Santee City Council
June 11, 2025

Effects of Wildfire on Wildlife and Connectivity
Prepared by: Megan K. Jennings, Ph.D.

January 23, 2018
Introduction
In southern California, where human impacts from development are limiting habitat connectivity for wide-ranging vertebrate species, fire is a disturbance regime that may also fragment habitats, further impacting those species. Although fire is a natural process in the southwestern U.S., increasing human development near open spaces has led to unnatural fire regimes with increased fire starts and an increased potential for vegetation-type conversion as a result. In the biodiversity hotspot of southern California, many studies have focused on the effects of urbanization and landscape fragmentation on wildlife. However, there has been relatively little attention to how human-mediated landscape fragmentation may influence natural disturbance processes, like wildfire, and how these synergistic disturbances impact wildlife populations.

Both fire frequency and size are increasing in southern California and are correlated with increasing anthropogenic development and human population growth in the region (Syphard et al. 2007, 2009). These studies suggest that at high human population densities, fire is eliminated from the ecosystem when contiguous vegetation necessary to carry fire is broken up by asphalt, concrete, and buildings. However, at intermediate human densities, housing developments and roadways are a source of increased fire ignitions which then spread into wildlands (Syphard et al. 2007, 2009). Both scenarios (too little fire, too frequent fire) present potential threats for species and community dynamics in southern California as shifts in the natural fire regime, coupled with increasing habitat fragmentation, have the potential to impact wildlife populations, communities, and entire ecosystems. In the highly urbanized landscape of southern California, long-term impacts such as habitat fragmentation and loss and shifts in disturbance regimes like the natural fire cycles, have resulted in persistent landscape changes (Syphard et al. 2009).

This report focuses on the impacts to wildlife connectivity posed by the proposed Newland Sierra project in the context of wildfires and the need for corridor redundancy. The Newland Sierra project proposes to build more than 2,100 homes on the I-15 corridor in the unincorporated portion of San Diego County between Escondido and Temecula. The project would be located in the area proposed for the North County Multiple Species Conservation Program (NCMSCP) on a site that has been identified as pre-approved mitigation area (PAMA).

As described in my previous reports (Jennings 2017a, 2017b), this project poses risks to wildlife connectivity in the area and could compromise overall design objectives of the NCMSCP. The proposed Newland Sierra project will significantly affect high quality core habitat and wildlife movement for both more common and sensitive and protected species to a degree that is not mitigated by the project design. The proposed project will have long-term direct and indirect impacts on wildlife from roadways, increased human activity, edge effects, human activity, and increasing fire frequency on wildlife movement. Due to the risks of wildfire and the numerous cumulative projects proposed along the I-15 corridor in northern San Diego County and southern Riverside County, it is particularly important to account for corridor redundancy in considering the Newland Sierra project. Regional connectivity plans must provide corridor redundancy to serve the range of species that may need to move between patches of habitat (Pinto and Keitt

2009, McRae et al. 2012), and to buffer against landscape disturbances, such as wildfires (Mcrae et al. 2008, McRae et al. 2012, Cushman et al. 2013, Olson and Burnett 2013). The biological analysis in the project’s draft environmental impact report lacked sufficient consideration of these issues.

Impacts of Wildfires and Shifting Fire Frequencies on Wildlife Disturbances that occur at large spatial scales, such as Santa Ana wind-driven fires in southern California, like the recent Lilac Fire in San Diego County, are most likely to change landscape configuration, or pattern, which can lead to change in resource availability, environmental features, and corresponding responses in the structure of populations and communities, all key metrics to landscape integrity (Sousa 1984, Pickett and White 1985, Fraterrigo and Rusak 2008, Turner 2010). Large-scale landscape changes, particularly fragmentation (Gardner et al. 1993), have been shown to alter biotic interactions, and lead to a loss of connectivity evidenced by a decline in dispersal, reduced survival rates (Riley et al. 2003), and limited gene flow (Riley et al. 2006). In southern California, the two disturbances that overlap and interact, fire and human development, are the predominant drivers of the landscape. In this region, empirical evidence suggests a shift is underway in the disturbance regime (Keeley and Fotheringham 2003, Safford and Van de Water 2014).

Shifts in fire regime typically involve changes to fire intensity, size, frequency, type, seasonality, and severity (Flannigan et al. 2000). Fire-return intervals, the average time between two fire events, in the shrubland habitats like the areas where the Lilac Fire occurred and the NewlandSierra development is proposed were historically 30 to 100 years. In similar areas of the County, fires are 33% more frequent now than pre-settlement, due in large part to increased development and roadways (Figure 1; Keeley et al. 1999, Safford and Van de Water 2014). This shifting disturbance regime with shortened intervals between fires interrupts the successional cycle, reduces plant diversity, and results in vegetation and habitat type change to non-native and grass dominated landscapes (Keeley 2005), reducing habitat suitability and connectivity for species dependent on intact shrubland landscape. Shifting weather patterns resulting from climate change may also contribute to the alteration of fire regimes in southern California. Climate models predict that temperatures will increase and humidity will decrease (Miller and Schlegel 2006).

Under these conditions, Santa Ana winds, the hot, dry winds from the deserts in the east, may occur more often and later in the season when fuels loads are highest (Miller and Schlegel 2006, Guzman-Morales et al. 2016). The concurrent disturbances of expanding human development and a shifting climate may alter how fire structures the landscape. Extensive development, particularly in exurban areas, results in increases in human-caused ignitions and fires of large spatial extents (Syphard and Keeley 2015), as well as an overall increase in fire threat (Figure 2), which can have long-lasting impacts on the landscape and wildlife habitat.

Many wildlife species that occur in the Mediterranean-type ecosystems of southern California have adapted to wildfires. Wildlife exhibit differential responses to wildfires depending on the availability of refugia and species’ mobility, which determine their susceptibility to impacts from the direct effects of the fire. Habitat and diet breadth, population size and growth rates, as well as landscape connectivity can affect post-fire colonization and overall resilience to these types of stochastic events. While some research efforts in southern California have taken advantage of the

natural experiment presented by San Diego’s 2003 and 2007 wildfires to gather information about bird (Mendelsohn et al. 2008), small mammal (Brehme et al. 2011, Diffendorfer et al. 2012), large mammal (Schuette et al. 2014), and herpetofauna (Rochester et al. 2010) responses to wildfire, there is much to learn about individual- and population-level responses, in particular as it relates to increasing fire frequency. Linking the effects of shifting fire regimes on wildlife where frequent fire may result in vegetation type conversion from shrublands to grass-dominated habitats (Keeley 2005, Keeley and Brennan 2012) is a significant challenge. There is evidence of the effect of increasing fire frequency on some species, such as the iconic coastal sage scrub species, the threatened California gnatcatcher (Polioptila californica californica). Already challenged by habitat loss and fragmentation in the coastal regions of southern California, frequent fires have degraded habitat for the gnatcatcher (Winchell and Doherty 2014) as California sagebrush (Artemesia californica), laurel sumac (Malosma laurina), and white sage (Salvia apiana), key habitat elements for the bird, have been replaced by non-native annual grasses in areas that have experienced repeated fires. Habitat specialists and small species are not the only ones subject to the impacts of increasing fire frequency. Despite the fact that mountain lions (Puma concolor) are highly mobile and able to move away from fires, the species is potentially at risk from vegetation-type conversion to non-native annual grasslands (Jennings et al. 2016). Although this species may tolerate grasslands when moving between habitats (Zeller et al. 2014), habitat fragmentation between San Diego County and the Santa Ana Mountains to the north has limited gene flow and resulted in inbreeding for the southern California population (Ernest et al. 2014), a situation which further habitat degradation, particularly as a result of increasing fire frequency, could worsen.

Wildfire and Connectivity
Habitat connectivity is essential to climate-smart landscape strategies (Heller and Zavaleta 2009) and strengthens ecosystem resilience to additional stressors such as habitat fragmentation (Beier and Gregory 2012), and other disturbances, e.g., fire and disease (Noss 1991, Hilty et al. 2006).

Across much of southern California, the state’s Natural Community Conservation Planning (NCCP) program and the federal Habitat Conservation Plan (HCP) have been used to establish conservation networks to protect natural communities and prevent further habitat fragmentation (Ogden Environmental and Energy Services 1996, Riverside County 2003). Although the direct effects of anthropogenic landscape alteration, namely habitat loss and fragmentation, are paramount in this region (Soulé 1991, Crooks 2002, Beier et al. 2006), the indirect effects of intense human development such as changing patterns of natural disturbance regimes, e.g. wildfire, may present an equally large risk to landscape integrity. As human populations in southern California have grown dramatically over the last century, particularly in coastal areas, short fire-return intervals paired with habitat fragmentation, may have synergistic and long-term impacts on landscape connectivity that present a formidable conservation challenge. Given that these disturbances exert measurable impacts individually (Lindenmayer et al. 2008, Turner 2010), it is likely that the synergistic effects of shifting disturbance regimes and fragmentation present a serious threat to landscape connectivity (Turner 2010).

Given the importance of landscape connectivity to ensuring population viability and persistence, accurate assessments of physical and functional connectivity are critical. Dynamic landscape processes, like wildfires, may impede movement for many species in the short-term, but an altered fire regime may permanently alter landscape linkages. In particular, shifting disturbance

regimes, like the increase in fire frequency and size reported in southern California, may have synergistic impacts that erode landscape connectivity if efforts are not made to buffer the number or impacts of fire on landscape linkages. New approaches to identifying factors that impair physical and functional connectivity are needed to develop mitigation strategies to maintain landscape connectivity if urbanization is considered on fire-frequent landscapes, with a particular focus on the coastal areas that are most impacted by development, and foothills and valleys where the wildland-urban interface is most at risk for increases in fire frequencies and consequential type conversion.

Building resilience into these networks of conserved lands can be approached from two perspectives: 1) reducing ignitions in fire-prone areas, and 2) account for these altered disturbance dynamics in conservation planning efforts like the Draft NCMSCP. Robust measures to reduce ignitions should be employed. However, reducing ignitions alone is unlikely to protect San Diego County’s open spaces from fire and must be paired with complementary approaches to provide for habitat and connectivity when fires do occur. This includes planning for redundancy in linkages connecting habitat patches (Pinto and Keitt 2009). Because a single path is unlikely to equally serve all individuals of a species, let alone all potential species that may need to move between patches of habitat, multiple corridors between landscape blocks are often necessary (Pinto and Keitt 2009, McRae et al. 2012). Furthermore, this redundancy can also buffer against uncertainty and dynamic processes, particularly disturbances, on the landscape (Mcrae et al. 2008, McRae et al. 2012, Cushman et al. 2013, Olson and Burnett 2013). As discussed in my previous comments on the Newland Sierra draft environmental impact report, the project’s biological analysis is deficient in its consideration of corridor redundancy. Threats to wildlife connectivity from wildfire emphasize the need to consider corridor redundancy with respect to Newland Sierra and the NCMSCP.

Furthermore, the assessment of connectivity and redundancy to provide for wildlife movement under a variety of conditions must be conducted at a regional scale. For San Diego County, this means consideration of conservation planning efforts and acquisitions as well as development projects in Orange and Riverside Counties. In particular, the Santa Ana-Palomar landscape linkage that has been identified as a critical movement corridor for many species (South Coast Wildlands 2008), most notably the mountain lion (Vickers et al. 2015), spans both San Diego and Riverside Counties and could be affected by several development projects that could limit functional connectivity in northern San Diego County.

Together with the cumulative projects proposed in this region, Newland Sierra could restrict wildlife movement in northern San Diego County as well as any opportunities to build resilience into a regional connectivity plan by providing for corridor redundancy. The Merriam Mountains area is currently one of the few shrub-dominated open spaces in San Diego County that has not experienced overly frequent wildfires which lead to increased risk of vegetation-type conversion from shrublands to non-native annual grasslands (Figure 1). Preserving a relatively intact landscape, such as the Merriam Mountains, is critical to developing a functional preserve system for the NCMSCP. However, the proposed Newland-Sierra Project’s new roadways, increased traffic, homes, and increased wildland-urban edge are all known sources of fire ignitions in southern California (Syphard and Keeley 2015) that will threaten to increase the fire frequency in this area, which is already at high risk of fire due to the site’s vegetation and terrain features

(Figure 2), as well as the risk of vegetation-type conversion. These same project elements will also further restrict wildlife movement in an area where movement is already constrained. The synergistic effects of restricted movement and habitat degradation caused by increasing fire frequency could greatly reduce connectivity in this region and threaten the functionality of the proposed preserve network under the NCMSCP. Figure 1 illustrates that few linkages remain in San Diego County that are not challenged by crossing urban development or areas that have burned repeatedly and are at risk for weed incursion, habitat degradation, and vegetation-type conversion. When dynamic landscape processes are considered, effective planning for landscape connectivity will require consideration of all potential projects that may affect wildlife movement as well as the synergistic disturbances that also affect landscape connectivity. The NCMSCP provides an opportunity for this type of regional wildlife connectivity planning, but individual development proposals considered in isolation and without adequately accounting for regional corridor redundancy could threaten the overall effectiveness of the planning process.

Figure 1. Map of fire-return interval departure (Safford and Van de Water 2014) for northern San Diego County and linkages identified in the Management Strategic Plan Connectivity documents for San Diego’s NCCP areas.

Figure 2. Map of fire threat for northern San Diego County as classified by California’s Fire and Resource Assessment Program and linkages identified in the Management Strategic Plan Connectivity documents for San Diego’s NCCP areas.

References
Beier, P., and A. J. Gregory. 2012. Desperately seeking stable 50-year-old landscapes with patches and long, wide corridors. PLoS Biology 10.

Beier, P., K. Penrod, C. Luke, W. Spencer, and C. Cabañero. 2006. South Coast missing linkages: restoring connectivity to wildlands in the largest metropolitan area in the United States. Pages 555–586 in K. R. Crooks and M. A. Sanjayan, editors. Connectivity conservation. Cambridge University Press, Cambridge.

Brehme, C. S., D. R. Clark, C. J. Rochester, and R. N. Fisher. 2011. Wildfires alter rodent community structure across four vegetation types in southern California, USA. Fire Ecology 7:81–98.

Crooks, K. R. 2002. Relative sensitivities of mammalian carnivores to habitat fragmentation.

Conservation Biology 16:488–502.

Cushman, S. A., B. Mcrae, F. Adriaensen, P. Beier, M. Shirley, and K. Zeller. 2013. Biological corridors and connectivity. Pages 384–404 in D. W. Macdonald and K. J. Willis, editors.

Key Topics in Conservation Biology 2. First edition. John Wiley & Sons, Ltd.

Diffendorfer, J., G. M. Fleming, S. Tremor, W. Spencer, and J. L. Beyers. 2012. The role of fire severity, distance from fire perimeter and vegetation on post-fire recovery of small-mammal communities in chaparral. International Journal of Wildland Fire 21:436–448.

Ernest, H. B., T. W. Vickers, S. A. Morrison, M. R. Buchalski, and W. M. Boyce. 2014.

Fractured Genetic Connectivity Threatens a Southern California Puma (Puma concolor) Population. PLoS One 9.

Flannigan, M. ., B. . Stocks, and B. . Wotton. 2000. Climate change and forest fires. Science of The Total Environment 262:221–229.

Fraterrigo, J. M., and J. A. Rusak. 2008. Disturbance-driven changes in the variability of ecological patterns and processes. Ecology Letters 11:756-770.

Gardner, R. ., R. V. O’Neill, and M. G. Turner. 1993. Ecological implications of landscape fragmentation. Pages 208–226 in S. T. A. Pickett and M. J. McDonnell, editors. Humans as components of ecosystems: subtle human effects and the ecology of populated areas.

Springer-Verlag, New York.

Guzman-Morales, J., A. Gershunov, J. Theiss, H. Li, and D. Cayan. 2016. Santa Ana Winds of Southern California: Their climatology, extremes, and behavior spanning six and a half decades. Geophysical Research Letters 43:2827–2834.

Heller, N. E., and E. S. Zavaleta. 2009. Biodiversity management in the face of climate change:

A review of 22 years of recommendations. Biological Conservation 142:14–32.

Hilty, J. A., W. Z. Lidicker, and A. M. Merenlender. 2006. Corridor ecology : the science and practice of linking landscapes for biodiversity conservation. Island Press.

Jennings, M. K. 2017a. Landscape Connectivity Issue Review, Newland Sierra June 2017 DEIR. 1 August 2017. 25pp. Technical Report.

Jennings, M. K. 2017b. Merriam Mountains Wildlife Connectivity Review. 18 April 2017. 14pp.

Technical Report.

Jennings, M. K., R. L. Lewison, T. W. Vickers, and W. M. Boyce. 2016. Puma response to the effects of fire and urbanization. The Journal of Wildlife Management 80:221–234.

Keeley, J. E. 2005. Fire as a threat to biodiversity in fire-type shrublands. USDA Forest Service General Technical Report PSW-GTR-195.

Keeley, J. E., and T. J. Brennan. 2012. Fire-driven alien invasion in a fire-adapted ecosystem.

Oecologia 169:1043–1052.

Keeley, J. E., and C. J. Fotheringham. 2003. Impact of past, present, and future fire regimes on North American Mediterannean shrublands. Pages 218–262 in T. T. Veblen, W. L. Baker, G. Montenegro, and T. W. Swetnam, editors. Fire and climatic change in temperate ecosystems of the western Americas. Springer-Verlag, New York.

Keeley, J. E., C. J. Fotheringham, and M. Morais. 1999. Reexamining Fire Suppression Impacts on Brushland Fire Regimes. Science 284:1829–1832.

Lindenmayer, D., R. J. Hobbs, R. Montague-Drake, J. Alexandra, A. Bennett, M. Burgman, P.

Cale, A. Calhoun, V. Cramer, P. Cullen, D. Driscoll, L. Fahrig, J. Fischer, J. Franklin, Y.

Haila, M. Hunter, P. Gibbons, S. Lake, G. Luck, C. MacGregor, S. McIntyre, R. Mac Nally, A. Manning, J. Miller, H. Mooney, R. Noss, H. Possingham, D. Saunders, F. Schmiegelow, M. Scott, D. Simberloff, T. Sisk, G. Tabor, B. Walker, J. Wiens, J. Woinarski, and E.

Zavaleta. 2008. A checklist for ecological management of landscapes for conservation.

Ecology Letters 11:78–91.

McRae, B. H., B. G. Dickson, T. H. Keitt, V. B. Shah, B. H. Mcrae, B. G. Dickson, T. H. Keitt, and V. B. Shah. 2008. Using Circuit Theory to Model Connectivity in Ecology, Evolution, and Conservation. Ecology 89:2712–2724.

McRae, B. H., S. A. Hall, P. Beier, and D. M. Theobald. 2012. Where to Restore Ecological Connectivity? Detecting Barriers and Quantifying Restoration Benefits. PLoS ONE 7:e52604.

Mendelsohn, M. B., C. S. Brehme, C. J. Rochester, D. C. Stokes, S. A. Hathaway, and R. N.

Fisher. 2008. Responses in bird communities to wildland fires in southern california. Fire Ecology Special Issue 4.

Miller, N. L., and N. J. Schlegel. 2006. Climate change projected fire weather sensitivity:

California Santa Ana wind occurrence. Geophysical Research Letters 33.

Noss, R. F. 1991. Landscape connectivity: different functions at different scales. Pages 27–39in W. E. Hudson, editor.Landscape linkages and biodiversity. Island Press.

Ogden Environmental and Energy Services. 1996. Biological Monitoring Plan for the Multiple

Species Conservation Plan. San Diego, CA.

Olson, D. H., and K. M. Burnett. 2013. Geometry of forest landscape connectivity: Pathways for persistence. USDA Forest Service - General Technical Report PNW-GTR:220–238.

Pickett, S. T. A., and P. S. White. 1985. The Ecology of Natural Disturbance and Patch Dynamics. Page The Ecology of Natural Disturbance and Patch Dynamics.

Pinto, N., and T. H. Keitt. 2009. Beyond the least-cost path: Evaluating corridor redundancy using a graph-theoretic approach. Landscape Ecology 24:253–266.

Riley, S. P. D., J. P. Pollinger, R. M. Sauvajot, E. C. York, C. Bromley, T. K. Fuller, and R. K.

Wayne. 2006. A southern California freeway is a physical and social barrier to gene flow in carnivores. Molecular Ecology 15:1733–1741.

Riley, S. P. D., R. M. Sauvajot, T. K. Fuller, E. C. York, D. A. Kamradt, C. Bromley, and R. K.

Wayne. 2003. Effects of urbanization and habitat fragmentation on bobcats and coyotes in southern California. Conservation Biology 17:566–576.

Riverside County. 2003. Western Riverside Multiple Species Habitat Conservation Plan Documents.

Rochester, C. J., C. S. Brehme, D. R. Clark, D. C. Stokes, S. A. Hathaway, and R. N. Fisher. 2010. Reptile and Amphibian Responses to Large-Scale Wildfires in Southern California.

Journal of Herpetology 44:333–351.

Safford, H. D., and K. M. Van de Water. 2014. Using Fire Return Interval Departure (FRID) analysis to map spatial and temporal changes in fire frequency on National Forest lands in California. Pacific Southwest Research Station - Research Paper PSW-RP-266:1–59.

Schuette, P. A., J. E. Diffendorfer, D. H. Deutschman, S. Tremor, and W. Spencer. 2014.

Carnivore distributions across chaparral habitats exposed to wildfire and rural housing in southern California. International Journal of Wildland Fire 23:591–600.

Soulé, M. E. 1991. Conservation: tactics for a constant crisis. Science 253:744–750.

Sousa, W. P. 1984. The Role of Disturbance in Natural Communities. Annual Review of Ecology and Systematics 15:353–391.

South Coast Wildlands. 2008. South Coast Missing Linkages: A Wildland Network for the South Coast Ecoregion.

Syphard, A. D., and J. E. Keeley. 2015. Location, timing and extent of wildfire vary by cause of ignition. International Journal of Wildland Fire 24:37–47.

Syphard, A. D., V. C. Radeloff, T. J. Hawbaker, and S. I. Stewart. 2009. Conservation Threats Due to Human-Caused Increases in Fire Frequency in Mediterranean-Climate Ecosystems.

Conservation Biology 23:758–769.

Syphard, A. D., V. C. Radeloff, J. E. Keeley, T. J. Hawbaker, M. K. Clayton, S. I. Stewart, and R. B. Hammer. 2007. Human influence on California fire regimes. Ecological Applications

17:1388–1402.

Turner, M. G. 2010. Disturbance and landscape dynamics in a changing world. Ecology 91:2833–2849.

Vickers, T. W., J. N. Sanchez, C. K. Johnson, S. A. Morrison, R. Botta, T. Smith, B. S. Cohen, P.

R. Huber, H. B. Ernest, and W. M. Boyce. 2015. Survival and mortality of pumas (Puma concolor) in a fragmented, urbanizing landscape. PLoS ONE 10.

Winchell, C. S., and P. F. Doherty. 2014. Effects of habitat quality and wildfire on occupancy dynamics of Coastal California Gnatcatcher ( Polioptila californica californica). The Condor 116:538–545.

Zeller, K. A., K. McGarigal, P. Beier, S. A. Cushman, T. W. Vickers, and W. M. Boyce. 2014.

Sensitivity of landscape resistance estimates based on point selection functions to scale and behavioral state: pumas as a case study. Landscape Ecology 29:541–557.

Fire as a Threat to Biodiversity in Fire-Type
Shrublands1
Jon E. Keeley2
Abstract
Chaparral and coastal sage scrub communities have a disproportionately high number of rare and endangered plants and thus are of particular conservation concern. Unnaturally high fire frequency has been a leading cause of degradation of chaparral and coastal sage scrub ecosystems. Although these shrublands are fire-adapted, below a certain threshold of fire frequency, resilience is inversely related to the fire return interval: this threshold is 3-5 years in coastal sage scrub and 10-20 years in chaparral, with the higher values more typical of interior sites. High fire frequency depletes the native flora and increases the proportion of non-native herbaceous species. Resilience to different fire regimes varies across growth forms, and thus it is of particular significance that the growth form distribution of rare species is significantly different from the proportions of growth forms in these communities.

Key words: chaparral, endangered species, high fire frequency, resilience, sage scrub Introduction
The California landscape has been altered in many ways, with the potential for profound impacts on biodiversity and ecosystem functioning (Keeley and Swift 1995).

Other than direct development, one of the more important changes in shrubland ecosystems has been the anthropogenic alteration of the natural fire regime. Despite a long-standing policy of fire suppression, the primary impact has been a dramatic acceleration of fire occurrence (Keeley and Fotheringham 2002, Keeley and others 1999, Moritz 1997). Although species in these shrublands are “fire adapted,” they are not adapted to all fire regimes, and one can distinguish species differences and broad growth form differences in resilience to increased fire occurrence. This is of particular concern because both chaparral and coastal sage scrub communities have a disproportionately high number of rare and endangered plants.

Shrubland Biodiversity
The California Native Plant Society (CNPS) list of rare and endangered plant species (Skinner and Pavlick 1994) places chaparral first in number of taxa that are of concern (table 1). Although coastal sage scrub ranks fifth, both shrub communities have a much higher number of rare taxa than expected based on their area occupied.

If all the rare and endangered taxa in the top five habitats were distributed randomly,

1 An abbreviated version of this paper was presented at Planning for Biodiversity: Bringing Research and Management Together, a Symposium for the California South Coast Ecoregion, February 29-March 2, 2000, California State Polytechnic University, Pomona, CA. 2 U.S. Geological Survey, Sequoia National Park, Three Rivers, CA 93271; tel 559-565-3170 (email:
[email protected])
USDA Forest Service Gen. Tech. Rep. PSW-GTR-195. 2005. 97 Fire as Biodiversity Threat in Shrublands—Keeley
Table 1—Top-ranking habitats of California’s rare plants according to all CNPS lists (data from Skinner and Pavlik 1994). These observed values are contrasted with the values expected based upon the amount of land area occupied by each habitat (data from Jones and Stokes 1987). ____________________________________________________________________________________ Rank Habitat Taxa ____________________________________________________________________________________ Observed Expected
1 Chaparral 516 432
2 Lower coniferous forests 359 294
3 Cismontane woodland 311 362
4 Valley/foothill grassland 247 431
5 Coastal scrub 211 132
(p < 0.001; Π2 = 164.2 >>> Π2

0.999[3] = 16.3) ____________________________________________________________________________________ one would expect chaparral to top the list based just on areal extent of this vegetation, but chaparral contains 18 percent more species than expected. Coastal sage scrub contains nearly 40 percent more taxa than predicted by its areal extent.

In both chaparral and coastal sage scrub, herbaceous perennials, typically geophytes, top the list of growth forms that are rare or endangered (Skinner and Pavlick 1994). If we contrast these numbers just for southern California chaparral and coastal sage vegetation with those expected, based on extensive surveys of postfire peak of diversity (Keeley 1998a), we see that rarity is not randomly distributed across growth forms (table 2). Annuals are very under-represented. If rarity were randomly distributed with respect to growth form we would expect three times as many annuals as observed in the lists. On the other hand, there are three times as many herbaceous perennials as expected and double the number of shrub species.

Resilience to different fire regimes varies across growth forms (Keeley 2000), and thus it is of interest to consider rarity in terms of fire specialization. Fire is a frequent ecological factor and has played an obvious evolutionary role in these communities. Most non-woody plant species reach their greatest population sizes, and thus greatest potential for spread, in the postfire environment. However, not all species or even all growth forms are equally specialized towards fire. Annuals comprise a rich diversity of species that range from extreme specialization—strictly fire-stimulated germination (Keeley and Fotheringham 1998)—to generalized opportunistic species that occupy many types of disturbance. Cursory examination of the rare annuals listed in Skinner and Pavlick (1994) indicates they are not highly restricted to burned sites. Herbaceous perennials, particularly geophytes, in these shrublands lack obvious specialization to fire, and their life cycle is not substantively different from that of geophytes in other less fire-prone habitats (Rundel 1996). At the time of fire most geophytes are dormant and nearly all have transient seed banks that do not require specialized fire cues (Keeley1991). Thus, most rare herbaceous species do not appear to be highly specialized for postfire recruitment.

This pattern is not evident in the rare shrubs, where a large proportion are species of Ceanothus and Arctostaphylos that have seedling recruitment strictly tied to postfire conditions (Keeley 1998b). Thus, it is to be expected that the direct effects of alterations in fire regime would affect some components of the rare plant flora more than others.
98 USDA Forest Service Gen. Tech. Rep. PSW-GTR-195. 2005.

Fire as Biodiversity Threat in Shrublands—Keeley
Table 2—Growth form distribution of CNPS rare and endangered taxa in southern California chaparral and coastal sage scrub (data from Skinner and Pavlik 1994). Expected values based on distribution of growth forms following fire in 90 0.1 ha sites in southern California (Keeley 1998a) ____________________________________________________________________________________ Annual Herbaceous Suffrutescent Subshrub/
perennial shrub/tree ____________________________________________________________________________________ Observed 43 79 17 63
Expected 131 27 16 28
(p < 0.001; Π2 = 202.5 >>> Π2

0.999[2] = 13.8) ____________________________________________________________________________________ This is well illustrated by the extreme event of back-to-back wildfires studied by Zedler and others (1983). At four Adenostoma fasciculatum dominated sites, postfire frequency of this shrub increased following a fire in mature vegetation but decreased dramatically following the second fire. The impact was most profound on the nonresprouting shrub Ceanothus oliganthus, which was present from seed in threefourths of the plots after the first fire but nearly absent after the second fire. In contrast, these repeat fires had little impact on the herbaceous perennials Calochortus weedii and Dichelostemma pulchella.

Fire suppression has been frequently cited as a major threat to fire type rare species, for example, the rare locoweed Astragalus brauntonii that is restricted to sites around the Los Angeles Basin (Skinner and Pavlick 1994). This idea is a logical extension of the well-documented threat of fire suppression in many western coniferous forests (SNEP 1996). However, fire records show that in southern California fire suppression has not effectively excluded fire (Conard and Weise 1998, Keeley and others 1999). Indeed, urban mountain ranges such as the Santa Monica Mountains in Los Angeles and Ventura Counties have a fire rotation interval of less than 15 years, and this is likely many times shorter than the natural regime (Keeley 2002a). Lack of fire is not likely a threat to the persistence of these shrublands or rare species within them. A far greater threat in many areas, particularly the Los Angeles Basin, is habitat degradation due to increased fire frequency.

Fire-Induced Habitat Degradation
Unnaturally high wildfire frequency has long been a leading cause of degradation of chaparral and coastal sage scrub ecosystems, second only to land development. While these shrublands are fire-adapted, below a certain threshold of fire frequency resilience is inversely related to the fire return interval. This threshold is probably about 3 years in coastal sage scrub (but longer for interior sage scrub sites) and 10-20 years in chaparral (Keeley 2000). Generally speaking, as fire frequency increases, herbaceous vegetation is favored over woody growth forms (Sauer 1975, Wells 1962). Numerous studies have shown that unnaturally high fire frequency depletes the native flora and increases the proportion of non-native herbaceous species (Haidinger and Keeley 1993, Zedler and others 1983). In California it is quite likely that, except on fine-textured argillaceous soils, grasslands are degraded shrubland sites (Cooper 1922, Keeley 1990, 1993; Wells 1962), and even grasslands on certain argillaceous soils may have been dominated by the shallow rooted summerdeciduous coastal sage scrub (Kirkpatrick and Hutchinson 1980, Wells 1962).

USDA Forest Service Gen. Tech. Rep. PSW-GTR-195. 2005. 99 Fire as Biodiversity Threat in Shrublands—Keeley

0.01 1.34

2.67 4.00
L
o
g
D
e
n
sity (#/h
a)
Pre-fire 1990 1994
89 fire 93 fire
C. megacarpus
C. spinosus
Figure 1—Effect of repeat fires on the populations of an obligate-seeding shrub, Ceanothus megacarpus, and a facultatively-seeding shrub, C. spinosus, in the Santa Monica Mountains. The first fire occurred in mature chaparral in 1989, the second in 1993 (from Keeley 2000).

The impact of frequent fires on the native shrub populations is well illustrated by the back-to-back wildfires that Zedler and others (1983) discussed, above. This extreme event resulted in the near extirpation of one obligate-seeding shrub population, and this can occur with even longer intervals between fires. Figure 1 shows the extirpation of a non-sprouting shrub with fires four years apart, whereas a resprouting congener survived this high fire frequency, albeit at a lower density.

As native shrub cover is reduced due to high fire frequency, there is typically a type conversion to an herbaceous community dominated by non-native species. This is illustrated in the study by Haidinger and Keeley (1993) showing vegetation changes on adjacent sites experiencing different fire regimes. For example, two fires in six years reduced shrub populations of Adenostoma fasciculatum and Salvia mellifera but favored the suffrutescent Lotus scoparius (fig. 2). Three fires in six years were detrimental to most natives but conducive to the spread of non-native invasives such as Brassica nigra and species of Bromus. The results of this chronosequence study are corroborated by similar observations on a single site over time (fig. 3). In this study a mature coastal sage scrub site burned in 1993, and over the following three growing seasons the exotic cover declined as the native shrubs recovered. Following a second fire in 1996 the exotics exploded; within two years they dominated the site.

Such type conversions of shrublands to grasslands are not always the result of wildfires but have long been a goal of prescription burning programs (CDF 1978, Sampson 1944). Sometimes these are done for “range improvement,” to increase deer and livestock range, but other times for reducing fire hazard. These herbaceous (grassland) associations generate far less intense fires than shrub associations, and these grassland fires are more safely suppressed than chaparral fires (Keeley 2002a). 100 USDA Forest Service Gen. Tech. Rep. PSW-GTR-195. 2005.

Fire as Biodiversity Threat in Shrublands—Keeley

2.6 2.8

3.0 3.2
L
o
g
D
e
n
sity (#
h/ a)
Adenostoma fasciculatum
resprouts

3.0 3.5

4.0 4.5
L
o
g
D
e
n
sity (#
h/ a)
A. fasciculatum
seedlings

3.0 3.6

4.2 4.8
L
o
g
D
e
n
sity (#/h
a)
Salvia mellifera
seedlings

3.0 3.5

4.0 4.5

5.0 5.5
Lotus scoparius

2.50 2.96

3.42 3.88

4.34 4.80
Brassica nigra

3.0 3.5

4.0 4.5
Bromus spp
123 1 23
Number of fires in 6 years
Figure 2—First growing season shrub and herb density in adjacent stands of mixed chaparral and coastal sage scrub burned once, twice, and three times in 6 years (redrawn from Haidinger and Keeley 1993).

Type conversion not only alters fire intensity, but also it often leads to increased fire frequency (Keeley 2002a). This results from several factors. These herbaceous species dry rapidly during the late spring and thus greatly expand the seasonal window of opportunity for fire. In addition, they constitute fine fuels that ignite readily and spread fire both horizontally through the stand and vertically into the shrubs with little wind. As the extent of herbaceous cover increases, it sets the stage for repeat fires in a self feed-back process where more fires thin the shrub overstory and increase the presence of a persistent herbaceous layer. Over time, repeating this process will type-convert shrublands to annual grasslands dominated by non-native species, which in turn alters the fire regime by increasing further expansion of USDA Forest Service Gen. Tech. Rep. PSW-GTR-195. 2005. 101 Fire as Biodiversity Threat in Shrublands—Keeley
the seasonal window for fire because non-native species dry earlier in the spring than most native herbaceous species (Keeley and Fotheringham 2003). 93 94 95 96 97 98 99
0 20
40 60
80 100
C
o
v
er b
y
N
o
n-n
at vi e
s
%( )
Fire Fire
Figure 3—Percentage ground surface covered by non-native species on a coastal sage scrub site in the Santa Monica Mountains burned at 21 years of age in 1993 and reburned three years later in 1996 (data from Keeley and Fotheringham, unpublished data).

Sites differ greatly in their propensity for repeat fires. In general, interior sites are far more vulnerable to frequent fires than coastal sites due to the slower rate of shrub recovery (fig. 4). In another study of 90 sites burned during the same week in 1993, total herbaceous cover was positively correlated with parameters such distance from the coast, soil phosphorous, organic matter, and sand content (table 3). Postfire herbaceous cover was negatively correlated with stand age prior to the fire and estimated annual insolation. In short, herbaceous cover sufficient to carry a repeat fire is most likely when fires occur in young stands on fertile sites in the interior.

History of Type Conversion
Cooper (1922) believed there was abundant evidence, based upon his observations in the Coast Ranges, to say that burning by Indians accounted for a shift from woody vegetation to grasslands. He contended that this process of type conversion continued with the colonization by Europeans. Bauer (1930) believed the same applied to the grasslands of the Tehachapi Mountains farther south and cited examples of relictual patches of shrubland as evidence.
102 USDA Forest Service Gen. Tech. Rep. PSW-GTR-195. 2005.

Fire as Biodiversity Threat in Shrublands—Keeley
10 20
30 40
50 60
70 80
C
o
v
er %( G
S
C)
a) Coastal sage scrub shrubs
herbs
-1 0 1 2 3 4 5 6
Years Postfire
0 10
20 30
40 50
60 70
80 90
C
o
v
er %( G
S
C)
b) Interior sage scrub herbs
shrubs
Prefire
Figure 4—Foliar cover before and after fire at coastal and interior sage scrub sites (redrawn from O’Leary and Westman 1988).

Today the coastal ranges south of San Francisco have 25 percent of their landscape dominated by alien grasses, and there is reason to believe this derives from shrubland type conversion beginning with Native Americans (Keeley 2002b). Brown and Show (1944) recounted the history of rural land use in California, which included the use of fire to convert “useless” brush to “more productive” grasslands. This type conversion process accelerated in the latter part of the 19th century with increasing competition for suitable grazing land. During the latter part of the last century such type conversion was officially sanctioned by the issuing of brush burning permits for “range improvement” by the California Division of Forestry (CDF 1978).

Throughout the 20th century, type conversion of shrublands occurred from other types of disturbance as well (Stylinksi and Allen 1999).

Presently, we have relatively limited information on the extent of such type conversions. However, there are reasons to believe this was done on a massive scale.

One is the extensive distribution of non-native grasslands in the Coast Ranges of central California and the lack of any obvious correlation with environmental parameters such as soil type or slope exposure (Wells 1962). Another is the widespread distribution of grasslands in the foothills east of San Francisco and the demonstration of their rapid conversion to woody vegetation upon cessation of grazing (McBride and Heady 1968). Finally, quantitative measures of type conversion in southern California coastal sage scrub indicate that over the past 60 years more than half of the vegetation has been partially or completely type converted to grassland (Minnich and Dezzani 1998).

USDA Forest Service Gen. Tech. Rep. PSW-GTR-195. 2005. 103 Fire as Biodiversity Threat in Shrublands—Keeley
Table 3—Stepwise multiple regression of postfire herbaceous cover vs. environmental variables at 90 sites of coastal sage scrub and chaparral (from Keeley and Fotheringham unpublished data).
_____________________________________________
r
2 = 0.42 P < 0.001
2-tail P
Positively correlated:

Distance from coast 0.000
Percentage sand 0.002
Soil phosphorus level 0.003
Soil organic matter 0.009
Negatively correlated:

Prefire stand age 0.002
Annual solar insolation 0.033
_____________________________________________
Conclusions
Besides human development, the greatest threat to biodiversity in Mediterraneanclimate shrublands of California is changes in the natural fire regime. Fire suppression alters the fire regime by increasing the fire return interval and potentially threatens the persistence of species with fire-dependent recruitment. However, this impact is offset by the abnormally high number of fire occurrences due to humancaused ignitions. Consequently, the fire return intervals in southern California shrublands are higher than was historically the case, and the greatest threat to the persistence of these vegetation types is type conversion to herbaceous species more resilient to and more conducive to frequent fires. In general, non-native grasses and forbs are most favored by the current fire regime of frequent fires.

In general the majority of rare herbaceous species show no specialized response to fire. However, areas heavily affected by human-induced acceleration of fire return intervals may not be suitable sites for these natives because of the intense competition from non-native invasives that are favored by high fire frequency.

References
Atwood, J.L. 1993. California gnatcatchers and coastal sage scrub: The biological basis for endangered species listing. In: Keeley, J.E., editor. Interface between ecology and land development in California. Los Angeles: Southern California Academy of Sciences; 149-170.

Bauer, H.L. 1930. On the flora of the Tehachapi Mountains, California. Bulletin of the Southern California Academy of Sciences 29: 96-99.

Brown, W.S.; Show, S.B. 1944. California rural land use and management: A history of the use and occupancy of rural lands in California. Berkeley, CA: USDA Forest Service, California Region.

California Department of Forestry (CDF). 1978. Brushland range improvement. Annual report 1974-1977 inclusive. Sacramento: California Department of Forestry.

Conard, S.G.; Weise, D.R. 1998. Management of fire regime, fuels, and fire effects in southern California chaparral: Lessons from the past and thoughts for the future.

Tall Timbers Fire Ecology Conference Proceedings 20: 342-350. 104 USDA Forest Service Gen. Tech. Rep. PSW-GTR-195. 2005.

Fire as Biodiversity Threat in Shrublands—Keeley
Cooper, W.S. 1922. The broad-sclerophyll vegetation of California: An ecological study of the chaparral and its related communities. Stanford, CA: Carnegie Institution of Washington, Publication No. 319. 124 p.

Haidinger, T.L.; Keeley, J.E. 1993. Role of high fire frequency in destruction of mixed chaparral. Madroño 40: 141-147.

Jones and Stokes Associates. 1987. Sliding toward extinction: The state of California's natural heritage, 1987. San Francisco, CA: The California Nature Conservancy.

Keeley, J.E. 1990. The California valley grassland. In: Schoenherr, A.A., editor.

Endangered plant communities of southern California. Fullerton, CA: Southern California Botanists, Special Publication No. 3, 2-23.

Keeley, J.E. 1991. Seed germination and life history syndromes in the California chaparral. Botanical Review 57: 81-116.

Keeley, J.E. 1993. Assessing suitable sites for grassland restoration. In: Keeley, J.E., editor. Interface between ecology and land development in California. Los Angeles, CA:

Southern California Academy of Sciences; 277-281.

Keeley, J.E. 1998a. Postfire ecosystem recovery and management: The October 1993 large fire episode in California. In: Moreno, J.M., editor. Large forest fires. Leiden, The Netherlands: Backhuys Publishers, 69-90.

Keeley, J.E. 1998b. Coupling demography, physiology and evolution in chaparral shrubs.

In: Rundel, P.W.; Montenegro, G.; Jaksic, F.M., editors. Landscape diversity and biodiversity in Mediterranean-type ecosystems. New York: Springer; 257-264.

Keeley, J.E. 2000. Chaparral. In: Barbour, M.G.; Billings, W.D., editors. North American terrestrial vegetation. Cambridge, U.K.: Cambridge University Press; 203-253.

Keeley, J.E. 2002a. Fire management of California shrubland landscapes. Environmental Management 29: 395-408.

Keeley, J.E. 2002b. Native American impacts on fire regimes in California coastal ranges.

Journal of Biogeography 29: 303-320.

Keeley, J.E.; Fotheringham, C.J. 1998. Smoke-induced seed germination in California chaparral. Ecology 79: 2320-2336.

Keeley, J.E.; Fotheringham, C.J. 2001. The historic fire regime in southern California shrublands. Conservation Biology 15: 1536-1548.

Keeley, J.E.; Fotheringham, C.J. 2003. Impact of past, present, and future fire regimes on North American Mediterranean shrublands. In: Veblen, T.T.; Baker, W.L.; Montenegro, G.; Swetnam, T.W., editors. Fire and climatic change in temperate ecosystems of the western Americas. New York: Springer; 218-261.

Keeley, J.E.; Fotheringham, C.J.; Morais, M. 1999. Reexamining fire suppression impacts on brushland fire regimes. Science 284: 1829-1832.

Keeley, J.E.; Swift, C.C. 1995. Biodiversity and ecosystem functioning in Mediterraneanclimate California. In: Davis, G.; Richardson, D.M., editors. Biodiversity and function in Mediterranean-type ecosystems. New York: Springer-Verlag; 121-183.

Kirkpatrick, J.B.; Hutchinson, C.F. 1980. The environmental relationships of Californian coastal sage scrub and some of its component communities and species. Journal of Biogeography 7: 23-28.

McBride, J.R.; Heady, H.F. 1968. Invasion of grassland by Baccharis pilularis DC. Journal of Range Management 21: 106-108.

USDA Forest Service Gen. Tech. Rep. PSW-GTR-195. 2005. 105 Fire as Biodiversity Threat in Shrublands—Keeley
Minnich, R.A.; Dezzani, R.J. 1998. Historical decline of coastal sage scrub in the Riverside-Perris Plain, California. Western Birds 29: 366-391.

Moritz, M.A. 1997. Analyzing extreme disturbance events: Fire in the Los Padres National Forest. Ecological Applications 7: 1252-1262.

O'Leary, J.F.; Westman, W.E. 1988. Regional disturbance effects on herb succession patterns in coastal sage scrub. Journal of Biogeography 15: 775-786.

Rundel, P.W. 1996. Monocotyledonous geophytes in the California flora. Madroño 43:
355-368.

Sampson, A.W. 1944. Plant succession and burned chaparral lands in northern California. Berkeley, CA: University of California, Berkeley, Agricultural Experiment Station, Bulletin 685.

Sauer, C.O. 1975. Man's dominance by use of fire. Geoscience and Man 10: 1-13.

Skinner, M.W.; Pavlik, B.M., editors. 1994. California Native Plant Society's inventory of rare and endangered vascular plants of California. Fifth edition. Sacramento, CA:

California Native Plant Society, Special Publication No. 1.

SNEP. 1996. Sierra Nevada Ecosystem Project: Final report to Congress. Status of the Sierra Nevada. Wildland Resources Center Report No. 39. Davis, CA: Centers for Water and Wildland Resources, University of California. Online see http://ceres.ca.gov/snep/ pubs/es.html
Stylinski, C.D.; Allen, E.D. 1999. Lack of native species recovery following severe exotic disturbance in southern Californian shrublands. Journal of Applied Ecology 36: 544-554.

Wells, P.V. 1962. Vegetation in relation to geological substratum and fire in the San Luis Obispo quadrangle, California. Ecological Monographs 32: 79-103.

Zedler, P.H.; Gautier, C.R.; McMaster, G.S. 1983. Vegetation change in response to extreme events: The effect of a short interval between fires in California chaparral and coastal scrub. Ecology 64: 809-818.
106 USDA Forest Service Gen. Tech. Rep. PSW-GTR-195. 2005.

Review
Fire Management Impacts on Invasive Plants in the
Western United States
JON E. KEELEY
U.S. Geological Survey, Western Ecological Research Center, Sequoia-Kings Canyon National Parks, Three Rivers, CA 93271-9651, U.S.A., email jon [email protected] and Department of Ecology and Evolutionary Biology, University of California, Los Angeles, CA 90095, U.S.A.

Abstract: Fire management practices affect alien plant invasions in diverse ways. I considered the impact of six fire management practices on alien invasions: fire suppression, forest fuel reduction, prescription burning in crown-fire ecosystems, fuel breaks, targeting of noxious aliens, and postfire rehabilitation. Most western United States forests have had fire successfully excluded for unnaturally long periods of time, and this appears to have favored the exclusion of alien plant species. Forest fuel reduction programs have the potential for greatly enhancing forest vulnerability to alien invasions. In part this is due to the focus on reestablishing pre-EuroAmerican fire regimes on a landscape that differs from pre-Euro-American landscapes in the abundance of aggressive non-native species. We may be forced to choose between restoring “natural” fire regimes or altering fire regimes to favor communities of native species. Intensive grazing in many western forests may exacerbate the alien problem after fire and temporally decoupling grazing and fire restoration may reduce the alien threat. Many shrubland ecosystems such as the Intermountain West sagebrush steppe or California chaparral have a natural, high-intensity crown fire regime that is less amenable to forest restoration tactics. Historical use of prescribed fire for type conversion of shrublands to more useful grazing lands has played some role in the massive annual grass invasion that threatens these shrublands. Fuel breaks pose a special invasive plant risk because they promote alien invasion along corridors into wildland areas. Use of prescription burning to eliminate noxious aliens has had questionable success, particularly when applied to disturbance-dependent annuals, and success is most likely when coupled with ecosystem restoration that alters the competitive balance between aliens and natives. Artificial seeding of alien species as a form of postfire stabilization appears to cause more problems than it solves and may even enhance alien invasion.

Key Words: exotic plants, fire suppression, fuel breaks, fuel reduction, non-native plants, postfire rehabilitation, prescription burning
Impactos de la Gesti´on de Fuego sobre Plantas Invasoras en el Oeste de Estados Unidos Resumen: Las practicas de gesti ´ on de fuego afectan de diversas maneras a las invasiones de plantas. Con- ´ sider´e el impacto de seis practicas de manejo de fuego sobre las invasiones: supresi ´ on de fuego, reducci ´ on´ de combustible forestal, quema prescrita en ecosistemas con fuego de dosel, guardarrayas, eliminacion de ´ invasoras daninas y rehabilitaci ˜ on post fuego. En la mayor ´ ´ıa de los bosques del oeste de Estados Unidos el fuego ha sido excluido exitosamente por largos per´ıodos de tiempo no naturales y esto parece haber favorecido la exclusion de especies de plantas ex ´ oticas. Los programas de reducci ´ on de combustible forestal tienen el ´ potencial para incrementar la vulnerabilidad de bosques a las invasiones de plantas exoticas. En parte, esto se ´ debe al enfoque en el reestablecimiento de reg´ımenes de fuego pre-Euroamericanos en un paisaje que difiere de paisajes pre-Euroamericanos en la abundancia de especies no nativas agresivas. Podremos ser forzados a elegir entre la restauracion de reg ´ ´ımenes de fuego “naturales” o la alteracion de reg ´ ´ımenes de fuego para favorecer a comunidades de especies nativas. El pastoreo intensivo en muchos bosques occidentales puede exacerbar el problema de invasoras despu´es del fuego y la reduccion temporal de pastoreo y gesti ´ on de incendios ´ puede reducir la amenaza de las invasoras. Muchos ecosistemas con matorrales como la estepa de artemisa Paper submitted October 26, 2004; revised manuscript accepted July 7, 2005. 375
Conservation Biology Volume 20, No. 2, 375–384
C 2006 Society for Conservation Biology
DOI: 10.1111/j.1523-1739.2006.00339.x
376 Fire Management and Invasive Plants Keeley
West Intermountain o el chaparral California tienen un r´egimen natural de fuego de alta intensidad que es menos docil a las t ´ acticas de restauraci ´ on de bosques. El uso hist ´ orico de quemas prescritas para la conversi ´ on´ de terrenos con matorrales a tierras de pastoreo mas ´ utiles ha jugado un papel en la invasi ´ on masiva an- ´ ual de pastos que amenaza a estos terrenos con matorrales. Las guardarrayas constituyen un riesgo especial porque promueven la invasion de ´ areas silvestres a lo largo de corredores. El ´ ´exito del uso de quemas prescritas para eliminar invasoras daninas es cuestionable, particularmente cuando se aplica a anuales dependientes ˜ de perturbacion, y el ´ ´exito es mas probable cuando se combinan con restauraci ´ on de ecosistemas que altera ´ el balance competitivo entre invasoras y nativas. La diseminacion artificial de semillas de especies invasoras ´ como una forma de estabilizacion posterior al fuego parece causar m ´ as problemas que los que resuelve e ´ incluso puede favorecer la invasion de ex ´ oticas. ´
Palabras Clave: guardarrayas, plantas ex´oticas, plantas no nativas, quema prescrita reducci´on de combustible, rehabilitaci´on post fuego, supresi´on de fuego
Introduction
U.S. federal policies incorporate alien plant concerns into management of public lands. For example, the U.S. National Park Service policy (U.S. Department of Interior 2001) mandates that “exotic species will not be allowed to displace native species if displacement can be prevented.” The chief of the U.S. Department of Agriculture Forest Service (USFS) has identified invasive species as one of the four significant threats to U.S. forest and rangeland ecosystems (U.S. Department of Agriculture Forest Service 2004), and the USFS manual states that “determining the risk of noxious weed introduction or spread as part of the NEPA process for proposed actions, especially for ground-disturbing and canopy-altering activities” is the explicit responsibility of managers (U.S. Department of Agriculture Forest Service 1995). Here I explore how these policies may be complicated, and sometimes compromised, by fire management practices. I examined impacts from six fire management practices: (1) fire suppression, (2) fuel reduction in forests, (3) prescription burning in shrublands, (4) fuel breaks, (5) prescription burning to target noxious aliens, and (6) postfire rehabilitation.

Fire Suppression
Fire suppression policy over the past century has worked toward excluding fires from forests. For some forest types, such as Southwest ponderosa pine (Pinus ponderosa
Laws.), the natural fire regime of frequent, low-intensity surface fires has been particularly amenable to fire suppression tactics. Consequently fires have been excluded over a significant portion of the landscape for much of the twentieth century (Allen et al. 2002). There is little debate about the critical nature of the fire hazard due to unnatural accumulation of understory fuels in these and many other western U.S. forest types. These fuels increase the probability of large, high-intensity wildfires and pose a threat to the long-term sustainability of these ecosystems (Graham et al. 2004).

Under this management policy of fire suppression,
however, forests appear to have fared well in terms of minimal alien plant invasion (Pierson & Mack 1990a, 1990b; Weaver et al. 2001; Keeley et al. 2003). One of the major reasons for the resilience to invasion of undisturbed forests is that the closed forest canopy is highly inhibitory to aliens, most of which require high light levels (Rejmanek 1989; Pierson et al. 1990; Charbonneau & Fahrig 2004).

Other factors that potentially play a role are the accumulation of surface litter, which diminishes sites for alien establishment, and reduced propagule sources (dense,
closed canopy forests have little herbaceous growth to
attract livestock).

Fuel Reduction in Forests
The National Fire Plan (U.S. Department of Agriculture
Forest Service 2001) addresses the threat of catastrophic fires by reducing fuels with prescription burning or mechanical thinning. The Healthy Forests Restoration Act of 2003 (House Resolution 1904) increases the ability of resource managers to perform necessary fuel reduction
projects and is called forest restoration because one of its goals is to return forests to their prefire-suppression-era structure and function. Fire lines and firefighting equipment associated with prescription burning directly favor alien species by creating soil disturbances and introducing alien propagules (Harrod & Reichard 2001; Backer et al. 2004), but the impact is potentially much broader. There is growing evidence that these fuel reduction projects alter ecosystem structure in ways that promote alien plant invasion.

Ponderosa pine forests in the Cedar Grove section of
Kings Canyon National Park in the southern Sierra Nevada of California have been managed with prescription burning for more than two decades. The primary goal is to return a quasi-natural fire cycle for the resource benefit of Conservation Biology
Volume 20, No. 2, April 2006
Keeley Fire Management and Invasive Plants 377
these forests. In 1998, however, fire management voluntarily halted this program because of the recognition that associated with prescription burning was an explosion
of cheatgrass (Bromus tectorum L.) in the burned forests (Caprio et al. 1999). Results of experiments on the interaction between cheatgrass and fire show that burning stimulates cheatgrass populations, regardless of whether it is late spring or early fall (T. McGinnis & J.E.K., unpublished data). Based on these studies, the only parameter with potential for inhibiting cheatgrass is accumulation of pine-needle litter, which suggests that lengthening the fire-return interval to significantly exceed the natural cycle may be one of the few options for controlling this alien invader.

Restoration includes restoring not only natural processes such as fire but also natural structure through mechanical thinning of forests, and these practices also may enhance alien invasion. Extensive forest restoration is currently under way in many western U.S. ponderosa pine forests. These treatments alone or in combination with
burning of slash increase both the diversity and abundance of alien plant species (Griffis et al. 2001; Dodson 2004; Wienk et al. 2004). Longer-term studies are needed, however, to determine whether this is a short-lived invasion or whether such practices provide an opportunity for invasives to gain a foothold that will allow long-term persistence in these forests.

These examples suggest a potential conundrum. Forest restoration often has as one of its goals returning the system to historical fire regimes of high fire frequency (Covington & Moore 1994). These historical fires, however, occurred on a landscape that lacked a background of diverse alien species poised to take advantage of such disturbance regimes. This situation may force a choice
between restoring “natural” fire regimes or altering those fire regimes to favor communities of native species. In reality, though, the question is not that simple because reducing the incidence of fire in these ecosystems has
long-term impacts on forest structure, with potential cascading effects on alien species.

Many western U.S. forests have historically had rather
complex fire regimes that included a mixture of surface fires and localized crown fires (Odion et al. 2004b). Lowintensity surface fires removed dead wood and thinned the sapling population, and localized patches of crown
fire created gaps that were essential for reproduction
(Keeley & Stephenson 2000). A century of fire suppression, coupled with other management activities such as grazing and logging, has added greatly to the amount and continuity of understory fuels such that now these perturbed forests face the reality that gaps created by highintensity crown fire will be potentially orders of magnitude larger (Fig. 1). These canopy gaps are sinks for alien invasion (Keeley et al. 2003). Crawford et al. (2001) reported more than a dozen alien species in gaps produced by high-severity wildfires in northern Arizona ponderosa Figure 1. Hypothetical distribution of fire-generated
gaps expected for natural fire regimes and future fire
regimes in Sierra Nevada mixed conifer forests
perturbed by a century of fire exclusion ( from Keeley
& Stephenson 2000).
forests, and these aliens constituted more than a quarter of the understory cover. These invasive species change
the fuel structure of forests (Brooks et al. 2004) and are capable of setting back both natural and artificial regeneration of the dominant forest trees.

There are perhaps ways to minimize effects of alien
species in fuel reduction projects. For example, many
of the aliens Crawford et al. (2001) recorded in their
burned sites were weeds that are often transported by
cattle (Arnold 1950; Wuerthner & Matteson 2002); thus,
prescription burning or logging, when coupled with grazing, may be a dangerous combination, exacerbating the alien invasion problem. This is supported by the report that wildfires in ungrazed ponderosa forests of northern Arizona have relatively few alien species (Laughlin et al. 2004). If there is a connection, then it could be rather large because 70% of the western United States is grazed, including wilderness areas, national forests, and some national parks (Fleischner 1994). I suggest that rotating grazing areas so that livestock are removed for an extended period of time before prescription burning might be one means of reducing alien species’ response to necessary
fuel reduction treatments.

Manipulating fire severity during prescription burning
can also affect the alien response because high-severity gaps are more vulnerable to invasion than low-severity
gaps (Keeley et al. 2003). This, however, is complicated by the requirement of many dominant trees in highseverity gaps for successful seedling recruitment (Keeley & Stephenson 2000).

Manipulating treatment patch size may be another way
of altering the invasive threat. For example, the size of burned patches affects postfire colonization by opportunistic species (Turner et al. 1997). Small patches have a greater perimeter-to-area ratio, making the burned area Conservation Biology
Volume 20, No. 2, April 2006
378 Fire Management and Invasive Plants Keeley
more vulnerable to invasion, whereas large burn patches have a smaller ratio, making the bulk of the burned area less susceptible to colonization from outside alien invaders. The landscape pattern of alien distribution, however, complicates drawing conclusions about community vulnerability to invasion. For example, forest patches adjacent to open habitat are much more susceptible to invasion than forests surrounded by more closed canopy forest (Charbonneau & Fahrig 2004). If aliens are sparsely distributed across the landscape, then small burn patches, despite their high perimeter-to-area ratio, are less likely to encounter alien populations, whereas large patches, with a greater absolute perimeter size, would have a higher
probability of encountering alien populations.

In short, grazing history, alien distribution patterns, treatment size, and fire severity are all factors that might be manipulated to reduce the alien threat linked to necessary fuel-reduction projects. Roads and recreational use are other parameters that interact with fire and invasives (e.g., Gelbard & Belnap 2003) and could be manipulated
in conjunction with fuel treatments to reduce alien invasion.

Prescription Burning in Shrublands
Many shrubland ecosystems such as the Intermountain
West sagebrush steppe or California chaparral have a natural fire regime of high-intensity crown fires. These ecosystems provide fewer options for fuel reduction because mechanical treatments are both expensive and unlikely
to provide commercial profit. Prescription burning is one of the more economically feasible treatments but there
are increasing constraints on its widespread use in shrubland ecosystems because of the hazards of high-intensity fires on populated landscapes. One of the realities of doing prescription burning in crown-fire ecosystems is the difficulty of defining controllable prescriptions (Keeley 2002a). This is particularly problematic for burns in the normal late summer through autumn fire season. One approach is to conduct burns outside the normal fire season, but such manipulations have the potential for extreme resource damage, as illustrated by the poor recovery of the native community and massive alien invasion following a winter burn in one California park (Fig. 2).

For shrublands as well as forests, prescription burning is justified if it provides either resource benefits to the ecosystem or reduces fire hazard for people. In California chaparral, prescription burning is primarily justified on the basis of fire-hazard reduction, whereas in the Intermountain West sagebrush, the primary justification is benefit to ecosystem resources. The most commonly cited resource benefits are improved rangeland for wildlife
(Beardall & Sylvester 1976; Holechek 1981) or livestock (Pechanec 1944; Sapsis & Kaufmann 1991). Other justifications include returning these ecosystems to their hisFigure 2. Alien-grass-dominated scar in chaparral shrublands 10 years after an out-of-season winter
burn in chaparral at Pinnacles National Monument
(central coastal California) (photo by J. Keeley). A
similar effect was also reported for another
cool-season chaparral prescription burn in northern
California (Parker 1987).
torical structure, which is considered by some to have
been a landscape of more open sagebrush steppe vegetation. Indeed, rangeland literature commonly refers to the unnaturally dense stands of sagebrush in need of prescription burning (Blaisdell et al. 1982; Miller et al. 1994).

In light of the massive cheatgrass invasion across much of this landscape (Mack 1981), coupled with the potential for burning to favor cheatgrass expansion (Harnis & Murray 1973; Knapp 1997; Young & Allen 1997), there is
need for a closer examination of prescription burning in these Intermountain West ecosystems.

Prescription burning in sagebrush ecosystems is a
highly effective method of improving rangelands for livestock grazing. The dominant shrub, Artemisia tridentata Nutt., is immediately replaced by more palatable herbaceous plants and recovers slowly over a period of decades (Stewart & Young 1939; Pechanec 1944; Ralphs & Busby
1979). On the other hand, prescription burning for enhancement of wildlife habitat appears to be justifiable in very few cases, and generally the loss of sagebrush following burning represents important habitat loss (Miller & Eddleman 2001; Welch & Criddle 2003). Restoring historical fire regimes is perhaps the weakest justification for prescription burning because many lines of evidence suggest fire-rotation intervals are currently at the low end of the historical range of variability (Menakis et al. 2003).

The natural fire regime in sagebrush ecosystems appears to have been one of infrequent fires at 60- to 110-year intervals (Whisenant 1990; Welch & Criddle 2003; W. Baker, personal communication), although at the mesic end of
the gradient it may have been shorter (Winward 1984).

Thus, except on rangelands where livestock production
Conservation Biology
Volume 20, No. 2, April 2006
Keeley Fire Management and Invasive Plants 379
is the only goal, prescription burning may not be a desirable fire-management treatment because of the potential threat of exacerbating the cheatgrass invasion.

In California chaparral and sage scrub shrublands, a similar annual grass invasion has also occurred, although firemanagement practices for rangeland improvement appear to have played a much bigger role. This began with burning by the Native Americans, largely to favor herbaceous vegetation over shrublands, which set much of the landscape in a quasi-disequilibrium vulnerable to rapid annual plant invasion upon the arrival of Europeans (Keeley 2002b). By the late nineteenth century rangelands were in short supply, widespread burning expanded the grazing
lands, and the coastal analogues of cheatgrass, specifically Bromus madritensis L., B. hordeaceous L., and B. diandrus Roth., and forbs such as Erodium cicutarium (L.) L’Her., rapidly expanded to fill the void created by removing natural shrub dominants (Keeley 1990, 2001, 2004b).

Initially these burning practices were unregulated, but in the mid-twentieth century organized efforts at rangeland expansion into shrublands was a state-sanctioned practice that resulted in substantial conversion to alien grasslands (Keeley & Fotheringham 2003).

Typically a repeat fire within the first postfire decade is sufficient to provide an initial foothold for aliens (Fig. 3).

With the first entry of alien annuals into these shrubland ecosystems, there is a potential shift from a crown-fire regime to a mixture of surface and crown fires, where
highly combustible grass fuels carry fire between shrub patches that have not yet attained a closed canopy capable of carrying crown fire under most weather conditions.

As fire frequency increases there is a threshold beyond which the native shrub cover cannot recover (Zedler et al. 1983; Haidinger & Keeley 1993; Jacobson et al. 2004). Not only do alien grasses increase the probability of burning, but also the shift from crown fires to a mixture of surface and crown fires increases the probability of alien seedFigure 3. Model of fire and alien species interactions in California chaparral.

Figure 4. Type conversion recorded for Malibu
Canyon, Santa Monica Mountains, California: left,
natural chaparral landscape representative of
chaparral in Malibu Canyon (photo by Anna
Jacobsen); right, landscape dominated by alien
annual grass after three fires in 12 years (based on
Jacobson et al. 2004; photo by Steve Davis).
bank survivorship (Keeley et al. 2005) because grass fuels generate lower temperatures (Zschaechner 1985). In these shrublands and in other ecosystems, alien grasses alter fire regimes in ways that enhance their own success, in what has been described as a “grass/fire cycle” (D’Antonio & Vitousek 1992), “niche construction” (Keeley 2001), or “invasive engineering” (Cuddington & Hastings 2004).

In recent years ineffective fire prevention has allowed an unnaturally high number of wildfires on chaparral
landscapes, which has resulted in conversion to aliendominated grasslands (Fig. 4). Such type conversions not only affect biodiversity, but replacing slopes dominated by natural shrublands with grasslands also makes these
landscapes highly vulnerable to major changes in hydrological processes. For example, experimental type conversions performed for fire hazard reduction have resulted in soil slips and other major geomorphological changes (Keeley 2002a).

On shrubland landscapes where the excessive load of
anthropogenic fires has stressed natural ecosystems to the point of collapse, fire managers need to be prudent about adding further fire in the form of prescription burning.

Currently this applies to much of the Great Basin and
all of the lower-elevation foothills in southern California, where type conversion to alien grasslands is happening
at an alarming rate ( J.K., personal observations). To be avoided are prescription burning at fire-return intervals of 5 years in southern California chaparral (Loomis et al. 2003; Gonzalez-Caban et al. 2006), which are likely to lead to type conversion to alien grassland and even exacerbate the sedimentation problems they are supposed to reduce
(Keeley et al. 2004).

Conservation Biology
Volume 20, No. 2, April 2006
380 Fire Management and Invasive Plants Keeley
Fuel Breaks
Forests and shrublands, particularly in California, have had a long history of experimentation with different types of fuel breaks. They are constructed to create barriers to fire spread and to provide access and defensible space for fire-suppression crews during wildfires. These activities have the potential for creating suitable sites for alien plant invasion, and invasion is closely tied to the loss in overstory cover. In a recent study of 24 fuel breaks distributed throughout California, alien plants constituted as much as 70% of the plant cover and the proportion of aliens varied significantly with distance to roads, fuel break age, construction method, and maintenance frequency (Merriam et al. 2006). The association of alien species with fuel breaks raises two critical concerns. One is that the linear connectedness of these disturbance zones acts as corridors for alien invasion into wildland areas. Another is that these zones of reduced fuels produce lower temperatures and thus safe sites for alien propagules during wildfires, ensuring survivorship of seed banks (Keeley
2001, 2004b). Consequently, following fires these fuel
breaks represent a major source area for alien invasion of adjacent wildlands (Fig. 5).

Prescription Burning to Target Noxious Aliens
Fire has diverse effects on alien species, and except for a small handful of cases, it generally promotes persistence of aliens (e.g., Grace et al. 2001; Harrod & Reichard 2001; Brooks et al. 2004). Invasive species in the western United States that seem to be controlled by fire include Mediterranean Basin macchi shrubs known collectively
as “brooms.” Some of these are vigorous resprouters after fire and thus are not readily controlled by burning. OthFigure 5. Interaction between number of fires and distance from the fuel break. Error bars represent + 1
SE ( from Merriam et al. 2006).
ers (e.g., Scot’s broom [Cytisus scoparius (L.) Link]) are weak resprouters, and burning shows promise of control.

All have dormant, fire-stimulated seed banks; thus several repeat fires appear to be required to extirpate brooms
from a site (Tveten & Fonda 1999; Alexander & D’Antonio 2003; Odion & Haubensak 2004), not unlike what happens to native shrublands in the face of repeat fires (Figs. 3 & 4). Burning, however, typically replaces these noxious woody aliens with herbaceous alien species (Keeley 2001).

Several lines of evidence point to precisely timed prescription burning as an effective treatment for eliminating certain noxious alien annuals with transient seed banks that are vulnerable to fire during spring seed dispersal.

One example widely cited in recent alien plant review articles as a demonstration of such success is the application of spring burning in the control of yellow starthistle (Centaurea solstitialis L.). This European pest is distributed from Idaho to California and has been targeted as a particularly noxious alien because it alters range conditions and severely reduces soil water resources (Gerlach 2004).

Confidence in prescribed burning treatment as a control for this species is based on the results of annual burning for 3 consecutive years in very dense stands that demonstrated 90–100% reduction in starthistle (DiTomaso et al. 1999; Odion et al. 2004a). Burn plans written by agencies undertaking prescribed burns in annual grasslands often use this as one of their primary goals (e.g., East Bay Regional Parks, http://www.ebparks.org/fire/rxfire).

This species, however, like many aliens, has a relatively long-lived seed bank (Callihan et al. 1993), and longerterm study shows that this thistle rapidly reestablishes once burning is halted (Fig. 6). Clearly, prescribed burning provides only temporary reduction, does not effect sustainable control of this alien, and may exacerbate the alien situation.

Most alien herbs are opportunistic species that capitalize on disturbance. I offer the hypothesis that when it comes to eliminating such noxious aliens, control is most likely under conditions that limit the use of further disturbances such as fire (or grazing, mowing, or herbicides).

In some cases prescribed fire may be appropriate if applied in a manner that affects the noxious target species more than potential native competitors and if coupled
with active ecosystem restoration that alters the competitive balance between aliens and natives. Sustainable control of these aggressive weeds is most likely going to occur only when natural, intact ecosystems are restored. In the case of yellow starthistle, it invades annual grasslands that owe their origin to disturbance, either displacement of native perennial grassland or type conversion of shrublands and woodlands (Huenneke 1989; Keeley 1990; Hamilton
1997). In the absence of community restoration, prescription burning is likely to provide only temporary control of this, and other, noxious annual weeds, and not be costeffective.

Conservation Biology
Volume 20, No. 2, April 2006
Keeley Fire Management and Invasive Plants 381
Figure 6. Yellow starthistle (a) cover, (b) seed, and (c) seedling production following three consecutive
annual burns applied to extremely dense populations
of this noxious alien weed. Immediate postfire results
were promising (DiTomaso et al. 1999), but follow-up
studies indicate that burning destabilized these
grasslands and allowed subsequent reinvasion once
burning was stopped (Kyser & DiTomaso 2002).

Postfire Rehabilitation
Propagule source is often the limiting step in the invasion process (D’Antonio et al. 2001) and thus postfire management practices such as site stabilization by seeding of non-natives must be considered a potential influence on alien plant invasion. These postfire rehabilitation projects illustrate well the Severide Principle, after the newscaster Eric Severide, who is quoted as saying, “Most problems
begin as solutions.”
Early efforts at such revegetation projects may have
played a role in the spread of some noxious weeds. For example, postfire seeding in southern California chaparral in the 1940s aerially seeded black mustard (Brassica nigra [L.] Koch and possibly related taxa) on steep southern California watersheds (Gleason 1948). These aggressive
weeds soon found their way into citrus orchards and other agricultural fields and were eventually abandoned by fire managers as a suitable slope stabilizer. These species, however, produce polymorphic seed banks with dormant
fire-stimulated germination (Went et al. 1952), and decades later on many of the previously seeded slopes in the Los Angeles Basin this species still figures prominently in the postfire flora as a ghost of seedings past (Keeley et al. 2005). Eventually postfire seeding projects replaced mustard with various grass species (e.g., ryegrass [Lolium multiflorum Lam.], zorro fescue [Vulpia myuros (L.) C.

Gmelin], crested wheatgrass [Agropyron cristatum (L.)
Gaertn.]) that appear to lack persistent seed banks. Although these grasses are not persistent on chaparral or forested slopes (Barclay et al. 2004; Beyers 2004), they are capable of invading adjacent grassland and savanna
communities.

Because they lack an ability to invade communities,
“sterile” or “nonpersistent” cereal grains have been considered a more desirable species for reseeding (Beyers 2004). Although seeding of these species may have achieved some of the intended goals of slowing soil erosion, they have introduced other problems. In one study in the Sierra Nevada the success of wheat seeding was so extraordinary (Fig. 7) that it resulted in the loss of substantial native plant diversity and pine reproduction (Keeley 2004a), a pattern common in many seeding projects (Beyers 2004). Seeding nonpersistent species also carries with it the problem that a marked loss of plant cover in the second postfire year will create an ecological vacuum, and aggressive alien invaders are well suited to exploit this situation.

Increasingly it is apparent that mechanical rehabilitation treatments, including straw mulch and hay bales, are more predictable means of reducing soil erosion and other postfire hydrological problems (Robichaud et al.
2000). Mulching treatments, however, are particularly
hazardous in terms of introducing and promoting alien
establishment (Kruse et al. 2004). In fact, accidental introduction of alien propagules is possible with any “burned Figure 7. Postfire ponderosa pine forest reseeded with
a nonpersistent variety of wheat after fire in the Giant Sequoia National Monument, Fresno County,
California (photo by J. Keeley).

Conservation Biology
Volume 20, No. 2, April 2006
382 Fire Management and Invasive Plants Keeley
area emergency rehabilitation” project. For example, following the 2000 Cerro Grande Fire it is estimated that contamination of aerial seeding sources was responsible for inadvertently broadcasting more than 1 billion cheatgrass seeds on recently burned sites (Keeley et al. 2006).

Conclusions
Fire management practices could have widespread effects on invasions of alien species. This linkage is best understood when these problems are placed in a context of community ecology theory. Fire suppression and prefire
fuel manipulations have ecological equivalents in that the former attempts to maintain ecosystem equilibrium by
preventing disturbance and the latter introduces disequilibrium.

In western U.S. forests, a century of successful fire suppression policy has shifted the competitive balance in favor of long-lived trees that create ecosystem conditions unfavorable to alien invasion. Although greater ecosystem equilibrium appears to exclude alien plants, fire exclusion has set these forests on a trajectory of undesirable conditions for both forest sustainability and human fire hazard.

Thus, forest thinning, fuel breaks, and prescribed burning are necessary and inevitable. But accompanying these management activities is a shift in ecosystem properties that favor early successional species, and when done in the context of a landscape with alien species it is likely to alter the balance of native and non-native species. The impact of these management practices may be altered by
considering management practices that decouple grazing
and burning practices and manipulate burning patterns
in light of prefire alien presence.

In many western U.S. shrubland ecosystems, fire suppression policy—despite valiant efforts—has not kept up with an ever-increasing frequency of fires. These communities exhibit weak resilience to major deviations from the natural crown-fire regime and often the dominant
life forms are lost, creating an ecological vacuum that is rapidly filled by alien weeds. In both the Intermountain West sagebrush and California chaparral (including sage scrub) this alien invasion has historically been exacerbated by fire management practices that included prescription burning for range improvement. Current infestations of annual grasses in both regions require enhanced efforts at fire prevention, fire suppression, and avoidance of prescribed burning under many situations.

Acknowledgments
I thank T. McGinnis and K. Merriam for sharing results
of their field work that contributed insights into these problems, and J. Wade, World Conservation Union, for
bringing to my attention the appropriateness of the Severide Principle to postfire seeding efforts.

Literature Cited
Alexander, J. M., and C. M. D’Antonio. 2003. Seed bank dynamics of French broom in coastal California grasslands: effects of stand age and prescribed burning on control and restoration. Restoration Ecology 11:185–197.

Allen, C. D., M. Savage, D. A. Falk, K. F. Suckling, T. W. Swetnam, T.

Schulke, P. B. Stacey, P. Morgan, M. Hoffman, and J. T. Klingel. 2002.

Ecological restoration of southwestern ponderosa pine ecosystems:
a broad perspective. Ecological Applications 12:1418–1433.

Arnold, J. F. 1950. Changes in ponderosa pine bunchgrass ranges in northern Arizona resulting from pine regeneration and grazing. Journal of Forestry 48:118–126.

Backer, D. M., S. E. Jensen, and G. R. McPherson. 2004. Impacts of firesuppression activities on natural communities. Conservation Biology 18:937–946.

Barclay, A. D., J. L. Betancourt, and C. D. Allen. 2004. Effects of seeding ryegrass (Lolium multiflorum) on vegetation recovery following fire in a ponderosa pine (Pinus ponderosa) forest. International Journal of Wildland Fire 13:183–194.

Beardall, L. E., and V. E. Sylvester. 1976. Spring burning for removal of sagebrush competition in Nevada. Tall Timbers Fire Ecology Conference Proceedings 14:539–547.

Beyers, J. L. 2004. Postfire seeding for erosion control: effectiveness and impacts on native plant communities. Conservation Biology 18:947– 956.

Blaisdell, J. P., R. B. Murray, and E. D. McArthur. 1982. Managing Intermountain rangelands—sagebrush-grass ranges. U.S. Department of Agriculture Forest Service, Intermountain Forest and Range Experiment Station, Provo, Utah.

Brooks, M. L., C. M. D’Antonio, D. M. Richardson, J. M. DiTomaso, J. B.

Grace, R. J. Hobbs, J. E. Keeley, M. Pellant, and D. Pyke. 2004. Effects of invasive alien plants on fire regimes. BioScience 54:677–688.

Callihan, R. H., T. S. Prather, and F. E. Northam. 1993. Longevity of yellow starthistle (Centaurea solstitialis) achenes in soil. Weed Technology 7:33–35.

Caprio, A., S. Haultain, M. B. Keifer, and J. Manley. 1999. Problem evaluation and recommendations: invasive cheatgrass (Bromus tectorum) in Cedar Grove, Kings Canyon National Park. Pages 88–107 in 1999 annual fire report on research, monitoring, and inventory. U.S. National Park Service, Sequoia National Park, Three Rivers, California.

Charbonneau, N. C., and L. Fahrig. 2004. Influence of canopy cover and amount of open habitat in the surrounding landscape on proportion of alien plant species in forest sites. Ecoscience 11:278–281.

Covington, W. W., and M. M. Moore. 1994. Southwestern ponderosa forest structure. Changes since Euro-American settlement. Journal of Forestry 92:39–47.

Crawford, J. A., C.-H. A. Wahren, S. Kyle, and W. H. Moir. 2001. Responses of exotic plant species to fires in Pinus ponderosa forests in northern Arizona. Journal of Vegetation Science 12:261–268.

Cuddington, K., and A. Hastings. 2004. Invasive engineers. Ecological Modeling 178:335–347.

D’Antonio, C., J. Levine, and M. Thomsen. 2001. Ecosystem resistance to invasion and the role of propagule supply: a California perspective.

Journal of Mediterranean Ecology 2:233–245.

D’Antonio, C. M., and P. M. Vitousek. 1992. Biological invasions by exotic grasses, the grass/fire cycle, and global change. Annual Review of Ecology and Systematics 23:63–87.

DiTomaso, J. M., G. B. Kyser, and M. S. Hastings. 1999. Prescribed burning for control of yellow starthistle (Centaurea solstitialis) and enhanced native plant diversity. Weed Science 47:233–242.

Dodson, E. K. 2004. Monitoring change in exotic plant abundance after Conservation Biology
Volume 20, No. 2, April 2006
Keeley Fire Management and Invasive Plants 383
fuel reduction/restoration treatments in ponderosa pine forests of western Montana. M.S. thesis. Department of Forestry, University of Montana, Missoula.

Fleischner, T. L. 1994. Ecological costs of livestock grazing in western North America. Conservation Biology 8:629–644.

Gelbard, J., and J. Belnap. 2003. Roads as conduits for exotic plant invasions in a semiarid landscape. Conservation Biology 17:420–432.

Gerlach, J. D. Jr. 2004. The impacts of serial land-use changes and biological invasions on soil water resources in California, USA. Journal of Arid Environments 57:365–379.

Gleason, C. H. 1948. How to sow mustard in burned watersheds of southern California. California Forest and Range Experiment Station, Berkeley.

Gonzalez-Caban, A., P. Wohlgemuth, J. B. Loomis, and D. R. Weise. 2006.

Costs and benefits of reducing sediment production from wildfires through prescribed burning: the Kinneloa fire case study. In press in general technical report. U.S. Department of Agriculture Forest Service Pacific Southwest Experiment Station, Albany, California.

Grace, J. B., M. D. Smith, S. L. Grace, S. L. Collins, and T. J. Stohlgren. 2001. Interactions between fire and invasive plants in temperate grasslands of North America. Pages 40–65 in K. E. M. Galley and T.

P. Wilson, editors. Proceedings of the invasive species workshop:
the role of fire in the control and spread of invasive species. Miscellaneous publication 11. Tall Timbers Research Station, Tallahassee, Florida.

Graham, R. T., S. McCaffrey, and T. B. Jain. 2004. Science basis for changing forest structure to modify wildfire behavior and severity. General technical report RMRS-GTR-120. U.S. Department of Agriculture Forest Service, Rocky Mountain Research Station, Missoula, Montana.

Griffis, K. L., J. A. Crawford, M. R. Wagner, and W. H. Moir. 2001. Understory response of management treatments in northern Arizona ponderosa pine forests. Forest Ecology and Management 146:239– 245.

Haidinger, T. L., and J. E. Keeley. 1993. Role of high fire frequency in destruction of mixed chaparral. Madrono˜ 40:141–147.

Hamilton, J. G. 1997. Changing perceptions of pre-European grasslands in California. Madrono˜ 44:311–333.

Harnis, R. O., and R. B. Murray. 1973. 30 years of vegetal change following burning of sagebrush-grass range. Journal of Range Management 26:322–325.

Harrod, R. J., and S. Reichard. 2001. Fire and invasive species within the temperate and boreal coniferous forests of western North America.

Pages 95–101 in K. E. M. Galley and T. P. Wilson, editors. Proceedings of the invasive species workshop: the role of fire in the control and spread of invasive species. Miscellaneous publication 11. Tall Timbers Research Station, Tallahassee, Florida.

Holechek, J. L. 1981. Brush control impacts on rangeland wildlife. Journal of Soil & Water Conservation 36:265–269.

Huenneke, L. F. 1989. Distribution and regional patterns of Californian grasslands. Pages 1–12 in L. F. Huenneke and H. A. Mooney, editors. Grassland structure and function. California annual grasslands.

Kluwer Academic, Dordrecht, Netherlands.

Jacobson, A. L., S. D. Davis, and S. L. Babritius. 2004. Fire frequency impacts non-sprouting chaparral shrubs in the Santa Monica Mountains of southern California. In M. Arianoutsou and V. P. Panastasis, editors. Ecology, conservation and management of mediterranean climate ecosystems. Millpress, Rotterdam, Netherlands.

Keeley, J. E. 1990. The California valley grassland. Pages 2–23 in A.

A. Schoenherr, editor. Endangered plant communities of southern California. Southern California Botanists, Fullerton.

Keeley, J. E. 2001. Fire and invasive species in mediterranean-climate ecosystems of California. Pages 81–94 in K. E. M. Galley and T. P. Wilson, editors. Proceedings of the invasive species workshop: the role of fire in the control and spread of invasive species. Miscellaneous publication 11. Tall Timbers Research Station, Tallahassee, Florida.

Keeley, J. E. 2002a. Fire management of California shrubland landscapes.

Environmental Management 29:395–408.

Keeley, J. E. 2002b. Native American impacts on fire regimes in California coastal ranges. Journal of Biogeography 29:303–320.

Keeley, J. E. 2004a. Ecological impacts of wheat seeding after a Sierra Nevada wildfire. International Journal of Wildland Fire 13:73–78.

Keeley, J. E. 2004b. Invasive plants and fire management in California mediterranean-climate ecosystems. In M. Arianoutsou, editor. 10th MEDECOS—international conference on ecology, conservation and management. Rhodes, Greece.

Keeley, J. E., and C. J. Fotheringham. 2003. Impact of past, present, and future fire regimes on North American mediterranean shrublands.

Pages 218–262 in T. T. Veblen, W. L. Baker, G. Montenegro, and T. W.

Swetnam, editors. Fire and climatic change in temperate ecosystems of the western Americas. Springer-Verlag, New York.

Keeley, J. E., and N. L. Stephenson. 2000. Restoring natural fire regimes in the Sierra Nevada in an era of global change. Pages 255–265 in D. N. Cole, S. F. McCool, and J. O’Loughlin, editors. Wilderness science in a time of change conference. RMRS-P-15, Volume 5. U.S. Department of Agriculture Forest Service, Rocky Mountain Research Station, Missoula, Montana.

Keeley, J. E., C. J. Fotheringham, and M. A. Moritz. 2004. Lessons from the October 2003 wildfires in southern California. Journal of Forestry 102(7):26–31.

Keeley, J. E., M. Baer-Keeley, and C. J. Fotheringham. 2005. Alien plant dynamics following fire in mediterranean-climate California shrublands of California. Ecological Applications 15:2109–2125.

Keeley, J. E., D. Lubin, and C. J. Fotheringham. 2003. Fire and grazing impacts on plant diversity and alien plant invasions in the southern Sierra Nevada. Ecological Applications 13:1355–1374.

Keeley, J. E., C. D. Allen, J. Betancourt, G. W. Chong, C. J. Fotheringham, and H. D. Safford. 2006. A 21st century perspective on postfire seeding. Journal of Forestry 104:1–2.

Knapp, P. A. 1997. Cheatgrass (Bromus tectorum L.) dominance in the Great Basin Desert. Global Environmental Change 6:37–52.

Kruse, R., E. Bend, and P. Bierzychudek. 2004. Native plant regeneration and introduction of non-natives following post-fire rehabilitation with straw mulch and barley seeding. Forest Ecology and Management 196:299–310.

Kyser, G. B., and J. M. DiTomaso. 2002. Instability in a grassland community after the control of yellow starthistle (Centaurea solstitialis) with prescribed burning. Weed Science 50:648–657.

Laughlin, D. C., J. D. Bakker, M. T. Stoddard, M. L. Daniels, J. D. Springer, C. N. Gilar, A. M. Green, and W. W. Covington. 2004. Toward reference conditions: wildfire effects on flora in an old-growth ponderosa pine forest. Forest Ecology and Management 199:137–152.

Loomis, J., P. Wohlgemuth, A. Gonzalez-Caban, and D. English. 2003.

Economic benefits of reducing fire-related sediment in southwestern fire-prone ecosystems. Water Resources Research 39:1–8.

Mack, R. N. 1981. Invasion of Bromus tectorum L. into western North America: an ecological chronicle. Agro-Ecosystems 7:145–165.

Menakis, J. P., D. Osborne, and M. Miller. 2003. Mapping the cheatgrasscaused departure from historical natural fire regimes in the Great Basin, USA. Pages 281–287 in Proceedings RMRS-P-29. U.S. Department of Agriculture Forest Service, Rocky Mountain Research Station, Missoula, Montana.

Merriam, K. E., J. E. Keeley, and J. L. Beyers. 2006. Fuel breaks affect nonnative species abundance in California plant communities.

Ecological Applications 16:515–527.

Miller, R. F., and L. L. Eddleman. 2001. Spatial and temporal changes of sage grouse habitat in the sagebrush biome. Technical bulletin 151.

Agricultural Experiment Station, Oregon State University, Corvallis.

Miller, R. F., T. J. Svejcar, and N. E. West. 1994. Implications of livestock grazing in the intermountain sagebrush region: plant composition.

Pages 101–146 in M. Vavra, W. A. Laycock, and R. D. Pieper, editors.

Ecological implications of livestock herbivory in the West. Society for Range Management, Denver, Colorado.

Odion, D. C., and K. A. Haubensak. 2004. Response of French broom to fire. Pages 296–307 in N. G. Sugihara, M. E. Morales, and T. J.

Conservation Biology
Volume 20, No. 2, April 2006
384 Fire Management and Invasive Plants Keeley
Morales, editors. Proceedings of the symposium: fire in California ecosystems: integrating ecology, prevention and management. Miscellaneous publication 2. Association for Fire Ecology, Berkeley, California.

Odion, D. C., J. Alexander, and M. Swezy. 2004. Use of short rotation burning to combat non-natives and their seed banks in California North Coastal Prairie. Pages 46–57 in N. G. Sugihara, M. E. Morales, and T. J. Morales, editors. Proceedings of the symposium: fire management: emerging policies and new paradigms. Miscellaneous publication 2. Association for Fire Ecology, Berkeley, California.

Odion, D. C., E. J. Frost, J. R. Strittholt, H. Jiang, D. A. DellaSala, and M. A. Moritz. 2004. Patterns of fire severity and forest conditions in the western Klamath Mountains, California. Conservation Biology 18:927–936.

Parker, V. T. 1987. Effects of wet-season management burns on chaparral vegetation: implications for rare species. Pages 233–237 in T. S. Elias, editor. Conservation and management of rare and endangered plants.

California Native Plant Society, Sacramento.

Pechanec, J. F. 1944. Sagebrush burning—good and bad. Farmers’ Bulletin No. 1948. U.S. Department of Agriculture, Washington, D.C.

Pierson, E. A., and R. N. Mack. 1990a. The population biology of Bromus tectorum in forests: distinguishing the opportunity for dispersal from environmental restriction. Oecologia 84:519–525.

Pierson, E. A., and R. N. Mack. 1990b. The population biology of Bromus tectorum in forests: effect of disturbance, grazing, and litter on seedling establishment and reproduction. Oecologia 84:526– 533.

Pierson, E. A., R. N. Mack, and R. A. Black. 1990. The effect of shading on photosynthesis, growth, and regrowth following defoliation for Bromus tectorum. Oecologia 84:534–543.

Ralphs, M. H., and F. E. Busby. 1979. Prescribed burning: vegetative change, forage production, cost, and returns on six demonstration burns in Utah. Journal of Range Management 32:267–270.

Rejmanek, M. 1989. Invasibility of plant communities. Pages 369–388 in J. A. Drake, H. A. Mooney, F. diCastri, R. H. Groves, F. J. Kruger, and M. Rejmanek, editors. Biological invasions: a global perspective.

John Wiley & Sons, New York.

Robichaud, P. R., J. L. Beyers, and D. G. Neary 2000. Evaluating the effectiveness of postfire rehabilitation treatments. General technical report RMRS-GTR-63. U.S. Department of Agriculture, Rocky Mountain Research Station, Fort Collins, Colorado.

Sapsis, D. B., and J. B. Kauffman. 1991. Fuel consumption and fire behavior associated with prescribed fires in sagebrush ecosystems.

Northwest Science 65:173–179.

Stewart, G., and A. E. Young. 1939. The hazard of basing permanent grazing capacity on Bromus tectorum. Journal of the American Society of Agronomy 31:1002–1015.

Turner, M. G., W. H. Romme, R. H. Gardner, and W. W. Hargrove. 1997.

Effects of fire size and pattern on early succession in Yellowstone National Park. Ecological Monographs 67:411–433.

Tveten, R. K., and R. W. Fonda. 1999. Fire effects on prairies and oak woodlands on Fort Lewis, Washington. Northwest Science 73:145– 158.

U.S. Department of Agriculture Forest Service (USFS). 1995. U.S. Forest Service manual. USFS, Washington, D.C.

U.S. Department of Agriculture Forest Service (USFS). 2001. National fire plan. USFS, Washington, D.C.

U.S. Department of Agriculture Forest Service (USFS). 2004. National strategy and implementation plan for invasive species management.

USFS, Washington, D.C.

U.S. Department of the Interior (USDI). 2001. National Park Service policy. USDI, National Park Service, Washington, D.C.

Weaver, T., D. Gustafson, and J. Lichthardt. 2001. Exotic plants in early and late seral vegetation of fifteen northern Rocky Mountain environments (HTs). Western North American Naturalist 61:417–427.

Welch, B. L., and C. Criddle. 2003. Countering misinformation concerning big sagebrush. Research paper RMRS-RP-40. U.S. Department of Agriculture Forest Service, Rocky Mountain Research Station, Fort Collins, Colorado.

Went, F. W., G. Juhren, and M. C. Juhren. 1952. Fire and biotic factors affecting germination. Ecology 33:351–364.

Whisenant, S. G. 1990. Changing fire frequencies on Idaho’s Snake River plains: ecological and management implications. Pages 4–10 in D.

D. McArthur, E. M. Romney, S. O. Smith, and P. T. Tueller, editors.

Proceedings, symposium on cheatgrass invasion, shrub die-off and other aspects of shrub biology and management. General technical report INT-276. U.S. Department of Agriculture Forest Service, Intermountain Research Station, Ogden, Utah.

Wienk, C. L., C. H. Sieg, and G. R. McPherson. 2004. Evaluating the role of cutting treatments, fire and soil seed banks in an experimental framework in ponderosa pine forests of the Black Hills, South Dakota. Forest Ecology and Management 192:375–393.

Winward, A. H. 1984. Fire in the sagebrush-grass ecosystem—the ecological setting. Pages 2–6 in K. Sanders and J. Durban, editors. Rangeland fire effects, a symposium. U.S. Department of Interior, Bureau of Land Management, Boise, Idaho.

Wuerthner, G., and M. Matteson. 2002. Welfare ranching. The subsidized destruction of the American West. Island Press, Covelo, California.

Young, J. A. and F. L. Allen. 1997. Cheatgrass and range science: 1930– 1950. Journal of Range Management 50:530–535.

Zedler, P. H., C. R. Gautier, and G. S. McMaster. 1983. Vegetation change in response to extreme events: the effect of a short interval between fires in California chaparral and coastal scrub. Ecology 64:809–818.

Zschaechner, G. A. 1985. Studying rangeland fire effects: a case study in Nevada. Pages 66–84 in K. Sanders and J. Durham, editors. Rangeland fire effects, a symposium. Bureau of Land Management, Boise, Idaho.

Conservation Biology
Volume 20, No. 2, April 2006
1 Rural and non-rural digital divide persists in older
adults: Internet access, usage, and attitudes
toward technology
Hee Yun Lee PhD MSWa, Shaheen Kanthawala PhDb, Eun Young Choi MAc, Young Sun Kim PhDd,*
a
School of Social Work, the University of Alabama, Tuscaloosa, USA; bDepartment of Journalism & Creative Media, the University of Alabama, Tuscaloosa, USA; cSchool of Gerontology, University of Southern California, Los Angeles, USA; dDepartment of Gerontology, AgeTech-Service Convergence Major, Kyung Hee University, Suwon, South Korea; *Corresponding author: [email protected]
Abstract
Background: The digital age divide remains persistent despite the recent increase in internet use among older adults. Additionally, older rural residents are at greater risk of being digitally disconnected.

Objective: Guided by the social determinants of health framework, our study aims to examine how one’s residential area relates to (1) internet use, (2) subtypes of usage patterns, and (3) attitudes toward technology use in later life.

Method: Cross-sectional data were drawn from the 2012 Health and Retirement Study.

The analytic sample consisted of 1,566 older adults aged 50 and above. Chi-square tests and logistic regression analyses were performed.

Results: Internet access rates were significantly lower in rural residents (54%) compared to the urban (66%) or the suburban group (61%). Compared to urban residents, those residing in suburban areas were less likely to use health technology while those living in rural areas had lower odds of communication, financial, and media technology use. Furthermore, the association between urban-rural residence and attitudes toward technology was compared among non-users (N = 633). Older adults in rural areas showed more unfavorable perceptions of technology than urban residents. They were less likely to conceive technology as “easily available,” but more likely to perceive it as “too complicated” and “too hard to learn.” No significant differences were found between rural and suburban residents.

Conclusion: Our findings suggest that older adults in rural areas, notably lag in using and adopting digital technology. Comprehensive intervention efforts are needed to narrow the digital divide for rural communities.

Keywords: technology, digital divide, rural, urban, attitude, older adults Original Research
Introduction
With the pervasive spread of technology and
digital technologies used in every aspect of life,
research has noted the development of a digital
divide (Ball et al., 2019; van Dijk, 2006). As elaborated through these studies, a digital divide refers to a division in access to digital technologies for
certain populations. More specifically, van Dijk
and Hacker (2003) proposed four access-related
gaps that might lead to a digital divide. The first
was a mental gap, which refers to emotional or
psychological gaps that emerge from people’s
lack of experience with digital technologies. Next
was the material access gap, referring to the more
traditional lack of access, such as lack of access
to technologies. Third, they referred to a lack of
skills, referring to how experience with technologies would increase an individual’s technology skills. Lastly, they mention the usage gap, referring to differences in usage that lead to various usage patterns (such as simpler communication
tasks only vs. more complex, social capital building tasks, for instance). Despite these multiple aspects of digital technology use, most previous
studies on the digital divide have been primarily
focused on accessibility (whether the participants
can use the technology or not), leaving usage patterns and personal attitudes underexplored.

The digital divide literature has conceptualized
three levels of the digital divide: (1) the first-level divide focusing on disparities in Internet access;
(2) the second-level divide referring to unequal
Internet skills and usage; (3) the third-level divide related to the outcomes of Internet use (van Dijk, 2020). While more recent research attention shifts towards the second-and third-level divide as mobile broadband becomes nearly
ubiquitous, the first-level digital divide still calls
for further investigation in the context of socioLee et al., Gerontechnology (2021) 20:2 https://doi.org/10.4017/gt.2021.20.2.32-472.12
2 Urban/rural digital divide among older adults
economic disparities (Gonzales, 2016). On the
other hand, theoretical frameworks have suggested two distinct reasons for the non-use of the Internet. First, a lack of material resources may
push individuals involuntarily away from Internet
use (Livingstone & Helsper, 2007). Second, psychological factors such as motivation, attitudes, and interest may drive individuals’ voluntary decisions not to use digital technologies (van Dijk, 2005). Studies found that inequalities in Internet
access are likely to be attributable to involuntary
exclusion, while disparities in Internet skills and
usage are also shaped by personal motivations
and attitudes (Eynon & Helsper, 2011; Yu et al.,
2016). The current study aims to investigate the
role of one’s residential area in relation to the
first- and second-level digital divides using a
representative sample of older Americans. We
consider internet access and specific types of
technology usage as outcome variables. In addition, we focus on non-users of communication technologies to examine what contributes to
their attitudes toward technology.

It is important to note that the digital divide is nuanced and shows different patterns of change depending on which aspects are considered: some digital gaps are narrowing down over the last
few decades while others remain. For example,
the rate of Americans who don’t use the internet
has decreased rapidly from 48% in 2000 to 7%
in 2021 (Perrin & Atske, 2021). Similarly, recent
studies found that the vast majority of people
have a smartphone or technological divide (Barrantes & Vargas, 2017). While the general access to the internet or technology has been improved,
individuals with low-income, rural residents,
and minorities have a limited number of digital
devices for online access and rely primarily on
smartphones (Vogels, 2021a). They are also less
likely to have broadband internet at home, creating other disparities such as educational and health inequalities (Ong, 2020; Vogels, 2021b).

The digital divide leads to the exclusion of populations based on different factors. Age has been a major leading factor in the digital divide, placing older adults into one of the largest affected populations (Tsai et al., 2015). As of 2019, 27%
of Americans aged 65 and older and 12% of
adults aged 50 to 64 were not using the internet,
which is a sharp contrast to the nearly ubiquitous internet usage among young adults; 100% of 18–29-year-olds and 97% of 30–49-year-olds
reported using the internet (Pew Research Center,
2019). Although one of the largest populations affected by the digital divide is older adults (Tsai et al., 2015), digital technologies may be most beneficial to this population. Older adults’ technology use spans various avenues of life. Using information and communication technologies (ICTs) has shown to have increased feelings of mattering
among older adults (Francis et al., 2019), connecting them with loved ones (Heo et al., 2015; Quan-Haase et al., 2017; Sum et al., 2008), and
assisting with managing their health through medical and health-related technologies (Levine et al., 2016; van Deursen, 2020). With increasing medical treatment improvements, older adults account for large percentages of both the U.S. and global
populations. By 2030, 20% of the U.S. population will be older adults – the largest population of older adults in U.S. history (Centers for Disease
Control and Prevention., 2013). Given the beneficial effects of technology use and demographic changes, more research efforts are needed to
close the older population's digital divide.

In addition to the generally low technology
adoption rates among older adults, it is vital
to understand the within-group heterogeneity.

Those with lower socioeconomic status and poor
health have been reported to be at greater risk of
being excluded from technology use (Gell et al.,
2015; Silver, 2014). Amongst various predictors
of older adults’ technology use, a recent growing body of literature has shown that rural living is significantly associated with limited access to
the internet and electronic devices (Berner et al.,
2015; Calvert et al., 2009). Moreover, in qualitative studies, older adults living in suburban or rural areas reported negative perceptions of using technologies, such that lack of knowledge about
how technologies function or difficulty in learning how to use them greatly limited their use (Marston et al., 2019; O’Brien et al., 2014). Prior
studies suggested low rural residents' socioeconomic status levels partly contributing to the urban-rural differences in technology use (Hale et al., 2010). However, underdeveloped infrastructure and broadband services in rural communities, which are related to high-speed internet access, can also result in their residents’ limited
technology use (Anderson, 2018; Greenberg et
al., 2018; Korupp, 2005). Indeed, having access
to technology, particularly high-speed internet,
is a critical facilitator of older adults’ technology
use (Hanson, 2010; Marston et al., 2019).

The observed digital divide by age and residential area is well encapsulated in the social determinants of health framework (World Health Organization [WHO], 2010), which posits that
the source of health inequities originates from
“the conditions in which people are born, grow,
work, live, and age, and the wider set of forces
and systems shaping the conditions of daily life.”
(WHO, 2020, p.1). According to the framework,
the digital divide could occur due to broader
systemic and socioeconomic contexts and individual-level sociodemographic factors (i.e., age and gender). Thus, we argue that a geographic
3 Urban/rural digital divide among older adults
residence where older adults live would be an
important social determinant of digital technology use by offering different opportunities and challenges embedded in regional infrastructure
and community support.

The present study
Although prior studies suggested some valuable insights into the urban-rural digital divide for the older population, most relied on small
convenient samples or measured a single aspect
of technology use (i.e., internet access). In addition, there were few studies focused on different usage patterns or psychological factors. The current study aims to investigate whether one’s
residential area relates to (1) their internet use, (2) the subtypes of their technology usage patterns,
and (3) their attitudes toward technology use.

We conducted a series of regression analyses on
data of a nationally representative sample of U.S.
adults aged 50 or older. Building on the previous
literature and the social determinants of health
framework, we hypothesized the following:

Hypothesis 1: Older adults living in suburban
and rural areas will use the internet less than
their urban counterparts.

Hypothesis 2: Older adults living in suburban
and rural areas will be less likely to use different
technologies (communication, financial, health,
and media) than urban residents.

Hypothesis 3: Older adults living in suburban
and rural areas will have more unfavorable attitudes toward technologies than those residing in urban communities.

Methods
Data and sample
The Health and Retirement Study (HRS) is a nationally representative panel survey of approximately 20,000 Americans aged 50 or older. The HRS has been repeated biannually since 1992.

Participants are asked a wide range of questions
on aging (e.g., health, jobs, retirement, and social
relationships) through face-to-face or telephone
interviews. The HRS’s sampling design, survey
procedures, and questionnaires are described
in detail at the HRS website (http://hrsonline.isr.
umich.edu). We retrieved the respondent’s socioeconomic and health information from the RAND HRS data file, which is cleaned and processed by
the RAND Center for the Study of Aging (Santa
Monica, California). The RAND HRS data set
provides imputed values for income and health
status that were generated using all information
available with a consistent imputation method.

In 2012, a randomly selected subsample of participants was administered to an experimental module that included questions about technology
use and perceptions of the barriers and benefits.

A total of 1,620 community-living adults aged 50
or older completed the module. We excluded 54
respondents (3.33%) who had missing data on
covariates, internet access, and internet usage
variables, which resulted in an analytic sample of
1,566 individuals. Among the respondents, those
who did not use any communication technologies such as email and social network sites (N = 633) were further asked about their attitudes toward technology. We use this subsample of respondents for the analyses of non-users ’ attitudes toward technology. Ethical approval for the HRS
was obtained from the University of Michigan Institutional Review Board and the National Institute on Aging. All respondents have given their written
informed consent before the data collection.

Measures
Internet access
Participants’ internet access was assessed with
a single item: "Do you regularly use the Internet (or the World Wide Web) for sending and receiving email or for any other purpose, such
as making purchases, searching for information,
or making travel reservations?’’ Responses were
coded as 1 = yes or 0 = no.

Technology usage
Technology usage was measured by a checklist of various kinds of electronic technologies.

The list included fifteen items (1 = yes for each),
which can be grouped into four broad types of
technology: (1) communication technology (i.e.,
email, social networks, online calls, online chatting, and smartphone); (2) financial technology (i.e., online bill payment and online banking); (3)
health technology (i.e., online wellness program,
online health information seeking, devise use of
health monitoring, and Wii fit use); and (4) media
technology (i.e., e-readers/tablets, MP3 players,
live-streaming radio, T.V., or movies, and video
games). Responses were dichotomized for each
technology category, where 0 = non-user and 1
= users who checked one or more technology
items in a given category.

Attitudes toward technology
Participants who reported no use of any communication technologies, or non-users, were further asked about their attitudes toward technology.

They answered yes or no to eight questions: (1) if
they would be interested in trying any communication technologies, (2) whether technology is too expensive, (3) easily available, (4) too complicated, (5) too hard to learn, (6) takes too much time to learn, (7) if they are opposed to learning
new technologies, and (8) whether it is difficult
to keep up with changes in technology.

Urban-rural residence
The residency was categorized as urban, suburban, and rural using the 2013 Beale Rural-Urban 4 Urban/rural digital divide among older adults
Continuum (RUC) codes. Developed by the
S.U.S. Department of Agriculture’s Economic
Research Service, the RUC codes represent the
degree of urbanization by population size and
adjacency to metropolitan areas. Urban refers to
metropolitan counties with a population size of
1 million or more (RUC code 1). Suburban indicates metropolitan counties of 250,000 to 1 million population (RUC code 2). Finally, rural
areas include counties with fewer than 250,000
population, both adjacent or not adjacent to a
metropolitan area (RUC code 3–9).

Covariates
Age (in years), gender (0 = male; 1 = female), race/
ethnicity (0 = non-Hispanic White; 1 = racial/
ethnic minority [non-Hispanic black, Hispanic,
American Indian, Alaskan Native, Asian, or Pacific Islander]), and marital status (0 = separated, divorced, widowed, or never married; 1 = married or partnered) were included as demographic variables. Educational attainment (in years) and
annual household income served as indicators of
socioeconomic status. Education was assessed
by the number of years of formal schooling completed (range: 0-17), and annual household income included income from all sources such as earnings, pensions, and social security. Because
the distribution of annual household income
was highly skewed, log-transformed values were
used for multivariate analyses. As for indicators
of health conditions, chronic diseases, functional limitations, self-rated health, and depressive symptoms were considered. Participants were
asked to report whether or not they had any of
the following chronic diseases diagnosed by a
physician: high blood pressure, diabetes, cancer,
lung disease, coronary heart disease, psychiatric problems, arthritis or rheumatism, and stroke.

The number of chronic diseases was calculated
by summing up all specific diseases (range 0-8).

Functional limitation was assessed by asking
participants’ difficulties in performing activities
of daily living (ADLs) such as bathing, dressing,
walking across a room, and getting in and out of
bed. The composite scores were dichotomized
into 0 = no functional limitation and 1 = one
or more functional limitations. Self-rated health
status was measured with a single item, “Would
you say your health is excellent, very good, good,
fair, or poor?” Responses were coded as 1 = poor
and 5 = excellent so that higher scores indicated
better subjective health status. Depressive symptoms were measured with the 8-item version of the Center for Epidemiological Studies Depression (CES-D) scale (Andresen et al., 1994; Radloff, 1977). Participants were asked to report whether
they have experienced eight depressive symptoms (e.g., lonely, feeling depressed, and sad) in the past week. Response options were 0 = no or
1 = yes. After reversing two positive mood items,
total scores were computed by adding responses
for each item. The CES-D scale's internal consistency was high in our sample, with the KuderRichardson 20 (KR-20) coefficient of .82.

Analytical strategy
Descriptive statistics were conducted to review
the sample’s demographic, socioeconomic, and
health characteristics by urban-rural residence.

Comparisons were made using adjusted Wald or
chi-square statistics. Next, group differences in
internet access and usage were compared across
urban, suburban, and rural residents through
chi-square tests. Separate multivariate logistic
regression analyses were performed for each
technology used to examine their association
with rural-urban residence (N = 1,566). Similarly,
we used multivariate logistic regression models
to investigate the association between rural-urban residence and attitudes toward technology among non-users (N = 633). For the analysis
of non-users’ attitudes toward technology, the
pairwise deletion was applied to make optimal
utilization of available data (Ns = 556–629). We
used sample weights to adjust for differential
5 Urban/rural digital divide among older adults
sampling probabilities and survey non-response.

All analyses were performed using Stata version

14.2. (StataCorp. College Station, TX).

Results
Sample characteristics
Table 1 summarizes the sample characteristics
by rural-urban residence. The sample was comprised of 52% urban, 23% sub-urban, and 25% rural residents. Compared to urban residents, the
rural group included a lower proportion of racial/ethnic minorities and those with functional limitations. In addition, rural residents had lower
annual household income levels and poorer selfrated health than their urban counterparts. Both suburban and rural residents had fewer years of
education and reported more chronic diseases
compared to the urban group. No significant
group differences were found for age, gender,
marital status, and depressive symptoms.

Urban-rural group differences in internet access and technology usage Table 2 presents a cross-tabulation of internet
access and different types of technology usage
by urban-rural residents. About 66% and 61% of
urban and suburban residents were regular internet users, whereas only 54% of rural residents regularly used the internet. In terms of communication technology use, significantly fewer individuals in the rural group (60%) reported use of at least one communication technology than
their urban counterparts (71%). For specific technology, significance was found in email use (55% in rural vs. 69% in urban), online chatting (17%
in rural vs. 23% in urban), and smartphone use
(16% in rural and 24% in sur-urban vs. 32% in
urban). Compared to urban residents, suburban
and rural residents had fewer individuals who
used one or more financial technologies (38%
in rural and 43% in suburban vs. 52% in urban).

About a third of suburban and rural residents
used online bill payment, while more than
40% of urban residents
did so. Rates of online
banking users were
also significantly lower
among the rural group
than their urban counterparts. Similarly, there
were fewer individuals from the suburban
group who used at least
one health technology
than the urban group.

More urban residents
sought online health information than the other
two groups (57% in urban vs. 43% in sub-urban
and 44% in rural). Finally, the rural group had
significantly fewer media technology users (37%)
compared to the urban group (53%). Specifically, significant differences between urban and rural residents were found in all sub-categories,
whereas the differences between urban and suburban were only found in the use of MP3 players and live-streaming radio, T.V., or movies.

Logistic regression models of technology usage
We further examined the observed group differences in technology usage with multivariate logistic regression analyses. Table 3 shows the association between rural-urban residence and four different technology use types after controlling for
demographic, socioeconomic, and health covariates. Compared to urban residents, rural residents had lower odds of using all technology types except health technology. On the other hand, the suburban group showed decreased odds of using
health technologies than the urban group. No significant associations were found in the use of communication, financial, and media technologies.

Logistic regression models of non-users’ attitudes
toward technology
Table 4 presents the models predicting attitudes
toward technology among non-users of communication technology. We found that three particular attitudes toward technology were significantly associated with the rural-urban residency.

Compared to urban residents, rural residents
were less likely to report that technology is easily
available and more likely to think that technology
is too complicated and too hard to learn. It was
noteworthy that no significant differences were
found between suburban and urban residents
in their attitudes. The associations of the ruralurban residence with other five types of attitudes χ
6 Urban/rural digital divide among older adults
toward technology (i.g., have interest in trying
technology, technology is too expensive, takes
too much time to learn, too difficult to keep up
with changes, and if they are opposed to learning new technologies) were not significant and thus not shown in the table for simplicity.

Discussion
Informed by the social determinants of health
framework, we analyzed the HRS data to explore whether one’s residence (urban, suburban, or rural) is associated with less use of the internet,
limited usage patterns of different technologies,
and more negative attitudes toward technology
in later life. Our bivariate results indicate that
individuals in rural communities use the internet at a lower rate (54%) as compared to urban populations (66%). Furthermore, rural living
older adults used technology at lower rates for
different technology usage sub-types, including
communication, financial, and media technologies. Compared to urban residents, older adults living in suburban areas also reported lower usage rates of certain types of technology use, such as smartphones, online bill payments, and online
health information seeking.

Our multivariate models of technology usage
further showed that rural residency (vs. urban)
was significantly associated with decreased odds
of using communication, financial, and media
technology, while suburban residency (vs. urban) was related to lower odds of health technology use. Overall, these findings add to the body of past research, pointing out a digital exclusion of rural populations in general (Greenberg et al., 2018; Perrin & Duggan, 2015) and in the
older populations (Berner et al., 2015; Calvert et
al., 2009). Importantly, our analyses on various
usage patterns of ICTs extend the scope of the
existing literature where a single aspect of technology use (i.e., internet use) was often focused.

Our findings documented that older rural residents particularly do not engage in online communication and media. Such exclusion could lead to many disadvantages, including connection with loved ones and personal mattering.

This mattering and connectedness often occur
through the usage of ICTs, which allow for the
maintenance of social connectedness and receiving social support from their networks (Francis et al., 2019). Mattering has been explained as occurring due to social interactions among
people, which helps reinforce their value to others (Fazio, 2009), but in situations where in-person interactions may not be possible, especially social interactions that reinforce ones’ value to
others (Fazio, 2009). Therefore, in lieu of and in
addition to face-to-face interactions, using ICTs
may be one way to enhance a sense of mattering
among older adults and potentially improve their
general well-being. This is even more relevant
in times like the current COVID-19 pandemic,
where people are required to socially distance
and avoid social interaction, especially for older
adults who are more susceptible to the illness.

We also found lower usage of technology for
financial reasons among rural residents. Today,
most financial institutions allow consumers to
perform tasks such as banking and bill payments
online, making transactions easier, quicker, and
even safer. Therefore, those residing in rural areas are more likely to lose all these advantages of being digitally connected.

Additionally, technology is being used in the
health sector in every way today. Following
our findings, older adults in both suburban and
rural areas use ICTs at a lower level for health
purposes (e.g., online health information seeking) than their urban counterparts. This is unfortunate, considering that the benefits of health technologies may be particularly pronounced
for non-urban communities. For example, telehealth can provide healthcare services with no traveling and maximizing efficiency (Heinz et al.,
2013). Furthermore, Ramsetty and Adams (2020)
describe how social determinants of health play
a role in the development of the digital divide
7 Urban/rural digital divide among older adults
(especially in the context of COVID-19) and can
reinforce inequity based on social factors, including health and healthcare access (along with education and economic stability). The pandemic has further increased how technology is being used for healthcare, and older adults can
now get some of the care they need inside their
homes – exclusion through the digital divide is
likely to increase adverse health outcomes.

Our analyses of non-users’ attitudes toward technology deserve particular attention. We showed that older adults in rural areas were more likely
to perceive that technology is less easily available, too complicated, and too hard to learn as compared to the urban group. These findings
are in line with previous qualitative work where
older suburban or rural residents reported that
their use of technologies was limited due to lack
of knowledge about technologies or difficulties in learning (Marston et al., 2019; O’Brien et al., 2014). However, there has been a dearth of
quantitative studies, particularly with a nationally representative sample, that examined perceptions or attitudes toward technology use among older adults living in non-urban communities.

Overall, the present study findings support that
older adults living in suburban and rural areas
not only have lower access to the Internet (Hypothesis 1), but they also use fewer Internetbased technologies (Hypothesis 2) compared to those from the urban. Thus, our work demonstrates the presence of the digital divide’s first level (inequal access to the Internet) and second
level (different technology skills and usage) in
the context of geographic location. Furthermore,
we showed that non-users of communication
technologies from suburban and rural areas have
more unfavorable attitudes toward technologies
than those in urban communities, supporting
Hypothesis 3. Importantly, our results pertain to
the conceptual discussion on the digital divide
(Gonzales, 2016; van Dijk, 2005, 2020). According to van Dijk and Hacker (2003), the digital divide is not only related to the simple lack of
access but also understood as usage patterns or
psychological gaps, as shown in our study. In
particular, rural non-users’ more negative attitudes towards technology may serve as critical determinants of their voluntary decision not to
use technology. Therefore, more research efforts
should be paid to explore multiple aspects of the
urban-rural digital divide.

Limitations
Despite our best efforts, we acknowledge certain
limitations of our research. First, the HRS dataset
is from 2012, so it is possible that this information
could be a little outdated as compared to current
technology usage patterns for older adults. For
similar reasons, the dataset does not include more
latest technologies such as wearable trackers,
for instance. Critically, our study utilized crosssectional data, which does not allow for making causal inferences. Future studies should use a longitudinal research design to provide a complete understanding of the relationship between one’s
residence and technology use in later life. In addition, although we included a wide range of information on specific types of technology use, data on frequency or duration of use was not obtained.

Based on our findings, a next step forward may
include those variables to examine whether the
urban-rural residence relates to how long or frequently one uses digital technologies.

Conclusions
Despite these limitations, however, we believe
that the results of our study provide valuable in-
8 Urban/rural digital divide among older adults
sights into the existing digital divide that excludes
older adults, often based on their geographic location. Our results identify an urgent need for interventions to introduce digital technologies to
older adults in rural areas. This is especially imperative in unprecedented isolating times, such as those created by the COVID-19 global pandemic. Digital technologies have helped people stay in touch with their loved ones and provide
medical assistance. Such technologies in this situation are more important for older adults, who are more susceptible to the novel coronavirus
and would benefit from integrating digitization
into their daily lives. Future work could include
interventions into marginalized communities to
better understand the causes of this divide in
greater detail, the community’s needs, and what
can be provided to them to prevent further exclusion. In addition, our work calls for further research attention to the role of geographic location. Although we focused on three categories from the Beale Rural-Urban continuum for this
study, future work can examine different typologies of spatial locations (e.g., urban networks) and their implications on the digital divide.

Conflict of interest
The authors declare that there is no conflict of interest.

Author statement
Hee Yun Lee: Funding acquisition, Conceptualization,
investigation, methodology, writing – review & editing.

Shaheen Kanthawala: Conceptualization, investigation,
writing – original draft. Eun Young Choi: Data curation, formal analysis, methodology, writing – original draft.

Young Sun Kim: Supervision, Funding acquisition, resources, writing – review & editing.

Acknowledgements
Data was from the Health and Retirement Study, which
was funded by the National Institute on Aging (NIA
U01AG009740) and the Social Security Administration.

The current study was not preregistered. Data, analytic methods, and study materials will be made available
upon request through the corresponding author.

Funding
This work was supported by the Ministry of Education
of the Republic of Korea and the National Research
Foundation of Korea (NRF-2018S1A3A2075402).

References
Anderson, M. (2018, September 10). About a quarter of
rural Americans say access to high-speed internet
is a major problem. Pew Research Center. https://
www.pewresearch.org/fact-tank/2018/09/10/
about-a-quarter-of-rural-americans-say-access-tohigh-speed-internet-is-a-major-problem/ Andresen, E. M., Malmgren, J., Carter, W. B., & Patrick, D. L. (1994). Screening for depression in well older
adults: Evaluation of a short form of the CES-D (Center for Epidemiologic Studies Depression Scale. American Journal of Preventive Medicine, 10, 77–84.

Ball, C., Francis, J., Huang, K.-T., Kadylak, T., Cotten, S. R., & Rikard, R. V. (2019). The Physical–Digital
Divide: Exploring the Social Gap Between Digital Natives and Physical Natives. Journal of Applied Gerontology, 38(8), 1167–1184. https://doi. org/10.1177/0733464817732518
Barrantes, R., & Vargas, E. (2017). Inequalities in the Appropriation of Digital Spaces in Metropolitan Areas of Latin America. In H. Kaur, E. Lechman, & A. Marszk
(Eds.), Catalyzing Development through ICT Adoption (pp. 207–226). Springer International Publishing. https://doi.org/10.1007/978-3-319-56523-1_11
Berner, J., Rennemark, M., Jogréus, C., Anderberg,
P., Sköldunger, A., Wahlberg, M., Elmståhl, S., &
Berglund, J. (2015). Factors influencing Internet usage in older adults (65 years and above) living in rural and urban Sweden. Health Informatics Journal, 21(3), 237–249. https://doi. org/10.1177/1460458214521226
Calvert, J. F., Kaye, J., Leahy, M., Hexem, K., & Carlson, N. (2009). Technology use by rural and urban oldest old. Technology and Health Care: Official
Journal of the European Society for Engineering
and Medicine, 17(1), 1–11. https://doi.org/10.3233/
THC-2009-0527
Centers for Disease Control and Prevention. (2013). The State of Aging and Health in America 2013. Centers for Disease Control and Prevention, US Dept of Health and Human Services. https://www.cdc.gov/
aging/pdf/state-aging-health-in-america-2013.pdf
Eynon, R., & Helsper, E. (2011). Adults learning online: Digital choice and/or digital exclusion? New Media & Society, 13(4), 534–551. https://doi.
org/10.1177/1461444810374789
Francis, J., Rikard, R. V., Cotten, S. R., & Kadylak, T. (2019). Does ICT Use matter? How information
and communication technology use affects perceived mattering among a predominantly female sample of older adults residing in retirement communities. Information Communication and Society, 22(9), 1281–1294. https://doi.org/10.1080/136911
8X.2017.1417459
Gell, N. M., Rosenberg, D. E., Demiris, G., LaCroix, A.

Z., & Patel, K. V. (2015). Patterns of Technology
Use Among Older Adults With and Without Disabilities. The Gerontologist, 55(3), 412–421. https:// doi.org/10.1093/geront/gnt166
Gonzales, A. (2016). The contemporary US digital
divide: From initial access to technology maintenance. Information, Communication & Society, 19(2), 234–248. https://doi.org/10.1080/13691 18X.2015.1050438
Greenberg, A. J., Haney, D., Blake, K. D., Moser, R. P., & Hesse, B. W. (2018). Differences in Access to
and Use of Electronic Personal Health Information
Between Rural and Urban Residents in the United
States: EPHI Use in Rural and Urban Patients. The
Journal of Rural Health, 34, s30–s38. https://doi.
org/10.1111/jrh.12228
Hale, T. M., Cotten, S. R., Drentea, P., & Goldner,
9 Urban/rural digital divide among older adults
M. (2010). Rural-Urban Differences in General
and Health-Related Internet Use. American Behavioral Scientist, 53(9), 1304–1325. https://doi. org/10.1177/0002764210361685
Hanson, V. L. (2010). Influencing technology adoption by older adults. Interacting with Computers, 22(6), 502–509. https://doi.org/10.1016/j. intcom.2010.09.001
Heinz, M., Martin, P., Margrett, J. A., Yearns, M., &
Franke, W. (2013). Perceptions of technology among
older adults. Journal of Gerontological Nursing,
39(1), 42–51. https://doi.org/10.3928/00989134-
20121204-04
Heo, J., Chun, S., Lee, S., Lee, K. H., & Kim, J. (2015).

Internet Use and Well-Being in Older Adults. Cyberpsychology, Behavior, and Social Networking, 18(5), 268–272. https://doi.org/10.1089/cyber.2014.0549 Korupp, S. E. (2005). Causes and Trends of the Digital
Divide. European Sociological Review, 21(4), 409–
422. https://doi.org/10.1093/esr/jci030
Levine, D. M., Lipsitz, S. R., & Linder, J. A. (2016).

Trends in Seniors’ Use of Digital Health Technology in the United States, 2011-2014. JAMA, 316(5), 538. https://doi.org/10.1001/jama.2016.9124
Livingstone, S., & Helsper, E. (2007). Gradations in digital inclusion: Children, young people and the digital divide. New Media & Society, 9(4), 671–696. https://doi.org/10.1177/1461444807080335
Marston, H. R., Genoe, R., Freeman, S., Kulczycki, C.,
& Musselwhite, C. (2019). Older Adults’ Perceptions of ICT: Main Findings from the Technology In Later Life (TILL) Study. 27.

O’Brien, T. R., Treiber, F., Jenkins, C., & Mercier, A. (2014). Use of Computer and Cellular Phone Technology by Older Rural Adults: CIN: Computers, Informatics, Nursing, 32(8), 390–396. https://doi.
org/10.1097/CIN.0000000000000080
Ong, P. M. (2020). COVID-19 and the Digital Divide
in Virtual Learning, Fall 2020. UCLA Center for
Neighborhood Knowledge. https://knowledge.luskin.ucla.edu/wp-content/uploads/2020/12/DigitalDivide-Phase2_brief_release_v01.pdf Perrin, A., & Atske, S. (2021). 7% of Americans don’t use the internet. Who are they? Pew Research Center. https://www.pewresearch.org/fact-tank/2021/04/02/7- of-americans-dont-use-the-internet-who-are-they/
Perrin, A., & Duggan, M. (2015, June 26). Americans
Internet Access: Percent of Adults 2000-2015.

Pew Research Center. https://www.pewresearch.
org/internet/2015/06/26/americans-internet-access-2000-2015/ Pew Research Center. (2019). Internet/Broadband Fact
Sheet. https://www.pewresearch.org/internet/factsheet/internet-broadband/ Quan-Haase, A., Mo, G. Y., & Wellman, B. (2017).

Connected seniors: How older adults in East York
exchange social support online and offline. Information, Communication & Society, 20(7), 967–983. https://doi.org/10.1080/1369118X.2017.1305428
Radloff, L. S. (1977). The CES-D Scale: A SelfReport Depression Scale for Research in the General Population. Applied Psychological Measurement, 1(3), 385–401. https://doi. org/10.1177/014662167700100306
Ramsetty, A., & Adams, C. (2020). Impact of the digital divide in the age of COVID-19. Journal of the American Medical Informatics Association, 27(7),
1147–1148. https://doi.org/10.1093/jamia/ocaa078
Silver, M. P. (2014). Socio-economic status over the
lifecourse and internet use in older adulthood.

Ageing and Society, 34(6), 1019–1034. https://doi.
org/10.1017/S0144686X12001420
Sum, S., Mathews, R. M., Hughes, I., & Campbell, A.
(2008). Internet Use and Loneliness in Older Adults.

CyberPsychology & Behavior, 11(2), 208–211. https://doi.org/10.1089/cpb.2007.0010 Tsai, H. S., Shillair, R., Cotten, S. R., Winstead, V., & Yost, E. (2015). Getting Grandma Online: Are Tablets the Answer for Increasing Digital Inclusion for Older Adults in the U.S.? Educational Gerontology,
41(10), 695–709. https://doi.org/10.1080/03601277
.2015.1048165
van Deursen, A. J. (2020). Digital Inequality During a
Pandemic: Quantitative Study of Differences in
COVID-19–Related Internet Uses and Outcomes
Among the General Population. Journal of Medical Internet Research, 22(8), e20073. https://doi. org/10.2196/20073
van Dijk, J. (2005). The deepening divide: Inequality in the information society. Sage.
van Dijk, J. (2006). Digital divide research, achievements and shortcomings. Poetics, 34(4–5), 221–235. https://doi.org/10.1016/j.poetic.2006.05.004
van Dijk, J. (2020). The digital divide. Polity Press.
van Dijk, J., & Hacker, K. (2003). The Digital Divide
as a Complex and Dynamic Phenomenon. The
Information Society, 19(4), 315–326. https://doi.
org/10.1080/01972240309487
Vogels, E. A. (2021a). Digital divide persists even as
Americans with lower incomes make gains in tech
adoption. Pew Research Center. https://www.pewresearch.org/fact-tank/2021/06/22/digital-dividepersists-even-as-americans-with-lower-incomesmake-gains-in-tech-adoption/ Vogels, E. A. (2021b). Some digital divides persist between rural, urban and suburban America. Pew Research Center. https://www.pewresearch.org/
fact-tank/2021/08/19/some-digital-divides-persistbetween-rural-urban-and-suburban-america/ World Health Organization. (2010). A conceptual
framework for action on the social determinants of health: Debates, policy & practice, case studies. http://apps.who.int/iris/bitstre
am/10665/44489/1/9789241500852_eng.pdf
World Health Organization. (2020). Social determinants of health. http://www.who.int/social_determinants/en/ Yu, R. P., Ellison, N. B., McCammon, R. J., & Langa, K.

M. (2016). Mapping the two levels of digital divide:

Internet access and social network site adoption
among older adults in the USA. Information, Communication & Society, 19(10), 1445–1464. https:// doi.org/10.1080/1369118X.2015.1109695
California wildfires are claiming more vulnerable victims - Los Angele... https://www.latimes.com/california/story/2025-02-25/california-wildfir... 1 of 12 3/28/2025, 1:04 PM
As California struggles to defend itself against increasingly destructive urban wildfires, recent fire deaths in Altadena highlight what researchers say is a growing trend in victim demographics.

Up until about the last decade, California wildfires traditionally have affected higherincome white households, with deaths skewing toward seniors and men. More recently, however, wildfires are increasingly touching more diverse communities as fires grow in intensity and extend into densely populated areas, according to researchers.

Of the 17 people killed by the Eaton fire, more than 70% were Black and 64% were women. More typically, however, those killed by the fire had a median age of 77 and at least a third of them suffered impairments that could affect their mobility, according to a Los Angeles Times analysis.

A 2023 study of California wildfires conducted by U.S. Forest Service researchers found that “new fire regimes are increasingly affecting more urban census tracts statewide, meaning greater numbers and more diverse groups of people are being and possibly will be affected by wildfires.” These changes, authors wrote, meant that more Latino, Asian and Black Californians were being impacted by wildfire than ever before.

The trend, they say, could become even more pronounced in the future and potentially have serious consequences for wildfire readiness.

California wildfires are claiming more vulnerable victims - Los Angele... https://www.latimes.com/california/story/2025-02-25/california-wildfir... 2 of 12 3/28/2025, 1:04 PM
“The increasing occurrence of wildfires in urban areas suggests a need for increased outreach to residents of more urbanized neighborhoods and communities that have never experienced wildfires,” the study’s authors wrote. “Even in those that have historically been sporadically affected by wildfire, their residents are likely to have lower perceptions of risk, believing that wildfires will not occur in their neighborhoods. Similarly, they are likely to be less prepared for wildfires, such as having a ‘go bag’ or completing home mitigation measures.” In the case of the Eaton fire, emergency management officials have been harshly criticized for significant delays in the issuing of electronic evacuation orders — delays that some say cost residents their lives.

California wildfires are claiming more vulnerable victims - Los Angele... https://www.latimes.com/california/story/2025-02-25/california-wildfir... 3 of 12 3/28/2025, 1:04 PM
Eastern Altadena received a warning text and evacuation orders within about the first hour after the fire began on the evening of Jan. 7. The western boundary for the alerts, North Lake Avenue, would come to be a fateful dividing line: Altadena homes east of North Lake received evacuation orders at least eight hours before homes on the west side.

All 17 Eaton fire deaths happened west of North Lake, in areas that never received evacuation warnings or received evacuation orders hours after homes had already been reported to be on fire, a Times investigation found. Many residents of western Altadena — which was shaped by discriminatory lending practices in the 1960s and ‘70s and has become known for its strong Black community — told The Times they felt forgotten. North Lake was the boundary line for these redlining efforts, and racial gaps have persisted: Altadena’s east side is whiter and incomes are higher than those of the west side, according to census data.

Nearly a third of those killed in the Eaton fire suffered some form of disability.

Among them was 56-year-old Carolyn Burns.

Burns, who used a wheelchair and walker to get around, lived with her 76-year-old mother.

California wildfires are claiming more vulnerable victims - Los Angele... https://www.latimes.com/california/story/2025-02-25/california-wildfir... 4 of 12 3/28/2025, 1:04 PM
Burns’ mother, who also is named Carolyn, told The Times that another of her daughters called the Los Angeles County Fire Department around 10 or 11 on the night of the fire to ask about conditions in their neighborhood. They were reportedly told that the street was safe at the moment and if they needed to be evacuated, someone would knock on the door.

At around 3:30 a.m. the mother woke to the sound of a neighbor banging on a window and yelling that the house was on fire. She jumped out of bed and saw flames in her kitchen. She managed to escape, but her daughter did not.

California wildfires are claiming more vulnerable victims - Los Angele... https://www.latimes.com/california/story/2025-02-25/california-wildfir... 5 of 12 3/28/2025, 1:04 PM
“I don’t think they did their jobs because they didn’t get over there in time or knock on our doors,” the mother told The Times. “I think it was too late. Even if they needed to get her out of there, they didn’t have enough time. We’re very angry inside and we’ll never be the same.”
The county Fire Department declined to comment on the evacuation orders or Burns’ death. In a statement, the agency said the L.A. County Board of Supervisors has opened an investigation into the evacuations and emergency notifications. Reports will be due back to the board every 90 days and shared with the public.

The Fire Department “acknowledges the immense loss and challenges faced by our communities during the wildfires and remain deeply committed to supporting those affected,” according to the statement.

The devastating fires killed at least 28 people,
destroying and damaging more than 18,000
buildings valued at more than $275 billion and
leaving a burn zone 2½ times the size of
Manhattan.

California wildfires are claiming more vulnerable victims - Los Angele... https://www.latimes.com/california/story/2025-02-25/california-wildfir... 6 of 12 3/28/2025, 1:04 PM
Mark Ghilarducci, who served as the director of the California Governor’s Office of Emergency Services for more than a decade, said it doesn’t “sound normal” if disabled residents were left to evacuate on their own.
“There’s generally an understanding of individuals who have access and functional needs where they can be supported,” he said. “Sometimes the 911 centers know that they exist. A lot of the time, it’s law enforcement going door to door to ensure that people are getting out and and being notified.”
Senior citizens tend to be the ones who die in wildfires, he said, because they aren’t as attuned with social media, could be more limited in their mobility or rely on others to drive them around.
“We need to figure out exactly what happened, why were decisions made the way they were, and if it was a failure in technology, we need to identify that as well and figure out, how do we address that?” Ghilarducci said. “It isn’t normal. We want people to get out of harm’s way.”
Fires with large death tolls are relatively rare in California but have been growing in frequency over the last 10 years, said Michele Steinberg, a spokesperson for the wildfire division of the National Fire Protection Assn.

In all major incidents — including the Eaton and Palisades fires — seniors are the most vulnerable.

Steinberg said older residents may have difficulty evacuating, due to potential mobility issues, or they may simply refuse to leave.

In some previous cases Steinberg has reviewed, fatalities among those who were younger — in their 50s or 60s — died while trying to help others.

California wildfires are claiming more vulnerable victims - Los Angele... https://www.latimes.com/california/story/2025-02-25/california-wildfir... 7 of 12 3/28/2025, 1:04 PM
“You’re trying to save pets, save people, take care of people,” Steinberg said. “You yourself might be able and willing and ready, but you have someone else to look after.”
Ultimately, life or death can come down to how quickly a person can react to a fastmoving and unexpected situation, she said. “People aren’t just independently jumping in their car wide awake — it’s chaos.”
In Altadena, where the fire spread in the “middle of the night,” residents probably had even less warning, Steinberg said, with the speed of the fire and possible issues with alerts making the situation more deadly.

Among the seniors killed in the Eaton fire were Dalyce “DeDe” Curry, 95, and Erliene Kelley, 83.

California wildfires are claiming more vulnerable victims - Los Angele... https://www.latimes.com/california/story/2025-02-25/california-wildfir... 8 of 12 3/28/2025, 1:04 PM
Curry was home alone on the night of the fire, while her granddaughter, Dalyce Kelley, was caring for a sick relative and checking in on her grandmother via text.

Kelley had fallen asleep at her home. At 6:38 a.m., she woke up and wrote to an Altadena neighborhood group chat: “My grandmother is still there. We got home around midnight. Have evacuation orders been implemented? … if they evacuate you all, I’ll come immediately, just please grab her.”
“Every one was evacuated at 3:30,” a neighbor responded. “Omg,” Kelley replied. “Getting dressed now. Her phone is going straight to voicemail.”
She rushed to her car and raced to her grandmother’s house but was stopped at a police barricade. Kelley gave a police officer her grandmother’s address and asked California wildfires are claiming more vulnerable victims - Los Angele... https://www.latimes.com/california/story/2025-02-25/california-wildfir... 9 of 12 3/28/2025, 1:04 PM
whether he could check on her. The officer called Kelley later, telling her that Curry’s home had burned down.

The Los Angeles Medical Examiner pronounced Curry dead Jan. 11 and notified her family that her remains had been found on her property, according to Kelley.

One of Curry’s neighbors, Ana Morales, 34, faults officials for failing to warn residents earlier. She said she and her husband decided not to wait for an official evacuation order and fled their home around 9 p.m.
“I don’t think there was enough notice for everybody,” Morales told The Times. “We were going off of our intuition and fear. They should’ve evacuated everybody when we left, told us to pack up everything we need and to leave. But nothing.” Briana Navarro, who lost her grandmother Erliene Kelley in the fire, agreed. She said her grandmother and others could have been saved with earlier notice from officials. “A lot of the lives that were lost were either elderly or disabled, which is unfortunate because they’re one of the vulnerable groups that need the most assistance from family or their support system,” she said. “I think with a notice it would have given enough time for some of us to go help our family members.” Times staff writer Ruben Vives contributed to this report.

California wildfires are claiming more vulnerable victims - Los Angele... https://www.latimes.com/california/story/2025-02-25/california-wildfir... 10 of 12 3/28/2025, 1:04 PM
Summer Lin is a reporter on the Fast Break Desk, the Los Angeles Times’ breaking news team. Before coming to The Times, she covered breaking news for the Mercury News and national politics and California courts for McClatchy’s publications, including the Miami Herald. An East Coast native, Lin moved to California after graduating from Boston College and Columbia University’s Graduate School of Journalism. Lin was among The Times’ staff members who covered the Monterey Park mass shooting in 2023, which was recognized by the Pulitzer Board as a finalist in breaking news.

Terry Castleman is a data reporter on the Fast Break Desk covering breaking news. In 2020, he was named alongside his colleagues as a Pulitzer Prize finalist in explanatory reporting. Previously, he worked at the New York Times and volunteered as a first responder for refugees arriving on the shores of Lesvos.

California wildfires are claiming more vulnerable victims - Los Angele... https://www.latimes.com/california/story/2025-02-25/california-wildfir... 11 of 12 3/28/2025, 1:04 PM
California wildfires are claiming more vulnerable victims - Los Angele... https://www.latimes.com/california/story/2025-02-25/california-wildfir... 12 of 12 3/28/2025, 1:04 PM
 

















 


 






  


  

!"#$%&'#()*"%+',-.*/01234510647
8-9:*2;<=9-:>?*23/19>
@A;1?*, =B7*CDE
F-=G*23H1*,6?/-9:

IJ;-4/1 J0-.KL34
MNOPQ*RSOTUMV*WXYZM O[ WT\V]X[MPV*NUYZ UV^PY*WT
YRW*QMNUV*WTRMY_WXOSYWXV`*NQMUWTOU^[ZY*NXOZ^US_NXMa
TWU_M*bNXY*] RSOV`*cXSdMUVYQ*WTeSU[XN*MfgUYV*NSOh
bMdUNP*TSUMV*ZNdM_WijSXMO*YWj^UX MNUPQ*kl` *N_UMV`*O YUWQSX[ YZW^VNXO*WTYU^_ MV*NXOmSPX[*NYPMVY*RWOnMX*gWPh
oZM*pNYSWXP*qMYZMU*b dS_*V^MO*U TPN[R UXS[V*NOZS[ RXO
RNUXS[V*N_UWV*b^YZMUX*rNPSTWUXN*W sWONQ*XOo^MVONQ*RSYZ
tgNUYS_^PNUQ*ONX[MUW^V*SY^NWXu*RNUXS[V*TWUiWVYWTZMTSUaNT_YO NUMhtvUQXMV*SYZMiNwWU*TN_YWU`*NXOYZSVTUMRNVMf_UjYO*jQYZMRSXOVYZN*UMW_^USX[* ZY*XWR`u*VNSOeMXmNYUiNX*\NmVZiS`*xWZX\h pMR_Wij*yUWTMV U*TzX[SMU[* ce]{V*vMgNUYiMX*WTrSdP*NXO
zXdSUWiMYNP*zX[SMU[h*toZMOUQ_WXOSYWXV*ZNdMUO^_ *YZ
dM[YNSWX*RNYMU*_WXYM *NO V UMV^PY`*ZMd[YNSWX*VdMUQ
V^_MgYSjPM*YWj^UXS[hu
toZM*YRWgUMdSW^V*QMNUV*RM d QR Y`SZNPiWVY*OW^jPM*YZ NdU[ NX^P*USXTPh*|XYZMV*RNYUaPSiSYMO*_WVQYMiV`*YZNPMOV*YW[UMNY [UWRYZ`*iWUMPNT[UWRYZ*NXOPMNTSYMU`*RZS_*OMVS_NYMV*^XOMU*YZ WY`  









 
















!"#$
% !"& #'&( )$&%*
)+',!%-#'.

&(! "/ 012345 6782719
:;<=
 '
&4>' 4?
@)&(A'
4B!.&4-
>)## 4CD$!
( D)+'%
 '
&' (
#)&( '
!.&
&)## / 7EFG913F2H4E9I3
>J&4K!'$

,4'(L
,4!-MB<N
&(! "4O
'

4<&.

,4PQ"M!(RJ
CD$!
( D&
$!'

,

'

&(! "'

&.

,S"

!(/
TUV4WXYTZ[\4]^UV4_^YT[4`abcZ4dUYeZf4dUYghieZ U4idYjXUhkef^lmnXeY [4Z^XTopqe4g^lXZTe4iXUe[4fq l^UreZf4hiqeihsU^nfXje4tl^ueZ[4tr^Y4WeZfhi ]hZ4bYrelZ4^YT hUfq4im^Yf4vhYXn^o4wftsUYeT4fhW^UT4v^lXts4^YTfhfqe4g^nXi4xe^Y[4WqXle4fq d^fhY4iXUetsYeT4y^UfZ4hig^ZTeY4 T blf^TeYo4mk^leU4iXeZ[4YhWkhZflV4nhYf^XeT[4^lZhtsUYXfqe^U o ]Xfle4hY U^XYqZileY4Xfqe^U ZXYne4l^ZfskkeU[4WXfql^e
^sfkY4^YTWXfeU4tXYr4fqesZ^l4WefZ^hY4iUfqeUrXhYo4pqe
z^fXhYl4{Uhsrqf4vXXr^hY4|efUcZ4T^fZqhW4fqe^U hsYT4]hZbYrelZ4khUyqeT4iUhk4YUk^l4nhYTXfhYZ4 xnfhteU4fhZejU4Thsrqf tV4}^Yo~pqeZ^ke4ZhsUne4ZqhW4fqXVe^UcZ4hXlkhZfsUe4nhYfe 4X fqe4UrXhY4teXr4fqeTUVZf4XYne4€‚o




 
 
 
 


 
! "

! 
 ! "
# $" 


 
 
! %
 
&
'"


(
)*+ ,- ./ 012-3 4+51,670 ,8/17*+8,05+-9 :;<+=>-1/ ,?@A)*+ 0,164+ 51,670 /B*,>- 39 CD63? 7C/>+B6/ 1*, -+C6E,8>1+9 F2-0>+678B,3? 08-+/>721@ )*+?87D60,1+ 37B,+B7/?1+D/9 ,/1*+C6-/ G >721BD>+167-/9H I,08/6@ A)*+B*,>- 3 /2-46+/ 1* 8-?9*71J+861+-,0 B36D,1+ /2DD+-/ F?/+B-1605 ,-+/60 1*,B7 / *+6- 3,4/9 -82B605 K,1+- 37/@ )*+-/606/ 4+-?73,16+ ,08F2-0/ *71@H  
 



!

 "
 
!

" 
 !
 "L
!
 "


!"



!

!
  &
 #!M
(
 



 
  





 

 


 







 





 

 

 


 




  

 




 
 

 









 




 

!" 

#$%& 

''
(





)* 


 





'


 

 


 

 





 


 








 

 








 


  +



 

  


 


 ,,







 
-

 (
.




 
 
 ' / 

  
 


  0 




 





 




  



 

  
 
! 

  










 !""#
$%"&' ()#
*+ ,-. +/ 0
12
3456 7
89:;<=>8 ?@ABCD< CD;=A?9E<BF:CD9:GFCHB@C9:G ;I<DF=<DJK  





 

 

!"#$%&
'&()'*(+ ,-./
,
$0
1234($
356

7 ' ()'6*(+ 6 4-./
64 $0
8  




 



 

!"#$%
&'#(
)"*
+',#-. /
#$(01'2

"3-,/
4"(,5
678',#( 7
9: ("#$ 9 $'#( 9;"8',#-. /91'29: "3-,/9#"(,5<
 


 

  





 

 !

"#$%&'(
)*+,$-".+
/$*#.0%1
$1" +!2
3%. #1
(
456$*1 5-7 8!
7+#$%&7*+$-+. 6$*#.0%17+!279%.8#1
:

Citation: Melton, C.C.; De Fries,
C.M.; Smith, R.M.; Mason, L.R.

Wildfires and Older Adults: A
Scoping Review of Impacts, Risks,
and Interventions. Int. J. Environ. Res.

Public Health 2023, 20, 6252. https://
doi.org/10.3390/ijerph20136252
Academic Editors: Juan Ramón
Molina Martínez, Francisco
Rodríguez y Silva and Jorge Antonio
De las Heras Ibáñez
Received: 20 April 2023
Revised: 15 June 2023
Accepted: 21 June 2023
Published: 29 June 2023
Copyright: © 2023 by the authors.

Licensee MDPI, Basel, Switzerland.

This article is an open access article
distributed under the terms and
conditions of the Creative Commons
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/

4.0/).

International Journal of
Environmental Research
and Public Health
Review
Wildfires and Older Adults: A Scoping Review of Impacts, Risks, and Interventions
Colleen Cummings Melton , Carson M. De Fries , Rebecca M. Smith and Lisa Reyes Mason * Graduate School of Social Work, University of Denver, Denver, CO 80210, USA; [email protected] (C.C.M.); [email protected] (C.M.D.F.); [email protected] (R.M.S.) * Correspondence: [email protected]
Abstract: Climate change is leading to worsening disasters that disproportionately impact older adults. While research has begun to measure disparities, there is a gap in examining wildfirespecific disasters. To address this gap, this scoping review analyzed literature to explore the nexus of wildfires and older adults. We searched peer-reviewed literature using the following inclusion criteria: (1) published in a peer-reviewed journal; (2) available in English; (3) examines at least one topic related to wildfires; and (4) examines how criterion three relates to older adults in at least one way. Authors screened 261 titles and abstracts and 138 were reviewed in full, with 75 articles meeting inclusion criteria. Findings heavily focused on health impacts of wildfires on older adults, particularly of smoke exposure and air quality. While many articles mentioned a need for communityengaged responses that incorporate the needs of older adults, few addressed firsthand experiences of older adults. Other common topics included problems with evacuation, general health impacts, and Indigenous elders’ fire knowledge. Further research is needed at the nexus of wildfires and older adults to highlight both vulnerabilities and needs as well as the unique experience and knowledge of older adults to inform wildfire response strategies and tactics.

Keywords: wildfires; climate change; disaster recovery; evacuation; adaptation; mitigation; older adults; elders

1. Introduction
The growing threat of climate change has been well-documented in recent years. Since 2011, concentrations of greenhouse gases in the atmosphere have soared, pushing global surface temperatures to an estimated 1.3 degrees Celsius above pre-industrial levels [1].

Human-induced climate change has accelerated impacts of ecological degradation, biodiversity loss, and extreme weather events. These include, but are not limited to, increases in areas burned in wildfires, cyclone intensity attributed to sea-level rise, severe and prolonged droughts, heavier precipitation, and substantial—and in some cases irreversible—damages to biodiversity and ecosystems [1]. These impacts are not felt evenly, with already vulnerable populations suffering the brunt of the crisis. Those living in poverty, women, children, older adults, outdoor workers, people with disabilities, Indigenous populations, and people of color are facing adverse health events. These include increased morbidity and mortality from disease connected to heat stress, exposure to air pollution and smoke, and vector-borne illnesses, in addition to ongoing human rights violations during this era of climate crisis [2].

The social and ecological consequences of wildfires are areas of growing concern, with recent wildfire seasons breaking precedents for frequency and intensity [3]. In the U.S. alone, wildfire events are increasing, with an average of 6.9 million acres burned annually, more than double the annual acreage burned in the 1990s, with the top five worst wildfire seasons in the U.S. all occurring since 2006 [3]. Record-breaking wildfire seasons from Australia to the Arctic and in North and South America are an ominous sign of the Int. J. Environ. Res. Public Health 2023, 20, 6252. https://doi.org/10.3390/ijerph20136252 https://www.mdpi.com/journal/ijerph Int. J. Environ. Res. Public Health 2023, 20, 6252 2 of 22 ever-growing duration, frequency, and intensity of wildfire seasons to come [4]. Even in the best-case scenarios for curbing emissions, the risk of global wildfire occurrence will still increase by 31–57% by the end of the century [4]. Environmental change related to wildfires is also unique in that wildfires are exacerbated by climate change and are also a contributing factor in the worsening of climate change through the release of greenhouse gasses (GHG) and destruction of carbon stored in trees.

Beyond the environmental impacts, increasing wildfires are also a grave threat to human health. Smoke from wildfires worsens air quality and increases exposure to and inhalation of smoke and small particulates from ash, referred to as particulate matter smaller than 2.5 microns (PM2.5) [5,6]. Wildfires lead to increased PM2.5 and decreased air quality—increasing the odds of respiratory and health concerns such as burning eyes, runny nose, scratchy throat, headaches, respiratory illness, and exacerbation of pre-existing conditions such as asthma and COPD [6,7]. Breathing wildfire smoke is associated with increased outpatient visits, emergency visits, hospitalization, and death from a myriad of respiratory issues, which is only complicated by the current COVID-19 pandemic, as breathing PM2.5 (the primary health concern related to wildfire smoke) is associated with increased morbidity and mortality of the novel coronavirus [7,8].

When a wildfire encroaches upon or destroys communities, emergency preparedness and response and mitigation strategies have also been investigated in conjunction with human health vulnerabilities during times of wildfire disaster. Studies concerning evacuation and emergency service systems in protecting human life and health have been carried out around the world [9]. Many studies indicate significant numbers of people delay evacuation during a wildfire event, often leading to increased evacuation danger [9]. In the immediate aftermath of a wildfire disaster, access to prescription medication, healthcare providers, and mental health services can be lacking [10]. Once these aftershocks have subsided, psychological distress following landscape and ecosystem loss—as well as personal loss or trauma—can be prevalent among the general populations [11].

The very same populations experiencing the most adverse health consequences from climate change are also vulnerable to impacts from other natural disasters, including wildfires, with older adults principal among them. Research has demonstrated that in addition to the usual concerns associated with natural disasters such as injuries and infectious disease outbreaks, older adults face added challenges due to functional or mobility limitations, decreased social supports, difficulty maintaining necessary health regimens, and limited access to information about disaster preparedness and recovery practices [12]. Due to the higher prevalence of chronic conditions among older adults, they often require specialized diets, medicine, and other medical treatments which can be more difficult to maintain or access following the trauma and disruptions caused by natural disasters [13]. Additionally, as people age, their social networks may shrink for a number of reasons, including spouses and close friends passing away or having their children move away, making it more difficult to reach out to others for help [14].

As a result of these age-related risks, older adults are disproportionately negatively impacted by natural disasters when compared to other age groups [15]. For example, while older adults made up only 15% of the New Orleans population, 71% of the people who died from Hurricane Katrina were over the age of 65 [15]. Studies have shown that older adults are often more likely to encounter life-threatening challenges while trying to evacuate during a natural disaster, are less likely to receive disaster warnings, and often experience greater financial losses following the destruction of natural disasters [16]. These disparate outcomes faced by older adults occur with all types of natural disasters, indicating that the needs of this population during these times of crisis need to be addressed [17].

While the disparate impact of natural disasters on older adults is well-documented in scholarly literature, most of this research has focused on hurricanes and flooding [18]. There is a gap in the literature examining the impact of wildfires on older adults [18]. While some findings from other natural disasters (e.g., evacuation, emergency communication, etc.) are relevant across disasters, wildfires have unique health impacts related to smoke and Int. J. Environ. Res. Public Health 2023, 20, 6252 3 of 22 heat exposure that may pose multiple burdens and harms for older adults. This study seeks to examine this gap in the literature through a systematic scoping review of scholarly literature to understand the existing knowledge base on the impact of wildfires on older adults, as well as identify other gaps in data and priorities and directions for interventions and future research.

2. Materials and Methods
Due to the lack of literature on wildfires and older adults, the scoping review methodology was chosen due to its usefulness to “determine the scope or coverage of a body of literature on a given topic and give clear indication of the volume of literature and studies available as well as an overview (broad or detailed) of its focus” [19] (p. 2). The scoping review methodological framework followed guidelines from Arksey and O’Malley [20], as well as recommendations by Levac et al. [21] and Cloquhoun et al. [22]. The PRISMA-ScR checklist was followed for documenting and reporting findings [23].

2.1. Inclusion Criteria
To answer the research question “What is the extent and scope of literature on wildfires and older adults?” the following inclusion criteria were used:

1. Published in a peer-reviewed journal;

2. Available in English language;

3. Examines at least one topic related to wildfires;

4. Examines how criterion (3) relates to older adults in at least one way.

For criterion (1), peer-reviewed journal publications were chosen to explore academic literature relating to older adults and wildfires to gain an understanding of relevant evidence, themes, needs, and gaps in the literature. For criterion (2), references were limited to the English language due to the research team’s inability to translate articles from other languages. For criterion (3), wildfires were specifically chosen as the disaster of focus due to gaps in the literature exploring the impacts of wildfires (versus other types of disasters such as hurricanes, flooding, etc.) on vulnerable populations, especially older adults. For criterion (4), we defined older adults as those who are 60 years and older, or who were referred to in the references as “older adults”, “elderly”, “elders”, etc. (see search terms below). This age was chosen based on literature indicating that 60 is a common parameter for identifying this age group [24]. Further, criterion (4) means that included articles must specifically connect wildfires to older adults in some way, excluding those that discussed older adults and wildfires separately.

2.2. Literature Search and Screening
Search terms and protocols were established in consultation with a university librarian.

Based on these discussions, the following databases were searched: PubMed, Web of Science, ProQuest (Agriculture and Environmental Sciences Collection, Sociological Abstracts, and Social Service Abstracts), and EBSCO Host (Academic Search Complete, Environment Complete, GreenFILE, PsycInfo, and SocINDEX).

In consultation with the librarian, the following search strings were created and run in each database:

1. “older adult*” OR senior* OR elder* OR “older person*” OR “older people” OR geriatric* OR gerontolog* OR “old age” OR “long term care” OR “nursing home*” OR “assisted living” OR “independent living” OR “skilled nursing facilit*” OR “memory care” OR “residential care” OR “retirement communit*”;
AND

2. wildfire* OR “wild fire*” OR bushfire* OR “bush fire*” OR bushfire* OR “forest fire*” OR “brush fire*” OR brushfire* OR “wildland fire*” OR “uncontrolled fire*” OR “fire season*”.

Int. J. Environ. Res. Public Health 2023, 20, 6252 4 of 22 Both search strings were searched “anywhere but full text (NOFT)” within the ProQuest database, and with the default search settings for other databases, which was the recommendation and guidance of the university librarian. The search, conducted in March of 2021, yielded 585 articles. After removing duplicate records, 261 remained.

The research team used Covidence systematic review software [25] to complete the screening process. Two authors independently reviewed the titles and abstracts of the 261 non-duplicate records. After title and abstract screening, 138 remained. Two authors then independently read the full text of these 138 remaining articles. Of these, 75 met the inclusion criteria (Figure 1) [18,26–99]. Throughout the screening and review process, any disagreements on inclusion/exclusion were discussed and reconciled as a team before making a final decision.

Int. J. Environ. Res. Public Health 2023, 20, x 4 of 23

2. wildfire* OR “wild fire*” OR bushfire* OR “bush fire*” OR bushfire* OR “forest fire*” OR “brush fire*” OR brushfire* OR “wildland fire*” OR “uncontrolled fire*” OR “fire season*”.

Both search strings were searched “anywhere but full text (NOFT)” within the ProQuest database, and with the default search settings for other databases, which was the recommendation and guidance of the university librarian. The search, conducted in March of 2021, yielded 585 articles. After removing duplicate records, 261 remained.

The research team used Covidence systematic review software [25] to complete the screening process. Two authors independently reviewed the titles and abstracts of the 261 non-duplicate records. After title and abstract screening, 138 remained. Two authors then independently read the full text of these 138 remaining articles. Of these, 75 met the inclusion criteria (Figure 1) [18,26–99]. Throughout the screening and review process, any disagreements on inclusion/exclusion were discussed and reconciled as a team before making a final decision.

Figure 1. Flowchart of search, screen, and review process.

2.3. Data Extraction and Analysis
Data collected on each article included: (1) article characteristics and type; (2) information related to environmental issues including the disaster recovery cycle, specific hazards, etc.; (3) information on how older adults were included and relevant findings; and Figure 1. Flowchart of search, screen, and review process.

2.3. Data Extraction and Analysis
Data collected on each article included: (1) article characteristics and type; (2) information related to environmental issues including the disaster recovery cycle, specific hazards, etc.; (3) information on how older adults were included and relevant findings; and (4) whether articles addressed problems, used responses or interventions, or suggested solutions, recommendations, or areas of future research. We created, pilot tested, and refined our data collection tool using Google Forms. Once the final form was created, two members of the research team independently recorded data from each article. Any Int. J. Environ. Res. Public Health 2023, 20, 6252 5 of 22 questions or disagreements were discussed and resolved as a team. During analysis, we also identified thematic topics arising from the literature.

First, basic characteristics included the year of publication, article title, author(s), journal title, country or geographic focus, study type, sample, and methods used. Second, information related to environmental issues included the hazards addressed (wildfires, air quality, heat, haze, or other types of hazards); specific disasters addressed; explicitly mentioning climate change or recommendations for climate adaptation and/or mitigation; focus on any part of the disaster recovery cycle (response, recovery, mitigation, preparedness); explicitly mentioning environmental justice or alluding to it; and the inclusion of Indigenous or Aboriginal traditional ecological knowledge (TEK). Third, questions related to older adults included whether the primary focus of the article was older adults and/or how older adults were included; focus on older adults in the community or in residential facilities; and relevant findings or recommendations related to older adults. Fourth, questions related to study focus included focus on problem description, measuring exacerbation of specific health problems, inclusion of responses or interventions, inclusion of Indigenous or Aboriginal knowledge of fire management, solutions or recommendations, areas of future research, and thematic topics arising in the literature. All criterion, except for thematic topics, were established during the creation and pilot testing of the data collection tool.

Thematic topics arose during data collection as patterns emerged in the literature.

3. Results

3.1. Basic Characteristics of the Literature
A total of 75 peer-reviewed journal articles met study inclusion criteria. There was no limit on year of publication in our initial search; the earliest article was published in 2001, with the frequency of publications increasing over time (Figure 2). Only 3 of the 75 articles (4%) were published between 2001 and 2006, 10 (13.3%) were published between 2007 and 2011, 23 (30.7%) were published between 2012 and 2016, and 39 (52%) were published between 2017 and 2021 (Figure 2).
(4) whether articles addressed problems, used responses or interventions, or suggested solutions, recommendations, or areas of future research. We created, pilot tested, and refined our data collection tool using Google Forms. Once the final form was created, two members of the research team independently recorded data from each article. Any questions or disagreements were discussed and resolved as a team. During analysis, we also identified thematic topics arising from the literature.

First, basic characteristics included the year of publication, article title, author(s), journal title, country or geographic focus, study type, sample, and methods used. Second, information related to environmental issues included the hazards addressed (wildfires, air quality, heat, haze, or other types of hazards); specific disasters addressed; explicitly mentioning climate change or recommendations for climate adaptation and/or mitigation; focus on any part of the disaster recovery cycle (response, recovery, mitigation, preparedness); explicitly mentioning environmental justice or alluding to it; and the inclusion of Indigenous or Aboriginal traditional ecological knowledge (TEK). Third, questions related to older adults included whether the primary focus of the article was older adults and/or how older adults were included; focus on older adults in the community or in residential facilities; and relevant findings or recommendations related to older adults.

Fourth, questions related to study focus included focus on problem description, measuring exacerbation of specific health problems, inclusion of responses or interventions, inclusion of Indigenous or Aboriginal knowledge of fire management, solutions or recommendations, areas of future research, and thematic topics arising in the literature. All criterion, except for thematic topics, were established during the creation and pilot testing of the data collection tool. Thematic topics arose during data collection as patterns emerged in the literature.

3. Results

3.1. Basic Characteristics of the Literature
A total of 75 peer-reviewed journal articles met study inclusion criteria. There was no limit on year of publication in our initial search; the earliest article was published in 2001, with the frequency of publications increasing over time (Figure 2). Only 3 of the 75 articles (4%) were published between 2001 and 2006, 10 (13.3%) were published between 2007 and 2011, 23 (30.7%) were published between 2012 and 2016, and 39 (52%) were published between 2017 and 2021 (Figure 2).

Figure 2.

Figure 2.

Year of publication.

Year of publication.

Geographic regions discussed were diverse, but the majority were based in North America (44%) and Oceania and Australia (26.7%), and many were about wildfires or fire management on First Nations or Tribal land (18.7%) (Figure 3). The United States (U.S.) was the most represented country, representing 26 of the 33 total mentions of North America.

Most of these focused on the western U.S. (n = 12), specifically California (n = 8). Of the eight articles focused on Canada, six were about wildfires on First Nations land. The 20 articles focusing on Oceania and Australia were almost exclusively focused on Australia, Int. J. Environ. Res. Public Health 2023, 20, 6252 6 of 22 with 1 mentioning New Zealand and 6 of the 20 focusing on Aboriginal land. Seven articles focused on Northern and Western Europe (Spain, Portugal, Greece, and two from Finland), and five articles focused on South America, all of which were in Brazil’s Amazon region.

Five articles were based in Southeastern Asia (two in Malaysia, two in Indonesia, and one covering Singapore, Malaysia, Indonesia, Brunei, and Thailand). Finally, those that covered more than three countries were labeled as “global”, though these predominantly focused on countries above including the U.S., Australia, Malaysia, and Indonesia. Notably, no articles covered geographic regions of Africa, Central America, or North and Central Asia, though one global article mentioned “Asia, Latin America, and Africa” [42] (p. 99).

Geographic regions discussed were diverse, but the majority were based in North America (44%) and Oceania and Australia (26.7%), and many were about wildfires or fire management on First Nations or Tribal land (18.7%) (Figure 3). The United States (U.S.) was the most represented country, representing 26 of the 33 total mentions of North America. Most of these focused on the western U.S. (n = 12), specifically California (n = 8). Of the eight articles focused on Canada, six were about wildfires on First Nations land. The 20 articles focusing on Oceania and Australia were almost exclusively focused on Australia, with 1 mentioning New Zealand and 6 of the 20 focusing on Aboriginal land. Seven articles focused on Northern and Western Europe (Spain, Portugal, Greece, and two from Finland), and five articles focused on South America, all of which were in Brazil’s Amazon region. Five articles were based in Southeastern Asia (two in Malaysia, two in Indonesia, and one covering Singapore, Malaysia, Indonesia, Brunei, and Thailand). Finally, those that covered more than three countries were labeled as “global”, though these predominantly focused on countries above including the U.S., Australia, Malaysia, and Indonesia.

Notably, no articles covered geographic regions of Africa, Central America, or North and Central Asia, though one global article mentioned “Asia, Latin America, and Africa” [42] (p. 99).

Figure 3. Geographic focus of articles. Note: Percentage exceeds 100% as some articles covered multiple regions. * Articles focusing on First Nations and/or Tribal lands were based in North America (Canada, n = 7; United States, n = 3) and Oceania and Australia (Australia, n = 6) Of the 75 articles, 63 (84%) were empirical research articles or evaluations (Table 1).

Most of these were quantitative (44%) or qualitative (26.7%), with a few being mixed methods (4%) or systematic reviews (9.3%). The 12 non-empirical articles (16%) were conceptual, descriptive, or commentaries. Methods used in empirical articles varied, with secondary data (37.3%) being the most prevalent. A large majority of articles focused on measuring morbidity and mortality related to wildfire smoke, with 22 articles (29.3%) using emergency room and hospital admissions or mortality rates as secondary data. The second most common method was remote-sensed environmental measures (29.3%), measuring air quality and pollution, particularly of PM2.5 levels and other particulate matter.

Interviewing was the third most prevalent method (21.3%). Other methods included systematic reviews (9.3%), surveys (9.3%), focus groups (8%), case studies (6.7%), field Figure 3. Geographic focus of articles. Note: Percentage exceeds 100% as some articles covered multiple regions. * Articles focusing on First Nations and/or Tribal lands were based in North America (Canada, n = 7; United States, n = 3) and Oceania and Australia (Australia, n = 6).

Of the 75 articles, 63 (84%) were empirical research articles or evaluations (Table 1).

Most of these were quantitative (44%) or qualitative (26.7%), with a few being mixed methods (4%) or systematic reviews (9.3%). The 12 non-empirical articles (16%) were conceptual, descriptive, or commentaries. Methods used in empirical articles varied, with secondary data (37.3%) being the most prevalent. A large majority of articles focused on measuring morbidity and mortality related to wildfire smoke, with 22 articles (29.3%) using emergency room and hospital admissions or mortality rates as secondary data.

The second most common method was remote-sensed environmental measures (29.3%), measuring air quality and pollution, particularly of PM2.5 levels and other particulate matter. Interviewing was the third most prevalent method (21.3%). Other methods included systematic reviews (9.3%), surveys (9.3%), focus groups (8%), case studies (6.7%), field research (5.3%), biological data (5.3%), and other methods (8%; e.g., participatory action research, future modeling, ethnography, and Q methodology) (Table 1).

Int. J. Environ. Res. Public Health 2023, 20, 6252 7 of 22 Table 1. Basic characteristics of the literature (n = 75).

Characteristic n (%)
Paper Type
Quantitative 33 (44)
Qualitative 20 (26.7)
Mixed Methods 3 (4)
Systematic/Scoping Review 7 (9.3)
Conceptual Papers 9 (12)
Other (e.g., commentary, interview transcript) 3 (4)
Method
Secondary Data 28 (37.3)
Remote-Sensed Environmental Measures (Air Quality) 22 (29.3) Interviews 16 (21.3)
Systematic Review 7 (9.3)
Survey 7 (9.3)
Focus Group 6 (8)
Case Study 5 (6.7)
Field Research 4 (5.3)
Biological Data 4 (5.3)
Other Methods 6 (8)
Not Applicable (e.g., conceptual papers or other paper type) 12 (16) Note: Methods percentage exceeds 100% as some articles used multiple methods and/or data collection strategies.

3.2. Environmental: Hazards, Climate Change, and Disaster Recovery Cycle We reviewed articles for specific information related to environmental issues including specific wildfires, other hazards, and language or information about climate change, environmental justice, or the disaster recovery cycle (Table 2). All articles discussed wildfires, bushfires, or forest fires in some way. Some articles also discussed other types of disasters such as flooding and hurricanes, but due to the proliferation of literature on these topics, we only collected data on hazards related to wildfires. Of these related hazards, 41 articles discussed air quality (54.7%), 12 covered heat (16%), and five discussed haze (6.7%). Almost half of articles were either about a specific wildfire (17.3%) or a specified wildfire season or time period where wildfires occurred (25.3%). Wildfire events or seasons that were covered in more than one article included: wildfires and associated “haze disaster” in Indonesia in 1997 [49,53]; wildfires in San Diego, California in 2007 [31,33]; a 2011 wildfire impacting Sandy Lake First Nation in Canada [27,28]; California’s 2017–2018 wildfire season [43,47,97]; and the catastrophic 2019–2020 wildfire season in southeastern Australia [40,50,89].

Table 2. Environmental hazards, climate change, and disaster recovery cycle (N = 75).

Environmental Categories n (%) Examples
Hazards
Fire 75 (100) • See articles with asterisks in reference list [18,26–99] Air Quality 41 (54.7) • Impacts of air quality and/or particulate matter [26,29,32,34,37,41,42,44–49,51,54,56–62,65,66,71,72,74,78,88–90] Heat 12 (16) • Impacts of heat [18,29,36,39,40,43,57,66,72,80,87,88] Haze 5 (6.7) • Haze disasters and/or impacts of haze [48,53,54,82,85] Specific Wildfire(s)
Specific Wildfire 13 (17.3)
• 1997 wildfire and “haze disaster” in Indonesia [49,53] • 2007 wildfire in San Diego, CA [31,33]
• 2011 wildfire in Canada impacting Sandy Lake First Nation [27,28] Wildfire Season/
Time Period 19 (25.3) • California’s 2017–2018 wildfire season [43,47,97] • Australia’s 2019–2020 wildfire season [40,50,89]
Int. J. Environ. Res. Public Health 2023, 20, 6252 8 of 22 Table 2. Cont.

Environmental Categories n (%) Examples
Climate Change (CC)
Mentions 41 (54.7)
Adaptation 41 (54.7)
• Individual-focused adaptations (e.g., adapting to heat, addressing disease burden, air filtration systems, individual survival plans) [18,29,34,39,41,42,61,66,80,84,85,88,91,96]
• Facility or community-level emergency protocols (planning, preparation, evacuation, communication, etc.) [18,26,27,29,33,47,52,63,79,81,87,97] • Land-use management (including traditional ecological knowledge and burning practices) [29,36,38,50,55,67–70,75,83,94,98]
Mitigation 18 (24)
• Traditional ecological knowledge and burning practices [38,55,64,67–70,73,74,83,94,98]
• Mentions or addresses need to reduce greenhouse gas emissions [29,58,71,80,88,93]
Article Focuses on CC 7 (9.3) • Health impacts of climate change [29,72,80,88] • Disproportionate impact on older adults [18,66,71]
Disaster Recovery Cycle
Mentions 42 (56)
Recovery 9 (12)
• Needs of older adults in recovery period following wildfires (e.g., disruption in continuity of care, physical recovery, economic recovery, and trauma/mental health) [29,32,79,81,97]
• Community recovery [50]
• Debriefing sessions with facility staff following wildfire [31,86] Response 24 (32)
• Needs of older adults during acute wildfire disaster (e.g., life-support equipment such as oxygen during power outages, immediate interventions for air quality, etc.) [9,17,18,34,79]
• Evacuation (individuals, facilities, communities)
[27,28,30,31,40,47,52,76,86,87,91,96]
• Early warning systems, communication, and local response protocols [18,29,33,34,63,96]
• Social support needs (families, caregivers, etc.) [18,27–29,79,81] • Response of health care providers and/or facilities [32,42,52,66,86,92,97] Mitigation 25 (33.3)
• Building codes and updates, and facility emergency protocols [18,29,71] • Mitigating smoke exposure [34,41,88]
• Public outreach, local contingency planning, community risk-reduction etc. [47,63,77,81,87]
• Reintroducing “ecologically beneficial fire” [35] (p. 677) and Indigenous burning practices [35,38,55,64,67–70,73,83,94,98]
Preparation 22 (29.3)
• Barriers or facilitators to preparedness for older adults (e.g., socioeconomic factors, mobility and health issues, etc.) [39,77,79] • Incorporating needs of older adults into planning measures (recommendations, community-engagement, etc.) [29,32,63,66,79,87] • Recommendations for evacuation preparedness and/or facility emergency protocols [27,28,33,52,86,98]
• Individual preparation (survival plans, preparing personal property, evacuating, etc.) [40,76,91,97]
Int. J. Environ. Res. Public Health 2023, 20, 6252 9 of 22 Table 2. Cont.

Environmental Categories n (%) Examples
Environmental Justice (EJ)
Explicit mention of EJ 2 (3)
• Intersectional analysis of subgroups of older adults most impacted by wildfire smoke using an environmental justice lens (e.g., race, gender, education) [40]
• Mention of environmental justice as factor of vulnerability for respiratory disease [26]
Alludes to EJ 40 (53)
• Intersectional view of impacted older adults (more impacted based on race, socioeconomic status, gender, housing status, chronic disease, urban vs. rural, and/or) [18,29,32,34,36,39,64–66,77]
• Calls for more focus on vulnerable populations in future research [18,34,57,65,71,82]
• Connection of Indigenous sovereignty and knowledge, colonization, historical oppression, and resistance [38,55,64,69,70,75,83,94] No Mention of EJ 33 (44)
Indigenous or Aboriginal Peoples
Traditional
Fire Knowledge 12 (16)
• Co-management strategies and tensions between Indigenous elders and peoples and state, national, or other fire management groups [38,64,67–70,83]
• Western science’s need for Indigenous knowledge and tension between the two [38,64,75,94]
• Description of Indigenous fire knowledge, experiences, and/or history [28,38,55,73,75,94,98]
Focus on
Indigenous or
Aboriginal Lands
14 (18.7)
• In addition to the above 12 articles, 2 focused on community experiences and needs during evacuation for Sand Lake First Nation [27,28] Data have consistently shown that climate change is increasing the intensity and impact of wildfires [1,18,60]. However, not all disaster research makes the connection between climate change-related causes and impacts. We found that 41 articles (54.7%) mentioned climate change or global warming explicitly, but only 7 (9.3%) focused on climate change as a main topic. We also collected data on interventions, recommendations, or responses that may be climate mitigation or adaptation strategies, even if they were not named as such. We found that 41 articles (54.7%) addressed some form of adaptation strategies and 18 (24%) addressed mitigation strategies (Table 2).

In addition to climate change, data were collected on mentions of the disaster recovery cycle and specific phases including recovery, response, mitigation, and preparation (Table 2).

A majority of articles mentioned the disaster recovery cycle or at least one phase (56%).

Mitigation measures were the most prevalent phase discussed (33.3%), closely followed by response (32%) and preparation (29.3%). Recovery was the least discussed phase, addressed by eight articles (10.7%).

Because of the particular vulnerability of older adults to disasters, including wildfires, we noted whether articles specifically mentioned environmental justice. However, during analysis, we found that many articles alluded to environmental justice by discussing “disadvantaged and vulnerable populations” [65] or “vulnerable populations, including the elderly, socioeconomically disadvantaged groups, and those with underlying chronic disease . . . [who are] most affected [29]. While only two (3%) articles explicitly named environmental justice [46,60,64], more than half (53%) alluded to environmental justice by discussing disproportionate impacts or particularly vulnerable populations in some way (Table 2).

Int. J. Environ. Res. Public Health 2023, 20, 6252 10 of 22

3.3. Older Adult Findings
When reviewing how articles discussed older adults (Table 3), 39 articles defined older adults based on either an age cutoff (e.g., 65 or older) or by naming this population (e.g., elders, older adults, seniors, etc.). A large portion of articles (41.3%) included older adults as a population they were specifically interested in looking at in addition to others, while 29.3% focused solely on older adults, and the remaining 29.3% made mention of this age group but did not have them as their primary focus. The majority of articles (69.3%) based their findings on older adults using information that was collected about them, rather than from them firsthand (24%), with some (6.7%) doing both. Most articles did not explicitly state the living conditions of the older adults that were included; however, out of the 24 articles that did make this specification, 20 focused on older adults living in the community while only 4 focused on older adults living in long-term care communities.

Table 3. Older adult findings (N = 75).

Categories Related to Older
Adults n (%) Examples
Focus Demographic
Older Adults Sole Focus 22 (29.3) • Focus on Indigenous elders [28,68–70] • Focus on health impacts of older adults in disasters [18,32,54,60] Focus on Older Adults in
Addition to Others 31 (41.3)
• Focus on Indigenous elders in addition to others (other Indigenous people, non-Indigenous land management decision makers, etc.) [27,29,38]
• Participants were stratified by age or age groups and included both older adults and younger participants [34,65,71,88,90,99] Mentioned Older Adults,
but not Focus 22 (29.3)
• Mentioned older adults as another group that could be affected but was not specifically studied in the article [35,47,63,76] Data Sources
From Older Adults 18 (24) • Older adults participated in the study (e.g., completed survey, participated in interview, etc.) [70,75,83,91,94]
About Older Adults 52 (69.3)
• Medical records about older adults were obtained and
analyzed [26,32,48,72,74]
• Other individuals contributed information about older adults (e.g., caregivers, health professionals, first responders, etc.) [52,63] Both 5 (6.7)
• A combination of information shared by older adults and obtained about older adults was used concurrently in the study [27,28,35,40,78]
Living Environment
Community 20 (26.7) • Articles focus on older adults living in community, not in long-term care [27,40,41,55,97]
Long-Term Care 4 (5.3) • Articles focus on older adults living in long-term care communities [31,33,66,86]
Not Specified 51 (68) • Articles do not specify the living setting of the older adult[s] in the study [26,44,56,62,78,95]
With respect to the findings and recommendations made for older adults in the context of wildfires, articles discussed the various ways that older adults are impacted by and respond to wildfires. A majority of articles (60%) discussed the health impacts that wildfires had on older adults, describing increased hospitalization and death rates for cardiovascular and respiratory issues during or following wildfires for this population [26,53,66,82,90].

These negative outcomes were increasingly worse for older women and older adults of color [60]. While a meta-analysis of these impact estimates was beyond the scope of this study, some examples of specific findings include a 7.2% increase in respiratory hospital Int. J. Environ. Res. Public Health 2023, 20, 6252 11 of 22 admissions among Medicare enrollees in the Western U.S. during intense smoke days [59] and impairments to lung function, especially among the elderly, of 33.9% of participants at two-years post-exposure to smoke from a Montana wildfire in the U.S. [78].

Additionally, when it came to responding to a wildfire, most notably with evacuations, older adults faced a disproportionate amount of barriers and challenges including difficulty maintaining the level of care they needed, accessing medications, and staying connected with caregivers, demonstrating how the needs of older adults may not be fully considered and addressed during wildfire disasters [27,28]. Finally, findings illustrated the role that older adults play during wildfires in supporting their local community, family members, and friends. During evacuations, older adults offered additional support to one another by making meals for one another, helping with laundry, and providing emotional support [27,28].

3.4. Thematic Topics, Problem-Focus, Interventions, Recommendations, and Future Research While reviewing included articles, authors made note of recurring themes of interest that provided additional insight on the impacts and experiences felt by older adults due to wildfires (Table 4). With respect to the experiences of older adults, 17.3% (n = 13) of articles discussed animals/pets, 12% (n = 9) included caregivers, 34.7% (n = 26) touched on evacuation efforts and experiences, 14.7% (n = 11), focused on intergenerational relationships during wildfires, and 37.3% (n = 28) mentioned the effect of social support/social capital for this population during these disasters. Additionally, some articles discussed more specific impacts on older adults during wildfires, including 25.3% (n = 19) that looked at mental health associations, and 48% (n = 36) focused on morbidity and/or mortality of wildfires and associated hazards (air pollution, particulate matter, heat, etc.) from an epidemiological focus on population health. Finally, it should be noted that the onset of the COVID-19 pandemic brought about additional issues, especially as they relate to an older adult’s health and well-being, and 2.7% (n = 2) of articles discussed the added complexity to the impact of wildfires.

Table 4. Specific topics and themes (N = 75).

Thematic Topic n (%) Examples
Animals/Pets 13 (17.3)
• Traditional ecological knowledge including importance of animals in landscape, ecosystem, or relationality between humans and the more-than-human world [55,67–70,75,83,94,98]
• “Animal guardians” or “animal ownership” and its impact on evacuation, preparedness, and/or emergency response [76,91,96]
Caregivers 9 (12)
• Importance of having caregivers of older adults involved in and/or educated on preparedness protocol for disasters [18,27,81]
• How the presence of a caregiver can impact how well older adults do during wildfires [28,76]
COVID-19 2 (2.7) • Wildfires and older adults within the context of COVID-19 [40,47] Evacuation 26 (34.7)
• Individual/community evacuation preparedness and/or experiences (e.g., survival plans, etc.) [27,28,30,40,76,87,96]
• Medical facilities’ evacuation preparedness and/or experiences [31,33,52,86,97] • Needs of/impacts on older adults during evacuations (care disruption, communication, social support, etc.) [18,43,60,79,81,88] • Not focused on evacuation, but mention implications, needs, or considerations for evacuation [66,81,84,88]
Int. J. Environ. Res. Public Health 2023, 20, 6252 12 of 22 Table 4. Cont.

Thematic Topic n (%) Examples
Health Issues 45 (60)
• Wildfire impacts on respiratory and/or cardiovascular health [26,34,54,62,84] • Heat-related hospitalizations, illnesses and/or deaths [36,66,80] • Effects of wildfires on cancer [81,82]
• Complications for older adults with dementia [39,43]
Intergenerational 11 (14.7)
• How relationships between generations were impacted by wildfires or how these intergenerational relationships could be used as a protective factor against the negative impacts of these natural disasters [30,40]
• Intergenerational transmission of Indigenous knowledge [55,64,67–69,83,94] Mental Health 19 (25.3)
• General discussion of traumatic impact of disasters/wildfires, evacuation, etc. [27,30,40,86,96]
• Disproportionate impact of disasters/wildfires on older adults’ mental health [18,32,79,81]
• Vulnerability of individuals with mental health issues during disasters and/or heat [29,32,36,39]
• Gap in research on mental health impact of disasters/wildfires [81,82] Morbidity/Mortality 36 (48)
• Secondary data of mortality rates related to wildfire smoke-related exposures (PM2.5, PM10, heat, etc.) [26,36,43,51,56,74,85]
• Secondary data of hospital records measuring morbidity of diseases (respiratory, pulmonary, cardiovascular, cancer, etc.) [37,44,46,48,54,59–62,81,84,89,90,93,95] • Systematic reviews of morbidity and/or mortality from wildfire smoke and related exposures [32,34,57,82,99]
• Primary data of health impacts related to wildfire smoke and related exposures [45,49,53,78]
• Future projections of hospital admissions under climate forecasting scenarios [58] • General review of climate change impacts on morbidity and mortality [66,72,88] Social Support or
Social Capital 28 (37.3)
• Importance of shared social support networks on older adults’ well-being, especially in disasters [19,28,43]
• Social isolation as a risk factor for older adults during wildfires [18,39,79] • Use of social networks on a community level to prepare for and respond to wildfires [47,77,81]
Of the 75 articles, 56 (74.4%) were problem-focused, describing negative impacts of wildfires in some way (e.g., need for evacuation, impacts of air quality, needs of communities, etc.) (Table 5). Of the 56 that focused on problem description, 36 (48%) described problems of morbidity or mortality related to wildfires and/or wildfire smoke. Most of these used secondary, epidemiologic data such as hospital admissions and death rates to describe the health impacts of wildfire smoke and/or PM2.5 (particulate matter smaller than

2.5 microns). Aside from morbidity and mortality, other articles described problems with evacuation, displacement, and/or issues with disaster response [27,28,40,47,86,87,91,96,97]; inequalities and vulnerabilities of certain populations to wildfires [29,77,79]; and general descriptions of health impacts without epidemiologic data [35,42,80].

Int. J. Environ. Res. Public Health 2023, 20, 6252 13 of 22 Table 5. Responses or interventions, solutions and recommendations, and future research (N = 75).

Categories n (%) Examples
Problem
Description 56 (74.7)
• Health impacts of wildfires and related exposures (smoke, heat, etc.) [26,32,41–46,51,53,54,56–62,72,74,78,80–82,84,85,88–90,93,95,99] • Problems/lessons from evacuation and/or disaster response (individuals, communities, and/or facilities) [27,28,40,47,86,87,91,96,97]
• Disproportionate impact of wildfires and disasters on older adults [18,29,32,39,60,65,66,71,79]
Responses or
Interventions 31 (41.3)
• Evacuation at individual, organizational, facility, or community levels [27,28,30,31,33,52,86,87]
• Community-level disaster response (communication, first responders, coordination of services, increasing community-engagement and relationships, etc.) [31,33,47,63,87] • Individual level response/interventions (e.g., air filters, masking, survival plans, etc.) [34,41,53,76,84,91,96]
• Indigenous response/interventions including traditional burning, integrating TEK into conservation/land “management”, employing Indigenous peoples in land “management”, etc. [38,55,64,67–70,73,75,94,98]
• Community care, social capital, caring for one another during disasters [27,28,40,50] • Interventions by healthcare providers and/or facilities [42,52,86,92,97] Solutions and
Recommendations 61 (81.3)
• Need for community-level disaster response workers and coordinators to increase community engagement towards better response and preparedness (community-inclusiveness, responsiveness, education, trust building, outreach, etc. [36,39,43,47,63,77,79,81]
• Needs for elders leading up to, during, and following evacuation [27,28,43,63,79] • Recommendations for healthcare providers and/or facilities regarding clinical or organizational response to wildfires [31,33,66,88,90]
• Recommendations for public policy [34,60–62,71,78,81] • Recommendations for utilizing TEK into wildfire management agencies and tactics (in culturally sensitive, ethical ways) [38,67–70,75,83]
• Greater need for community care, increased social support, etc. [18,27,28,40,43,47,50,79] • Recommendations for climate mitigation and/or drawdown strategies [29,64,71,88,93] Future Research
Directions 61 (81.3)
• Future research should work to better understand the impacts of wildfires on the health of older adults [26,31,48,57,65,81]
• More research is needed on how to develop and evaluate community preparedness and response strategies and the effects of these strategies [27,28,47,84,87,97] • Additional research should look at and evaluate effective mitigation strategies [55,63,88] • More research should work to find ways to address specific needs of older adults and reduce risks faced by this population before, during, and following wildfires [18,66,77] Many articles moved beyond problem description with 32 of the 75 (41.3%) articles describing responses or interventions during, after, or in preparation for wildfires. Interventions and responses included individual, organizational, and community-level efforts.

Individual efforts included masking to avoid smoke exposure [53,84], installing in-home air filters [34,41,84], and creating survival plans [76,91,96]. Organizational interventions predominantly focused on organizations (e.g., long-term care facilities, rehabs, and hospitals) evacuation and/or disaster management plans [27,28,30,31,33,52,86,97], but also included treatment recommendations for providers [42,92]. Community-level responses included descriptions of families and neighbors caring for one another during acute disaster phases [27,28,30,40,50], and disaster management and coordinating systems at the community level [28,30,63,87]. Finally, many articles described traditional ecological knowledge (TEK) or Indigenous and Aboriginal elders as an important intervention for “hazard abatement” [55], as well as the opportunity for fire management institutions to listen to, learn from, and rematriate (e.g., return power to Indigenous peoples to reclaim ancestral traditions) [100] fire “management“ as well as the ethics of fire management agencies “using” this knowledge [26,38,64,67–70,73,75,94,98].

Int. J. Environ. Res. Public Health 2023, 20, 6252 14 of 22 In addition to interventions and solutions, 61 (81.3%) articles provided recommendations targeted at multiple levels and points of intervention including individuals, organizations, communities, policy, scholarly literature, and disaster and fire management agencies. Many recommendations intersected with other findings, such as recommended adaptation strategies [29,58,61,66,71,88] and the importance of individual survival plans, community evacuation plans, and organizational disaster management protocols and plans, especially in relation to communicating with older adults [28,39,52,54,76,77,79,81,97].

Community-centered disaster management planning and strategies were prolific across recommendations, with 26 of the 31 (83%) discussing community needs, community engagement, or community inclusion in disaster management planning in some way (Table 5).

Finally, most articles (81.3%) outlined areas for future research, describing the importance of utilizing more rigorous and longitudinal research methods to examine the long-term health effects on older adults due to wildfires, especially those from more minoritized communities (Table 5). Additionally, findings suggest community and local government officials need to consider the needs of older adults during wildfires and research should serve as a tool to evaluate the short- and long-term impacts of responses and interventions through all phases of the disaster recovery cycle [2,7].

3.5. Study Strengths and Limitations
One strength of this review is its systematic and rigorous approach to identifying relevant peer-reviewed literature, by using expansive search terms and searching more than 10 databases. This allowed a breadth of literature to be explored across geographic regions, fields of study, and disciplines. However, one limitation is the exclusion of gray literature (e.g., books, non-peer-reviewed articles, etc.) that may have had additional information related to the impact of wildfires on older adults and relevant recommendations or interventions. Further, our search was limited to publications available in English, which excluded two potential studies from full review, as well as other non-English publications that may have been excluded from our initial database search.

4. Discussion

4.1. Wildfires and Older Adults: Increased Engagement and Trends While there is prolific literature on the impact of extreme heat and hurricanes on older adults, there is a gap in the literature “on the vulnerability of older adults to other health-related climate impacts, such as . . . wildfire [and] changes in air quality” [18] (p. 21).

This review systematically synthesized scholarly literature focusing on older adults and wildfires to help identify priorities and directions for addressing gaps in the literature on the impact of wildfires on older adults, and recommendations for interventions and future research. In a global search with no restriction on publication date, only 75 articles were found and most (52%) were published within the past 5 years (2017–2021). This may indicate the impact of wildfires on older adults is a newer area of research that requires additional exploration and evaluation.

Wildfires may have unique health impacts that spread beyond a specific boundary where the disaster occurred, as smoke and air quality transcend boundaries, with smoke from large fires sometimes traveling thousands of miles, across countries and even continents [54,101,102]. This was seen in multiple articles, with some specifically addressing “long-range transboundary air pollution” [54,85] and others examining health-related impacts of air quality even when the source of the fire was in a different geographic location [53,82,85].

Findings from this review show the particular vulnerabilities of older adults to wildfires, particularly due to poor air quality and exposure to smoke and particulate matter (i.e., PM2.5). Many articles within the review explained that older adults are more susceptible to adverse health impacts of PM2.5 [29,34,37,42], as are those with pre-existing respiratory or cardiovascular diseases and those with lower socioeconomic status (SES) [29,34,72].

While older adults are named as specifically susceptible, they also often have pre-existing Int. J. Environ. Res. Public Health 2023, 20, 6252 15 of 22 conditions or may have lower incomes, exhibiting a double—or triple—burden related to poor air quality. While there is substantial research on health impacts related to particulate matter, some studies have found that PM2.5 exposure from wildfires may be more toxic than equal doses of ambient PM2.5 [59,103], highlighting the importance of examining wildfire-related air quality and health impacts, especially for older adults.

Impacts of air quality are compounded by heat exposure—another hazard related to wildfires. Many articles spoke to the health impacts of heat on older adults particularly, highlighting “the double burden that heat and socioeconomics play for low-income older adults who are unable to afford air conditioning or caregiver support during extreme heat” [66] (p. 7). Heat-related deaths are the most deadly “natural disaster”, and accompany wildfires—along with poor air quality—illustrating the impact of wildfires on older adults even if they are not directly exposed to the epicenter of a wildfire event [36].

Aside from indirect—albeit very real—impacts of wildfires through air quality and smoke, many articles discussed acute phases of the disaster recovery cycle when a wildfire occurs, namely the response phase (32%) and evacuation (34.7%). The findings showed that older adults are particularly vulnerable during evacuation phases, noting the importance of considering elders when planning for communitylevel communications for evacuation [47,79,81] and physical difficulties elders may have with evacuation, especially without social support [18,27,28,79]. Even if older adults are not evacuated, being in the geographic region of a wildfire event with power outages may affect life-sustaining equipment such as oxygen, ventilators, CPAP machines, refrigeration for medications, power wheelchairs, elevators, and heating and cooling systems for body regulation [18,79]. Wildfires may also pose a threat to the continuity of care for older adults who need ongoing medical treatment such as dialysis, cancer treatment, obtaining medications, or other medical needs [18,79,97].

4.2. Dominant Narratives: Secondary Data and Epidemiological Studies The most prolific finding in this review was the use of secondary data to measure morbidity and mortality from wildfires or associated hazards (e.g., heat, air quality, etc.). This aligns with findings from an included article stating, “in relation to extreme weather conditions, literature has highlighted the vulnerability of older adults as a cohort, though there is limited attention on how to prevent the cohort from experiencing increased risk” [39] (p. 974). The majority of articles (74.7%) focused on problem description, with 48% of all articles describing the problem of morbidity and mortality impacts—either focusing on older adults or whose findings skewed towards older adults. This illustrates the dominant narrative of wildfires and older adults, telling a story of risk and vulnerability. While many articles also discussed responses or interventions, these were still predominantly focused on describing problems within the intervention or response itself, such as lessons learned from evacuation or community responses or preparation. Epidemiologic findings are imperative to provide statistics to build a base of scientific knowledge about this issue, but they only tell a fraction of the story about older adults, leaving out vital information from older adults on their lived experiences and needs before, during, and after wildfires.

4.3. Older Adults: Lived Experiences and Primary Data Sources The results demonstrate how most of the information on the intersection of wildfires and older adults is primarily data collected about older adults from other sources rather than from this population firsthand. Medical and hospital records were one of the main sources of information that articles drew from, focusing on the negative physical health effects of wildfires on this population, but articles rarely focused on learning from what older adults went through or how they felt about wildfires and their role in relation to these disasters. To adequately address the disproportionately negative issues faced by older adults in the face of wildfires, it is essential to better understand their perspectives and what they find to be their greatest challenges and needs during these disasters. Articles also demonstrated how older adults can be a vital source of knowledge in knowing how to reduce or respond to wildfires, as evidenced by Int. J. Environ. Res. Public Health 2023, 20, 6252 16 of 22 the numerous articles on the role Indigenous elders have previously had in mitigation efforts (see also Section 4.4) [38,55,64,67–70,73,74,83,94,98]. It is important to understand that older adults are not simply victims of wildfires but can, in fact, play a major role in addressing these growing disasters.

4.4. Social Support and Community Focus
Articles demonstrated the importance of social support for older adults at both a community and individual level. Older adults who lacked social support were more likely to die during a wildfire as they did not receive adequate warning of the danger or were unable to evacuate on their own [43,79]. Caregivers were noted as a vital source of support for older adults but were still in need of the appropriate resources and financial assistance to prepare for and respond to wildfires [97]. Caregivers should be considered a valuable point of contact for older adults in providing needed public health and disaster response messaging to this population [81]. When formal institutional responses were not adequate in meeting the needs of older adults, articles stressed the importance and power of informal neighborhood and community responses to make up for this lack of support [43,50]. In fact, one study found the number of fatalities due to wildfires was reduced when communities supported their older adults [43]. A good social support network was also found to provide critical psychological and emotional support for older adults during wildfire evacuations, which older adults cited as the most prevalent and valuable support they received during this crisis [27,28].

Building on the importance of social support and community care, many articles discussed the need for community-engaged tactics within disaster management systems including first responders and emergency management agencies. Articles discussed the need for community-responsive practices, with one article asserting “community engagement to determine most appropriate strategies from the local level should become a focus of adaptation. For example, bushfire preparedness and management should incorporate knowledge of community, government, and industry groups to identify impacts on community safety” [29] (p. 754). Other articles reiterated this, highlighting the need to build partnerships between local, state, and federal emergency management and public health systems, and that these should be in conversation and relationship with community members and responsive to their needs [32,35]. Other findings highlighted the need for communication strategies to be developed in conjunction with communities [34,35] and a need for more education and “community activism . . . to promote outreach that assists vulnerable persons [e.g., older adults] during emerging hazardous weather situations” [43] (p. 383).

4.5. Elders and Traditional Ecological Knowledge (TEK)
An important finding of this review was the inclusion of Indigenous, First Nations, and/or Aboriginal elders’ experiences and knowledge of fire. Almost 20% (n = 14) of articles focused explicitly on Indigenous, First Nations, or Aboriginal elders, with 12 focusing on fire knowledge and TEK and 2 focusing on the impacts of evacuation during a fire event [27,28]. While it is beyond the scope of this paper to fully explore the relevance of TEK to wildfires, this emergent finding became salient during data collection and analysis due to the volume of related articles. These articles highlighted the importance of community in a different way, illustrating the deeply held community and cultural ties of Indigenous peoples to each other and the land. In contrast with other articles focusing on evacuation, those focused on the evacuation of Indigenous peoples highlighted a deeper sense of social cohesion and therefore social disruption when evacuations occurred. A participant from one article discussed the way evacuation broke up “communityness” stating, “the evacuation breaks up families, it breaks up that ‘communityness’, how you feel home. It breaks that up and you’re being sent to a strange land” [27] (p. 372). These findings illustrate not only recommendations for Indigenous elders during evacuations, Int. J. Environ. Res. Public Health 2023, 20, 6252 17 of 22 but also aspects of building “communityness” and social cohesion that other communities may learn from as a form of disaster preparation and response.

The majority of articles focused on Indigenous elders’ fire knowledge and how this contrasts with dominant “fire management” agencies, policies, and protocols. Fire knowledge included cultural and traditional burning practices that have been utilized by Indigenous peoples for generations, and how fire knowledge is a part of sacred and cultural practices of being in relationship with the land. Almost all of these articles discussed implications of fire knowledge for fire management agencies, and many included crosscultural dialogues or comparisons between Indigenous elders and other fire management agencies [38,55,64,67–69,75,83]. Many of these articles discussed the difference between Indigenous peoples’ ontological views of fire and those of fire management institutions, most of which are run by White settler nations (e.g., Australia, Canada, and the United States). One article explained the difference between TEK and scientific ecological knowledge (SEK) [64], explaining that TEK takes a relational view of nature whereas SEK views nature through lenses of control, domination, and subjugation. Other articles affirmed this, explaining the incongruity of “fire management” or “fire-fighting” with TEK’s view of fire and land as something to be in balance and relationship with rather than managed or fought [38,64,68,75].

Findings provided examples for collaborative co-management between Indigenous elders and fire management agencies, highlighting the importance and potential of TEK in fire “management” practices, while also naming the tension and ethics of non-Indigenous peoples “using” TEK for fire mitigation and/or adaptation measures [64,70]. One article explained that Indigenous elders have difficulty trusting fire management agencies run by the government due to generational trauma of genocide, relocation, and colonialism, with an Indigenous elder stating “science means not us” [83] (p. 26). Other articles provided recommendations for adaptive co-management strategies to build relationships between fire management agencies and Indigenous peoples to create “cross-cultural partnerships directed towards fostering resilience” [68–70]. These findings illustrate a nuanced and complex picture of the role of TEK in fire “management”. Indigenous people have been caretakers of the land for generations and TEK must be incorporated into any understanding of ecological care, including wildfire management. While ethics and use of TEK are beyond the scope of this paper, there is a breadth of literature that looks at the intersection of TEK and fire “management”, building on the articles related to TEK in this review. For the purpose of this paper, these findings illustrate not only the impact that wildfires have on older adults, but also the positive impact older adults can have on adaptation, mitigation, or responses to wildfires.

4.6. Increased Focus on (Un)Natural Disasters: Climate Change and Environmental Justice While a primary focus of this paper was the impact of wildfires on older adults, this impact cannot be understood without analyzing the causes of wildfires. While wildfires are not new, the frequency and intensity of wildfires have dramatically increased due to climate change, creating (un)natural disasters [1]. While only 7 (9.3%) articles had a primary focus on climate change, 41 (51.7%) mentioned climate change as a reason for increasing disasters, reaffirming the relationship between worsening wildfires and climate change.

Of the seven articles focused on climate change, three highlighted the disproportionate impact of climate change on older adults [18,66,71] while others had findings that skewed towards older adults [29,72,80,88]. However, the low number of articles focusing on the intersection of climate change and older adults illustrates a need for further research in this area, particularly the relationship between climate change, wildfires, and older adults. This intersection will only become more pertinent, as 8 of the 10 worst global wildfire seasons have happened in the past decade. Coupling the increased intensity and frequency of wildfires with the ongoing COVID-19 pandemic, further research is needed to examine these intersectional crises and their impacts on older adults.

Int. J. Environ. Res. Public Health 2023, 20, 6252 18 of 22 The findings from this review continually reiterated that the impact of disasters is not distributed equally. The disproportionate impact of environmental hazards on some groups of people more than others is known as environmental injustice. While only 2 (3%) articles named environmental justice specifically, 40 (53%) of articles alluded to environmental justice in some way. Most of these references were related to older adults as being particularly “vulnerable” to wildfires and associated hazards (e.g., heat, air quality), while others provided a more nuanced understanding of environmental justice with other intersecting identities such as race, class, ethnicity, gender, geographic location (i.e., urban versus rural), and socioeconomic status [18,29,64–66,77]. While environmental justice is well-documented within scholarly literature, these findings point to the importance of incorporating an environmental justice perspective into research on wildfires and older adults. Some articles that used the “dominant narrative” named above (i.e., epidemiological studies of morbidity and mortality) used variables to understand the impacts of intersectional identities, providing a framework to incorporate environmental justice, though others named the gap in understanding environmental justice through public health-related data [65]. Some articles explicitly named this as a limitation or need for future research [34,47,49,57,65]. Future studies of all kinds should incorporate environmental justice into their data collection, methods, or analysis to understand the nuanced and disproportionate burden or wildfires on vulnerable populations and ways to address these harms and uneven impacts. Studies may also build upon this literature base by incorporating climate justice into environmental justice, especially in the case of increasing and worsening wildfires [104].

5. Conclusions
Findings from this scoping review demonstrate how older adults can be an important source of knowledge for wildfire mitigation, response, recovery, and adaptation strategies and should be included in local community planning efforts. Additional efforts should be made to incorporate environmental justice and intersectionality to better understand the root causes in health disparities among older adults during and following wildfires.

Overall, the literature on the different ways older adults respond to or are impacted by wildfires is still relatively new and needs further development and exploration to better learn from and support this population in the face of worsening wildfire disasters.

Author Contributions: Conceptualization, methodology, C.C.M., C.M.D.F. and L.R.M.; data collection, C.C.M. and C.M.D.F.; data analysis, C.C.M., C.M.D.F. and R.M.S.; writing—original draft preparation, C.C.M., C.M.D.F. and R.M.S.; writing—review and editing, C.C.M., C.M.D.F., R.M.S. and L.R.M. All authors have read and agreed to the published version of the manuscript.

Funding: Open access publishing of this manuscript was supported by the University of Denver’s Open Access Publication Equity Fund.

Institutional Review Board Statement: Not applicable.

Informed Consent Statement: Not applicable.

Data Availability Statement: Not applicable. The included literature in references.

Acknowledgments: We thank Jenny Bowers for database and search consultation.

Conflicts of Interest: The authors declare no conflict of interest.

References

1. Intergovernmental Panel on Climate Change (IPCC). Climate Change 2021: The Physical Science Basis. Contribution of Working Group I to the Sixth Assessment of the Intergovernmental Panel on Climate Change; Masson Delmotte, V., Zhai, P., Pirani, A., Connors, S.L., Péan, C., Berger, S., Caud, N., Chen, Y., Goldfarb, L., Gomis, M.I., et al., Eds.; Cambridge University Press: Cambridge, UK, 2021.

2. Levy, B.S.; Patz, J.A. Climate Change, Human Rights, and Social Justice. Ann. Glob. Health 2015, 81, 310–322. [CrossRef]

3. Hoover, K.; Hanson, L.A. Wildfire Statistics. Congressional Research Service. Available online: https://fas.org/sgp/crs/misc/ IF10244.pdf (accessed on 8 December 2022).

Int. J. Environ. Res. Public Health 2023, 20, 6252 19 of 22

4. United Nations Environment Programme. Spreading Like Wildfire: The Rising Threat of Extraordinary Landscape Fires. 2022. Available online: http://www.unep.org/resources/report/spreading-wildfire-rising-threat-extraordinary-landscape-fires (accessed on 8 December 2022).

5. Reid, C.E.; Brauer, M.; Johnston, F.H.; Jerrett, M.; Balmes, J.R.; Elliott, C.T. Critical Review of Health Impacts of Wildfire Smoke Exposure. Environ. Health Perspect. 2016, 124, 1334–1343. [CrossRef] [PubMed]

6. Stone, S.L.; Anderko, L.A.; Butler, C.R.; Cascio, W.E.; Clune, A.; Damon, S.; Garbe, P.; Hauptman, M.; Haskell, W.E.; Hoshiko, S.; et al. Wildfire Smoke: A Guide for Public Health Officials|AirNow.Gov. 2019. Available online: https://www.airnow.gov/ publications/wildfire-smoke-guide/wildfire-smoke-a-guide-for-public-health-officials/ (accessed on 8 December 2022).

7. Henderson, S.B.; Brauer, M.; MacNab, Y.C.; Kennedy, S.M. Three Measures of Forest Fire Smoke Exposure and Their Associations with Respiratory and Cardiovascular Health Outcomes in a Population-Based Cohort. Environ. Health Perspect. 2011, 119, 1266–1271. [CrossRef] [PubMed]

8. Mehmood, K.; Iqbal, M.; Abrar, M.M. Can Exposure to PM2.5 Particles Increase the Incidence of Coronavirus Disease 2019 (COVID-19)? Sci. Total Environ. 2020, 741, 140441. [CrossRef]

9. McCaffrey, S. Community Wildfire Preparedness: A Global State-of-the-Knowledge Summary of Social Science Research. Curr.

For. Rep. 2015, 1, 81–90. [CrossRef]

10. Jenkins, J.L.; Hsu, E.B.; Sauer, L.M.; Hsieh, Y.-H.; Kirsch, T.D. Prevalence of Unmet Health Care Needs and Description of Health Care–seeking Behavior Among Displaced People After the 2007 California Wildfires. Disaster Med. Public Health Prep. 2009, 3, S24–S28. [CrossRef] [PubMed]

11. Eisenman, D.; McCaffrey, S.; Donatello, I.; Marshal, G. An Ecosystems and Vulnerable Populations Perspective on Solastalgia and Psychological Distress After a Wildfire. EcoHealth 2015, 12, 602–610. [CrossRef] [PubMed]

12. Ashida, S.; Robinson, E.L.; Gay, J.; Slagel, L.E.; Ramirez, M.R. Personal Disaster and Emergency Support Networks of Older Adults in a Rural Community: Changes After Participation in a Preparedness Program. Disaster Med. Public Health Prep. 2017, 11, 110–119. [CrossRef] [PubMed]

13. Durant, T.J. The Utility of Vulnerability and Social Capital Theories in Studying the Impact of Hurricane Katrina on the Elderly.

J. Fam. Issues 2011, 32, 1285–1302. [CrossRef]

14. Somes, J. The Loneliness of Aging. J. Emerg. Nurs. 2021, 47, 469–475. [CrossRef]

15. Aldrich, N.; Benson, W.F. Disaster Preparedness and the Chronic Disease Needs of Vulnerable Older Adults. Prev. Chronic Dis. 2008, 5, A27. [PubMed]

16. Acierno, R.; Ruggiero, K.J.; Kilpatrick, D.G.; Resnick, H.S.; Galea, S. Risk and Protective Factors for Psychopathology Among Older versus Younger Adults After the 2004 Florida Hurricanes. Am. J. Geriatr. Psychiatry 2006, 14, 1051–1059. [CrossRef]

17. Gibson, A.; Kusmaul, N.; Leedahl, S.; Rizzo, V.M. Create Social Responses to a Changing Environment. In Gerontological Social Work and the Grand Challenges; Sanders, S., Kolomer, S.R., Spellman, C.W., Rizzo, V.M., Eds.; Springer: Berlin/Heidelberg, Germany, 2019; pp. 93–110.

18. Gamble, J.L.; Hurley, B.J.; Schultz, P.A.; Jaglom, W.S.; Krishnan, N.; Harris, M. Climate Change and Older Americans: State of the Science. Environ. Health Perspect. 2013, 121, 15–22. [CrossRef] [PubMed]

19. Munn, Z.; Peters, M.D.J.; Stern, C.; Tufanaru, C.; McArthur, A.; Aromataris, E. Systematic Review or Scoping Review? Guidance for Authors When Choosing between a Systematic or Scoping Review Approach. BMC Med. Res. Methodol. 2018, 18, 143. [CrossRef]

20. Arksey, H.; O’Malley, L. Scoping Studies: Advancing the Methodology. Int. J. Soc. Res. Methodol. 2005, 8, 19–32. [CrossRef]

21. Levac, D.; Colquhoun, H.; O’Brien, K.K. Scoping Studies: Advancing the Methodology. Implement. Sci. 2010, 5, 69. [CrossRef]

22. Colquhoun, H.L.; Levac, D.; O’Brien, K.K.; Straus, S.; Tricco, A.C.; Perrier, L.; Kastner, M.; Moher, D. Scoping Reviews: Time for Clarity in Definition, Methods, and Reporting. J. Clin. Epidemiol. 2014, 67, 1291–1294. [CrossRef] [PubMed]

23. Tricco, A.C.; Lillie, E.; Zarin, W.; O’Brien, K.K.; Colquhoun, H.; Levac, D.; Moher, D.; Peters, M.D.J.; Horsley, T.; Weeks, L.; et al.

PRISMA Extension for Scoping Reviews (PRISMA-ScR): Checklist and Explanation. Ann. Int. Med. 2018, 169, 467–473. [CrossRef]

24. World Health Organization: Ageing and Health. Available online: https://www.who.int/news-room/fact-sheets/detail/ageingand-health (accessed on 1 February 2023).

25. Covidence. Covidence Systematic Review Software; Veritas Health Information: Melbourne, Australia, 2022.

26. Analitis, A.; Georgiadis, I.; Katsouyanni, K. Forest Fires Are Associated with Elevated Mortality in a Dense Urban Setting. Occup.

Environ. Med. 2012, 69, 158–162. [CrossRef] [PubMed]

27. Asfaw, H.W.; McGee, T.; Christianson, A.C. The Role of Social Support and Place Attachment during Hazard Evacuation: The case of Sandy Lake First Nation, Canada. Environ. Hazards 2019, 18, 361–381. [CrossRef]

28. Asfaw, H.W.; McGee, T.K.; Christianson, A.C. Indigenous Elders’ Experiences, Vulnerabilities and Coping during Hazard Evacuation: The Case of the 2011 Sandy Lake First Nation Wildfire Evacuation. Soc. Nat. Resour. 2020, 33, 1273–1291. [CrossRef]

29. Bambrick, H.J.; Capon, A.G.; Barnett, G.B.; Beaty, R.M.; Burton, A.J. Climate change and health in the urban environment:

Adaptation opportunities in Australian cities. Asia Pac. J. Public Health 2011, 23, 67S–79S. [CrossRef] [PubMed]

30. Barber, C.E. Full Article: Forest Fire As a Shared Intergenerational Experience: Perceived Short-term Impacts on the GrandparentGrandchild Relationship. J. Intergener. Relatsh. 2014, 12, 128–140. [CrossRef]

31. Barnett, J.; Dennis-Rouse, M.; Martinez, V. Wildfire Disaster Leads to Facilities Evacuation. Orthop. Nurs. 2009, 28, 17–20. [CrossRef]
Int. J. Environ. Res. Public Health 2023, 20, 6252 20 of 22

32. Bell, S.A.; Horowitz, J.; Iwashyna, T.J. Health Outcomes After Disaster for Older Adults with Chronic Disease: A Systematic Review. Gerontologist 2020, 60, e535–e547. [CrossRef]

33. Blake, S.C.; Howard, D.H.; Eiring, H.; Tarde, S. San Diego’s Area Coordinator System: A Disaster Preparedness Model for US Nursing Homes| Disaster Medicine and Public Health Preparedness|Cambridge Core. Disaster Med. Public Health Prep. 2012, 6, 424–427. [CrossRef]

34. Cascio, W.E. Wildland Fire Smoke and Human Health—ScienceDirect. Sci. Total. Environ. 2018, 624, 586–595. [CrossRef]

35. Cisneros, R.; Alcala, E.; Schweizer, D.; Burke, N. Smoke Complaints Caused by Wildland Fire in the Southern Sierra Nevada Region, California. Int. J. Wildland Fire 2018, 27, 677. [CrossRef]

36. Coates, L.; Haynes, K.; O’brien, J.; McAneney, J.; de Oliveira, F.D. Exploring 167 Years of Vulnerability: An Examination of Extreme Heat Events in Australia 1844–2010. Environ. Sci. Policy 2014, 42, 33–44. [CrossRef]

37. DeFlorio-Barker, S.; Crooks, J.; Reyes, J.; Rappold, A.G. Cardiopulmonary Effects of Fine Particulate Matter Exposure among Older Adults, during Wildfire and Non-Wildfire Periods, in the United States 2008–2010. Environ. Health Perspect. 2019, 127, 37006. [CrossRef]

38. Eriksen, C.; Hankins, D.L. The Retention, Revival, and Subjugation of Indigenous Fire Knowledge through Agency Fire Fighting in Eastern Australia and California. Soc. Nat. Resour. 2014, 27, 1288–1303. [CrossRef]

39. Farugia, T.L.; Cuni-Lopez, C.; White, A.R. Potential Impacts of Extreme Heat and Bushfires on Dementia. J. Alzheimer’s Dis. 2021, 79, 969–978. [CrossRef]

40. Fergie, D.; Lucas, R.; Harrington, M. Take My Breath Away: Transformations in the Practices of Relatedness and Intimacy through Australia’s 2019–2020 Convergent Crises. Anthr. Action 2020, 27, 49–62. [CrossRef]

41. Fisk, W.J.; Chan, W.R. Health Benefits and Costs of Filtration Interventions That Reduce Indoor Exposure to PM2.5 during Wildfires. Indoor Air 2017, 27, 191–204. [CrossRef] [PubMed]

42. Forsberg, N.T.; Longo, B.M.; Baxter, K.; Boutté, M. Wildfire Smoke Exposure: A Guide for the Nurse Practitioner. J. Nurse Pract. 2012, 8, 98–106. [CrossRef]

43. Garner, J.M.; Iwasko, W.C.; Jewel, T.D.; Charboneau, B.R.; Dodd, A.A.; Zontos, K.M. A Multihazard Assessment of Age-Related Weather Vulnerabilities. Weather Clim. Soc. 2020, 12, 367–386. [CrossRef]

44. Haikerwal, A.; Akram, M.; Del Monaco, A.; Smith, K.; Sim, M.R.; Meyer, M.; Tonkin, A.M.; Abramson, M.J.; Dennekamp, M.

Impact of Fine Particulate Matter (PM2.5) Exposure During Wildfires on Cardiovascular Health Outcomes. J. Am. Heart Assoc. 2015, 4, e001653. [CrossRef]

45. Huttunen, K.; Siponen, T.; Salonen, I.; Yli-Tuomi, T.; Aurela, M.; Dufva, H.; Hillamo, R.; Linkola, E.; Pekkanen, J.; Pennanen, A.; et al. Low-Level Exposure to Ambient Particulate Matter Is Associated with Systemic Inflammation in Ischemic Heart Disease Patients—ScienceDirect. Environ. Res. 2012, 116, 44–51. [CrossRef]

46. Ignotti, E.; Valente, J.G.; Longo, K.M.; Freitas, S.R.; Hacon, S.D.S.; Netto, P.A. Impact on Human Health of Particulate Matter Emitted from Burnings in the Brazilian Amazon Region. Rev. Saude Publica 2010, 44, 121–130. [CrossRef]

47. Jia, S.; Kim, S.H.; Nghiem, S.V.; Doherty, P.; Kafatos, M.C. Patterns of Population Displacement during Mega-Fires in California Detected Using Facebook Disaster Maps. Environ. Res. Lett. 2020, 15, 074029. [CrossRef]

48. Johnston, F.H.; Purdie, S.; Jalaludin, B.; Henderson, S.B.; Morgan, G.G. Air Pollution Events from Forest Fires and Emergency Department Attendances in Sydney, Australia 1996–2007: A case-crossover analysis. Environ. Health 2014, 13, 105. [CrossRef] [PubMed]

49. Kim, Y.; Knowles, S.; Manley, J.; Radoias, V. Long-Run Health Consequences of Air Pollution: Evidence from Indonesia’s Forest Fires of 1997. Econ. Hum. Biol. 2017, 26, 186–198. [CrossRef] [PubMed]

50. Kitson, A. Rising from the Ashes: Affirming the Spirit of Courage, Community Resilience, Compassion and Caring. J. Clin. Nurs. 2020, 29, 2765–2766. [CrossRef] [PubMed]

51. Kollanus, V.; Tiittanen, P.; Niemi, J.V.; Lanki, T. Effects of Long-Range Transported Air Pollution from Vegetation Fires on Daily Mortality and Hospital Admissions in the Helsinki Metropolitan Area, Finland. Environ. Res. 2016, 151, 351–358. [CrossRef] [PubMed]

52. Kreinin, A.; Lyansberg, A.; Yusupov, M.; Moran, D.S. Fire Disaster Readiness: Preparation for the Evacuation of Medical Facilities During Fires in Haifa, Israel. Disaster Med. Public Health Prep. 2019, 13, 375–379. [CrossRef]

53. Kunii, O.; Kanagawa, S.; Yajima, I.; Hisamatsu, Y.; Yamamura, S.; Amagai, T.; Ismail, I.T.S. The 1997 Haze Disaster in Indonesia:

Its Air Quality and Health Effects. Arch. Environ. Health Int. J. 2002, 57, 16–22. [CrossRef]

54. Le, G.E.; Breysse, P.N.; McDermott, A.; Eftim, S.E.; Geyh, A.; Berman, J.D.; Curriero, F.C. Canadian Forest Fires and the Effects of Long-Range Transboundary Air Pollution on Hospitalizations among the Elderly. ISPRS Int. J. Geo. Inf. 2014, 3, 713–731. [CrossRef]

55. Lewis, M.; Christianson, A.; Spinks, M. Return to Flame: Reasons for Burning in Lytton First Nation, British Columbia. J. For. 2018, 116, 143–150. [CrossRef]

56. Linares, C.; Carmona, R.; Tobias, A.; Mirón, I.J.; Diaz, J. Influence of Advections of Particulate Matter from Biomass Combustion on Specific-Cause Mortality in Madrid in the Period 2004–2009. Environ. Sci. Pollut. Res. 2015, 22, 7012–7019. [CrossRef]

57. Liu, J.C.; Pereira, G.; Uhl, S.A.; Bravo, M.A.; Bell, M.L. A systematic Review of the Physical Health Impacts from Non-Occupational Exposure to Wildfire Smoke. Environ. Res. 2015, 136, 120–132. [CrossRef] Int. J. Environ. Res. Public Health 2023, 20, 6252 21 of 22

58. Liu, J.C.; Mickley, L.J.; Sulprizio, M.P.; Yue, X.; Peng, R.D.; Dominici, F.; Bell, M.L. Future Respiratory Hospital Admissions from Wildfire Smoke under Climate Change in the Western US. Environ. Res. Lett. 2016, 11, 124018. [CrossRef]

59. Liu, J.C.; Wilson, A.; Mickley, L.J.; Dominici, F.; Ebisu, K.; Wang, Y.; Sulprizio, M.P.; Peng, R.D.; Yue, X.; Son, J.-Y.; et al. Wildfirespecific Fine Particulate Matter and Risk of Hospital Admissions in Urban and Rural Counties. Epidemiology 2017, 28, 77–85. [CrossRef] [PubMed]

60. Liu, J.C.; Wilson, A.; Mickley, L.J.; Ebisu, K.; Sulprizio, M.P.; Wang, Y.; Peng, R.D.; Yue, X.; Dominici, F.; Bell, M.L. Who Among the Elderly Is Most Vulnerable to Exposure to and Health Risks of Fine Particulate Matter From Wildfire Smoke? Am. J. Epidemiol. 2017, 186, 730–735. [CrossRef] [PubMed]

61. Machado-Silva, F.; Libonati, R.; de Lima, T.F.M.; Peixoto, R.B.; França, J.R.D.A.; Magalhães, M.D.A.F.M.; Santos, F.L.M.; Rodrigues, J.A.; DaCamara, C.C. Drought and Fires Influence the Respiratory Diseases Hospitalizations in the Amazon. Ecol. Indic. 2020, 109, 105817. [CrossRef]

62. Machin, A.B.; Nascimento, L.F.; Mantovani, K.; Machin, E.B. Effects of Exposure to Fine Particulate Matter in Elderly Hospitalizations Due to Respiratory Diseases in the South of the Brazilian Amazon. Braz. J. Med. Biol. Res. 2019, 52, e8130. [CrossRef] [PubMed]

63. Madsen, R.S.; Haynes, H.J.G.; McCaffrey, S.M. Wildfire Risk Reduction in the United States: Leadership Staff Perceptions of Local Fire Department Roles and Responsibilities. Int. J. Disaster Risk Reduct. 2018, 27, 451–458. [CrossRef]

64. Mason, L.; White, G.; Morishima, G.; Alvarado, E.; Andrew, L.; Clark, F.; Durglo, M.; Durglo, J.; Eneas, J.; Erickson, J.; et al.

Listening and Learning from Traditional Knowledge and Western Science: A Dialogue on Contemporary Challenges of Forest Health and Wildfire. J. For. 2012, 110, 187–193. [CrossRef]

65. Mazdiyasni, O.; AghaKouchak, A. Natural Disasters Are Prejudiced Against Disadvantaged and Vulnerable Populations: The Lack of Publicly Available Health-Related Data Hinders Research at the Cusp of the Global Climate Crisis. Geohealth 2020, 4, e2019GH000219. [CrossRef]

66. McDermott-Levy, R.; Fick, D.M. Advancing Gerontological Nursing Science in Climate Change. Res. Gerontol. Nurs. 2020, 13, 6–12. [CrossRef]

67. McKemey, M.; Ens, E.; Rangers, Y.M.; Costello, O.; Reid, N. Indigenous Knowledge and Seasonal Calendar Inform Adaptive Savanna Burning in Northern Australia. Sustainability 2020, 12, 995. [CrossRef]

68. Miller, A.M.; Davidson-Hunt, I. Agency and Resilience: Teachings of Pikangikum First Nation Elders, Northwestern Ontario.

Ecol. Soc. 2013, 18, e2019GH000219. [CrossRef]

69. Miller, A.M.; Davidson-Hunt, I.J.; Peters, P. Talking about Fire: Pikangikum First Nation Elders Guiding Fire Management. Can. J.

For. Res. 2010, 40, 2290–2301. [CrossRef]

70. Miller, A.M.; Davidson-Hunt, I. Fire, Agency and Scale in the Creation of Aboriginal Cultural Landscapes. Hum. Ecol. 2010, 38, 401–414. [CrossRef]

71. Mills, D.; Jones, R.; Wobus, C.; Ekstrom, J.; Jantarasami, L.; Juliana, A.S.; Crimmins, A. Projecting Age-Stratified Risk of Exposure to Inland Flooding and Wildfire Smoke in the United States under Two Climate Scenarios. Environ. Health Perspect. 2018, 126, 47007. [CrossRef]

72. Mirsaeidi, M.; Motahari, H.; Khamesi, M.T.; Sharifi, A.; Campos, M.; Schraufnagel, D.E. Climate Change and Respiratory Infections. Ann. Am. Thorac. Soc. 2016, 13, 1223–1230. [CrossRef]

73. Morgan, G.W.; Tolhurst, K.G.; Poynter, M.W.; Cooper, N.; McGuffog, T.; Ryan, R.; Wouters, M.A.; Stephens, N.; Black, P.; Sheehan, D.; et al. Prescribed burning in south-eastern Australia: History and future directions. Aust. For. 2020, 83, 4–28. [CrossRef]

74. Nunes, K.V.R.; Ignotti, E.; de Souza Hacon, S. Circulatory Disease Mortality Rates in the Elderly and Exposure to PM2.5 Generated by Biomass Burning in the Brazilian Amazon in 2005. Cad. Saúde Pública 2013, 29, 589. [CrossRef] [PubMed]

75. O’Brien, L.; Watson, I. In Conversation with Uncle Lewis: Bushfires, Weather-Makers, Collective Management. Altern. Int. J. Indig.

Peoples 2014, 10, 450–461. [CrossRef]

76. O’Dwyer, L.; Thompson, K. Attachment, Bushfire Preparedness, Planning, and Response among Animal Guardians: A South Australian Case Study. PLoS Curr. 2018, 10. Available online: https://pubmed.ncbi.nlm.nih.gov/30191080/ (accessed on 20 June 2023).

77. Oliveira, S.; Zêzere, J.L.; Queirós, M.; Pereira, J.M. Assessing the Social Context of Wildfire-Affected Areas. The Case of Mainland Portugal. Appl. Geogr. 2017, 88, 104–117.

78. Orr, A.; Migliaccio, C.A.L.; Buford, M.; Ballou, S.; Migliaccio, C.T. Sustained Effects on Lung Function in Community Members Following Exposure to Hazardous PM2.5 Levels from Wildfire Smoke. Toxics 2020, 8, 53. [CrossRef]

79. Pekovic, V.; Seff, L.; Rothman, M. Planning for and Responding to Special Needs of Elders in Natural Disasters. Generations 2007, 31, 37–41.

80. Pennisi, E. Living with heat. Science 2020, 370, 778–781. [CrossRef]

81. Prohaska, T.R.; Peters, K.E. Impact of Natural Disasters on Health Outcomes and Cancer Among Older Adults. Gerontologist 2019, 59 (Suppl. 1), S50–S56. [CrossRef] [PubMed]

82. Ramakreshnan, L.; Aghamohammadi, N.; Fong, C.S.; Bulgiba, A.; Zaki, R.A.; Wong, L.P.; Sulaiman, N.M. Haze and Health Impacts in ASEAN Countries: A Systematic Review. Environ. Sci. Pollut. Res. 2018, 25, 2096–2111. [CrossRef]

83. Ray, L. Using Q-Methodology to Identify Local Perspectives on Wildfires in Two Koyukon Athabascan Communities in Rural Alaska. Sustain. Sci. Pract. Policy 2011, 7, 18–29. [CrossRef] Int. J. Environ. Res. Public Health 2023, 20, 6252 22 of 22

84. Reisen, F.; Brown, S.K.; Reisen, F.; Brown, S.K. Implications for Community Health from Exposure to Bushfire Air Toxics. Environ.

Chem. 2006, 3, 235–243. [CrossRef]

85. Sahani, M.; Zainon, N.A.; Mahiyuddin, W.R.W.; Latif, M.T.; Hod, R.; Khan, M.F.; Tahir, N.M.; Chan, C.-C. A Case-Crossover Analysis of Forest Fire Haze Events and Mortality in Malaysia. Atmos. Environ. 2014, 96, 257–265. [CrossRef]

86. Scrymgeour, G.C.; Smith, L.; Maxwell, H.; Paton, D. Nurses Working in Healthcare Facilities during Natural Disasters: A Qualitative Enquiry. Int. Nurs. Rev. 2020, 67, 427–435. [CrossRef] [PubMed]

87. Shahparvari, S.; Chhetri, P.; Abareshi, A.; Abbasi, B. Multi-Objective Decision Analytics for Short-Notice Bushfire Evacuation: An Australian Case Study. Australas. J. Inf. Syst. 2015, 19, 133–151. [CrossRef]

88. Shankar, H.M.; Rice, M.B. Update on Climate Change: Its Impact on Respiratory Health at Work, Home, and at Play. Clin. Chest Med. 2020, 41, 753–761. [CrossRef]

89. Sherwen, A.; Paratz, E.; La Gerche, A.; Prior, D.; MacIsaac, A. Smoke but No Fire: Increased Myocardial Infarction with Normal Coronary Arteries in the Australian Bushfires? Int. Med. J. 2020, 50, 1439–1440. [CrossRef] [PubMed]

90. Stowell, J.D.; Geng, G.; Saikawa, E.; Chang, H.H.; Fu, J.; Yang, C.-E.; Zhu, Q.; Liu, Y.; Strickland, M.J. Associations of Wildfire Smoke PM2.5 Exposure with Cardiorespiratory Events in Colorado 2011–2014. Environ. Int. 2019, 133, 105151. [CrossRef] [PubMed]

91. Thompson, K.; Trigg, J.; Smith, B. Animal Ownership Among Vulnerable Populations in Regional South Australia: Implications for Natural Disaster Preparedness and Resilience. J. Public Health Manag. Pract. 2017, 23, 59–63. [CrossRef] [PubMed]

92. Tilney, P. Extensive Burns in an Elderly Man Found in a Brush Fire. Air Med. J. 2010, 29, 98–103. [CrossRef] [PubMed]

93. Uriarte, M.; Yackulic, C.B.; Cooper, T.; Flynn, D.; Cortes, M.; Crk, T.; Cullman, G.; McGinty, M.; Sircely, J. Expansion of Sugarcane Production in São Paulo, Brazil: Implications for Fire Occurrence and Respiratory Health. Agric. Ecosyst. Environ. 2009, 132, 48–56. [CrossRef]

94. Vaarzon-Morel, P.; Gabrys, K. Fire on the Horizon: Contemporary Aboriginal burning Issues in the Tanami Desert, Central Australia. GeoJournal 2008, 74, 465. [CrossRef]

95. Weichenthal, S.; Kulka, R.; Lavigne, E.; van Rijswijk, D.; Brauer, M.; Villeneuve, P.J.; Stieb, D.; Joseph, L.; Burnett, R.T. Biomass Burning as a Source of Ambient Fine Particulate Air Pollution and Acute Myocardial Infarction. Epidemiology 2017, 28, 329–337. [CrossRef]

96. Westcott, R.A.N.; Ronan, K.; Bambrick, H.; Taylor, M. “Don’t Just Do Something . . . Stand There!” Emergency Responders’ Peri-Incident Perceptions of Animal Owners in Bushfire. Front. Vet. Sci. 2017, 4, 34. [CrossRef]

97. Wyte-Lake, T.; Claver, M.; Johnson-Koenke, R.; Dobalian, A. Home-Based Primary Care’s Role in Supporting the Older Old During Wildfires. J. Prim. Care Community Health 2019, 10, 2150132719846773. [CrossRef] [PubMed]

98. Yibarbuk, D.; Whitehead, P.J.; Russell-Smith, J.; Jackson, D.; Godjuwa, C.; Fisher, A.; Cooke, P.; Choquenot, D.; Bowman, D.M.J.S. Fire Ecology and Aboriginal Land Management in Central Arnhem Land, Northern Australia: A Tradition of Ecosystem Management. J. Biogeogr. 2001, 28, 325–343. [CrossRef]

99. Youssouf, H.; Liousse, C.; Roblou, L.; Assamoi, E.-M.; Salonen, R.O.; Maesano, C.; Banerjee, S.; Annesi-Maesano, I. Non-Accidental Health Impacts of Wildfire Smoke. Int. J. Environ. Res. Public Health 2014, 11, 11772–11804. [CrossRef] [PubMed] 100. Urbanski, C. Rematriate the Land: How the Indigenous Women of the Sogorea Te’ Land Trust Help Solve Climate Change.

Available online: https://sogoreate-landtrust.org/2020/05/01/rematriate-the-land/ (accessed on 20 June 2023). 101. Hung, W.-T.; Lu, C.-H.; Shrestha, B.; Lin, H.-C.; Lin, C.-A.; Grogan, D.; Hong, J.; Ahmadov, R.; James, E.; Joseph, E. The Impacts of Transported Wildfire Smoke Aerosols on Surface air Quality in New York State: A Case Study in Summer 2018. Atmos. Environ. 2020, 227, 117415. [CrossRef]
102. Langford, A.O.; Senff, C.J.; Alvarez, R.J.; Brioude, J.; Cooper, O.R.; Holloway, J.S.; Lin, M.Y.; Marchbanks, R.D.; Pierce, R.B.; Sandberg, S.P.; et al. An Overview of the 2013 Las Vegas Ozone Study (LVOS): Impact of Stratospheric Intrusions and Long-Range Transport on Surface air Quality. Atmos. Environ. 2015, 109, 305–322. [CrossRef] 103. Aguilera, R.; Corringham, T.; Gershunov, A.; Benmarhnia, T. Wildfire Smoke Impacts Respiratory Health More than Fine Particles from Other Sources: Observational evidence from Southern California. Nat. Commun. 2021, 12, 1493. [CrossRef] 104. Boyd, E.; Chaffin, B.C.; Dorkenoo, K.; Jackson, G.; Harrington, L.; N’Guetta, A.; Johansson, E.L.; Nordlander, L.; De Rosa, S.P.; Raju, E.; et al. Loss and damage from climate change: A new climate justice agenda. One Earth 2021, 4, 1365–1370. [CrossRef] Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Ecological Applications, 17(5), 2007, pp. 1388–1402
 2007 by the Ecological Society of America

HUMAN INFLUENCE ON CALIFORNIA FIRE REGIMES
ALEXANDRA D. SYPHARD,
1,6 VOLKER C. RADELOFF,1 JON E. KEELEY,2 TODD J. HAWBAKER,1 MURRAY K. CLAYTON,3 SUSAN I. STEWART,
4 AND ROGER B. HAMMER5
1 Department of Forest Ecology and Management, University of Wisconsin, Madison, Wisconsin 53706 USA 2 U.S. Geological Survey, Western Ecological Research Center, Sequoia Field Station, Three Rivers, California 93271-9651 USA, and Department of Ecology and Evolutionary Biology, University of California, Los Angeles, California 90095 USA 3 Departments of Plant Pathology and Statistics, University of Wisconsin, Madison, Wisconsin 53706 USA 4 USDA Forest Service, Northern Research Station, Evanston, Illinois 60201 USA 5 Department of Sociology, Oregon State University, Corvallis, Oregon 97331 USA Abstract. Periodic wildfire maintains the integrity and species composition of many ecosystems, including the mediterranean-climate shrublands of California. However, human activities alter natural fire regimes, which can lead to cascading ecological effects. Increased human ignitions at the wildland–urban interface (WUI) have recently gained attention, but fire activity and risk are typically estimated using only biophysical variables. Our goal was to determine how humans influence fire in California and to examine whether this influence was linear, by relating contemporary (2000) and historic (1960–2000) fire data to both human and biophysical variables. Data for the human variables included fine-resolution maps of the WUI produced using housing density and land cover data. Interface WUI, where development abuts wildland vegetation, was differentiated from intermix WUI, where development intermingles with wildland vegetation. Additional explanatory variables included distance to WUI, population density, road density, vegetation type, and ecoregion. All data were summarized at the county level and analyzed using bivariate and multiple regression methods. We found highly significant relationships between humans and fire on the contemporary landscape, and our models explained fire frequency (R2 ¼ 0.72) better than area burned (R2 ¼ 0.50).

Population density, intermix WUI, and distance to WUI explained the most variability in fire frequency, suggesting that the spatial pattern of development may be an important variable to consider when estimating fire risk. We found nonlinear effects such that fire frequency and area burned were highest at intermediate levels of human activity, but declined beyond certain thresholds. Human activities also explained change in fire frequency and area burned (1960– 2000), but our models had greater explanatory power during the years 1960–1980, when there was more dramatic change in fire frequency. Understanding wildfire as a function of the spatial arrangement of ignitions and fuels on the landscape, in addition to nonlinear relationships, will be important to fire managers and conservation planners because fire risk may be related to specific levels of housing density that can be accounted for in land use planning. With more fires occurring in close proximity to human infrastructure, there may also be devastating ecological impacts if development continues to grow farther into wildland vegetation.

Key words: California, USA; fire; fire history; housing density; nonlinear effects; regression; wildland– urban interface.

INTRODUCTION
Fire is a natural process in many biomes and has
played an important role shaping the ecology and
evolution of species (Pyne et al. 1996, Bond and Keeley 2005). Periodic wildfire maintains the integrity and
species composition of many ecosystems, particularly
those in which taxa have developed strategic adaptations to fire (Pyne et al. 1996, Savage et al. 2000, Pausas et al. 2004). Despite the important ecosystem role played by
fire, human activities have altered natural fire regimes relative to their historic range of variability. To develop effective conservation and fire management strategies to deal with altered fire regimes, it is necessary to
understand the causes underlying altered fire behavior
and their human relationships (DellaSalla et al. 2004).

Nowhere is this more critical in the United States than in California, which is the most populous state in the
nation, with roughly 35 3 106 people. Most of the
population lives in lower elevations dominated by
hazardous chaparral shrublands susceptible to frequent
high-intensity crown fires.

In California, as elsewhere, the two primary mechanisms altering fire regimes are fire suppression, resulting in fire exclusion, and increased anthropogenic ignitions, resulting in abnormally high fire frequencies (Keeley and Manuscript received 3 July 2006; revised 7 December 2006; accepted 4 January 2007. Corresponding Editor (ad hoc): K. A.

Hibbard. 6 E-mail: [email protected]
1388
Fotheringham 2003), though climate change, vegetation
manipulation, and other indirect factors may also play a role (Lenihan et al. 2003, Sturtevant et al. 2004). For most of the 20th century, fire suppression effectively
excluded fire from many western U.S. forest ecosystems, such as ponderosa pine. In these ecosystems, fire
exclusion contributed to unnatural fuel accumulation
and increased tree density (Veblen et al. 2000, Allen
et al. 2002, Gray et al. 2005). Recently, when wildfires have hit many of these forests, hazardous fuel loads have contributed to high-intensity crown fires that are
considered outside the historical range of variability
(Stephens 1998). While these patterns are widely
applicable to many forested landscapes in the western
United States, California chaparral shrublands have
experienced such substantial human population growth
and urban expansion that the increase in ignitions,
coupled with the most severe fire weather in the country (Schroeder et al. 1964), have acted to offset the effects of suppression to the point that fire frequency exceeds the historic range of variability (Keeley et al. 1999). Because anthropogenic ignitions tend to be concentrated near
human infrastructure, more fires now occur at the urban fringe than in the backcountry (Pyne 2001, Keeley et al. 2004). Profound impacts on land cover condition and
community dynamics are possible if a disturbance
regime exceeds its natural range of variability, and
altered fire regimes can lead to cascading ecological
effects (Landres et al. 1999, Dale et al. 2000). For
example, too-frequent fire can result in habitat loss and fragmentation, shifting forest composition, reduction of small-mammal populations, and accompanying loss of
predator species (Barro and Conard 1991, DellaSalla
et al. 2004).

Landscape-level interactions between human activities
and natural dynamics tend to be spatially concentrated
at the wildland–urban interface (WUI; see Plate 1),
which is the contact zone in which human development
intermingles with undeveloped vegetation (Radeloff
et al. 2005). The WUI has received national attention
because housing developments and human lives are
vulnerable to fire in these locations and because
anthropogenic ignitions are believed to be most common
there (Rundel and King 2001, USDA and USDI 2001).

The majority of WUI fire research has focused on
strategies to protect lives and structures (e.g., Cohen 2000, Winter and Fried 2000, Winter et al. 2002,
Shindler and Toman 2003) or on the assessment of fire
risk using biophysical or climate variables that influence fire behavior (Bradstock et al. 1998, Fried et al. 1999, Haight et al. 2004). However, it is also important to
understand how the WUI itself (or other indicators of
human activity) affects fire and to quantify the spatial relationships between human activities and fire (Duncan and Schmalzer 2004).

The influence of proximity to the WUI and other
human infrastructure appears to vary markedly with
region. In the northern Great Lakes states, areas with
higher population density, higher road density, and
lower distance to nonforest were positively correlated
with fire (Cardille et al. 2001). Also, in southern
California, a strong positive correlation between population density and fire frequency was reported (Keeley et al. 1999). However, no relationship between housing
count and fire was found in northern Florida counties
(Prestemon et al. 2002); population density and unemployment were positively related, and housing density and unemployment were negatively related to fire in a
different analysis of Florida counties (Mercer and
Prestemon 2005). A negative relationship between
housing density and fire was also found in the Sierra
Nevada Mountains of California (CAFRAP 2001).

In addition to potential regional differences, it is also difficult to draw general conclusions from these studies because they used different indicators of human
activities, their data sets differed in spatial and temporal scale, and they were conducted in small areas where
ranges of variability in both fire frequency and level of development were limited. Human–fire relationships
may also vary based on factors that were not accounted
for, such as pattern of development. Another explanation for the discrepancy is that relationships between human activities and fire may be nonlinear in that
humans may affect fire occurrence positively or negatively, depending on the level of influence. These nonlinear effects were apparent in data from a recent
study in the San Francisco Bay region, where population growth was positively related to fire frequency over time up to a point, but then fire frequency leveled off as
population continued to increase (Keeley 2005).

Whether positive or negative, the significance of the
relationships between human activities and fire that were detected in previous studies stresses the importance of further exploring links between anthropogenic and
environmental factors and their relative influence on
wildfire patterns across space and time. Therefore, our research objective was to quantify relationships between human activities and fire in California counties using
temporally and spatially rich data sets and regression
models. Although fire regimes encompass multiple
characteristics, including seasonality, intensity, severity, and predictability, we restricted our analysis to questions about fire frequency and area burned to determine:
(1) what the contemporary relationship between human
activities and fire is; (2) how human activities have
influenced change in fire over the last 40 years; and (3) whether fire frequency and area burned vary nonlinearly in response to human influence.

Humans are responsible for igniting the fires that burn the majority of area in California (Keeley 1982);
therefore, we expected our anthropogenic explanatory
variables to significantly explain fire activity on the current landscape and over time. In addition to
population density (which simply quantifies the number
of people in an area), we expected the spatial pattern of human development (indicated by housing density and
July 2007 HUMAN INFLUENCE ON CALIFORNIA FIRE 1389
land cover combinations and distance variables) to be an important influence on fire because we assumed that
anthropogenic ignitions are most likely to occur where
human presence is greatest. We also expected that the
relationships between human activities and fire would be both positive and negative because humans ignite fires, but development patterns affect fuel continuity and the accessibility of fire suppression resources. Finally, we included several environmental variables in the analysis because we expected the human relationships to be
mediated by these other biophysical variables that shape the pattern and frequency of fire (Wells et al. 2004).

METHODS
Study area
California is the second largest state in the continental United States and is the most populous and physically
diverse. Most of the state has a mediterranean climate, which, along with a heterogeneous landscape, contributes to tremendous biodiversity (Wilson 1992). Because the state contains a large proportion of the country’s
endangered species, it is considered a ‘‘hotspot’’ of
threatened biodiversity (Dobson et al. 1997). There is
extensive spatial variation in human population density:
large areas in the north are among the most sparsely
populated in the country, but metropolitan regions in the south are growing at unprecedented rates (Landis and
Reilly 2004). Much of the landscape is highly fire-prone, but fire regimes vary, and fire management is divided
among many institutions. Humans have altered California’s fire regimes, and its fire-related financial losses are among the highest in the country (Halsey 2005).

Data
Dependent variables: fire statistics.—We assembled
our fire statistics from the California Department of
Forestry and Fire Protection (CDF; Sacramento,
California, USA) annual printed records, which included information on all fires for which the CDF took action between 1931 and 2004. For all state responsibility areas (SRA; Fig. 1), fire statistics are recorded by county and include numbers by size class, total area
burned, vegetation type, and cause. Because the
statistics did not include spatially explicit information on individual fires, we weighted the data by the area
within the SRA in each county by calculating proportions to use as our dependent variables. These fire statistics were substantially more comprehensive than
the readily available electronic Statewide Fire History Database, which excludes most fires ,40 ha, which in
many counties represents .90% of the fires. Although
both anthropogenic and lightning ignitions would be
important to consider for fully understanding fire
patterns in other regions (e.g., Marsden 1982), humans
were responsible for ;95% of both the number of fires
and area burned in California in the last century. We
restricted our analysis to these anthropogenic fires
because our focus was on human relationships with fire.

Although the fire statistics were not spatially explicit, we developed GIS grids at 100-m resolution to derive data
for all of the explanatory variables. The data for these explanatory variables were only extracted and averaged
from within the SRA boundaries corresponding to the
fire data.

Out of the 58 counties in California, we had fire
statistics for 54 of them for the year 2000. Therefore, to assess the contemporary relationship between fire and
human activities (hereafter referred to as the ‘‘contemporary analysis’’), we analyzed the data from these counties using the annual number of fires and area
burned as our dependent variables (Table 1).

Based on a preliminary exploration of the fire history
data (averaged across all counties), we observed two
distinct trends during the last 50 years. First, the number of fires substantially increased until 1980 and then
decreased until 2000; and second, the average area
burned changed inversely to the number of fires, but the differences over time were less dramatic and not
statistically significant (Fig. 2). Considering these trends, we broke the historic analysis into two equal time
periods (1960–1980 and 1980–2000) to compare the
relative influence of the explanatory variables on both the increase (i.e., from 1960 to 1980) and decrease (from 1980 to 2000) in fire activity. The year 1980 is used to compute differences for both time periods because the
census data that formed the basis for many of our
explanatory variables were only available by decade. We averaged the number of fires and the area burned for 10- FIG. 1. Map of California Department of Forestry and Fire Protection (CDF) state responsibility areas (SRAs) within county boundaries of California, USA.
1390 ALEXANDRA D. SYPHARD ET AL. Ecological Applications Vol. 17, No. 5
year time periods that bracketed the dates of the census data (e.g., 1955–1964 [1960], 1975–1984 [1980], 1995–
2004 [2000]) and then calculated the difference in
averages from the 1960–1980 and 1980–2000 periods
for our dependent variables (Table 1). By averaging the fire data, we smoothed some of the annual variability
that may have occurred due to stochastic factors such as weather.

Explanatory variables: housing data.—Data for most
of the anthropogenic variables were available through a nationwide mapping project that produced maps of the
WUI in the conterminous United States using housing
density data from the 1990 and 2000 U.S. Census (U.S.

Census Bureau 2002) and land cover data from the
USGS National Land Cover Dataset (Radeloff et al.
2005). The maps were produced at the finest demographic spatial scale possible, the 2000 decennial census blocks. The vegetation data were produced at 30-m
resolution. These maps delineated two types of WUI in
accordance with the Federal Register definition (USDA
and USDI 2001). ‘‘Intermix WUI’’ is defined as the
intermingling of development with wildland vegetation;
the vegetation is continuous and occupies .50% of the
area. ‘‘Interface WUI’’ is defined as the situation in
which development abuts wildland vegetation; there is
,50% vegetation in the WUI, but it is within 2.4 km of
an area that has .75% vegetation. In both types of WUI
communities, housing must meet or exceed a density of
more than one structure per 16 ha (6.17 housing
units/km2
). Interface WUI tends to occur in buffers
surrounding higher-density housing, whereas intermix
WUI is more dispersed across the landscape (Fig.
3A, B).

The WUI data were only produced for 1990 and 2000
due to the lack of historic land cover data, but housing density data were available from 1960 to 2000. Historic housing density distribution was estimated using backcasting methods to allocate historic county-level housing unit counts into partial block groups (as described in
Hammer et al. 2004). We used both intermix and
interface WUI as explanatory variables (proportions
within the county SRAs) in the current analysis to
evaluate how these different patterns of vegetation and housing density affected fire activity. We also used lowdensity housing (housing density 6.17 housing units/km2 and ,49.42 housing units/km2
) to determine
whether it could act as a substitute for WUI as an
explanatory variable in the historic analysis (Table 1).

TABLE 1. Variables analyzed in the regression models.

Variable Source Processing
2000 data
Dependent variables
Number of fires CDF proportion in SRA, square-root transformed Area burned CDF proportion in SRA, square-root transformed Explanatory variables
Human
Intermix WUI SILVIS proportion in SRA
Interface WUI SILVIS proportion in SRA
Low-density housing SILVIS proportion in SRA
Distance to intermix WUI SILVIS mean Euclidean distance in SRA Distance to interface WUI SILVIS mean Euclidean distance in SRA Population density SILVIS proportion in SRA
Road density TIGER mean km/km2 in SRA
Distance to road TIGER mean Euclidean distance in SRA
Biophysical
Ecoregion CDF discrete class
Vegetation type CDF area burned in vegetation type/area burned in SRA Historic data, 1960–1980 and 1980–2000
Dependent variables
Change in number of fires CDF difference between decadal averages, proportion in SRA, square-root transformed
Change in area burned CDF difference between decadal averages, proportion in SRA, square-root transformed
Explanatory variables
Human
Change in housing density SILVIS difference between decades Change in distance to low-density housing SILVIS difference between mean Euclidean distance in SRA Initial housing density SILVIS mean housing density in either 1960 or 1980 Initial distance to low-density housing SILVIS mean Euclidean distance in SRA in either 1960 or 1980 Biophysical
Ecoregion CDF discrete class
Vegetation type CDF mean area burned in vegetation type/area burned in SRA over time period
Notes: Key to abbreviations: WUI, wildland–urban interface; SRA, state responsibility area. Sources are as follows: CDF, California Department of Forestry and Fire Protection, Sacramento, California, USA, unpublished data; SILVIS, Radeloff et al. (2005); TIGER, U.S. Census Bureau (2000).

July 2007 HUMAN INFLUENCE ON CALIFORNIA FIRE 1391
Looking at an overlay of fire perimeters from the
electronic Statewide Fire History Database (from the
last 25 years; available online)
7 on the WUI data, it was
apparent that many fires occurred close to the WUI, but not necessarily within the WUI (Fig. 3C, D). Therefore, we calculated the mean distance to intermix and
interface WUI to evaluate as explanatory variables
(Table 1). These means were calculated by iteratively
determining the Euclidean distances from every grid cell in the county SRA boundaries and then averaging the
distances across all cells to determine means for the
counties. We also included population density data from the 2000 Census.

For the historic analysis, we calculated changes in
mean housing density and mean distance to low-density
housing between the 1960–1980 and 1980–2000 periods
to relate to change in the dependent variables. We
excluded the proportion of low-density housing from
our analysis because it was highly correlated with mean housing density (r ¼ 0.84). Unlike the historical fire data that switched in their direction of change over time,
housing density continued to increase while the mean
distance to low-density housing continued to decline
(Fig. 4). We included the initial values of these data (e.g., 1960 and 1980) to account for the fact that the same
magnitude of change may have different effects on the
dependent variables depending on the starting value of
the explanatory variables (Table 1).

Explanatory variables: road data.—The quality of
road data can vary according to data source (Hawbaker
and Radeloff 2004), so we compared the U.S. Geological Survey digital line graph (DLG; U.S. Geological Survey 2002) and the US TIGER 2000 GIS (U.S.

Census Bureau 2000) layers of roads to determine
whether there were substantial differences that could
affect the interpretation of the results. After calculating and summarizing road density by county, we found a
strong positive correlation (r ¼ 0.97). Therefore, we used the TIGER data because they were produced in 2000,
the same year as the contemporary analysis. The more
current TIGER data generally capture new development
that might not be included in the DLG data. We
evaluated mean road density and mean distance to roads
in the current analysis (Table 1), but road data were
unavailable for the historic analysis.

Explanatory variables: environmental.—In the absence
of human influence, fire behavior is primarily a function of biophysical variables (Pyne et al. 1996, Rollins et al. 2002). These can vary widely across a county, but
ecoregions capture broad differences by stratifying
landscapes into unique combinations of physical and
biological variables (ECOMAP 1993). Our ecoregion
data were the geographic subdivisions of California
defined for The Jepson Manual (Hickman 1993),
designated through broadly defined vegetation types
and geologic, topographic, and climatic variation
(Fig. 5).

Because vegetation type influences the ignitability of
fuel and the rate of fire spread (Bond and van Wilgen
1996, Pyne et al. 1996), we also evaluated the proportion of area burned within three broad vegetation types:
shrubland, grassland, and woodland (Fig. 5). Differences in fire regimes between broadly defined vegetation types can be striking, particularly between shrubland
and woodland in southern California (Wells et al. 2004).

The CDF fire statistics included information on the
proportion of area burned in these vegetation types. For the historic analysis, we averaged the proportion of fires burned within different vegetation types over the entire decade (Table 1).

Analytical methods
Diagnostics and data exploration.—Before developing
regression models, we examined scatter plots for each
variable. Nonlinear trends were apparent (e.g., Fig. 6), suggesting that we needed to include quadratic terms for the explanatory variables in the regressions. Unequal
variances in the residual plots prompted us to apply a
square-root transformation to the dependent variables.

We also plotted semivariograms of the models’ residuals (using centroids from the SRA boundaries) and found
no evidence of spatial autocorrelation. To check for
FIG. 2. Trends in number of fires and area burned for all land in the state responsibility areas (SRAs) in California from 1960 to 2000.
7 hhttp://frap.cdf.ca.gov/data/frapgisdata/select.aspi
1392 ALEXANDRA D. SYPHARD ET AL. Ecological Applications Vol. 17, No. 5
multicollinearity, we calculated the correlation coefficients between all of the explanatory variables and only included noncorrelated variables (r  0.7) in the multiple regression models.

The areas of CDF jurisdiction for each county varied
slightly over time. Therefore, we compared separate
regressions from the full historic data set (n ¼ 37) to a subset of the data excluding counties that experienced a greater than 20% change in area over time (n ¼ 23). For both the 1960–1980 regressions and the 1980–2000
regressions, every one of the explanatory variables that was significant in the subset was also significant in the full data set, with very similar R2 values; therefore, we felt confident proceeding with the full data set for the historic analysis because we had greater power with the larger sample size.

FIG. 3. The wildland–urban interface (WUI) in 2000 with and without fire perimeter overlays (from 1979 to 2004) in (A, C) California and (B, D) southern California. Housing density is defined as follows: very low, .0–6.17 housing units/km2 ; low, 6.17–

49.42 housing units/km2
; medium, 49.42–741.31 housing units/km2; and high, .741.31 housing units/km2 (USDA and USDI 2001). ‘‘Fires 25y’’ refers to 25 years of fire perimeters, from 1980 to 2005.

July 2007 HUMAN INFLUENCE ON CALIFORNIA FIRE 1393
Statistical analysis
We used the same regression modeling approach for
both the current and historic analyses. First, we
developed bivariate regression models for all of the
explanatory variables and their quadratic terms so that we could evaluate their independent influence on fire
frequency and area burned. To account for the
interactions between variables (and their quadratic
terms), we also built multiple regression models using
the R statistical package (R Development Core Team
2005). For all models, we first conducted a full stepwise selection analysis (both directions) using Akaike Information Criteria to identify the best combination of predictor variables (Burnham and Anderson 2002).

Some of the models retained a quadratic term without
including the lower-order variable. In these models, we added the lower-order term, rebuilt the model, and then proceeded with a backwards elimination process until all predictor variables in the model were significant with P values  0.05.

RESULTS
Current analysis
Bivariate regressions.—Many of the anthropogenic
variables were highly significant in explaining the
number of fires in 2000. The quadratic term for each
of these variables was also significant, and the direction of influence was both positive and negative (Fig. 7).

Compared to the other variables, population density
explained the greatest amount of variability. The
proportion of intermix WUI and low-density housing
in the counties also explained significant variation in the number of fires; but the proportion of interface WUI
was insignificant. The number of fires was significantly related to the mean distance to both types of WUI, but
neither of the road variables was significant. All three vegetation types, particularly shrubland, significantly influenced the number of fires, but ecoregion was
insignificant.

For the anthropogenic variables, the number of fires
was highest at intermediate levels of population density (from ;35 to 45 people/km2
; Fig. 6), proportion of
intermix WUI (;20–30% in the county), and proportion
of low-density housing (;25–35% in the county). It was
also highest at the shortest distances to intermix and
interface WUI, but started to level off at ;9–10 km for intermix (Fig. 6) and 14–15 km for interface WUI.

Unlike the number of fires, none of the anthropogenic
variables were significantly associated with the area
burned in 2000. In fact, shrubland was the only variable that explained significant variation in area burned.

Multiple regression.—When all of the variables were
modeled in the multiple regressions, the resulting model for number of fires in 2000 included population density, the proportion of intermix WUI and its quadratic term,
grassland and its quadratic term, and shrubland
(Table 2). The model was highly significant with an
adjusted R2 value of 0.72.

The multiple regression model for area burned in 2000
included distance to road, shrubland, and woodland,
and all three variables had significant positive relationships (no quadratic terms were retained). This model was also highly significant with an adjusted R2 of 0.50.

Historical analysis 1960–1980
Bivariate regressions.—Change in the number of fires
(net increase) from 1960 to 1980 was significantly
explained by each of the human-related variables except for change in the mean distance to low-density housing
(Fig. 8). The quadratic term was also significant in the separate models, except for the initial distance to lowdensity housing (in 1960), which had a negative influence on the change in number of fires. Change in number of
fires was also significantly related to ecoregion and
shrubland vegetation.

The only three variables with significant influence on
the change in area burned (net decrease) were the three vegetation types.

Multiple regression.—The explanatory variables that
were retained in the multiple regression model for
change in the number of fires from 1960 to 1980
included mean housing density in 1960 and its quadratic term, grassland vegetation, and ecoregion (Table 2). The adjusted R2 value was highly significant at 0.72.

FIG. 4. Trends in housing density and distance to lowdensity housing (6.17–49.42 housing units/km2 ) for all land in
the state responsibility areas (SRAs) in California from 1960 to 2000.
1394 ALEXANDRA D. SYPHARD ET AL. Ecological Applications Vol. 17, No. 5
Mean housing density in 1960 was positively associated with change in area burned from 1960 to 1980, and the distance to low-density housing had first a positive, then a negative influence because the quadratic term was included. Other variables retained in the multiple
regression model included shrubland and its quadratic
term, grassland, woodland, and ecoregion.

Historical analysis 1980–2000
Bivariate regressions.—Initial housing density (in
1980) was the only significant explanatory variable
explaining change in number of fires (net decrease) from 1980 to 2000 (Fig. 9). Woodland vegetation was the only significant variable out of the separate models explaining change in area burned from 1980 to 2000 (net increase). The quadratic terms were significant for both of these models.

Multiple regression.—The multiple regression model
explaining change in number of fires from 1980 to 2000
included change in housing density, initial housing
density (in 1980), and woodland vegetation; the quadratic term was also significant for these three variables (Table 2). Although the model was significant, the R2 was substantially lower than the 1960–1980 model, at 0.26.

The multiple regression model explaining change in
area burned included initial housing density (in 1980)
and its quadratic term, initial distance to low-density FIG. 5. Maps showing ecoregion boundaries and the proportion of area burned in shrubland, grassland, and woodland in 2000.

FIG. 6. The relationships between (A) the proportion of the number of fires and population density and (B) the proportion of the number of fires and mean distance to intermix wildland– urban interface (WUI).

July 2007 HUMAN INFLUENCE ON CALIFORNIA FIRE 1395
housing, woodland vegetation and its quadratic, and
ecoregion. This model had better explanatory power
than the number of fires model, with an R2 of 0.41.

DISCUSSION
The expression of fire on a landscape is influenced by
a combination of factors that vary across spatial and
temporal scales and involve both physical and biological characteristics. Fire behavior has long been viewed as a largely physical phenomenon illustrated by the
classic fire environment triangle that places fire as a function of weather, fuels, and topography (Countryman 1972), but clearly the human influence on modern fire regimes must also be understood to meet fire
management needs (DellaSalla et al. 2004). We first
asked what the current relationship is between human
activities and fire in California and found that humans and their spatial distribution explained a tremendous
proportion of the variability in the number of fires, but that area burned was more a function of vegetation
type. Anthropogenic ignitions are the primary cause of
fire in California and were the focus of our analysis, so we were not surprised by the strong human influence.

Nevertheless, the high explanatory power of the models
underscores the importance of using locally relevant
anthropogenic factors as well as biophysical factors in fire risk assessments and mapping. The models also
identify which indicators of human activity are most
strongly associated with fire in California. For number of fires, the proportion of intermix WUI explained more variation than any other variable except for population density, suggesting that the spatial pattern of housing development and fuel are important risk factors for fire starts.

Human-caused ignitions frequently occur along transportation corridors (Keeley and Fotheringham 2003, Stephens 2005), so it was surprising that neither road
density nor average distance to road were significant in explaining fire frequency. Although roads are important in local-scale ignition modeling, detecting their influence on fire ignitions may be difficult at an aggregated, county level since they are narrow, linear features. On the other hand, distance to roads was the only anthropogenic
variable associated with area burned, having a positive influence when grassland and shrubland were also
accounted for in the multiple regression model, which
may reflect the difficulty of fire suppression access
contributing to fire size.

Humans influence fire frequency more than area
burned because anthropogenic ignitions are responsible
FIG. 7. R2 values and significance levels for the explanatory variables in the bivariate regression models for number of fires and area burned in 2000.
* P , 0.05; ** P , 0.01; *** P , 0.001.
1396 ALEXANDRA D. SYPHARD ET AL. Ecological Applications Vol. 17, No. 5
for fire initiation, but fire spread and behavior is
ultimately more a function of fuel availability and type (Bond and van Wilgen 1996, Pyne et al. 1996). Yet
humans do have some control over fire size through
suppression and, indirectly, through fuel connectivity
(Sturtevant et al. 2004), although fires are extremely
difficult to suppress in California shrublands under
high-wind conditions that typify the most destructive
fires (Keeley and Fotheringham 2003). Therefore,
human effects on area burned may cancel one another
out to some extent because fire suppression can
minimize the increase in area burned that would result
from increased ignitions, at least at the WUI. Fire
suppression resources are more likely to be concentrated on structural protection in developed areas (Calkin et al. 2005), which would explain the positive relationship
between area burned and distance to road. Roads can
serve as firebreaks and can also provide access routes for firefighters.

The inclusion of vegetation type in the multiple
regression models illustrates that, despite the strong
influence of humans, fire occurrence remains a function TABLE 2. Variables retained in the multiple regression models for the current and historic analyses.

Analysis and
explanatory variable
Coefficient
and intercept P
Current
2000
No. fires
Population density 0.0006 ,0.01
Proportion intermix 0.0702 ,0.01
(Proportion intermix)2 0.2629 ,0.01
Grassland 0.0496 ,0.01
(Grassland)2 0.0441 ,0.01
Shrubland 0.0093 0.02
Overall model (adjusted R2
: 0.72) 0.0001 ,0.01
Area burned
Distance to road 0.00004 ,0.01
Shrubland 0.0833 ,0.01
Woodland 0.0559 ,0.01
Overall model (adjusted R2
: 0.50) 0.0052 ,0.01
Historic
1960–1980
No. fires
Initial housing 2.7649 ,0.01
(Initial housing)2 0.1523 ,0.01
Grassland 4.6311 0.05
Ecoregion ... ,0.01
Overall model (adjusted R2
: 0.72) 0.6443 ,0.01
Area burned
Initial housing 0.0188 ,0.01
Initial distance 0.00002 ,0.01
(Initial distance)2 2 3 1010 ,0.01
Shrubland 0.3641 0.12
(Shrubland)2 0.8778 0.01
Grassland 0.0371 ,0.01
Woodland 0.0449 0.01
Ecoregion ... 0.03
Overall model (adjusted R2
: 0.51) 0.373 ,0.01
1980–2000
No. fires
Change housing 3.0666 0.01
(Change housing)2 0.2661 0.01
Initial housing 1.8269 0.01
(Initial housing)2 0.0505 0.03
Woodland 38.1957 0.03
(Woodland)2 107.0112 0.02
Overall model (adjusted R2
: 0.26) 1.894 0.01
Area burned
Initial housing 0.0114 0.01
(Initial housing)2 0.0003 0.05
Initial distance 0.000003 ,0.01
Woodland 0.0292 0.18
(Woodland)2 1.2831 0.02
Ecoregion ... 0.05
Overall model (adjusted R2
: 0.41) 0.0409 ,0.01
 Coefficients are not listed for categorical variables.

July 2007 HUMAN INFLUENCE ON CALIFORNIA FIRE 1397
FIG. 8. R2 values and significance levels for the explanatory variables in the bivariate regression models for number of fires and area burned from 1960 to 1980.
* P , 0.05; ** P , 0.01; *** P , 0.001.

FIG. 9. R2 values and significance levels for the explanatory variables in the bivariate regression models for number of fires and area burned from 1980 to 2000.
* P , 0.05; ** P , 0.01; *** P , 0.001.
1398 ALEXANDRA D. SYPHARD ET AL. Ecological Applications Vol. 17, No. 5
of multiple interacting social and environmental variables. For number of fires and area burned, shrubland had the strongest explanatory power of the vegetation
types. Chaparral and coastal sage scrub are both
extremely fire-prone vegetation types and high human
population density tends to be distributed in these types; other studies have shown that they have experienced a
higher rate of burning than other vegetation types in the southern part of the state in the last century (Keeley
et al. 1999, Keeley 2000, Wells et al. 2004). Increased ignitions in highly flammable vegetation types can lead to very hazardous conditions (Halsey 2005).

The second question we asked was ‘‘How do human
activities relate to change in fire?’’ In the last 40 years, the most substantial change was the increase in number
of fires from 1960 to 1980. The decrease in number of
fires was less dramatic between 1980 and 2000; and the
change in area burned was relatively small in both time periods. Housing development patterns were most
influential when change was greatest, from 1960 to
1980, and for trends in fire frequency (vs. area burned).

Although anthropogenic influence was partially responsible for the change in area burned, the apparent inverse relationship between change in fire frequency
and change in area burned may be spurious. In other
words, the explanation for a decrease in number of fires may be independent of the concurrent increase in area
burned. Trends in area burned are naturally cyclic due
to broad-scale factors such as climate. Recent research has shown that change in climate was a major factor
driving fire activity in the western United States in the last several decades (Westerling et al. 2006); however, that research was restricted to large montane fire events on federally owned land above 1370 m. Therefore, while
climate change may have played some role in our
observed change in area burned, we cannot extend those
results to our analysis because we included fires of all sizes under multiple land ownership classes, and
historical fire patterns in the lower elevations do not correspond to patterns in montane forests (Halsey
2005).

Fire both constrains and is constrained by the fuel
patterns it creates, resulting in cycles of fire activity and temporal autocorrelation in area burned, in part because young fuels are often less likely to burn (Malamud et al. 2005). Temporal autocorrelation effects vary with
ecosystem, fuel type, and the area of analysis; but in
all vegetation types, temporal dependence diminishes
over time due to post-fire recovery. Therefore, we
assumed that the effects would be low in our study
because we were looking at change over 20-year time
periods. Furthermore, the chaparral vegetation that
dominates much of California recovers very quickly
following fire, meaning that the effect of temporal
autocorrelation in this vegetation type would last for
only brief periods of time. Also, under extreme weather conditions, young age classes are capable of carrying
fires in the southern portions of California (Moritz 1997, Moritz et al. 2004).

In general, the anthropogenic influence on fire
frequency and extent was complicated through the
combination of positive and negative effects, which
helps to answer our third question: ‘‘Do fire frequency and area burned vary nonlinearly in response to human
influence?’’ Nonlinear effects were evident in the scatter plots and confirmed by the significance of quadratic
terms in most of the models. The regression models
indicate that humans were responsible for first increasing and then decreasing fire frequency and area burned.

These dual influences may explain why prior studies
presented conflicting results, because a positive or
negative response was dependent on the level of human
presence. Aside from the fact that we intentionally tested hypotheses regarding nonlinear relationships, our data
also contained a wide range of human presence due to
the large extent and diversity of the state of California.

The scatter plots illustrate how these human–fire
relationships occurred. For both the number of fires and area burned, and in the current and historic analyses, the PLATE 1. (Left) Wildland–urban interface (WUI) and (right) burned-over fuel break, both at the eastern end of Scripps Ranch (San Diego County, California, USA) after the autumn 2003 Cedar Fire (largest fire in California since the beginning of the 20th century). Photo credits: J. E. Keeley.

July 2007 HUMAN INFLUENCE ON CALIFORNIA FIRE 1399
maximum fire values occurred at intermediate levels of
human presence (as in Fig. 6A); and when human
activity was either lower or higher, fire activity was
lower. Initial increase in fire occurrence with increasing population is reasonable since human presence results in more ignitions. However, it appears that when human
population density and development reach a certain
threshold density, ignitions decline, and this is likely the result of diminished and highly fragmented open space
with fuels insufficient to sustain fire. In addition, above a certain population threshold, fire suppression resources are likely to be more concentrated in the WUI. Inverse
relationships were evident in the scatter plots of distance (Fig. 6B). In these, fire frequency and area burned were greatest at short distances to WUI; and at longer
distances, the trend lines leveled off. These distance
relationships indicate that more fires would be expected in close proximity to settled areas where ignitions are likely to occur.

The inclusion of quadratic terms in the multiple
regression models supports the concept that fire
frequency and area burned were dependent on the level
of human activity. Initial housing density was important in all four historic multiple regression models, and initial distance to low-density housing was important in both
of the historic area-burned models. The change in
number of fires for both periods was also related to
change in housing density, in bivariate regression models for the earlier period and in the multiple regression
model for the later period (1980–2000). These results
further emphasize that fire activity was a function of a certain level of human presence. In addition to the
strong influence of human presence, ecoregion and
vegetation types were also highly significant in the
multiple regression models, suggesting that the particular level of human activity that was most influential in explaining fire activity was dependent upon biophysical context.

The primary value of the multiple regression models
was to identify the most influential variables and their direction of influence when accounting for other factors.

While they explained how fire activity varied according to context-dependent interactions, their purpose was not to provide a formula for determining fire risk at a
landscape scale. Environmental and social conditions
differ from region to region, and processes such as fire and succession are controlled by a hierarchy of factors, with different variables important at different scales
(Turner et al. 1997). Nevertheless, these models provide strong evidence about the strength and nature of
human–fire relationships. That these relationships are
significant across a state as diverse as California suggests that human influence is increasingly overriding the
biophysical template; yet, managers must account for
the interactions with ecoregion and vegetation type
when making management decisions. Determining the
conditions (e.g., thresholds) for nonlinear anthropogenic relationships will be important to understand how fire
risk is distributed across the landscape.

At the coarse scale of our analysis, we can estimate
these thresholds based on the nonlinear relationships in our scatter plots (as in Fig. 6) and suggest that fire
frequency is likely to be highest when population density is between 35 and 45 people/km2
, proportion of intermix
WUI is ;20–30%, proportion of low-density housing is
;25–35%, the mean distance to intermix WUI is ,9 km,
and the mean distance to interface WUI is ,14 km. Our
next step is to more precisely define these relationships at scales finer than the county level (where management
decisions often occur) and to understand the conditions under which human activities positively or negatively
influence fire.

These results imply that fire managers must consider
human influence, together with biophysical characteristics such as those represented in the LANDFIRE database, when making decisions regarding the allocation of suppression and hazard mitigation resources. If human presence is not explicitly included in decision
making, inefficiencies may result, because fire occurrence is related to human presence on the landscape. In
particular, we identify an intermediate level of housing density and distance from the WUI at which the effects
of human presence seem to be especially damaging, i.e., a point at which enough people are present to ignite
fires, but development has not yet removed or fragmented the wildland vegetation enough to disrupt fire spread. This intermediate level of development is one
that large areas of the lower 48 states, particularly in the West and Southwest, will achieve in the coming decade.

Hence, the WUI’s location, extent, and dynamics will
continue to be essential information for wildland fire
management.

CONCLUSION
In addition to the risk to human lives and structures,
changing fire regimes may have substantial ecological
impacts, and the results in this analysis support the
hypothesis that humans are altering both the spatial and temporal pattern of the fire regime. Although the overall area burned has not changed substantially, the distribution of fires across the landscape is shifting so that the majority of fires are burning closer to developed areas, and more remote forests are no longer burning at their
historic range of variability (Pyne 2001). In either case, the ecological impacts may be devastating. Due to lack
of dendrochronological information, historic reference
conditions are difficult to determine in stand-replacing chaparral shrublands. Although chaparral is adapted to
periodic wildfire, there is substantial evidence that fires are burning at unprecedented frequencies, and this
repeated burning (at intervals closer than 15–20 years
apart) exceeds many species’ resilience and has already resulted in numerous extirpations (Zedler et al. 1983,
Haidinger and Keeley 1993, Halsey 2005).
1400 ALEXANDRA D. SYPHARD ET AL. Ecological Applications Vol. 17, No. 5
If present trends continue in California, the population may increase to 90 3 106 residents in the next 100 years. Recent trends in housing development patterns
also indicate that growth in area and number of houses
in intermix WUI has far outpaced the growth in
interface WUI (Radeloff et al. 2005; Hammer et al., in
press). Our results showing that fire frequency and area burned tend to be highest at intermediate levels of
development (more typical of intermix than interface)
suggest that fire risk is a function of the spatial
arrangement of housing development and fuels. Therefore, in addition to more people in the region that could ignite fires, future conditions that include continued
growth of intermix WUI may also contribute to greater
fire risk. Land use planning that encourages compact
development has been advocated to lessen the general
impacts of growth on natural resources (Landis and
Reilly 2004), and we suggest that reducing sprawling
development patterns will also be important to the
control of wildfires in California.

ACKNOWLEDGMENTS
We thank Ayn Shlisky, Julie Yee, and an anonymous
reviewer for insightful comments that improved our manuscript. We are also grateful for the support from the U.S. Forest Service Northern Research Station and the Pacific Northwest Research Station.

LITERATURE CITED
Allen, C. D., M. Savage, D. A. Falk, K. F. Suckling, T. W.

Swetnam, T. Schulke, P. B. Stacey, P. Morgan, M. Hoffman, and J. T. Klingel. 2002. Ecological restoration of southwestern ponderosa pine ecosystems: a broad perspective. Ecological Applications 12:1418–1433.

Barro, S. C., and S. G. Conard. 1991. Fire effects on California chaparral systems: an overview. Environment International 17:135–149.

Bond, W. J., and J. E. Keeley. 2005. Fire as a global ‘herbivore’:
the ecology and evolution of flammable ecosystems. Trends in Ecology and Evolution 20:387–394.

Bond, W. J., and B. van Wilgen. 1996. Fire and plants.

Chapman and Hall, London, UK.

Bradstock, R. A., A. M. Gill, B. J. Kenny, and J. Scott. 1998.

Bushfire risk at the urban interface estimated from historical weather records: consequences for the use of prescribed fire in the Sydney region of south-eastern Australia. Journal of Environmental Management 53:259–271.

Burnham, K. P., and D. R. Anderson. 2002. Model selection and inference: a practical information-theoretic approach.

Springer-Verlag, New York, New York, USA.

CAFRAP [California Fire Resource and Assessment Program]. 2001. Comparison of areas burned in developed and
undeveloped wildland areas in the northwestern Sierra
Nevada foothills. California Department of Forestry and Fire Protection, Sacramento, California, USA.

Calkin, D. E., K. M. Gebert, K. M. Jones, and R. P. Neilson. 2005. Forest Service large fire area burned and suppression expenditure trends, 1970–2002. Journal of Forestry 103:
179–183.

Cardille, J. A., S. J. Ventura, and M. G. Turner. 2001.

Environmental and social factors influencing wildfires in the upper Midwest, United States. Ecological Applications 11:
111–127.

Cohen, J. D. 2000. Preventing disaster: home ignitability in the wildland–urban interface. Journal of Forestry 98:15–21.

Countryman, C. M. 1972. The fire environment concept.

USDA Forest Service, Pacific Southwest Range and Experiment Station, Berkeley, California, USA.

Dale, V. H., L. A. Joyce, S. McNulty, and R. P. Neilson. 2000.

The interplay between climate change, forests and disturbances. Science of the Total Environment 262:201–204.

DellaSalla, D. A., J. E. Williams, C. D. Williams, and J. F.

Franklin. 2004. Beyond smoke and mirrors: a synthesis of fire policy and science. Conservation Biology 18:976–986.

Dobson, A. P., J. P. Rodriguez, W. M. Roberts, and D. S.

Wilcove. 1997. Geographic distribution of endangered
species in the United States. Science 275:550–553.

Duncan, B. W., and P. A. Schmalzer. 2004. Anthropogenic influences on potential fire spread in a pyrogenic ecosystem of Florida, USA. Landscape Ecology 19:153–165.

ECOMAP. 1993. National hierarchical framework of ecological units. USDA Forest Service, Washington, D.C., USA.

Fried, J. S., G. Winter, and J. K. Gilless. 1999. Assessing the benefits of reducing fire risk in the wildland–urban interface:
a contingent valuation approach. International Journal of Wildland Fire 9:9–20.

Gray, A. N., H. S. J. Zald, R. A. Kern, and N. Malcolm. 2005.

Stand conditions associated with tree regeneration in Sierran mixed-conifer forests. Forest Science 51:198–210.

Haidinger, T. L., and J. E. Keeley. 1993. Role of high fire frequency in destruction of mixed chaparral. Madron˜o 40:
141–147.

Haight, R. G., D. T. Cleland, R. B. Hammer, V. C. Radeloff, and T. S. Rupp. 2004. Assessing fire risk in the wildland– urban interface. Journal of Forestry 104:41–48.

Halsey, R. W. 2005. Fire, chaparral, and survival in southern California. Sunbelt, San Diego, California, USA.

Hammer, R. B., V. C. Radeloff, J. S. Fried, and S. I. Stewart. In press. Wildland–urban interface growth during the 1990s in California, Oregon and Washington. International Journal of Wildland Fire.

Hammer, R. B., S. I. Stewart, R. Winkler, V. C. Radeloff, and P. R. Voss. 2004. Characterizing spatial and temporal
residential density patterns across the U.S. Midwest, 1940– 1990. Landscape and Urban Planning 69:183–199.

Hawbaker, T. J., and V. C. Radeloff. 2004. Road and landscape pattern in northern Wisconsin based on a comparison of four road data sources. Conservation Biology 18:1233–1244.

Hickman, J. C. 1993. The Jepson manual: higher plants of California. University of California Press, Berkeley, California, USA.

Keeley, J. E. 1982. Distribution of lightning and man-caused wildfires in California. Pages 431–437 in C. E. Conrad and W. C. Oechel, editors. Proceedings of the International Symposium on the Dynamics and Management of Mediterranean-type Ecosystems. USDA Forest Service, Pacific Southwest Forest and Range Experiment Station, Berkeley, California, USA.

Keeley, J. E. 2000. Chaparral. Pages 202–253 in M. G. Barbour and W. D. Billings, editors. North American terrestrial vegetation. Cambridge University Press, New York, New
York, USA.

Keeley, J. E. 2005. Fire history of the San Francisco East Bay region and implications for landscape patterns. International Journal of Wildland Fire 14:285–296.

Keeley, J. E., and C. J. Fotheringham. 2003. Impact of past, present, and future fire regimes on North American
Mediterannean shrublands. Pages 218–262 in T. T. Veblen, W. L. Baker, G. Montenegro, and T. W. Swetnam, editors.

Fire and climatic change in temperate ecosystems of the western Americas. Springer-Verlag, New York, New York,
USA.

Keeley, J. E., C. J. Fotheringham, and M. Morais. 1999.

Reexamining fire suppression impacts on shrublandland fire regimes. Science 284:1829–1832.

July 2007 HUMAN INFLUENCE ON CALIFORNIA FIRE 1401
Keeley, J. E., C. J. Fotheringham, and M. A. Moritz. 2004.

Lessons from the 2003 wildfires in southern California.

Journal of Forestry 102:26–31.

Landis, J. D., and M. Reilly. 2004. How we will grow: baseline projections of the growth of California’s urban footprint through the year 2100. Working Paper 2003-04. Institute of Urban and Regional Development, University of California, Berkeley, California, USA.

Landres, P. B., P. Morgan, and F. J. Swanson. 1999. Overview of the use of natural variability concepts in managing
ecological systems. Ecological Applications 9:1179–1188.

Lenihan, J. L., R. Drapek, D. Bachelet, and R. P. Neilson. 2003. Climate change effects on vegetation distribution, carbon, and fire in California. Ecological Applications 13:
1667–1681.

Malamud, B. D., J. D. A. Millington, and G. L. W. Perry. 2005.

Characterizing wildfire regimes in the United States. Proceedings of the National Academy of Sciences (USA) 102:
4694–4699.

Marsden, M. A. 1982. A statistical analysis of the frequency of lightning-caused forest fires. Journal of Environmental Management 14:149–159.

Mercer, D. E., and J. P. Prestemon. 2005. Comparing
production function models for wildfire risk analysis in the wildland–urban interface. Forest Policy and Economics 7:
782–795.

Moritz, M. A. 1997. Analyzing extreme disturbance events: fire in the Los Padres National Forest. Ecological Applications 7:
1252–1262.

Moritz, M. A., J. E. Keeley, E. A. Johnson, and A. A.

Schaffner. 2004. Testing a basic assumption of shrubland fire management: Does the hazard of burning increase with the age of fuels? Frontiers in Ecology and the Environment 2:
67–72.

Pausas, J. G., R. A. Bradstock, D. A. Keith, and J. E. Keeley. and GCTE (Global Change of Terrestrial Ecosystems) Fire Network. 2004. Plant functional traits in relation to fire in crown-fire ecosystems. Ecology 85:1085–1100.

Prestemon, J. P., J. M. Pye, D. T. Butry, T. P. Holmes, and D. E. Mercer. 2002. Understanding broadscale wildfire risks in a human-dominated landscape. Forest Science 48:685–693.

Pyne, S. J. 2001. Fire in America. Princeton University Press.

Princeton, New Jersey, USA.

Pyne, S. J., P. L. Andrews, and R. D. Laven. 1996. Introduction to wildland fire. John Wiley and Sons, New York, New York, USA.

Radeloff, V. C., R. B. Hammer, S. I. Stewart, J. S. Fried, S. S.

Holcomb, and J. F. McKeefry. 2005. The wildland–urban
interface in the United States. Ecological Applications 15:
799–805.

R Development Core Team. 2005. R: a language and
environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. hhttp://www.

R-project.orgi
Rollins, M. G., P. Morgan, and T. Swetnam. 2002. Landscapescale controls over 20th century fire occurrence in two large Rocky Mountain (USA) wilderness areas. Landscape Ecology 17:539–557.

Rundel, P. W., and J. A. King. 2001. Ecosystem processes and dynamics in the urban/wildland interface of Southern
California. Journal of Mediterranean Ecology 2:209–219.

Savage, M., B. Sawhill, and M. Askenazi. 2000. Community dynamics: What happens when we rerun the tape? Journal of Theoretical Biology 205:515–526.

Schroeder, M. J., et al. 1964. Synoptic weather types associated with critical fire weather. U.S. Forest Service, Pacific Southwest Range and Experiment Station. Berkeley, California, USA.

Shindler, B., and E. Toman. 2003. Fuel reduction strategies in forest communities. Journal of Forestry 101:8–14.

Stephens, S. L. 1998. Evaluation of the effects of silvicultural and fuels treatments on potential fire behaviour in Sierra Nevada mixed-conifer forests. Forest Ecology and Management 105:21–35.

Stephens, S. L. 2005. Forest fire causes and extent on United States Forest Service lands. International Journal of Wildland Fire 14:213–222.

Sturtevant, B. R., P. A. Zollner, E. J. Gustafson, and D. T.

Cleland. 2004. Human influence on the abundance and
connectivity of high-risk fuels in mixed forests of northern Wisconsin, USA. Landscape Ecology 19:235–253.

Turner, M. G., W. H. Romme, R. H. Gardner, and W. W.

Hargrove. 1997. Effects of fire size and pattern on early succession in Yellowstone National Park. Ecological Monographs 67:411–433.

U.S. Census Bureau. 2000. Census 2000 TIGER line files. U.S.

Census Bureau, Washington, D.C., USA.

U.S. Census Bureau. 2002. Census 2000 summary file 3A
technical documentation. U.S. Census Bureau, Washington, D.C., USA.

USDA and USDI. 2001. Urban wildand interface communities within vicinity of federal lands that are at high risk from wildfire. Federal Register 66:751–777.

U.S. Geological Survey. 2002. Digital line graph (DLG)
availability, 7.5 minute transportation overlay. U.S. Geological Survey, Washington, D.C., USA.

Veblen, T. T., T. Kitzberger, and J. Donnegan. 2000. Climatic and human influences on fire regimes in ponderosa pine
forests in the Colorado Front Range. Ecological Applications 10:1178–1195.

Wells, M. L., J. F. O’Leary, J. Franklin, J. Michaelson, and D. E. McKinsey. 2004. Variations in a regional fire regime related to vegetation type in San Diego County, California (USA). Landscape Ecology 19:139–152.

Westerling, A. L., H. G. Hidalgo, D. R. Cayan, and T. W.

Swetnam. 2006. Warming and earlier spring increase western U.S. forest wildfire activity. Science 313:940–943.

Wilson, E. O. 1992. The diversity of life. Norton, New York, New York, USA.

Winter, G., and J. S. Fried. 2000. Homeowner perspectives on fire hazard, responsibility, and management strategies at the wildland–urban interface. Society and Natural Resources 13:
33–49.

Winter, G. J., C. Vogt, and J. S. Fried. 2002. Fuel treatments at the wildland–urban interface: common concerns in diverse regions. Journal of Forestry 100:15–21.

Zedler, P. H., C. R. Gautier, and G. S. McMaster. 1983.

Vegetation change in response to extreme events: the effect of a short interval between fires in California chaparral and coastal scrub. Ecology 64:809–818.
1402 ALEXANDRA D. SYPHARD ET AL. Ecological Applications Vol. 17, No. 5
Contributed Paper
Conservation Threats Due to Human-Caused
Increases in Fire Frequency in
Mediterranean-Climate Ecosystems
ALEXANDRA D. SYPHARD,∗‡ VOLKER C. RADELOFF,∗ TODD J. HAWBAKER,∗ AND SUSAN I. STEWART†
∗Department of Forest and Wildlife Ecology, University of Wisconsin, Madison, WI 53706, U.S.A. †USDA Forest Service, Northern Research Station, Evanston, IL 60201, U.S.A.

Abstract: Periodic wildfire is an important natural process in Mediterranean-climate ecosystems, but increasing fire recurrence threatens the fragile ecology of these regions. Because most fires are human-caused, we investigated how human population patterns affect fire frequency. Prior research in California suggests the relationship between population density and fire frequency is not linear. There are few human ignitions in areas with low population density, so fire frequency is low. As population density increases, human ignitions and fire frequency also increase, but beyond a density threshold, the relationship becomes negative as fuels become sparser and fire suppression resources are concentrated. We tested whether this hypothesis also applies to the other Mediterranean-climate ecosystems of the world. We used global satellite databases of population, fire activity, and land cover to evaluate the spatial relationship between humans and fire in the world’s five Mediterranean-climate ecosystems. Both the mean and median population densities were consistently and substantially higher in areas with than without fire, but fire again peaked at intermediate population densities, which suggests that the spatial relationship is complex and nonlinear. Some land-cover types burned more frequently than expected, but no systematic differences were observed across the five regions. The consistent association between higher population densities and fire suggests that regardless of differences between land-cover types, natural fire regimes, or overall population, the presence of people in Mediterranean-climate regions strongly affects the frequency of fires; thus, population growth in areas now sparsely settled presents a conservation concern. Considering the sensitivity of plant species to repeated burning and the global conservation significance of Mediterranean-climate ecosystems, conservation planning needs to consider the human influence on fire frequency. Fine-scale spatial analysis of relationships between people and fire may help identify areas where increases in fire frequency will threaten ecologically valuable areas.

Keywords: fire, land cover, Mediterranean, MODIS, population density, remote sensing Amenazas a la Conservacion Debido a Incrementos en la Frecuencia de Incendios Causados por Humanos en ´ Ecosistemas de Clima Mediterr´aneo
Resumen: El fuego periodico es un proceso natural importante en los ecosistemas de clima mediterr ´ aneo, ´ pero el incremento de la recurrencia de fuego amenaza la fragil ecolog ´ ´ıa de esas regiones. Debido a que la mayor´ıa de los incendios son causados por humanos, investigamos el efecto de los patrones de poblacion´ humana sobre la frecuencia del fuego. Investigaciones previas en California sugieren que la relacion entre la ´ densidad poblacional y la frecuencia de incendios no es lineal. Hay pocas igniciones humanas en areas con ´ baja densidad poblacional, as´ı que la frecuencia de incendios es baja. A medida que aumenta la densidad poblacional, los incendios causados por humanos y la frecuencia de incendios tambi´en incrementa; pero al ‡Current address: Conservation Biology Institute, 7916 Deerfield Street, San Diego, CA 92120, U.S.A., email [email protected] Paper submitted February 5, 2008; revised manuscript accepted January 16, 2009. 758
Conservation Biology, Volume 23, No. 3, 758–769
C 2009 Society for Conservation Biology
DOI: 10.1111/j.1523-1739.2009.01223.x
Syphard et al. 759
llegar a un umbral de densidad, la relacion se vuelve negativa ya que los combustibles son escasos y se con- ´ centran recursos para la supresion de fuego. Probamos s ´ ´ı esta hipotesis tambi ´ ´en aplica a los otros ecosistemas de clima mediterraneo en el mundo. Utilizamos bases de datos de sat ´ ´elite de poblacion, actividad de fuego ´ y cobertura de suelo para evaluar la relacion espacial entre humanos y fuego en los cinco ecosistemas de ´ clima mediterraneo en el mundo. Tanto las densidades medias y medianas fueron consistente y sustancial- ´ mente mas altas en ´ areas con fuego como sin fuego, pero los incendios alcanzaron su m ´ aximo en densidades ´ poblacionales intermedias, lo que sugiere que la relacion espacial es compleja y no lineal. Algunos tipos de ´ cobertura de suelo tuvieron incendios mas frecuentemente de lo esperado, pero no se observaron diferencias ´ significativas en las cinco regiones. La asociacion consistente entre mayores densidades poblacionales y fuego ´ sugiere que, independientemente de las diferencias entre tipos de cobertura de suelo, los reg´ımenes de fuego naturales o la poblacion total, la presencia de gente en regiones de clima mediterr ´ aneo afecta fuertemente a ´ la frecuencia de incendios; por lo tanto, el crecimiento poblacional en areas escasamente pobladas es preocu- ´ pante para la conservacion. Considerando la sensibilidad de las especies de plantas a incendios recurrentes y ´ la significancia para la conservacion de los ecosistemas de clima mediterr ´ aneo, la planificaci ´ on de la conser- ´ vacion requiere que se considera la influencia humana sobre la frecuencia de incendios. El an ´ alisis espacial ´ a fina escala de las relaciones entre gente y fuego puede ayudar a identificar areas en las que el incremento ´ en la frecuencia de fuego amenazaraa ´ areas valiosas ecol ´ ogicamente. ´ Palabras Clave: cobertura de suelo, densidad poblacional, fuego, Mediterr´aneo, MODIS, percepcion remota ´ Introduction
The biodiversity of Mediterranean-climate ecosystems is among the highest of any biome in the world. The five
regions in the world with Mediterranean climates (the
Mediterranean Basin, central Chile, the Cape Region of
South Africa, southwestern Australia, and parts of California and northern Baja California in North America) collectively occupy <5% of the Earth’s unglaciated land surface, yet they contain 20% of the world’s flora (Cowling et al. 1996), and many species are endemic (Mittermeier et al. 1998). Because of rapid global change and increasing anthropogenic pressure, all Mediterranean regions are of high global conservation concern (Medail & ´ Quezel 1999; Olson & Dinerstein 2002; Vogiatzakis et al. ´ 2006).

Although Mediterranean-climate ecosystems are geographically disjunct, they are classic examples of convergence in ecosystem structure and dynamics (Cody & Mooney 1978). The Mediterranean climate is characterized by cool, wet winters and warm to hot, dry summers, and the summer drought produces water stress that affects the seasonal distribution of wildfires. Vegetation in Mediterranean-climate regions is dominated by evergreen, woody, sclerophyllous shrubs that are very
flammable and support crown fires (Christensen 1985).

Nevertheless, specialized postfire persistence traits (e.g., seed banking in the soil and canopy and resprouting)
make plant species resilient to periodic wildfire (Naveh 1975). The presence of fire-stimulated reproduction indicates an adaptive response to fire, and seed banking evolved independently in all Mediterranean-climate ecosystems except Chile (Bond & van Wilgen 1996).

Nevertheless, all the woody shrubs in Chile resprout in response to fire, which is now frequent due to anthropogenic ignitions (Montenegro et al. 2004).

Fire in Mediterranean-climate ecosystems predates humans (except in Chile), and natural fire frequencies have varied between and among regions over time and in response to climate fluctuations (Rundel 1998). The history of human impact on fire regimes also differs among regions. For example, humans ignited fires in the Mediterranean Basin for thousands of years to support agropastoral activities (Lozano et al. 2008), Native Americans ignited fires in California since the early Holocene (Keeley 2002), and small populations of hunter-gatherers ignited fires in other regions until a few centuries ago (Rundel 1998). Evidence regarding early human influence on fire is circumstantial and controversial, but human activity is now thought to be a major determinant of the timing and location of fire. In fact, humans ignite most fires in Mediterranean regions (Bond & van Wilgen 1996).

Current human influence on fire regimes and the potential ecological impact of their influence on fire is similar among Mediterranean-climate regions and differs strongly from fire problems in other forested systems.

In dry coniferous forests, like those in the western
United States, the primary concern is a lack of fire primarily due to 20th-century fire suppression. Lower fire frequency in forests that naturally experienced highfrequency, low-intensity surface fires resulted in high accumulation of surface and canopy fuels (Parsons & Landres 1998). Fuel accumulation increases the likelihood fires will become uncharacteristically large and intense, which can kill even large, surface-fire-resistant trees.

Conservation threats and changes in fire regimes in
Mediterranean-climate regions, however, are different.

The shrublands are adapted to fire-return intervals that Conservation Biology
Volume 23, No. 3, 2009
760 Fire in Mediterranean-Climate Ecosystems
are generally longer than those historically experienced in conifer forests (Sugihara et al. 2006). Despite their capacity for rapid postfire regeneration, many shrubland plant species are sensitive to repeated burning. Serotinous species are particularly vulnerable (e.g., Wark et al. 1987; Pausas 1999; Syphard et al. 2006), but repeated
burning may also extirpate resprouting species by reducing their capacity to regenerate and constraining their reproductive ability (e.g., Haidinger & Keeley 1993; Montenegro et al. 2004; Espelta et al. 2008). A related issue is that exotic species may facilitate fire and may expand under frequent fire (Mack & D’Antonio 1998). In California biodiversity is critically threatened by shrubland conversion to exotic annual grasses caused by atypically frequent fire (Keeley et al. 2005). Therefore, where the primary concern in dry coniferous forests is fire exclusion, the problem in Mediterranean-climate regions is repeated fires in the same location (Montenegro et al.
2004; Badia-Perpiny`a & Pallares-Barbera 2006; Forsyth & van Wilgen 2008), although the intensity of fires may
vary from region to region because of differences in prescribed management practices. Thus, understanding the causes and spatial distribution of altered fire regimes in Mediterranean-climate ecosystems has become a major
research priority with strong conservation implications (Lavorel et al. 1998) and is particularly important given population growth in Mediterranean-climate ecosystems.

Studies in California show that area burned and number of fires are highest when population and housing densities are intermediate (Keeley 2005; Syphard et al. 2007). Fires initially increase with population and housing density and then decline where a threshold density is reached. There are several interrelated reasons for this.

Ninety-five percent of California’s fires are human caused; therefore, anthropogenic ignitions are lower in areas with low population density. As population and housing densities increase, fuels are still abundant and contiguous enough to carry fire, and the number and frequency of
fires increase (Syphard et al. 2007). As population density increases further and an area is developed, wildland fuel is reduced and fragmented and fire-suppression resources
are concentrated, resulting in lower fire frequencies at high population densities. Finally, even if fire frequency remains stable, fires may cluster in certain areas (e.g., human settlements) or land-cover types (Nunes et al. 2005; Table 1. Number of Bailey’s ecoregions, total area, and biogeographic characteristics∗ of Mediterranean-climate regions.

Number of Total area Number of native Endemic Threatened ecoregions (km2) vascular plants species (%) species (%) Mediterranean Basin 25 2,392,048 23,300 50 18
North America 5 407,654 4,300 35 17
Chile 2 74,863 2,100 23 unknown
South Africa 1 69,401 8,550 68 15
Southwest Australia 1 118,882 8,000 75 18
∗Biogeographic characteristics based on Calow (1998) and Vogiatzakis et al. (2006).

Forsyth & van Wilgen 2008), resulting in high fire frequency in localized areas.

Although the relationship between human population
densities and fires has been studied in California, less is known about fire trends and patterns in other Mediterranean ecosystems. In recent years, fire frequency has escalated because of population growth and human ignitions in Chile (e.g., Montenegro et al. 2004) and South Africa (Forsyth & van Wilgen 2008), and fires increased exponentially in many areas in the Mediterranean Basin, in part due to the abandonment of traditional land-use
practices (Pausas & Vallejo 1999). Interactions between fire and exotic species have been exacerbated by recurrent human-caused fires in Chile (Montenegro et al. 2004), South Africa (Bond & van Wilgen 1996), the
Mediterranean Basin (Kark & Sol 2005; Vogiatzakis et
al. 2006), and Australia (Offor 1990). In Spain fire ignitions cluster near urban areas (Badia-Perpiny`a & PallaresBarbera 2006), and population density has been correlated with the number of fires and area burned (V´azquez de la Cueva et al. 2006). Results of previous studies thus suggest that the relationship between human populations and fire frequency may be similar in all Mediterraneanclimate ecosystems, but this idea has not been examined systematically across the different areas. Whether fire frequencies consistently peak at intermediate densities of human population is unclear. Nor is it clear whether certain land-cover types are more likely to burn.

Our objective was to quantify the relationship between
humans and fire in Mediterranean-climate ecosystems
across the globe. We asked, Are population densities
higher in places where fires occur than in places without fires? Are fires consistently most frequent at intermediate population densities? Are certain land-cover types in each region more prone to fires?
Methods
Study Area
We used Bailey’s ecoregion boundaries to demarcate Mediterranean-climate ecoregions (Bailey 1989). (Table 1). This is a hierarchical system with four levels (domains, divisions, provinces, and sections). For all five Conservation Biology
Volume 23, No. 3, 2009
Syphard et al. 761
Figure 1. For the Mediterranean
Basin, (a) MODIS active-fire
detections in 2005, (b) LandScan
population density in 2005, and
(c) MODIS land-cover data. Fire
and population density values
are averaged across 225-km2
pixels.
continents, we selected all ecoregions classified as either the Mediterranean Division or the Mediterranean Regime
Mountains. To ensure comparability of area calculations, all spatial data were projected into an Albers equal area projection.

Processing of Population Data
We used population data from the LandScan Global Population Product because it has the finest resolution (<1 km) of any global population data set (Dobson et al.
2000). The LandScan database represents ambient population, accounting for diurnal movement and travel patterns. Every grid cell is allocated a population count based on a distribution model that incorporates the best available data on human population for every country, proximity of people to roads, land cover, nighttime lights, and urban density.

Because the accuracy and precision of LandScan are
continually being improved, we restricted our analysis to 2005, the year with the most current data (Fig. 1). For comparison purposes, we divided the population counts
by area and analyzed population density.

Processing of MODIS Fire Data
We used fire data from the Moderate Resolution Imaging Spectroradiometer (MODIS) to assess fire activity in Mediterranean-climate ecoregions because of its unmatched spatial and temporal detail (Justice 2002). With two polar-orbiting satellites, the MODIS active-fire product provides daily global information on fires. These data show actively burning fires based on radiant energy and comparisons of target pixels with surrounding pixels
(Giglio et al. 2003).

Instead of mapping individual fires and area burned,
MODIS indicates pixels in which fire activity was detected. Thus, there could be more than one fire active within a 1-km2 MODIS pixel (Csiszar et al. 2006). In
addition, fires occupying only a portion of a pixel can Conservation Biology
Volume 23, No. 3, 2009
762 Fire in Mediterranean-Climate Ecosystems
be detected (Dozier 1981). Although many small fires
are missed, MODIS consistently detects larger fires that are ecologically relevant (Hawbaker et al. 2008), and the number of contiguous MODIS fire pixels tends to correlate with fire size (Giglio et al. 2006).

We analyzed MODIS fire data from the Land Processes
Distributed Active Archive Center (LPDAAC, http://
edcdaac.usgs.gov/modis/dataproducts.asp) for both sensors every day in 2005 to match the date of the population data. Using the boundaries of the Mediterranean ecoregions, we put all images into a mosaic (i.e., joined them together to form daily continuous tiles) for both sensors and summarized the daily data to create annual maps of
fire for each region (Fig. 1). We included fire detections from all classified confidence levels because detection accuracy varies little whether fires are classified as low or high confidence (Hawbaker et al. 2008).

Processing of MODIS Land-Cover Data
In addition to the active-fire product, we used the 2003 MODIS 1 km Land Cover Dataset (Friedl et al. 2002) to analyze fire activity by land-cover class (Fig. 1). We used the LAI/fPAR Biome land-cover classification scheme because it was designed to capture differences in vegetation structural types (grasslands and cereal crops, shrubs, broadleaf crops, savannah, broadleaf forest, needle leaf forest, unvegetated, and urban; Myneni et al. 1997).

Analysis
In California fires are most likely to occur when the distance to housing is <15 km (Syphard et al. 2007). Because scale dependencies of ecological patterns and processes vary by region (Shugart 1998) and because people are
mobile and affect their surroundings, we conducted our
analysis of humans and fire at three levels of resolution (1, 15, and 45 km). Land-cover analyses were conducted
only at the 1-km resolution, however, because we did
not consider relationships between land cover and population measures.

We conducted a moving-window GIS analysis to summarize data across the entire land area. Within each window and at each resolution, we summarized the population density and the number of fires. Satellite fire detections can be obscured by clouds, and the MODIS
active-fire product explicitly masks cloud cover in every daily image (Giglio et al. 2003). Therefore, we excluded cloud pixels, calculated the number of “observable days” within each window, and used this number to calculate
average fire frequency. Uncertainty due to land-cover misclassification, undetected fires, and errors in population distribution was assumed to be consistent among the
Mediterranean-climate ecoregions.

To determine whether population densities were
higher in areas with fires, we selected all pixels and
windows where there was one or more fires and calculated the mean and median population densities. We compared those with mean and median population densities in pixels and windows where no fires occurred.

If there is a relationship between humans and fire, the proportion of fire should be higher where population is higher and lower where population is lower. We did not
conduct a statistical test to determine whether the distributions differed because our data represent a complete enumeration, not a sample, and any difference would be
statistically significant. Instead, we distributed the population data into 25 equally spaced categories and plotted the proportion of fires that occurred within each category for the three window sizes. The resulting bar charts showed whether more fires occurred at low, intermediate, or high population densities.

To determine whether fires burned more often (selectively) in different land-cover types, we calculated the total proportion of land-cover types in each region, then selected only the pixels with fires and recalculated the proportion. We calculated the ratio of the proportion of fires in the land-cover types and the proportion of the land-cover types in the landscape. A ratio of 1.0 means fire occurred in a land-cover type as often as would be expected by chance, >1.0 means fire occurred in the landcover type more often, and <1.0 means fire occurred less often than expected by chance.

Results
We observed substantial differences in population density among the regions. Both the mean and median population densities in southwestern Australia were lowest of all the regions, and those in the Mediterranean Basin were highest. Although median population densities were substantially lower than mean population densities for all regions, the difference in North America was so substantial that mean population density was highest among the regions, but median population density was equal to that in southwestern Australia.

Pixels or windows with fires typically had higher population densities than pixels or windows without fires (Fig. 2). The only exception was in the 1-km pixels
in North America, where mean population density was
higher in the pixels without fires. Median population densities were nearly equal with and without fire in 1-km pixels in North America, South Africa, and southwestern Australia.

The relationship between population density and fire
was more pronounced at 15 km than at 1 km, and at 45 km the mean population densities in areas with fires were
much higher than where there were no fires (Fig. 2a). The median population density with fire was almost 3 times
larger than the population density without fire at 45-km resolution.

Conservation Biology
Volume 23, No. 3, 2009
Syphard et al. 763
Figure 2. (a) Mean and (b)
median population densities in
areas with and without fires for
1-, 15-, and 45-km resolution
windows. The y-axis scales differ.

Although population densities were, on average,
higher where there were fires, the largest proportion of fires peaked at intermediate population densities (Fig. 3).

Patterns of variation and peak population densities varied from region to region though, particularly at the 1- and 15-km window sizes. In addition, the peak in proportion of fires occurred in areas of lower population densities in North America at the 1-km resolution. In Chile and southwestern Australia, peak in proportion of fires occurred at the higher end of the population density distribution in the 1- and 45-km window sizes. The most consistent trend was apparent at the 45-km window size, where the highest proportion of fires occurred between 100 and 250 people per 45 km2.

Land cover in the five regions included grasslands and
cereal crops, shrubs, and savanna, with lower proportions of broad-leaf crops, broad-leaf forest, needle-leaf forest, unvegetated, and urban cover (Fig. 4). Distribution of these land-cover types, however, varied widely from region to region. Grasslands and cereal crops accounted for 40% of land cover in South Africa and southwestern Australia, but in Chile and North America they were just 20% of land cover. Substantially more needle-leaf forest was present in North America (21%) than in the other regions (<10%), and much of Chile was unvegetated (23%).

Some land-cover classes burned proportionately more
than expected by chance given their areal distribution in the regions, but patterns were not consistent (Table 2; Fig. 4). In North America and Chile grasslands and cereal crops burned substantially more than expected but only
as much as expected in the other three regions. Broadleaf forest burned more than expected in southwestern Australia but not in the other regions. In North America shrubs burned more than expected and needle-leaf forest burned less than expected, but in the Mediterranean Basin, shrubs burned less than expected and needle-leaf forest burned more. In all regions, except for North
America, more fires occurred in savannah than expected.

Overall, very little fire occurred in unvegetated or urban areas.

Conservation Biology
Volume 23, No. 3, 2009
764 Fire in Mediterranean-Climate Ecosystems
Figure 3. Proportion of fires within population density classes for 1-, 15-, and 45-km resolution windows.

Discussion
We found strong evidence that people are associated with the frequency and spatial distribution of fire similarly in all five Mediterranean-climate regions. Both mean and
median population densities were consistently and substantially higher in areas with fire than in areas that did not burn; fires in Mediterranean-climate regions tended to occur close to people. Despite their convergence in
ecosystem structure and function, Mediterranean-climate regions do vary in fire history, land-use history, or socioeconomic and political conditions (Pignatti et al. 2002; Carmel & Flather 2004; Vogiatzakis et al. 2006). Because of these differences, variations among the regions in
population densities and land cover are not surprising.

But these differences make the consistency of spatial
relationships between people and fire across the five
regions even more striking. The spatial pattern of fires Conservation Biology
Volume 23, No. 3, 2009
Syphard et al. 765
Figure 3. (continued)
in any region depends on complex interactions between
ignition sources, landscape characteristics, and fuel continuity (Whelen 1995). So the consistent relationship between fire and population density suggests that the presence of people in Mediterranean-climate regions overrides these other factors.

Understanding the distribution of fire in
Mediterranean-climate ecosystems is critical due to
the vulnerability of its unique vegetation to repeated
burning. Unlike other ecoregions in which decreased
fire frequency threatens some species (Allen et al.
2002), in Mediterranean-climate ecoregions, the conservation concern is increased fire frequency (e.g., Keeley et al. 1999; Montenegro et al. 2004; Badia-Perpiny`a&
Pallares-Barbera 2006). The persistence of native plants is threatened and may have cascading ecological effects (Barro & Conard 1991; DellaSalla et al. 2004). Because
Mediterranean regions are highly heterogeneous, the
Conservation Biology
Volume 23, No. 3, 2009
766 Fire in Mediterranean-Climate Ecosystems
Figure 3. (continued)
sensitivity of different plant species to specific fire frequencies will vary (Public Library of Science ONE
DOI:10.1371/journal.pone.0000938. 2007). Nevertheless, identifying where the landscape is likely to burn frequently is an important step in identifying areas
vulnerable to the extirpation of native species.

The association of people with the spatial distribution of fire occurrence is likely due to the fact that humans now cause the majority of ignitions in all
five Mediterranean-climate regions (Bond & van Wilgen
Table 2. Ratio of the proportion of fires by land-cover type and proportion of land-cover type in the landscape.∗
Land-cover Mediterranean North South SW
type Basin America Chile Africa Australia
Grass/cereal 0.79 1.76 1.72 1.09 0.85
Broad crops 1.07 1.70 1.65 0.55 0.49
Shrubs 0.42 1.35 1.00 0.79 0.43
Savannah 2.01 0.72 1.51 1.46 1.35
Broad leaf 0.80 0.45 1.02 1.62 1.90
Needle leaf 2.01 0.54 1.03 0.94 2.64
Unvegetated 0.06 0.17 0.03 0.13 0.06
Urban 1.92 0.89 1.88 1.41 0.96
∗A ratio of 1.0 means fire occurred in a land-cover type as often as would be expected by chance, >1.0 means that fire occurred more often than expected, and <1.0 less often than expected by chance. 1996), and human ignitions are likely to occur close
to roads and human infrastructure (e.g., Yang et al.
2007; Syphard et al. 2008). Nevertheless, our results also showed that fire occurrence consistently peaked where
population densities were intermediate, which suggests
that fire patterns in Mediterranean-climate regions are related to the spatial arrangement between people, urban development, and fuel. When population density is lowest, human ignitions are also low but increase with
population density. Nevertheless, there appears to be a threshold above which fire occurrence declines, possibly due to less open space and fuel fragmentation caused by urban development or other land-use change. Firesuppression resources also tend to be concentrated near urban areas (Calkin et al. 2005), and intermediate-density housing when located within wildland vegetation is classified as the wildland–urban interface (WUI) in the United States and given special fire-management considerations (Radeloff et al. 2005).

The relationship between people and fire in our study
was most pronounced at the 15- and 45-km scales of analysis. Many ecological processes and spatial relationships have characteristic scales or space and time intervals over which the process can be detected (Shugart 1998). One
explanation for the scale effect in our results is that analysis with the 15- and 45-km window sizes could include pixels where fires did, and did not, burn. The observed relationship and scale dependence of the results may
therefore have been related to the relative proportion
of burned cells within a window. At the 1-km resolution, the pixel either burned or it did not, and the analysis did not account for neighborhood effects.

Although our primary focus was to assess the relationship between population density and fire, other researchers have shown that land use and land cover may be important covariates of fire patterns due to their
effects on fuel types, flammability, and human use of fire (e.g., Viedma et al. 2006; Baeza et al. 2007). In our analysis some land-cover types burned more frequently than expected, but no systematic differences were observed.

Conservation Biology
Volume 23, No. 3, 2009
Syphard et al. 767
Figure 4. Aerial proportion of
land-cover classes in the
ecoregions and within pixels with
an active fire in 2005.

Therefore, the patterns we observed in land-cover types were likely related to unique combinations of human land use and management practices within each region. For
example, in North America, needle-leaf forest burned less than expected, whereas shrublands burned more. Fire
suppression has successfully excluded fire from California’s high-elevation-mixed conifer forests. On the other hand, the disproportionately high level of fire in shrubs is likely due to housing development and increased human ignitions in low-elevation areas where these shrubs (i.e., chaparral) are common (Keeley et al. 1999). More fires than expected in needle-leaf forests in the Mediterranean Basin may be due to land abandonment, which has resulted in substantial increases of fire in pine forests (Pausas & Vallejo 1999).

In North America and Chile fire burned more in grasslands and cereal crops than expected. Grasslands can sustain and even promote higher fire frequencies than
other land-cover types (Mack & D’Antonio 1998), a major conservation concern in southern California, where exotic annual grasses have replaced native shrublands under unnaturally high fire frequencies (Haidinger & Keeley 1993). Problems with exotic annual grasses have also been reported in Chile and Australia (Pignatti et al. 2002) and may become more pronounced if fire frequency continues to increase.

Conclusions
Mediterranean-climate ecosystems are among the most
biologically diverse regions in the world with rates of endemism ranging from 23% (Chile) to 75% (southwestern Australia), and at least 15% of the taxa in Mediterranean-climate ecosystems are threatened (Calow
1998). Our results suggest that conservation planners in Mediterranean-climate regions should seriously consider human alteration of fire patterns. Although we used fire data for only 1 year, the consistency in our results demonstrates that, regardless of the overall fire frequency in a region and its annual weather-driven variations, it may be possible to predict where fires are concentrated. Our results therefore provide a foundation for further research and planning to identify where frequent fire threatens
vulnerable Mediterranean-climate plant species.

Future research should identify regionally specific
ranges of population densities where fire occurrence is highest, be conducted at the scales most relevant to planning and management, and incorporate other drivers of fire pattern, such as biophysical variables. Finally, compact development should be studied for its potential to mitigate the effects of human presence by limiting expansion into undeveloped vegetation. Education efforts to reduce human-caused ignitions were once the foundation of outreach programs, such as Smokey Bear; perhaps the time has come to bring the bear back from semiretirement.

Acknowledgments
We gratefully thank the U.S. Forest Service Northern Research Station for supporting this research. Three anonymous reviewers and the assigning editor, B. McComb made valuable suggestions, which greatly strengthened
our manuscript.

Literature Cited
Allen, C. D., M. Savage, D. A. Falk, K. F. Suckling, T. W. Swetnam, T.

Schulke, P. B. Stacey, P. Morgan, M. Hoffman, and J. T. Klingel. 2002.

Ecological restoration of southwest Ponderosa pine ecosystems: a broad perspective. Ecological Applications 12:1418–1433.

Conservation Biology
Volume 23, No. 3, 2009
768 Fire in Mediterranean-Climate Ecosystems
Badia-Perpiny`a, A., and M. Pallares-Barbera. 2006. Spatial distribution of ignitions in Mediterranean periurban and rural areas: the case of Catalonia. International Journal of Wildland Fire 15:187–196.

Baeza, M. J., A. Valdecantos, J. A. Alloza, and R. Vallejo. 2007. Human disturbance and environmental factors as drivers of long-term post-fire regeneration patterns in Mediterranean forests. Journal of Vegetation Science 2:243–252.

Bailey, R. G. 1989. Explanatory supplement to ecoregions map of the continents. Environmental Conservation 16:307–309.

Barro, S. C., and S. G. Conard. 1991. Fire effects on California chaparral systems: an overview. Environment International 17:135–149.

Bond, W. J., and B. van Wilgen. 1996. Fire and plants. Chapman & Hall, London.

Calkin, D. E., K. M. Gebert, K. M. Jones, and R. P. Neilson. Forest service large fire area burned and suppression expenditure trends, 1970–2002. Journal of Forestry 103:179–183.

Calow, P. 1998. The encyclopedia of ecology & environmental management. Blackwell Science, Oxford, United Kingdom.

Carmel, Y., and C. H. Flather. 2004. Comparing landscape scale vegetation dynamics following recent disturbance in climatically similar sites in California and the Mediterranean basin. Landscape Ecology 19:573–590.

Christensen, N. L. 1985. Shrubland fire regimes and their evolutionary consequences. Pages 85–100 in S. T. Pickett and P. S. White, editors.

Ecology of natural disturbance and patch dynamics. Academic Press, Orlando, Florida.

Cody, M. L., and H. A. Mooney. 1978. Convergence versus nonconvergence in Mediterranean-climate ecosystems. Annual Review of Ecology and Systematics 9:265–321.

Cowling, R. M., P. W. Rundel, B. B. Lamont, M. K. Arroyo, and M.

Arianoutsou. 1996. Plant diversity in Mediterranean-climate regions.

Trends in Ecology & Evolution 11:362–366.

Csiszar, I., J. Morisette, and L. Giglio. 2006. Validation of active fire detection from moderate resolution satellite sensors: the MODIS example in Northern Eurasia. IEEE Transactions on Geoscience and Remote Sensing 44:1757–1764.

DellaSala, D. A., J. E. Williams, C. D. Williams, and J. F. Franklin. 2004.

Beyond smoke and mirrors: a synthesis of fire policy and science.

Conservation Biology 18:976–986.

Dobson, J. E., E. A. Bright, P. R. Coleman, R. C. Durfee, and B. A. Worley. 2000. LandScan: a global population database for estimating populations at risk. Photogrammetric Engineering and Remote Sensing 66:849–857.

Dozier, J. 1981. A method for satellite identification of surfacetemperature fields of subpixel resolution. Remote Sensing of Environment 11:221–229.

Espelta, J. M., I. Verkaik, M. Eugenio, and F. Lloret. 2008. Recurrent wildfires constrain long-term reproductive ability in Pinus halepensis Mill. International Journal of Wildland Fire 17:579–585.

Forsyth, G. G., and B. W. van Wilgen. 2008. The recent fire history of the Table Mountain National Park and implications for fire management.

Koedoe 50:3–9.

Friedl, M. A., et al. 2002. Global land cover mapping from MODIS:
algorithms and early results. Remote Sensing of the Environment 83:287–302.

Giglio, L., G. R. van der Werf, J. T. Randerson, G. J. Collatz, and P.

Kasibhatla. 2006. Global estimation of burned area using MODIS active fire observations. Atmospheric Chemistry and Physics 6:957– 974.

Giglio, L., J. Descloitres, C. O. Justice, and Y. K. Kaufman. 2003. An enhanced contextual fire detection algorithm for MODIS. Remote Sensing of Environment 87:273–282.

Haidinger, T. L. and J. E. Keeley. 1993. Role of high fire frequency in destruction of mixed chaparral. Madrono˜ 40:141–147.

Hawbaker, T. J., V. C. Radeloff, A. D. Syphard, Z.-L. Zhu, and S. I.

Stewart. 2008. Detection rates of the MODIS active fire product.

Remote Sensing of Environment 112:2656–2664.

Justice, C. O., L. Giglio, S. Korontzi, J. Owens, J. T. Morisette, D. Roy, J. Descloitres, S. Alleaume, F. Petitcolin, and Y. Kaufman. 2002. The MODIS fire products. Remote Sensing of Environment 83:244–262.

Kark, S., and D. Sol. 2005. Establishment success across convergent Mediterranean ecosystems: an analysis of bird introductions. Conservation Biology 19:1519–1527.

Keeley, J. E. 2002. Native American impacts on fire regimes of the California coastal ranges. Journal of Biogeography 29:303– 320.

Keeley, J. E. 2005. Fire history of the San Francisco East Bay region and implications for landscape patterns. International Journal of Wildland Fire 14:285–296.

Keeley, J. E., C. J. Fotheringham, and M. Morais. 1999. Reexamining fire suppression impacts on brushland fire regimes. Science 284:1829– 1832.

Keeley, J. E., M. Keeley, and C. J. Fotheringham. 2005. Alien plant dynamics following fire in Mediterranean-climate California shrublands. Ecological Applications 15:2109–2125.

Lavorel, S., J. Canadell, S. Rambal, and J. Terradas. 1998. Mediterranean terrestrial ecosystems: research priorities on global change effects.

Global Ecology and Biogeography 7:157–166.

Lozano, F. J., S. Su´arez-Seoane, M. Kelly, and E. Luis. 2008. A multi-scale approach for modeling fire occurrence probability using satellite data and classification trees: a case study in a mountainous Mediterranean region. Remote Sensing of Environment 112:708–719.

Mack, M. C., and C. M. D’Antonio. 1998. Impacts of biological invasions on disturbance regimes. Trends in Ecology & Evolution 13:195–198.

Medail, F., and P. Qu ´ ezel. 1999. Biodiversity hotspots in the Mediter- ´ ranean basin: setting global conservation priorities. Conservation Biology 13:1510–1513.

Mittermeier, R. A., N. Myers, J. B. Thomsen, G. A. B. da Fonseca, and S.

Olivieri. 1998. Biodiversity hotspots and major tropical wilderness areas: approaches to setting conservation priorities. Conservation Biology 12:516–520.

Montenegro, G., R. Ginocchio, A. Segura, J. E. Keeley, and M. Gomez. ´ 2004. Fire regimes and vegetation responses in two Mediterraneanclimate regions. Revista Chilena de Historia Natural 77:455–464.

Myneni, R. B., R. R. Nemani, and S. W. Running. 1997. Estimation of global leaf area index and absorbed PAR using radiative transfer model. IEEE Transactions on Geoscience and Remote Sensing 35:1380–1393.

Naveh, Z. 1975. The evolutionary influence of fire in the Mediterranean region. Vegetatio 29:199–208.

Nunes, M. C. S., M. J. Vasconcelos, J. M. Pereira, N. Dasgupta, R. J.

Alldredge, and F. C. Reglo. 2005. Land cover type in Portugal:
do fires burn land cover selectively? Landscape Ecology 20:661– 673.

Offor, T. 1990. What future for the sandy heaths of Wilson’s Promontory. Victorian Naturalist 107:120–123.

Olson, D. M., and E. Dinerstein. 2002. The Global 200: priority ecoregions for global conservation. Annals of the Missouri Botanical Garden 89:125–126.

Parsons, D. J., and P. B. Landres. 1998. Restoring natural fire to wilderness. How are we doing? Pages 366–372 in T. Pruden and L. A. Brennan, editors. Fire in ecosystem management: shifting the paradigm from suppression to prescription. Tall Timbers, Tallahassee, Florida.

Pausas, J. G. 1999. The response of plant functional types to changes in the fire regime in Mediterranean ecosystems. A simulation approach.

Journal of Vegetation Science 10:717–722.

Pausas, J. G., and R. Vallejo. 1999. The role of fire in European Mediterranean ecosystems. Pages 3–16 in E. Chuvieco, editor. Remote sensing of large wildfires in the European Mediterranean basin. SpringerVerlag, New York.

Pignatti, E., S. Pignatti, and P. G. Ladd. 2002. Comparison of ecosystems in the Mediterranean Basin and Western Australia. Plant Ecology 163:177–186.

Conservation Biology
Volume 23, No. 3, 2009
Syphard et al. 769
Radeloff, V. C., R. B. Hammer, S. I. Stewart, J. S. Fried, S. S. Holcomb, and J. F. McKeefry. 2005. The wildland urban interface in the United States. Ecological Applications 15:799–805.

Rundel, P. W. 1998. Landscape disturbance in Mediterranean-type ecosystems: an overview. Pages 3–22 in P. W. Rundel, G. Montenegro, and F. M. Jaksic, editors. Landscape disturbance and biodiversity in Mediterranean-type ecosystems. Springer-Verlag, Berlin.

Shugart, H. H. 1998. Terrestrial ecosystems in changing environments.

Cambridge University Press, Cambridge, United Kingdom.

Sugihara, N. G., J. W. Van Wagtendonk, and J. Fites-Kaufman. 2006. Fire as an ecological process. Pages 58–74 in N. G. Sugihara, J. W. Van Wagtendonk, K. E. Shaffer, J. Fites-Kaufman, and A. E. Thode, editors. Fire in California’s ecosystems. University of California Press, Berkeley, California.

Syphard, A. D., K. C. Clarke, and J. Franklin. 2006. Simulating frequent fire and urban growth in southern California coastal shrublands.

Landscape Ecology 22:431–445.

Syphard, A. D., V. C. Radeloff, J. E. Keeley, T. J. Hawbaker, M. K.

Clayton, S. I. Stewart, and R. B. Hammer. 2007. Human influence on California fire regimes. Ecological Application 17:1388–1402.

Syphard, A. D., V. C. Radeloff, N. S. Keuler, R. S. Taylor, T. J. Hawbaker, S. I. Stewart, and M. K. Clayton. 2008. Predicting spatial patterns of fire on a southern California landscape. International Journal of Wildland Fire 17:602–613.

V´azquez de la Cueva, A., J. M. Garc´ıa del Barrio, M. O. Quero, and O. S.

Palomares. 2006. Recent fire regime in peninsular Spain in relation to forest potential productivity and population density. International Journal of Wildland Fire 15:397–405.

Viedma, O., J. M. Moreno, and I. Rieiro. 2006. Interactions between land use/land cover change, forest fires and landscape structure in Sierra de Gredos (Central Spain). Environmental Conservation 33:212–222.

Vogiatzakis, I. N., A. M. Mannion, and G. H. Griffiths. 2006. Mediterranean ecosystems: problems and tools for conservation. Progress in Physical Geography 30:175–200.

Wark, M. C., D. M. White, D. J. Robertson, and P. Marriott. 1987. Regeneration of heath and heath woodland in the north eastern Otway Ranges following the wildfire of February 1983. Proceedings of the Royal Society of Victoria 99:51–88.

Whelen, R. J. 1995. The ecology of fire. Cambridge University Press, Cambridge, United Kingdom.

Yang, J., H. S. He, S. R. Shifley, and E. J. Gustafson. 2007. Spatial patterns of modern period human-caused fire occurrence in the Missouri Ozark Highlands. Forest Science 53:1–15.

Conservation Biology
Volume 23, No. 3, 2009

DATE COUNTY ACRES STRUCTURES DEATHS
1 CAMP (Powerlines) November 2018 Butte 153,336 18,804 85 2 EATON (Under Investigation) * January 2025 Los Angeles 14,021 9,413 17 3 PALISADES (Under Investigation) * January 2025 Los Angeles 23,707 6,833 12 4 TUBBS (Electrical) October 2017 Napa & Sonoma 36,807 5,636 22 5 TUNNEL - Oakland Hills (Rekindle) October 1991 Alameda 1,600 2,900 25 6 CEDAR (Human Related) October 2003 San Diego 273,246 2,820 15 7 NORTH COMPLEX (Lightning) August, 2020 Butte, Plumas, & Yuba 318,935 2,352 15 8 VALLEY (Electrical) September 2015 Lake, Napa & Sonoma 76,067 1,955 4 9 WITCH (Powerlines) October 2007 San Diego 197,990 1,650 2 10 WOOLSEY (Electrical) November 2018 Ventura 96,949 1,643 3 11 CARR (Human Related) July 2018 Shasta County, Trinity 229,651 1,614 8 12 GLASS (Undetermined ) September 2020 Napa & Sonoma 67,484 1,520 0 13 LNU LIGHTNING COMPLEX
(Lightning/Arson) August 2020 Napa, Solano, Sonoma, Yolo, Lake, & Colusa 363,220 1,491 6
14 CZU LIGHTNING COMPLEX (Lightning) August 2020 Santa Cruz, San Mateo 86,509 1,490 1 15 NUNS (Powerline) October 2017 Sonoma 54,382 1,355 3
16 DIXIE (Powerline) July 2021 Butte, Plumas, Lassen, & Tehama 963,309 1,311 1 17 THOMAS (Powerline) December 2017 Ventura & Santa Barbara 281,893 1,063 2 18 CALDOR (Under Investigation) September 2021 Alpine, Amador, & El Dorado 221,774 1,003 1 19 OLD (Human Related) October 2003 San Bernardino 91,281 1,003 6 20 JONES (Undetermined) October 1999 Shasta 26,200 954 1 2/7/2025
Top 20 Most Destructive California Wildfires

FIRE NAME (CAUSE)
"Structures" include homes, outbuildings (barns, garages, sheds, etc) and commercial properties destroyed.

This list does not include fire jurisdiction. These are the Top 20 regardless of whether they were state, federal, local or tribal responsibility. *Numbers not final *DINS Disclaimer: These numbers are preliminary based on aerial assessments dedicating heat sources which can include chicken coops, outbuildings, sheds, water containers, etc. *Validated inspections are currently being ground-verified by Damage Assessment Teams. 







 !"#$
!%
&
'!() *
+

),
 )
)
-.
)
,/)
 /
,

,0
)1
)2
-
,*),/

,

),

.)/

3 4
,
5,
2,)
/)

,6)/
6)

2)
)/
!
"(789 !:

'3
;,
)<=*

2)0.

),2 
0>

,?! !(2
,
),@ 0

 
,
)
3 



0-
*)
,
),A)B)
-2)'C!"

D!E




-FG )/3
/)
)


) /
,

-
,
,.

),2)/
,
H)
A
3 I)

,)
4
.
/



)

 +
*
/,

,

,0
 
,
3


! '#CJ!:
&D(

$#
').,

, /**,K
-

)  *
,
)

,
/

*)
+


*)-0*,


),
)
5,
H
,
-3
LMNOPQRSN TU VWXYZ




 


 



 




  



 

 

 
 
 





 
!


  "
#








 



  $

%

  









 &
 &




' $





(
)





 

 






$




( 



 
* 

(  

$
$

+ 










' $


 

  ,
-





















  




.
/*

0






 
 
 

 





 
  










 

 
 



 




 

 



 
 




 
 





 










 
 


  

! 



 

 

 

 
""
 

 



 

 
 


#






 $%&'()*)+, -


.

/




 


 
 



 



0 




 
 







 




 1,234)(526785942







:
:

;<
 
 




 




0
 





 

=
  

 












 

 
 

 



  



>?@ABCDE @FGHIJKLBMKAIJ@ LN?OPKQCRSTUVMWNVX