides recommended by the U.S. Department of Agriculture (USDA) Livestock Insects Investigations Laboratory were tested.

The recommendations on agents and
VAMPIRE BAT CONTROL IN LATIN AMERICA 154
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. dosages were based on margins of safety for cattle. These compounds were tested in Mexico, and one of them, Famophos, gave positive results. In the laboratory, cattle were treated with a 13.2% pour-on formulation of Famophos recommended for the control of screwworm larvae and sucking flies. All vampire bats that fed on these treated cattle died (Mitchell et al., 1970, 1971).

However, research with Famophos was discontinued, because it was not available in Latin America and it was prohibitively expensive.

When it was shown that vampires could be killed with a systemic insecticide, DWRC scientists began looking at other substances that could be administered to cattle, especially anticoagulants. Because they bind to blood proteins, anticoagulants are excreted by the cow very slowly and are available longer than most other chemicals for the bat to ingest with a blood meal. Bats ingest about 50% of their body weight in blood at each feeding and concentrate it by excreting water (Breidenstein, 1982), so finding a dose of anticoagulant lethal to the bat but not harmful to cattle appeared feasible.

Treating cattle with the anticoagulant actually selected, diphenadione, at a concentration of 1.0 mg/kg gave a wide margin of safety for the cattle and killed all the vampires that fed on blood taken from the treated cattle up to 72 hours after treatment. Diphenadione was not retained in the liver of treated animals, as was another effective anticoagulant, chlorophacinone (Bullard et al., 1970, 1971).

In 1971, diphenadione was administered with a balling gun to cattle in Mexico. Results were good at Rancho Huichi (90.2% reduction in fresh bites) and poor at Rancho Don Tomas (20% reduction in fresh bites) (Mitchell et al., 1971). The poor results at Rancho Don Tomas were attributed to weather that reduced the normal foraging activities of the bats. In 1972, diphenadione was again field-tested in Mexico, but was administered intraruminally with a syringe; this test reduced biting by 93% (Thompson et al., 1972).

Topical Method. The idea of topical treatment was based on the likelihood that bats would ingest a toxicant while grooming. Similar behavior is a basis of common methods of rat control. Greenhall (1965) reported that vampires are thorough groomers and spend a considerable amount of time in this activity.

Captive vampires were observed to spend some 2 hours per day grooming (Flores Crespo et al., 1971 a) — enough time for a control compound on the fur to be ingested.

To determine whether there would be sufficient contact between the vampire and its prey for the vampire to receive the chemical from a cow's skin and whether the vampire has a preferred biting area, 49 feedings VAMPIRE BAT CONTROL IN LATIN AMERICA 155
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. were observed in a bat-proof corral under laboratory conditions. Of the 49 bites, 27 were made in the hoof region while the cattle were standing, and 22 on various parts of the body while they were lying down; moreover, each feeding entailed minimal contact between the vampire and its prey (Flores Crespo et al., 1971b). As a followup to the laboratory study, vampires were observed with Starlight night-vision telescopes, which electronically magnify available light, while they fed under natural conditions. Vampires were observed while feeding under normal conditions on three races of cattle — Brahma, Charolais, and Holstein. Approximately one-third of the bats landed on the Brahma before biting and feeding, and 31% on the Charolais, but 85% landed directly on the Holstein. Only on the Holstein was there a preferred area for biting: 74.5% of the bites were on the neck (Flores Crespo et al., 1974).

Almost all cattle in the geographical range of vampire bats are Brahma.

Because the Brahma are on open range and because only 21% of the vampires that fed on Brahma cattle fed on their necks, the idea of applying a control compound to an appropriate area of the prey animal was discarded as impractical.

An alternative topical method would be to apply the toxicant to the bat itself. While writing a manuscript on grooming, DWRC biologist Samuel B.

Linhart recognized that vampires are extensive groomers and roost in compact groups separate from beneficial bats in the same cave. Hence, a substance placed on the fur of a captured bat might be passed to other bats in the colony once the treated bat returned to the roost. These ideas led Linhart to develop the topical method now in use.

On the basis of the recommendations of DWRC specialists, four chemicals were tested for toxicity to vampire bats. Chlorophacinone, an anticoagulant, was the most toxic. Ten carriers were evaluated by mixing them with a dye and applying them to the dorsal surface of vampire bats. Bats were checked several times a day to determine the persistence of the carriers. Three carriers were selected for additional study: a saturated solution of acetone and an acrylic polyester fluorescent paint pigment; petroleum jelly; and a mixture of abalyn (a resin), mineral oil, and polyethylene (Epoline-10). For each candidate carrier, one vampire was treated with a mixture of carrier and a dye and then introduced into a simulated roost containing 19 other vampires. The roost was checked several times a day to determine the degree of transfer among colony members by grooming. Petroleum jelly was superior to the other carriers.

In the final test with the simulated roost, 50 mg of chlorophacinone was mixed with 1.5 ml of petroleum jelly and applied to the dorsal surface of one vampire. This vampire was put into the roost with 19 other bats. The carrier bat was found dead on the morning after reintroduction into VAMPIRE BAT CONTROL IN LATIN AMERICA 156
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. the roost, and 18 of the 19 other bats died 5-19 days after reintroduction (Linhart, 1970).

Several field tests confirmed these results. Vampire bats captured in two caves were treated with a mixture of 50 mg of chlorophacinone and 1.5 ml of petroleum jelly. Six vampires captured in one cave were treated and released; later, 94 dead vampires were found — a ratio of 15-16 vampires killed to each one treated. After 2 weeks, the caves contained only one live vampire (Linhart et al., 1972). In another test, bats were captured in mist nets as they flew in to feed on corralled cattle on two ranches, were treated with the control mixture, and were released. Two weeks later, vampire bites on the same cattle had decreased by 95% (Linhart et al., 1972).

At first, all vampires captured were treated with the chlorophacinone- petroleum jelly mixture. Later, it was determined that one treated bat would kill 19 others in the roost. There is a high correlation between the number of fresh bites on members of a herd and the number of vampires feeding on the herd. If a herd had 100 fresh bites, the vampire population was estimated to be 100 and only five bats were treated.

There were two reasons for changing the control chemical from chlorophacinone to diphenadione. First, other anticoagulants were as effective as chlorophacinone in controlling vampire bat populations with the topical method. Second, diphenadione was available for manufacture, because the patent had expired. The use patent is held by the U.S. government, which gives permission to anyone to use the chemical as a vampiricidal agent.

Although the topical method is more difficult to use than the systemic method, because it requires night work and the proper identification and treatment of only vampires, it is sometimes the preferred method. Injection of the anticoagulant into the rumen is the easiest way to administer it; but, of the animals commonly attacked by vampires, only cattle have a rumen and can be treated this way. If other livestock or humans are being attacked, the topical method is recommended.

Another control method, application of anticoagulants to roost walls, was evaluated in a pilot study in Mexico. Diphenadione in petroleum jelly was placed at vampire roost sites in a tunnel and in a mine (Mitchell et al., 1971, 1972). In both cases, all the vampires were killed. Four nonhematophagous bats also died in the tunnel. In view of these results, control of vampire bats by treating roosts was not recommended.

Control Techniques
Systemic Treatment of Cattle with Diphenadione. Diphenadione is injected into cattle intraruminally at 1.0 mg/kg of body weight. The drug VAMPIRE BAT CONTROL IN LATIN AMERICA 157
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. is absorbed and circulates in the blood. Any vampire bat that feeds on a properly treated animal within 72 hours after treatment receives a lethal dose of the drug. Treating all cattle in a herd reduces biting by 90-95% (Thompson et al., 1972).

Topical Treatment of Vampire Bats with Diphenadione. Vampire bats are captured with mist nets that are set around corralled cattle or at cave entrances.

Approximately 1.5 ml of a diphenadione-petroleum jelly mixture is placed on the dorsal surface of each captured bat, and the bat is released. The bats return to their roost, and, because they live in compact colonies, pass the chemical from one to another. The bats die after ingesting the chemical during grooming (Linhart et al., 1972). For every treated bat, approximately 20 vampires die at the roost.

Vaccines
Although topical or systemic use of anticoagulants can reduce the number of bites by up to 96%, and thus reduce secondary losses by a similar amount, rabies can still occur among unvaccinated cattle even when the bat population is reduced. Therefore, both cattle vaccination and bat population control are recommended if the aim is to eliminate the disease, rather than only to reduce cattle losses to an acceptable point (Lord, 1980).

Piccinini (1977) used both vaccines and anticoagulants in a 500-km 2 area
in the State of Pernambuco, Brazil. In the 2 years before control, 17,870 cattle were in the study area, 2,342 were not protected by vaccines, and 140 died of rabies. After application of both control methods, the vampire population was reduced by 96.1%. In the 2 years after control, 17,431 cattle were in the area, 4,848 were not vaccinated, and only 2 died of rabies. This shows that reduction of the vampire population gives cattle some protection from rabies without the use of vaccines. It should be noted that the treatments in Piccinini's study were made by scientists; the results might be less impressive under normal operational field conditions.

KINDS OF ECOLOGICAL KNOWLEDGE USED
Ecological Facts
Ecological and physiological facts were obtained from published literature and from the observations of experienced people in the field. Experiments were done in simulated and natural situations and yielded
VAMPIRE BAT CONTROL IN LATIN AMERICA 158
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. much knowledge of the life history and behavior of vampire bats. The ecological observations provided the basis of the control programs.

Pilot Studies
The studies of topical anticoagulants in simulated roosts constitute an example of the successful use of analog studies, and the pilot tests of both systemic and topical anticoagulants were important in the development of practical control programs. The studies did not all lead directly to successful control, but the ones that "failed" enabled workers to avoid expensive and ineffective methods.

Project as Experiment
Experience gained in the utilization phase led to improvements in the procedures. During that phase, one problem was encountered in Nicaragua, where 9 of 14 calves died after treatment with diphenadione (1.0 mg/ kg). These calves were less than 3 months old and lacked a fully functional rumen. Further tests showed that calves without a functioning rumen should not be treated (Elias et al., 1978).

UTILIZATION AND TRAINING
From 1974 to 1978, AID funded DWRC biologists to develop a utilization or training phase for vampire bat control. It wanted DWRC to take the control methods to Latin American cattlemen and to continue adaptive research to answer questions that could arise. Initially, DWRC biologists conducted training seminars in Latin America, often involving teams of veterinarians organized to vaccinate cattle. The basic training tool was a pamphlet, Chemical Control of Vampire Bats, published in Spanish, English, and Portuguese (Mitchell and Burns, 1973a,b, 1978). Later, the Peace Corps in Belize (Mitchell et al., 1975, 1976) and the Pan American Health Organization in Venezuela, Trinidad, and Surinam (Mitchell et al., 1974) became involved. In-country training also occurred. For example, 125 veterinarians from 19 states were trained in Brazil (Mitchell et al., 1976). In 1976, Rodrigo Gonzales presented results of the campaign in Nicaragua at an international conference (Gonzales and Mitchell, 1976), and the Nicaraguan government hosted a seminar on vampire bat control, attended by 46 people from 11 countries (Mitchell et al., 1976).

EFFECTIVENESS OF THE PROGRAM
After the publication of our control methods, new questions arose. The possibility of hazards to other bats from use of the topical method was suggested (Turner, 1975). Killing of nonhematophagous bats where vampires have been controlled topically has not been reported, and none was killed in the many field tests conducted. Brazilian workers controlled vampires in 3,062 caves and never found a nonhematophagous bat killed by the topical method (Mitchell et al., 1976).

The topical method was field-tested in every Latin American country exposed to vampire predation. Results were consistent: 90-95% reduction in the vampire population. The systemic method of treating cattle with diphenadione is even more specific in that it kills only the vampires that are attacking cattle. If cattle are treated with the recommended dosage, 1.0 mg/kg, no chemical is passed in the milk, and residues in tissue are negligible (Bullard and Thompson, 1977; Bullard et al., 1976, 1977). With sound management and husbandry practices in Ecuador, the killing of vampire bats had no effect on milk production. In Nicaragua, under more tropical conditions, milk production increased by 16% (Thompson et al., 1977).

These methods of vampire bat control are very effective even when applied by nonprofessionals. For example, in Nicaragua, 20 high-school graduates were trained to apply the control methods over 4 years. They examined 270,825 cattle on 2,124 ranches and found 123,376 fresh vampire bites. They treated 148,142 of the cattle and 2,696 vampires and reduced fresh bites by 90% to about 12,000. The annual benefit to farmers from the vampire bat control program in Nicaragua was US$2,414,158, and annual costs, $129,750, for a very favorable benefit-to-cost ratio of about 18.6:1 (Badger and Schmidt, 1979).

ACKNOWLEDGMENTS
Although I prepared this case study, the solution to the problem is based on an accumulation of many scientific experiments conducted by a host of scientists. Much of the work was done by DWRC scientists in an array of disciplines, including wildlife biology, pharmacology, animal psychology, statistics, physiology, electronics, chemistry, nutrition, and ecology.

Undoubtedly, this multidisciplinary approach was influential in solving the problem.

Thanks go to Bernardo Villa, Master Zoologist at the University of Mexico, for sharing his vast knowledge of bats.

I also wish to give special thanks to Charles Ladenheim (AID retired), VAMPIRE BAT CONTROL IN LATIN AMERICA 160
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. who had the insight to support the transition from a research project to an operational program. Research cannot be considered accomplished or successful until its results are made known and implemented — in this case, by the small producers of livestock in Latin America.

Funds for this research were provided to DWRC by AID under Participating Agency Service Agreements RA (ID) 01-67 and ID- TAB-000-1076.

Reference to trade names in this review does not imply U.S. government endorsement.

REFERENCES
Arteche, E. 1969. Rabia en Rio Grande do Sul, Brazil. Zoonosis 11:34-35 .

Badger, D. D., and K. M. Schmidt. 1979. Evaluation of the Vampire Bat Control Program.

Department of Agricultural Economics, Oklahoma State University, Stillwater, Okla.

Available from U.S. Agency for International Development, Washington, D.C.

Baer, G. M. 1975. Rabies in nonhematophagous bats. Pp. 79-98 in G. M. Baer, ed. The Natural History of Rabies. Vol. 2. Academic Press, New York.

Breidenstein, C. P. 1982. Digestion and assimilation of bovine blood by a vampire bat (Desmodus rotundus). J. Mammal. 63:482-484 .

Bullard, R. W., and R. D. Thompson. 1977. Efficacy and safety of the systemic method of vampire bat control. Interciencia 2:149-152 .

Bullard, R. W., S. R. Kilburn, and G. Holguin. 1970. Effectiveness of chlorophacinone (DRC-3776) as a livestock systemic for control of vampire bats. Pp. 33-34 in 1970 Annual Report, Denver Wildlife Research Center, Denver, Colo.

Bullard, R. W., S. R. Kilburn, and G. Holguin. 1971. Livestock systemic technique for control of vampire bats (Desmodus rotundus). Pp. 36-41 in 1971 Annual Report, Denver Wildlife Research Center, Denver, Colo.

Bullard, R. W., R. D. Thompson, and G. Holguin. 1976. Diphenadione residues in tissues of cattle.

J. Agric. Food Chem. 24:261-263.

Bullard, R. W., R. D. Thompson, and S. R. Kilburn. 1977. Diphenadione residues in milk of cattle.

J. Agric. Food Chem. 25:79-81 .

Burns, R. J. 1970. Twin vampire bats born in captivity. J. Mammal. 51:391-392 .

Burns, R. J., and R. Flores Crespo. 1975. Local movement and reproduction of common vampire bats in Colima, Mexico. Southwest. Nat. 19:446-449 .

Constantine, D. G. 1969. Trampa portatil para vampiros usada en programas de campana antirrabica. Bol. Of. Sanit. Panam. 67:39-42 .

Constantine, D. G. 1970. Bats in relation to health, welfare, and economy of man. Pp. 319-449 in W. A. Wimsatt, ed. Biology of Bats. Vol. 2. Academic Press, New York.

Dalquest, W. W. 1954. Netting bats in tropical Mexico. Trans. Kans. Acad. Sci. 57:1-10.

Elias, D. J., R. D. Thompson, and P. J. Savarie. 1978. Effects of the anticoagulant diphenadione on suckling calves. Bull. Environ. Contam. Toxicol. 20:71-78 .

Flores Crespo, R., S. B. Linhart, and R. J. Burns. 1971a. Comportamiento del vampiro ( Desmodus rotundus) en cautiverio. Southwest. Nat. 17:139-143.

Flores Crespo, R., R. J. Burns, and S. B. Linhart. 1971b. Comportamiento del vampiro ( Desmodus rotundus) durante su alimentacion en ganado bovino en cautiverio. Tecnica Pecuaria en Mexico 18:40-44 .

Flores Crespo, R., S. B. Linhart, R. J. Burns, and G. C. Mitchell. 1972. Foraging behavior of the common vampire bat related to moonlight. J. Mammal. 53:366-368.

Flores Crespo, R., S. Said Fernandez, R. J. Burns, and G. C. Mitchell. 1974. Observaciones sobre el comportamiento del vampiro comun ( Desmodus rotundus ) al alimentarse en condiciones naturales. Tecnica Pecuaria en Mexico 27:39-45.

Gonzales, R., and G. C. Mitchell. 1976. Vampire bat control programs in Latin America. Proc.

Vertebr. Pest Control Conf. 7:254-257.

Greenhall, A. M. 1963. Use of mist nets and strychnine for vampire bat control in Trinidad. J.

Mammal. 44:396-399.

Greenhall, A. M. 1965. Notes on behavior of captive vampire bats. Mammalia 29:441-451.

Greenhall, A. M. 1970. Vampire bat control: A review and proposed research programme for Latin America. Proc. Vertebr. Pest Control Conf. 4:41-54.

Irons, G. V., R. B. Eads, J. E. Grimes, and A. Conklin. 1957. The public health importance of bats.

Tex. Rep. Biol. Med. 15:292-298.

Kverno, N. B., and G. C. Mitchell. 1976. Vampire bats and their effect on cattle production in Latin America. World Anim. Rev. 17:1-7.

Linhart, S. B. 1970. Work Unit DF-101.1: Basic Research in Mammal and Bird Damage Control:

Chemicals. 1970 Annual Progress Report, Denver Wildlife Research Center, Denver, Colo.

Linhart, S. B. 1973. Age determination and occurrence of incremental growth lines in the dental cementum of the common vampire bat (Desmodus rotundus). J. Mammal. 54:493-496 .

Linhart, S. B. 1975. The biology and control of vampire bats. Pp. 221-241 in G. M. Baer, ed. The Natural History of Rabies. Vol. 2. Academic Press, New York.

Linhart, S. B., R. Flores Crespo, and G. C. Mitchell. 1972. Control of vampire bats by topical application of an anticoagulant, chlorophacinone. Bol. Of. Sanit. Panam. 6:31-38 .

Lord, R. D. 1980. An ecological strategy for controlling bovine rabies through elimination of vampire bats. Proc. Vertebr. Pest Control Conf. 9:170-175.

Mitchell, G. C., and R. J. Burns. 1973a. Combate Quimico de los Murcielagos Vampiros. Regional Technical Assistance Center, U.S. Embassy, Mexico City, Mex.

Mitchell, G. C., and R. J. Burns. 1973b. Chemical Control of Vampire Bats. Denver Wildlife Research Center, Denver, Colo. (mimeo)
Mitchell, G. C., and R. J. Burns. 1978. Combate Quimico aos Morcegos Hematofagos. Ministry of Agriculture, EMBRAPA,Coronel Pacheco, Braz.

Mitchell, G. C., R. Flores Crespo, R. J. Burns, S. Said Fernandez, and S. B. Linhart. 1970. Vampire Bats: Rabies Transmission and Livestock Production in Latin America. 1970 Annual Report of the Denver Wildlife Research Center's Mexican Field Station, Palo Alto, Mex.

Mitchell, G. C., R. Flores Crespo, R. J. Burns, and S. Said Fernandez. 1971. Vampire Bats: Rabies Transmission and Livestock Production in Latin America. 1971 Annual Report of the Denver Wildlife Research Center's Mexican Field Station, Palo Alto, Mex.

Mitchell, G. C., R. Flores Crespo, R. J. Burns, and S. Said Fernandez. 1972. Vampire Bats: Rabies Transmission and Livestock Production in Latin America. 1972 Annual Report of the Denver Wildlife Research Center's Mexican Field Station, Palo Alto, Mex.

Mitchell, G. C., R. J. Burns, and A. L. Kolz. 1973. Rastreo del comportamiento nocturno de los murcielagos vampiros por radio-telemetria. Tecnica Pecuaria en Mexico 24: 47-56 .

Mitchell, G. C., R. J. Burns, J. W. De Grazio, and D. J. Elias. 1974. Vampire Bats:

Rabies Transmission and Livestock Production in Latin America. 1974 Annual Report.

Denver Wildlife Research Center, Denver, Colo.

Mitchell, G. C., R. D. Thompson, D. J. Elias, C. E. Shuart, R. W. Bullard, P. J. Savarie, and R. J.

Burns. 1975. Vampire Bats: Rabies Transmission and Livestock Production in Latin America. 1975 Annual Report. Denver Wildlife Research Center, Denver, Colo.

Mitchell, G. C., R. D. Thompson, D. J. Elias, C. E. Shuart, and R. W. Bullard. 1976. Vampire Bats:

Rabies Transmission and Livestock Production in Latin America. 1976 Annual Report.

Denver Wildlife Research Center, Denver, Colo.

Molina Solis, J. F. 1896. Historia del Descubrimiento y Conquista de Yucatan, con una Resena de la Historia Antigua de la Peninsula. R. Caballero, Merida, Mex.

Piccinini, R. S. 1977. The Use of Diphenadione (2-Diphenylacetyl-1,3 indandione) for Vampire Bat Control in Endemic Rabies Areas, Northeastern Brazil. Master's thesis, University of California, Davis.

Thompson, R. D., G. C. Mitchell, and R. J. Burns. 1972. Vampire bat control by systemic treatment of livestock with an anticoagulant. Science 177:806-808.

Thompson, R. D., D. J. Elias, and G. C. Mitchell. 1977. Effects of vampire bat control on bovine milk production. J. Wildl. Manage. 41:736-739.

Trapido, H. 1946. Observations on the vampire bat with special reference to longevity in captivity.

J. Mammal. 27:217-219 .

Turner, D. C. 1975. The Vampire Bat. John Hopkins University Press, Baltimore.

Uieda, W. 1982. Aspectos do Comportamento Alimentar das Tres Especies do Morcegos Hematofagos (Chiroptera, Phyllostomidae). Master's thesis, Instituto de Biologia da Universidade Estadual de Campinas, Braz.

Venters, H. D., W. R. Hoffert, J. E. Scatterday, and A. V. Hardy. 1954. Rabies in bats in Florida.

Am. J. Public Health 44:182-185.

Villa, R. B. 1966. Los Murcielagos de Mexico. Institute of Biology, National Autonomous University, Mexico City, Mex.

Villa, R. B. 1969. The ecology and biology of vampire bats and their relationship to paralytic rabies.

Report to the Government of Brazil. Rep. No. TA 2656. U.N. Development Programme/ Food and Agriculture Organization. Rome.

Committee Comment
The control of pests is a common theme in humans' attempts to live in a world populated with animals and plants, and that theme is reflected in several of the cases discussed in this report (Chapters 14, 15, and 24). Many serious pests are controlled only poorly, despite great expense; those controlled successfully usually have idiosyncrasies of life history that make them vulnerable, and successful control depends on detailed knowledge of those idiosyncrasies.

Early attempts to control vampires did not rely on knowledge of the animals' natural history; they were ineffective and destructive. The resulting careful and thorough research into the natural history of vampires by the Denver Wildlife Research Center (DWRC) exemplifies an approach VAMPIRE BAT CONTROL IN LATIN AMERICA 163
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. that is usually needed for pest control; the idiosyncrasies discovered represent a certain amount of the luck that so often seems to accompany successful cases.

DWRC biologists studied the behavior, dispersal, physiology, and ecology of vampires, using field and laboratory experiments and observations. These studies and literature review revealed important aspects of vampires' natural history that made them susceptible to control. Two of these aspects commonly help in pest control: the bats have a low reproductive rate, which meant that less powerful control would be needed to keep their numbers down, and they do not migrate.

Successful control also depended on several idiosyncrasies. The bats, because of their diet of blood, are much more susceptible to anticoagulants than cattle are. They forage only on dark nights, so when the moon rises shortly after sunset or sets shortly before sunrise, their activity is confined to short periods, and that makes them easier to trap. They roost extremely close to each other and they groom each other — both characteristics that favor transfer of vampiricides between individuals and that make it unnecessary to treat more than a few members of each roost.

Thus, successful control in this case depended on thorough research into the pests' natural history, which is nearly always needed, and on the vulnerabilities that the research revealed. The vulnerable points might not exist, but in the absence of thorough research they will almost surely not be found.

VAMPIRE BAT CONTROL IN LATIN AMERICA 164
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. 14 Biological Control of California Red Scale
Interest in biological control of agricultural pests has been stimulated by the evolution of resistance of the target organisms to pesticides, by secondary outbreaks caused by losses of natural predators and parasites, and by the increasing seriousness of unwanted side effects of pesticides on many species, including humans. However, biological control is not yet a highly predictive branch of ecology. Most biological control efforts fail, either because the control agent does not survive or because, if it survives, it does not result in satisfactory control. The California red scale study illustrates features common to many instances of biological control, but it also shows how imaginative use of some new elements of ecological theory, such as those related to foraging and sex allocation, led to changes in project organization that illuminated the processes underlying effective biological control.

Case Study
ROBERT F. LUCK, Department of Entomology, University of California, Riverside

INTRODUCTION
California red scale, Aonidiella aurantii (Maskell) (Diaspididae:

Homoptera), is an insect pest of citrus in arid and semiarid regions (Bodenheimer, 1951; Ebeling, 1959). It inhabits all aboveground parts of a citrus plant. At moderate densities, the scale infests the fruit, which might then be culled in accordance with industry standards for appearance. At higher densities, it inhibits fruit production and kills branches. In California, it is one of three principal arthropod pests of citrus and infests all four major cultivars grown in the state: grapefruit, lemons, and two orange varieties.

California red scale was introduced into southern California between 1868 and 1875 on shipments of citrus nursery stock from Australia (Quayle, 1938).

Lindorus lophanthae (Blaisd.) (Coccinellidae: Coleoptera), a predatory ladybird beetle introduced from Australia, was the first of 52 predators and parasites (36 and 16, respectively) introduced against California red scale. Of the eight that became established, three were predators — L. lophanthae, introduced in 1889; Orcus chalybeus (Bvdl.), in 1892; and Chilocorus similis, in 1924-1925 — and five were parasitoids — Comperiella bifasciata Howard, in 1940; Habrolepis rouxi Compere, in 1937-1939; Aphytis lingnanensis Compere, in 1948; Encarsia (= Prospaltella) perniciosi Tower, in 1949; and A. melinus DeBach, in 1956-1957 (Rosen and DeBach, 1978). Aphytis chrysomphali (Mercet), a thelytokous (parthenogenetic, mostly female producing) parasitoid, was unknowingly introduced around 1900, probably in parasitized scale on infested nursery stock from the Mediterranean Basin (Rosen and DeBach, 1978, 1979).

Before the introduction of the wasp Aphytis lingnanensis (Aphelinidae:

Hymenoptera), in 1947, California red scale was economically controlled by A. chrysomphali in the Santa Barbara, California, area (DeBach and Sisojevic, 1960), perhaps with some help from ladybird beetles, so long as ants, dust, and insecticide interference were minimized. The scale remained a problem, however, in inland coastal valleys and other coastal areas, even though A. chrysomphali was found throughout these regions. When A. lingnanensis was introduced, biological control was achieved in
BIOLOGICAL CONTROL OF CALIFORNIA RED SCALE 166
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. most coastal areas. But only with the introduction of A. melinus in 1956-1957 was biological control achieved in inland valleys (Rosen and DeBach, 1979).

In contrast with the situation in southern California, economical biological control of California red scale does not exist in the San Joaquin Valley or the desert, even though both Comperiella bifasciata (Encyrtidae: Hymenoptera) and A. melinus are present. Remaining natural enemy species, both introduced and native, are scarce. Moreover, citrus production in the San Joaquin Valley is assuming increasing importance as that in southern California declines with increasing urbanization. In the absence of economic biological control of scale, growers have relied on synthetic organic insecticides (Flaherty et al., 1973).

THE ENVIRONMENTAL PROBLEM
Synthetic organic insecticides have several drawbacks. Their initial effectiveness, convenience, and low cost all promote their use (Luck et al., 1977). But these materials are general in their effects, so they also kill native predators and parasites of the target pests, as well as the pests themselves, and sometimes cause outbreaks of the target pests and secondary pests (phytophagous arthropods normally held at bay by their natural enemies).

Pesticide applications increase the nutritional value of plants for some prey species (Dittrich et al., 1974; Jones and Parrella, 1984; Maggi and Leigh, 1983).

Pest resurgence and secondary-pest outbreaks promote further pesticide use, and the grower becomes enmeshed in a pesticide treadmill in which still further pesticide use increases control costs and selects for resistance in primary or secondary pest species (Luck et al., 1977).

Citrus is one of many crops that have followed this scenario (Luck et al., 1977). California red scale populations in South Africa, Israel, and Lebanon have evolved resistance to one or more of the synthetic organic scalicides (Georghiou and Taylor, 1977). California populations have evolved resistance to hydrogen cyanide gas, used as a fumigant before World War II. Secondary- pest outbreaks have also occurred in citrus. Since the 1960s, several Lepidoptera species and citrus red mite, Panonychus citri McG. (Tetranichidae), have become increasingly frequent pests in California citrus because of insecticide use (Luck et al., 1977). Citrus red mite has also evolved resistance to many acaricides used against it (Georghiou and Taylor, 1977).

ECOLOGICAL APPROACH TO THE ENVIRONMENTAL

PROBLEM
We have had two objectives in our studies of California red scale, its host plants, and its natural enemies: the specific objective of improving the biological control of the scale, especially in the San Joaquin Valley, and the general objective of understanding how biological control of red scale works and what limits its success. The latter objective presupposes that understanding the reasons for success and failure and the associated processes and mechanisms in a specific case will improve our ability to predict the outcome both of this particular natural enemy-pest interaction and of such interactions in general.

The specific objective motivates a continuing biological control project, whose key components are the obtaining and manipulation of natural enemies of the scale, the importation of exotic predator and parasitoid species reared from California red scale or related species around the world, and the mass rearing and release of endemic and exotic natural enemies. We seek by these means to develop practical and economical ways of augmenting endemic natural enemy populations by releasing insectary-reared individuals. The duration of this project depends on the discovery of new species of natural enemies, on political conditions in the geographical region where we wish to collect natural enemies, and on the degree of biological control achieved by introducing enemies. Importation and release of natural enemies would cease once effective biological control were achieved throughout California.

We have pursued the general objective of using California red scale as a model system through a combination of field and laboratory studies to identify attributes and processes that characterize a successful biological control project.

Laboratory studies have sought to understand the host selection and foraging behavior of the several natural enemies of California red scale. The results suggested hypotheses to be tested with field populations of the scale. In addition to field tests of hypotheses (e.g., augmentative releases), comparative demographic studies of red scale populations were begun in three of California's four citrus regions; these studies sought to evaluate the effect of natural enemies on the dynamics of red scale populations (cf. Rosen and DeBach, 1978) and to provide a baseline for assessing new introductions.

USES OF KNOWLEDGE AND UNDERSTANDING
Ecosystem Definition
The ecosystem components of interest in these studies are several citrus cultivars grown in commercial and dooryard plantings in the state, honeydew- producing insects and ant species that tend them, natural enemies that attack California red scale and its primary natural enemies, California red scale, and climates in the three citrus regions. Also of interest are associated phytophagous arthropods that receive frequent or occasional insecticide treatments, whose consequences can affect red scale or its interaction with its natural enemies.

Chemical control of red scale might also affect densities of other phytophagous species or their natural enemies, including the citrus thrips, Scirtothrips citri (Thripidae: Thysanoptera), citrus red mite, and several lepidopterans (Luck, 1984). With increasing information and understanding, we might also need to include other components of the ecosystem, such as citrus nematode, phytophthora root rot, or Tristeza (seedling yellows), because they and their treatment influence the physiology of the tree or otherwise affect the interaction between red scale and its entomophages. The ecosystem components named earlier were identified because they were economically important or because previous research with citrus, with other agricultural ecosystems, or with natural ecosystems suggested that these components were biologically important.

Significance of Impact
If biological control of California red scale is achieved, chemical control will be required only infrequently when weather patterns, dust, insecticides (drifting from adjacent crops from treatments of an associated pest), or ants interfere with the scale's natural enemies and thus induce increases in local scale populations. Biological control should reduce red scale population densities, pest-control costs, and pesticide-related environmental and health problems.

These expectations are based on experience with biological control of pests in general and of red scale in both southern California (Rosen and DeBach, 1978, 1979) and other citrus-growing regions, such as Greece (DeBach and Argyriou, 1967), Australia (Furness et al., 1983), South Africa (Bedford and Georgala, 1978), and Cyprus (Orphanides, 1984).

Study Strategy and Monitoring
Our first specific goal associated with the second objective was to understand why Aphytis melinus replaced A. lingnanensis in southern California and what limits the degree of parasitism achieved by A. lingnanensis. A. lingnanensis was first introduced into California in 1947. It was not the first Aphytis present on red scale in California, however; A. chrysomphali had apparently been introduced before 1900 (Rosen and DeBach, 1979). A. lingnanensis replaced A. chrysomphali, probably because it was better adapted to the climate of California's citrus districts (DeBach and Sisojevic, 1960).

However, A. lingnanensis failed to achieve economic scale control in the more interior citrus districts, so additional natural enemies were sought. One of these, A. melinus, was established in the interior in 1956-1957 and quickly spread throughout the area, displacing A. lingnanensis. A melinus provided economic scale control in much of southern California. In the late 1970s, the three species were distributed as follows: A. melinus in the interior citrus areas, A. lingnanensis along the coast and inland approximately 10 km, and A. chrysomphali as a relict in a few coastal enclaves (Rosen and DeBach, 1979).

DeBach and his co-workers monitored establishment and spread of these introduced Aphytis species by sampling red scale populations in many places throughout southern California (DeBach, 1965). DeBach noticed A. melinus displacing A. lingnanensis. In some groves, the displacement occurred even when third-stage female scale, the putative host for the two Aphytis species, was abundant. DeBach concluded that hosts for the parasitoids were not limiting and that competition for scarce hosts could not explain the displacement (DeBach, 1966; DeBach and Sundby, 1963; Rosen and DeBach, 1979). Laboratory experiments designed to test this hypothesis, however, contradicted the field observation: A. lingnanensis displaced A. melinus in the laboratory — the opposite of what happened in the field (DeBach and Sundby, 1963). DeBach then suggested that A. melinus performed better in the field because it was both a better searcher and better adapted to California's climate than A. lingnanensis.

Established Boundaries
For demographic studies, we used the citrus tree as a sampling universe.

Such a universe is reasonable for characterizing an interaction between red scale and its natural enemies because red scale is mobile only during the crawler stage (first instar) and, in the case of male scale, briefly (less than 24 hours) as an adult while seeking mates. Between-tree movement depends on wind dispersal of crawlers and sometimes on crawlers that BIOLOGICAL CONTROL OF CALIFORNIA RED SCALE 170
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. walk between contiguous trees. Adult natural enemies can easily fly between trees, but we understand little of their behavior or patterns of movement.

However, because we wanted to know the effects of natural enemies on the red scale population of a tree, it seemed reasonable, at least initially, to consider the natural enemies on a within-tree basis also.

FIGURE 1 Climatic zones in which California citrus is grown.

For our studies of economic effects, we again used the tree as a universe, but we were also concerned with the economic effects of biological control.

Thus, we chose ''blocks" (management units of citrus groves) of about 2.5 hectares as our economic universe.

We limited our study to California citrus groves and dooryard citrus. We divided the state's citrus growing areas into zones ( Figure 1 ) on the basis of climate and industry experience. We used the differences between zones as manipulations with which to compare changes in the scale's age structure caused by seasonal temperature patterns and the influence of these changes on control of the scale.

Developing and Implementing a Study Strategy
We first tested DeBach's assertion that A. melinus and A. lingnanensis used the same host resource (scale stage and size) (DeBach, 1966; DeBach BIOLOGICAL CONTROL OF CALIFORNIA RED SCALE 171
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. and Sundby, 1963; Rosen and DeBach, 1979) by characterizing the oviposition behaviors of the two species. We used these behaviors to see what the parasitoids recognized as hosts and how many eggs of what gender they chose to allocate to those hosts (male, female, or mixed gender). An Aphytis egg is laid on the host's body beneath a scale cover where it cannot be seen, but the laying of an egg is reflected in obvious oviposition behavior (Luck et al., 1982) (Figure 2).

We next determined what size and stage of scale each wasp species used as hosts for its progeny. Most hymenopterans, including A. melinus and A. lingnanensis, have diploid females and haploid males. Thus, the parental female determines the gender of her offspring by "deciding" whether or not to fertilize the egg. Some hymenopterans lay male eggs on small hosts and female eggs on large hosts (Charnov, 1982; Charnov et al., 1981; Chewyreuv, 1913; Clausen, 1939). This behavior characterizes both A. melinus ( Figure 3 ) and A. lingnanensis ( Figure 4 ), although A. melinus allocates sons and daughters to smaller scale than does A. lingnanensis (Luck and Podoler, 1985). We hypothesized that A. melinus displaced A. lingnanensis because it usurped a substantial proportion of the scale population before scale attained the size that A. lingnanensis required for its offspring— especially the size needed to produce daughters. Both Aphytis species paralyze the scale before they deposit eggs on it, thus preventing further scale growth. As a scale grows, it attains the size used by A. melinus first. Use of the scale by A. melinus then eliminates the scale as a potential resource for A. lingnanensis. Each species tends to avoid depositing eggs on scale previously parasitized by the other (Luck and Podoler, 1985).

Thus, the initial laboratory studies on host selection by Aphytis led us to propose a hypothesis to explain how A. melinus displaced A. lingnanensis in southern California. Some of the predictions derived from this hypothesis were testable in the field.

Specific Predictions and Hypotheses
If our hypothesis is correct, A. melinus should eventually eliminate A. lingnanensis from southern California. Rosen and DeBach (1979) found that in 1965 A. melinus occurred in the interior coastal valleys and that A. lingnanensis occurred along the coast to a point about 10 km inland. By 1972, A. melinus had spread coastally while the area populated by A. lingnanensis had contracted (Luck and Podoler, 1985). In 1983, A. lingnanensis was restricted to a few locations, all within a kilometer or so of the coast (R. F. Luck and S. Warner, unpublished data).

The age structure of California red scale varies seasonally with location in California. Lower winter temperatures inhibit crawler production by BIOLOGICAL CONTROL OF CALIFORNIA RED SCALE 172
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved.

FIGURE 2 Behavioral sequence exhibited by Aphytis melinus and A. lingnanensis during oviposition.

FIGURE 3 a-d, size distribution of third-stage California red scale on which Aphytis melinus laid indicated numbers and sexes of eggs. e-f, size distribution of scale chosen by and offered to A. melinus for parasitization.

FIGURE 4 a-d, size distribution of third-stage California red scale on which Aphytis lingnanensis laid indicated numbers and sexes of eggs. e-f, size distribution of scale chosen by and offered to A. lingnanensis for parasitization.

BIOLOGICAL CONTROL OF CALIFORNIA RED SCALE 175
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. gravid and parturient female red scale. They also inhibit male scale emergence; hence, virgin (third-instar) female scale remain uninseminated.

First-instar and first-molt scale (those molting to second instar) are easily killed by lower winter temperatures (under 8 °C) (Abdelrahman, 1974). These stages are expected to die during the winter if below-average winter temperatures prevail in the interior coastal valleys or if normal winter temperatures prevail in the San Joaquin Valley. Finally, normal winters are warm enough to allow some development of second-stage scale. They usually attain the third instar, but the lack of males prevents them from developing further as they remain uninseminated. Development of the male scale is arrested at the unemerged imago or pupal stage. The degree to which the scale's age structure is dominated by one or two age classes thus depends on temperature and is correlated with the citrus zone: the more interior the zone, the more variable the seasonal temperature pattern and, thus, the more the scale's age structure is dominated by the older scale stages (virgin and gravid females).

With the higher spring temperatures, males inseminate the virgin females, and crawlers appear after a gestation period. The more extreme the scale's age structure (dominated by one or two scale classes), the more synchronized the pulse of crawlers that initiates the spring generation. This pulse can be observed passing through the successive scale stages (ages). It becomes less distinct with each successive generation and is usually unrecognizable by autumn (third generation). Such a pattern produces a host resource for the Aphytis species that is at first abundant and then scarce. The abundant third-stage (virgin female) scale present in the spring escapes parasitization by Aphytis, because the low temperatures (under 18 °C) inhibit searching by Aphytis. We therefore expected Aphytis to be ineffective as a biological control agent where the early-season scale population (spring generation) is dominated by one or two age classes.

Thus, a hypothesis that evolved initially to explain the competitive displacement of A. lingnanensis by A. melinus also suggested a testable explanation for the absence of economic biological control in the San Joaquin Valley. These expectations provided the rationale for the experimental design that compared the scale's and parasitoid's demographic characteristics in the different citrus zones.

Developing and Implementing a Monitoring Program
The comparative demographic studies of Aphytis species involved a sampling approach aimed at monitoring seasonal patterns in the scale's age structure, the scale size (stage) from which an Aphytis species emerges, the sex of the emerging wasp, the geographical distribution of the scale's BIOLOGICAL CONTROL OF CALIFORNIA RED SCALE 176
Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies Copyright National Academy of Sciences. All rights reserved. natural enemies, and the diurnal temperature pattern during the season. In addition to the work already performed, a survey will be conducted every 10 years to determine the distribution of natural enemies of California red scale.

However, if a new natural enemy becomes established, an extensive sampling program will be conducted more frequently.

Parasitized scale encountered in the samples were individually reared to determine the gender and size of the emerging parasitoid. Second-stage and older scale were measured, as were parasitized scale that retained their shape (Aphytis species are ectoparasitoids; hence, as they feed they cause the scale to shrivel). Natural-enemy field densities and female parasitoid sizes were also monitored. We expect to modify these sampling schemes as we gain experience and as our hypotheses change.

We monitored diurnal temperatures hourly in each study grove. Scale phenology was also monitored. Because adult male scale live less than 24 hours in the field, males captured on pheromone traps mirror the temporal pattern of virgin third-stage female scale, i.e., the scale's age structure relevant to female progeny production by A. melinus and A. lingnanensis.

Future Goals
Future research is planned along several lines. First, why are some parasitoid species that exist on red scale elsewhere unable to exist on the scale in California, and, if they are present there, why are their densities and distributions limited there? Numerous parasitoid species have been introduced into southern California to control red scale, but most have failed to become established. Others have become established, but remain scarce or geographically limited. For example, Aphytis africanus, a common parasitoid of California red scale in South Africa, has been introduced several times, but has not become established. Encarsia (= Prospaltella) perniciosi (red scale strain) is limited to coastal citrus and immediately adjacent areas; it is infrequently encountered in the interior coastal valleys. Similarly, Comperiella bifasciata, a parasitoid introduced into southern California in 1939, is rare in the interior coastal valleys and coastal citrus zones, but is common in the San Joaquin Valley.

Sacramento City CouncilDecember 2, 2025

Sacramento City Council
December 2, 2025