032123-C-8 Corresp. Elkins.pdf — Long Beach City Council, March 21, 2023 (part 3 of 3)

Gene ID Gene Log2(fold_change) q_value
ENSDARG00000011404 fen1 -0.63 1.01E-02
ENSDARG00000039345 drg1 -0.63 7.20E-03
ENSDARG00000043276 calr -0.63 1.42E-03
ENSDARG00000086370 apoea -0.63 2.27E-03
ENSDARG00000020956 pck2 -0.63 2.04E-03
ENSDARG00000009001 pdip5 -0.64 9.24E-04
ENSDARG00000071212 lepre1 -0.64 2.85E-03
ENSDARG00000056767 itgb3a -0.65 3.01E-03
ENSDARG00000009401 vcanb -0.65 2.81E-03
ENSDARG00000014091 osr1 -0.66 3.62E-02
ENSDARG00000005023 fkbp9 -0.66 2.87E-02
ENSDARG00000010962 fkbp7 -0.66 4.09E-02
ENSDARG00000042021 mapk12a -0.66 2.19E-02
ENSDARG00000038768 mrpl12 -0.66 6.56E-03
ENSDARG00000010487 sae1 -0.66 5.24E-03
ENSDARG00000092467 si:ch73-46j18.5 -0.66 2.88E-03
ENSDARG00000057683 mcm6 -0.67 6.51E-04
ENSDARG00000018846 dgat2 -0.67 2.42E-02
ENSDARG00000037158 rcc1 -0.67 3.94E-02
ENSDARG00000001993 myhb -0.67 3.74E-02
ENSDARG00000015088 dnajb11 -0.68 2.19E-02
ENSDARG00000055945 asph -0.68 2.19E-02
ENSDARG00000037846 hm13 -0.68 6.32E-04
ENSDARG00000005161 gpib -0.68 2.79E-03
ENSDARG00000020103 calrl -0.68 1.96E-04
ENSDARG00000033855 rqcd1 -0.69 1.54E-02
ENSDARG00000015911 mcm2 -0.69 1.45E-03
ENSDARG00000021004 c5 -0.69 6.70E-03
ENSDARG00000019507 mcm5 -0.69 1.12E-03
ENSDARG00000000796 nr4a1 -0.69 1.32E-02
ENSDARG00000037961 rcn3 -0.70 1.14E-03
ENSDARG00000074908 col6a1 -0.70 1.67E-04
ENSDARG00000027495 elovl4b -0.70 2.19E-02
ENSDARG00000039131 atp1a1a.3 -0.71 1.68E-02
ENSDARG00000010640 SLC12A2 (6 of 6) -0.71 4.65E-03
ENSDARG00000039041 sfrp5 -0.71 4.00E-02
ENSDARG00000044975 KRT23 (9 of 15) -0.72 2.88E-03
ENSDARG00000044261 si:ch211-243g18.2 -0.72 2.51E-02
ENSDARG00000040535 CSGALNACT1 (1 of 2) -0.73 3.93E-02
ENSDARG00000079233 E2F2 -0.73 2.67E-02
ENSDARG00000002968 a1cf -0.73 3.38E-02
ENSDARG00000075016 APOB (3 of 3) -0.73 5.84E-03
ENSDARG00000012016 HPGD -0.73 2.91E-02
ENSDARG00000069116 timm10 -0.73 1.16E-02
ENSDARG00000092155 apoc2 -0.73 3.84E-04
ENSDARG00000060797 pfkmb -0.74 8.38E-04
ENSDARG00000002071 adss -0.74 6.29E-04
ENSDARG00000062688 gpnmb -0.74 3.49E-03
ENSDARG00000069415 col17a1a -0.75 9.61E-05
157

Gene ID Gene Log2(fold_change) q_value
ENSDARG00000055493 hic1 -0.75 4.05E-02
ENSDARG00000069980 lman1 -0.75 5.98E-03
ENSDARG00000015495 klf3 -0.76 2.61E-04
ENSDARG00000002831 col4a4 -0.77 4.57E-02
ENSDARG00000071219 pik3r3a -0.77 3.11E-03
ENSDARG00000038153 lgals2b -0.77 4.62E-02
ENSDARG00000014594 anxa1b -0.77 6.94E-04
ENSDARG00000069823 PROCA1 -0.78 1.16E-02
ENSDARG00000057738 hells -0.78 6.94E-04
ENSDARG00000024928 ITIH4 -0.78 9.58E-03
ENSDARG00000035652 sat1a -0.78 1.37E-02
ENSDARG00000007221 pbk -0.79 4.10E-02
ENSDARG00000063177 manf -0.79 3.48E-03
ENSDARG00000056778 cfhl2 -0.79 2.78E-02
ENSDARG00000069048 serpinf1 -0.79 1.53E-04
ENSDARG00000012366 fbp2 -0.79 4.34E-03
ENSDARG00000079370 utp18 -0.79 3.40E-02
ENSDARG00000052039 caspb -0.80 2.38E-03
ENSDARG00000069261 metap2a -0.80 4.35E-02
ENSDARG00000055278 cfb -0.80 1.93E-05
ENSDARG00000069706 prmt6 -0.81 4.73E-02
ENSDARG00000091397 - -0.82 1.99E-02
ENSDARG00000018258 ADK (2 of 2) -0.82 9.14E-03
ENSDARG00000076624 ptprb -0.82 2.12E-02
ENSDARG00000027867 PAPLN (1 of 2) -0.82 8.10E-03
ENSDARG00000007377 odc1 -0.83 1.24E-03
ENSDARG00000018266 mthfd1a -0.83 5.67E-03
ENSDARG00000052470 igfbp2a -0.83 3.72E-04
ENSDARG00000090468 PPP1R3A (2 of 2) -0.83 4.34E-03
ENSDARG00000090802 MCM3 (2 of 2) -0.83 1.77E-02
ENSDARG00000033140 desi1a -0.84 6.28E-04
ENSDARG00000042641 cyp51 -0.84 1.19E-02
ENSDARG00000004665 hspa5 -0.84 1.86E-06
ENSDARG00000045141 aqp8a.1 -0.84 4.61E-02
ENSDARG00000029075 pfkfb4l -0.84 1.87E-04
ENSDARG00000056938 kera -0.85 5.80E-06
ENSDARG00000045843 apex1 -0.85 1.59E-03
ENSDARG00000073928 mrc1a -0.85 1.48E-04
ENSDARG00000009782 myh11a -0.85 3.55E-02
ENSDARG00000035891 acana -0.85 3.45E-02
ENSDARG00000041110 dnajc3 -0.87 3.51E-02
ENSDARG00000004282 zgc:77375 -0.88 2.07E-02
ENSDARG00000029204 TYRP1 (2 of 2) -0.88 1.87E-05
ENSDARG00000052738 hmgcs1 -0.89 3.84E-04
ENSDARG00000018491 pdia4 -0.90 8.75E-08
ENSDARG00000001975 hsd11b2 -0.90 3.62E-04
ENSDARG00000003820 nr1d2a -0.90 3.49E-02
ENSDARG00000057575 pnp4a -0.90 2.28E-04
ENSDARG00000003570 hsp90b1 -0.91 2.70E-07
158

Gene ID Gene Log2(fold_change) q_value
ENSDARG00000014488 ca2 -0.92 2.62E-08
ENSDARG00000090467 CABZ01074130.1 -0.92 2.53E-08
ENSDARG00000016491 aglb -0.93 2.23E-06
ENSDARG00000016718 mmp11b -0.94 4.33E-02
ENSDARG00000073699 - -0.94 2.44E-08
ENSDARG00000079111 zgc:86725 -0.95 4.68E-04
ENSDARG00000013670 hyou1 -0.96 1.50E-02
ENSDARG00000060345 apod -0.96 1.78E-04
ENSDARG00000069827 crygm2d11 -0.97 1.88E-04
ENSDARG00000041595 ces3 -0.98 2.37E-05
ENSDARG00000070597 prelp -0.98 7.57E-03
ENSDARG00000018351 hpda -0.98 1.79E-02
ENSDARG00000055172 BX470254.2 -0.98 1.04E-07
ENSDARG00000055100 cxcl12b -0.99 1.30E-05
ENSDARG00000053853 slc13a2 -1.00 2.37E-05
ENSDARG00000076357 CABZ01102039.1 -1.01 3.40E-02
ENSDARG00000010478 hsp90aa1.1 -1.03 2.90E-10
ENSDARG00000090268 KRT23 (12 of 15) -1.03 4.81E-08
ENSDARG00000039462 CABZ01076094.1 -1.03 2.73E-08
ENSDARG00000042780 APOB (1 of 3) -1.03 8.62E-07
ENSDARG00000026771 tmem41ab -1.04 3.45E-02
ENSDARG00000052917 im:7154842 -1.04 3.51E-09
ENSDARG00000091260 MYLK4 (1 of 2) -1.04 1.63E-04
ENSDARG00000035914 tmem167a -1.05 3.05E-06
ENSDARG00000038785 abcf2a -1.06 8.08E-10
ENSDARG00000090623 CR392352.1 -1.07 5.51E-03
ENSDARG00000007480 rpe65a -1.07 2.48E-05
ENSDARG00000036840 krt15 -1.09 1.68E-05
ENSDARG00000069988 ARID5A (2 of 2) -1.09 4.51E-02
ENSDARG00000043806 postna -1.10 6.50E-04
ENSDARG00000045180 acta2 -1.11 3.40E-02
ENSDARG00000039605 mat1a -1.11 2.39E-12
ENSDARG00000037278 lrata -1.11 2.81E-02
ENSDARG00000075161 defbl1 -1.13 4.37E-08
ENSDARG00000093774 rbp2b -1.14 3.05E-06
ENSDARG00000030215 matn1 -1.14 1.05E-04
ENSDARG00000045808 rlbp1b -1.16 2.91E-02
ENSDARG00000054753 col10a1 -1.17 2.44E-11
ENSDARG00000057064 enpep -1.20 2.46E-02
ENSDARG00000023324 rab11bb -1.23 7.79E-03
ENSDARG00000001760 tnxb -1.23 2.55E-09
ENSDARG00000095321 si:dkey-9l20.3 -1.26 0.00E+00
ENSDARG00000038424 C4A -1.26 1.62E-07
ENSDARG00000035309 entpd3 -1.28 2.33E-09
ENSDARG00000055118 mylipb -1.32 1.63E-02
ENSDARG00000077872 CR626907.1 -1.33 2.88E-03
ENSDARG00000095239 si:dkeyp-106c3.1 -1.35 2.28E-02
ENSDARG00000009018 rhbg -1.36 0.00E+00
ENSDARG00000076874 abhd5 -1.37 1.33E-02
159

Gene ID Gene Log2(fold_change) q_value
ENSDARG00000016412 agt -1.37 2.44E-08
ENSDARG00000079302 and2 -1.40 0.00E+00
ENSDARG00000076192 ankrd1b -1.43 2.23E-10
ENSDARG00000041685 BX663520.1 -1.47 1.07E-03
ENSDARG00000087345 CABZ01059415.2 -1.57 7.28E-07
ENSDARG00000088636 wu:fa03e10 -1.64 0.00E+00
ENSDARG00000052631 thbs4a -1.69 1.61E-07
ENSDARG00000030604 phkg1a -1.70 2.90E-03
ENSDARG00000062132 cyp4v8 -1.76 2.64E-02
ENSDARG00000052207 c3c -1.76 0.00E+00
ENSDARG00000067639 prpf4 -1.78 7.68E-04
ENSDARG00000002311 fabp11b -1.81 0.00E+00
ENSDARG00000079933 SLC46A3 (1 of 2) -1.84 6.91E-05
ENSDARG00000031952 mb -1.84 0.00E+00
ENSDARG00000035544 agxt2l1 -1.91 1.42E-02
ENSDARG00000042379 zgc:103681 -2.02 4.35E-02
ENSDARG00000056875 rgs2 -2.19 1.28E-02
ENSDARG00000071173 slc12a10.2 -2.22 1.48E-04
ENSDARG00000087359 c3b -2.26 1.58E-04
ENSDARG00000056587 cyp2r1 -2.29 2.15E-03
ENSDARG00000007024 uox -2.94 4.67E-02
ENSDARG00000068194 klf9 -3.13 4.33E-06
ENSDARG00000088589 ponzr3 -3.41 7.93E-10
ENSDARG00000028396 fkbp5 -4.23 0.00E+00
ENSDARG00000087440 ponzr4 -4.42 3.80E-09

160

Table 4-S4 BEZO significantly misexpressed transcripts
Zebrafish genome assembly Zv9 transcripts identified by Cuffdiff as significantly differentially expressed in BEZO-exposed embryos compared to control (1% FDR, q < 0.05). Log 2 fold change values represent mean BEZO-exposed compared to control.

Gene ID Gene Log2(fold_change) q_value
ENSDARG00000069375 zgc:162608 1.94 9.06E-09
ENSDARG00000014947 igfbp1a 1.91 0.00E+00
ENSDARG00000026039 cyp1a 1.83 1.27E-12
ENSDARG00000086047 CABZ01067657.1 1.72 0.00E+00
ENSDARG00000061481 CABZ01041812.1 1.72 0.00E+00
ENSDARG00000074971 DHRS13 (2 of 5) 1.68 1.43E-05
ENSDARG00000003902 ctsl.1 1.63 2.46E-02
ENSDARG00000044685 nr0b2a 1.53 1.36E-13
ENSDARG00000038025 cbx7a 1.49 1.21E-04
ENSDARG00000076221 zgc:198419 1.43 1.09E-02
ENSDARG00000007344 tcap 1.42 2.93E-12
ENSDARG00000091047 CABZ01045617.2 1.39 6.79E-08
ENSDARG00000070012 sesn2 1.36 5.73E-08
ENSDARG00000075524 CYLC2 1.35 0.00E+00
ENSDARG00000019236 gsr 1.33 1.45E-07
ENSDARG00000005713 ethe1 1.29 2.50E-07
ENSDARG00000037121 mat2ab 1.26 6.22E-06
ENSDARG00000023217 CREM 1.25 4.36E-04
ENSDARG00000005039 gstp1 1.25 0.00E+00
ENSDARG00000007823 atf3 1.23 6.38E-09
ENSDARG00000061120 slc43a2b 1.23 0.00E+00
ENSDARG00000061634 zgc:158614 1.22 2.65E-06
ENSDARG00000021149 cbr1l 1.21 0.00E+00
ENSDARG00000058734 prdx1 1.21 0.00E+00
ENSDARG00000089697 si:dkey-3d18.4 1.18 4.96E-02
ENSDARG00000094719 si:dkeyp-1h4.9 1.18 2.65E-06
ENSDARG00000057633 cxcr4a 1.18 2.51E-03
ENSDARG00000030872 cetp 1.17 2.12E-05
ENSDARG00000059914 SDR42E2 1.15 6.40E-09
ENSDARG00000078878 METTL21C (1 of 2) 1.14 2.06E-04
ENSDARG00000058980 cyp1c1 1.11 7.89E-11
ENSDARG00000051956 isca1 1.11 5.54E-10
ENSDARG00000060189 ESM1 1.09 4.13E-02
ENSDARG00000055974 TPMT (1 of 2) 1.07 1.12E-02
ENSDARG00000067857 pmt 1.04 5.14E-04
ENSDARG00000056638 pir 1.03 1.15E-04
ENSDARG00000000551 slc1a4 1.03 4.71E-12
ENSDARG00000091871 si:ch211-13o20.3 1.00 2.35E-06
ENSDARG00000039269 arg2 0.99 5.73E-08
ENSDARG00000031683 fos 0.98 4.03E-03
ENSDARG00000045708 adm2a 0.98 8.88E-03
ENSDARG00000002405 si:ch211-225b11.1 0.97 7.03E-04
ENSDARG00000056680 stc2a 0.96 8.50E-08
ENSDARG00000058476 stc1l 0.94 2.70E-02
ENSDARG00000052694 micall2a 0.94 1.55E-03
161

Gene ID Gene Log2(fold_change) q_value
ENSDARG00000016457 irf9 0.93 2.62E-03
ENSDARG00000071567 TSTD1 0.91 9.14E-03
ENSDARG00000055723 hsp70l 0.90 4.12E-04
ENSDARG00000078882 CR407583.2 0.89 9.16E-06
ENSDARG00000077559 NCOA7 (2 of 2) 0.89 4.17E-03
ENSDARG00000061081 arpp21 0.88 2.90E-03
ENSDARG00000021242 mvp 0.88 6.92E-08
ENSDARG00000079938 zgc:173594 0.88 4.80E-02
ENSDARG00000076386 epdl1 0.88 3.93E-03
ENSDARG00000070020 cyp2aa9 0.88 7.92E-05
ENSDARG00000031776 zgc:92066 0.85 1.49E-06
ENSDARG00000020645 slc7a3a 0.85 5.98E-08
ENSDARG00000086826 sult6b1 0.83 1.04E-07
ENSDARG00000074829 RASIP1 0.83 4.21E-02
ENSDARG00000091111 TIFA 0.83 4.69E-04
ENSDARG00000033539 paics 0.83 7.89E-11
ENSDARG00000077785 ATF5 (2 of 2) 0.81 1.59E-05
ENSDARG00000075014 sqstm1 0.80 2.15E-04
ENSDARG00000073786 cmbl 0.80 4.80E-03
ENSDARG00000052148 ptgs1 0.80 1.64E-02
ENSDARG00000057121 C7 (1 of 2) 0.79 2.19E-02
ENSDARG00000019274 rasd1 0.79 8.98E-05
ENSDARG00000054058 h1fx 0.79 1.22E-11
ENSDARG00000042725 cebpb 0.78 9.16E-06
ENSDARG00000035602 dao.1 0.78 2.70E-03
ENSDARG00000025338 hagh 0.77 1.26E-05
ENSDARG00000093156 si:ch73-21g5.7 0.76 8.39E-03
ENSDARG00000040907 gcgb 0.76 1.85E-02
ENSDARG00000030905 cited2 0.73 1.05E-04
ENSDARG00000092337 gas5 0.73 1.41E-05
ENSDARG00000094557 nupr1 0.72 4.72E-04
ENSDARG00000026359 PBLD (2 of 2) 0.70 1.40E-02
ENSDARG00000091350 MYO18B 0.69 3.57E-02
ENSDARG00000045568 bcat1 0.69 3.77E-02
ENSDARG00000045976 sidt2 0.68 3.41E-04
ENSDARG00000027529 hmox1 0.68 3.18E-04
ENSDARG00000032496 zgc:91887 0.68 2.40E-02
ENSDARG00000070000 txnipb 0.68 1.89E-03
ENSDARG00000060113 znf395a 0.66 1.40E-04
ENSDARG00000001873 phgdh 0.65 1.93E-06
ENSDARG00000056379 si:ch211-154o6.6 0.65 1.85E-02
ENSDARG00000060315 si:dkey-193b15.6 0.65 1.86E-02
ENSDARG00000035559 tp53 0.64 2.83E-02
ENSDARG00000026611 socs3b 0.64 1.81E-02
ENSDARG00000038559 h1f0 0.63 7.48E-08
ENSDARG00000068096 ATF5 (1 of 2) 0.63 1.47E-03
ENSDARG00000035519 histh1l 0.63 9.62E-08
ENSDARG00000090552 wu:fb12e11 0.62 9.72E-03
ENSDARG00000059202 TSPAN2 (2 of 2) 0.62 1.53E-02
ENSDARG00000020742 NAGA 0.61 2.59E-02
162

Gene ID Gene Log2(fold_change) q_value
ENSDARG00000021787 abcb5 0.61 7.61E-05
ENSDARG00000088885 PHGR1 0.61 1.02E-02
ENSDARG00000006019 tkt 0.60 5.66E-04
ENSDARG00000042874 phlda2 0.59 2.59E-02
ENSDARG00000070426 chac1 0.59 1.14E-06
ENSDARG00000016733 psat1 0.58 7.23E-06
ENSDARG00000035890 fuca1 0.58 1.17E-04
ENSDARG00000076838 APOM 0.57 6.17E-03
ENSDARG00000095147 KRT23 (2 of 15) 0.57 4.93E-02
ENSDARG00000033666 pi4k2a 0.56 1.03E-02
ENSDARG00000038147 hbbe3 0.56 3.51E-03
ENSDARG00000017388 gstt1b 0.56 1.82E-02
ENSDARG00000091816 CABZ01088039.1 0.56 3.58E-03
ENSDARG00000029011 xpnpep1 0.55 1.56E-02
ENSDARG00000052705 pkp1 0.55 3.21E-02
ENSDARG00000067652 im:7150988 0.54 1.92E-03
ENSDARG00000028618 KRT18 (2 of 3) 0.54 1.54E-04
ENSDARG00000041607 eif4ebp3l 0.53 1.26E-02
ENSDARG00000020232 eif6 0.53 4.29E-02
ENSDARG00000070047 rgs4 0.53 1.22E-03
ENSDARG00000018566 CU302436.1 0.53 1.78E-04
ENSDARG00000012987 gpia 0.53 9.23E-04
ENSDARG00000015164 mknk2b 0.52 3.54E-04
ENSDARG00000041394 dnajb1b 0.51 2.59E-02
ENSDARG00000007955 iars 0.51 5.19E-04
ENSDARG00000010312 cp 0.50 2.37E-04
ENSDARG00000016319 c9 0.50 1.02E-02
ENSDARG00000013968 psap 0.49 1.05E-04
ENSDARG00000036107 txnipa 0.49 1.49E-04
ENSDARG00000028957 maff 0.49 3.41E-02
ENSDARG00000094300 si:ch211-160e1.5 0.49 1.92E-02
ENSDARG00000016200 trib3 0.48 2.52E-02
ENSDARG00000056395 onecut3 0.48 2.13E-02
ENSDARG00000013430 bhmt 0.48 3.62E-03
ENSDARG00000089920 MLIP 0.48 1.22E-03
ENSDARG00000076667 ccng1 0.47 1.24E-03
ENSDARG00000017960 sfxn2 0.46 1.96E-02
ENSDARG00000054030 hoxb5b 0.46 2.17E-02
ENSDARG00000010946 cbsb 0.45 2.66E-02
ENSDARG00000069142 aars 0.45 3.18E-03
ENSDARG00000037910 FILIP1L (2 of 2) 0.44 3.98E-02
ENSDARG00000033609 map1lc3a 0.44 3.75E-02
ENSDARG00000071082 p4ha1b 0.44 1.13E-02
ENSDARG00000091996 si:ch211-117m20.5 0.43 2.01E-03
ENSDARG00000017180 npc1 0.42 3.47E-03
ENSDARG00000042934 ctgfa 0.41 1.49E-02
ENSDARG00000037618 ddit4 0.41 4.21E-02
ENSDARG00000076241 txlnbb 0.41 2.27E-02
ENSDARG00000044125 txn 0.38 4.93E-02
ENSDARG00000061100 nars 0.36 4.21E-02
163

Gene ID Gene Log2(fold_change) q_value
ENSDARG00000057698 ctsd 0.36 3.48E-02
ENSDARG00000058656 desma 0.35 4.67E-02
ENSDARG00000079745 si:ch211-166a6.5 0.34 3.72E-02
ENSDARG00000041811 rps25 0.34 4.95E-02
ENSDARG00000026726 anxa1a 0.33 4.81E-02
ENSDARG00000042245 MYL4 -0.34 4.67E-02
ENSDARG00000057052 DSCAML1 -0.34 4.29E-02
ENSDARG00000018259 atp1a3a -0.34 4.83E-02
ENSDARG00000019566 neurod -0.35 3.21E-02
ENSDARG00000055216 tuba1l -0.35 2.82E-02
ENSDARG00000009001 pdip5 -0.35 4.21E-02
ENSDARG00000074908 col6a1 -0.35 3.82E-02
ENSDARG00000017568 HNRNPAB (2 of 2) -0.36 4.37E-02
ENSDARG00000056151 tyrp1b -0.36 2.55E-02
ENSDARG00000073732 myh14 -0.36 4.58E-02
ENSDARG00000029058 rbb4 -0.36 3.10E-02
ENSDARG00000005551 hnrnph1l -0.37 2.96E-02
ENSDARG00000056725 hmgb3a -0.37 3.37E-02
ENSDARG00000019353 sparc -0.37 4.57E-02
ENSDARG00000037846 hm13 -0.37 4.64E-02
ENSDARG00000079772 hmgb1a -0.38 1.49E-02
ENSDARG00000090268 KRT23 (12 of 15) -0.38 3.91E-02
ENSDARG00000068507 crybb1 -0.39 4.82E-02
ENSDARG00000004665 hspa5 -0.40 7.57E-03
ENSDARG00000071353 AL929007.1 -0.40 1.04E-02
ENSDARG00000028524 col5a3b -0.40 5.37E-03
ENSDARG00000086222 NAT16 -0.40 1.81E-02
ENSDARG00000020103 calrl -0.40 5.12E-03
ENSDARG00000027355 slc25a4 -0.40 5.12E-03
ENSDARG00000013963 mipb -0.40 6.08E-03
ENSDARG00000016235 rbp1a -0.40 4.21E-02
ENSDARG00000003570 hsp90b1 -0.41 6.65E-03
ENSDARG00000088514 and1 -0.42 2.84E-03
ENSDARG00000054362 ccdc47 -0.42 4.46E-03
ENSDARG00000069737 pou4f2 -0.42 1.47E-02
ENSDARG00000031100 ivns1abpa -0.43 1.53E-02
ENSDARG00000063914 mt-nd3 -0.43 2.18E-03
ENSDARG00000008732 zgc:66479 -0.43 4.85E-03
ENSDARG00000058117 snap25b -0.43 3.78E-02
ENSDARG00000061124 srpr -0.43 1.60E-02
ENSDARG00000056292 vsx1 -0.44 1.45E-02
ENSDARG00000037284 ptges3a -0.44 1.45E-02
ENSDARG00000062688 gpnmb -0.44 4.90E-02
ENSDARG00000054807 sec13 -0.44 2.75E-02
ENSDARG00000076768 REPS2 -0.45 3.42E-02
ENSDARG00000018491 pdia4 -0.45 1.40E-03
ENSDARG00000011125 snrpb -0.45 3.08E-02
ENSDARG00000014179 pfkma -0.46 1.77E-03
ENSDARG00000037285 mipa -0.46 1.78E-03
ENSDARG00000039913 tmem147 -0.46 4.67E-02
164

Gene ID Gene Log2(fold_change) q_value
ENSDARG00000074169 GPAM -0.47 1.28E-02
ENSDARG00000030411 crygn2 -0.47 3.72E-04
ENSDARG00000090467 CABZ01074130.1 -0.48 2.73E-04
ENSDARG00000001889 tuba1l2 -0.48 2.55E-04
ENSDARG00000045843 apex1 -0.48 4.29E-02
ENSDARG00000092467 si:ch73-46j18.5 -0.48 7.23E-03
ENSDARG00000036840 krt15 -0.48 4.83E-02
ENSDARG00000012381 zgc:63663 -0.49 1.01E-03
ENSDARG00000018130 rhot1a -0.49 4.61E-02
ENSDARG00000056248 wu:fb15e04 -0.50 1.14E-04
ENSDARG00000045143 hbbe2 -0.50 4.25E-03
ENSDARG00000043257 ckbb -0.50 1.26E-02
ENSDARG00000073699 - -0.51 2.07E-04
ENSDARG00000071219 pik3r3a -0.51 3.64E-02
ENSDARG00000057738 hells -0.51 1.08E-02
ENSDARG00000002071 adss -0.51 3.61E-03
ENSDARG00000007576 crybb1l1 -0.51 1.03E-04
ENSDARG00000038056 FGFBP2 (1 of 2) -0.51 3.55E-04
ENSDARG00000042780 APOB (1 of 3) -0.51 2.40E-03
ENSDARG00000005161 gpib -0.52 4.36E-03
ENSDARG00000074752 hlfa -0.52 1.92E-03
ENSDARG00000063631 ch1073-291c23.1 -0.52 5.49E-03
ENSDARG00000053875 cryba1b -0.52 1.49E-04
ENSDARG00000001910 rorab -0.53 4.63E-04
ENSDARG00000042021 mapk12a -0.53 3.35E-02
ENSDARG00000018119 cox5ab -0.53 1.30E-02
ENSDARG00000053502 cryaa -0.53 4.55E-03
ENSDARG00000059357 SARNP -0.53 2.78E-04
ENSDARG00000044562 cycsb -0.54 2.37E-05
ENSDARG00000005643 gcat -0.54 4.60E-02
ENSDARG00000087765 si:ch211-212n6.17 -0.54 2.44E-04
ENSDARG00000057575 pnp4a -0.55 1.76E-02
ENSDARG00000031316 six6b -0.55 6.48E-03
ENSDARG00000086030 PCBP3 (2 of 2) -0.55 4.67E-02
ENSDARG00000038785 abcf2a -0.55 8.56E-05
ENSDARG00000090468 PPP1R3A (2 of 2) -0.56 4.64E-02
ENSDARG00000015495 klf3 -0.56 8.15E-04
ENSDARG00000037997 tubb5 -0.57 3.80E-05
ENSDARG00000036344 calb2b -0.57 2.31E-04
ENSDARG00000024548 cryba4 -0.58 2.77E-05
ENSDARG00000070386 KRTCAP2 -0.58 2.52E-02
ENSDARG00000061836 nfixb -0.59 2.33E-03
ENSDARG00000037588 bhlhe23 -0.59 4.51E-02
ENSDARG00000007697 fabp7a -0.60 1.38E-06
ENSDARG00000054804 anp32e -0.60 4.49E-04
ENSDARG00000044975 KRT23 (9 of 15) -0.61 1.30E-03
ENSDARG00000077341 PPP1R14C (1 of 2) -0.61 3.55E-02
ENSDARG00000007715 lgsn -0.61 2.13E-04
ENSDARG00000009401 vcanb -0.63 1.46E-05
ENSDARG00000036058 gnao1b -0.63 5.90E-05
165

Gene ID Gene Log2(fold_change) q_value
ENSDARG00000010717 chchd10 -0.63 1.66E-02
ENSDARG00000032929 cryba1l -0.64 7.06E-08
ENSDARG00000021720 COL4A6 -0.65 1.11E-05
ENSDARG00000039605 mat1a -0.65 4.51E-08
ENSDARG00000056938 kera -0.65 8.98E-05
ENSDARG00000093774 rbp2b -0.66 1.97E-03
ENSDARG00000033760 pmelb -0.66 3.90E-06
ENSDARG00000011166 cahz -0.66 4.72E-04
ENSDARG00000087324 crygm2d1 -0.66 1.72E-04
ENSDARG00000030349 cryba2a -0.66 5.73E-08
ENSDARG00000018846 dgat2 -0.67 6.40E-04
ENSDARG00000069415 col17a1a -0.67 1.21E-07
ENSDARG00000075161 defbl1 -0.67 1.12E-04
ENSDARG00000029689 TKT -0.67 1.05E-05
ENSDARG00000069823 PROCA1 -0.68 7.87E-03
ENSDARG00000032200 rgn -0.68 4.67E-02
ENSDARG00000095863 zgc:161979 -0.68 1.24E-08
ENSDARG00000014488 ca2 -0.69 7.78E-09
ENSDARG00000023181 pcp4l1 -0.69 1.96E-02
ENSDARG00000029019 epb41b -0.69 4.43E-03
ENSDARG00000086917 si:ch211-212n6.18 -0.71 1.14E-02
ENSDARG00000094760 si:dkey-125i10.3 -0.72 6.00E-03
ENSDARG00000004282 zgc:77375 -0.72 3.64E-02
ENSDARG00000076693 si:ch211-212n6.8 -0.74 5.29E-05
ENSDARG00000018258 ADK (2 of 2) -0.74 3.67E-03
ENSDARG00000012366 fbp2 -0.75 5.15E-04
ENSDARG00000014594 anxa1b -0.76 5.45E-06
ENSDARG00000011989 crx -0.76 4.71E-11
ENSDARG00000086658 si:ch211-212n6.16 -0.76 9.16E-06
ENSDARG00000079302 and2 -0.76 2.12E-11
ENSDARG00000040535 CSGALNACT1 (1 of 2) -0.77 3.95E-03
ENSDARG00000003820 nr1d2a -0.77 2.40E-03
ENSDARG00000052700 si:dkey-162b23.4 -0.77 2.33E-03
ENSDARG00000091260 MYLK4 (1 of 2) -0.78 1.15E-03
ENSDARG00000086281 zgc:112992 -0.78 1.37E-03
ENSDARG00000052039 caspb -0.78 1.56E-04
ENSDARG00000002193 rho -0.78 8.72E-04
ENSDARG00000074001 crygmxl2 -0.78 5.07E-10
ENSDARG00000045685 cntn1b -0.78 4.64E-02
ENSDARG00000002311 fabp11b -0.78 9.38E-05
ENSDARG00000019417 gadd45g -0.78 6.00E-03
ENSDARG00000090689 si:busm1-118j2.5 -0.79 2.41E-07
ENSDARG00000042641 cyp51 -0.79 3.32E-03
ENSDARG00000075270 CU896655.3 -0.80 3.42E-02
ENSDARG00000033140 desi1a -0.80 2.48E-05
ENSDARG00000060345 apod -0.82 9.09E-05
ENSDARG00000057206 nmt1b -0.82 3.28E-02
ENSDARG00000038643 alas2 -0.83 1.61E-08
ENSDARG00000016491 aglb -0.84 3.05E-08
ENSDARG00000087164 crygm2d4 -0.85 1.93E-06
166

Gene ID Gene Log2(fold_change) q_value
ENSDARG00000071488 AGL (3 of 3) -0.86 2.38E-02
ENSDARG00000052917 im:7154842 -0.86 5.59E-11
ENSDARG00000087390 hbbe1.1 -0.87 0.00E+00
ENSDARG00000015076 cx44.1 -0.87 2.78E-04
ENSDARG00000089963 hbbe1.1 -0.87 2.54E-13
ENSDARG00000043961 BX957322.1 -0.88 1.62E-04
ENSDARG00000087188 nfil3-6 -0.90 1.14E-03
ENSDARG00000088823 crygm2d3 -0.91 2.16E-05
ENSDARG00000045141 aqp8a.1 -0.91 6.26E-03
ENSDARG00000079305 hbae3 -0.94 2.54E-13
ENSDARG00000037371 dcun1d1 -0.94 3.88E-02
ENSDARG00000023537 ahr1b -0.97 5.06E-08
ENSDARG00000057460 crygm2d13 -0.98 1.22E-11
ENSDARG00000069792 crygm2d5 -0.98 1.92E-05
ENSDARG00000052631 thbs4a -0.98 3.93E-04
ENSDARG00000035309 entpd3 -0.98 1.27E-07
ENSDARG00000007655 crybb1l3 -0.98 2.41E-02
ENSDARG00000023082 krt1-19d -0.99 5.84E-03
ENSDARG00000069801 crygm2d12 -1.01 0.00E+00
ENSDARG00000030215 matn1 -1.03 7.50E-04
ENSDARG00000011640 syt5b -1.03 1.10E-10
ENSDARG00000067639 prpf4 -1.04 4.67E-02
ENSDARG00000001760 tnxb -1.04 2.09E-08
ENSDARG00000037921 gng13b -1.04 0.00E+00
ENSDARG00000045808 rlbp1b -1.06 2.41E-02
ENSDARG00000069988 ARID5A (2 of 2) -1.07 3.41E-02
ENSDARG00000078440 CCDC88A -1.08 1.70E-08
ENSDARG00000040321 rx2 -1.09 1.98E-03
ENSDARG00000092945 si:ch211-250g4.3 -1.11 0.00E+00
ENSDARG00000001976 si:ch211-13k12.1 -1.11 6.26E-03
ENSDARG00000037337 cnrip1b -1.12 4.05E-02
ENSDARG00000044212 CR385063.1 -1.12 0.00E+00
ENSDARG00000023324 rab11bb -1.13 8.54E-03
ENSDARG00000091148 zgc:162402 -1.13 7.16E-09
ENSDARG00000004358 gnb3a -1.14 0.00E+00
ENSDARG00000041382 si:dkey-283b15.2 -1.15 4.35E-03
ENSDARG00000088330 si:ch211-5k11.2 -1.16 2.09E-06
ENSDARG00000055118 mylipb -1.16 2.55E-02
ENSDARG00000068194 klf9 -1.16 1.05E-02
ENSDARG00000003991 fhl2b -1.18 3.42E-02
ENSDARG00000069451 cx50.5 -1.20 1.98E-03
ENSDARG00000087345 CABZ01059415.2 -1.20 9.59E-06
ENSDARG00000051981 STX3 (2 of 2) -1.20 8.23E-03
ENSDARG00000088636 wu:fa03e10 -1.21 0.00E+00
ENSDARG00000057629 slc30a8 -1.22 9.96E-07
ENSDARG00000093318 CRYGB -1.24 8.68E-05
ENSDARG00000031952 mb -1.25 0.00E+00
ENSDARG00000012504 rlbp1a -1.26 8.73E-10
ENSDARG00000076572 crygm2d7 -1.27 1.07E-12
ENSDARG00000033382 grifin -1.29 2.03E-02
167

Gene ID Gene Log2(fold_change) q_value
ENSDARG00000007788 atp2b1b -1.29 1.60E-02
ENSDARG00000062661 ABCA4 (1 of 2) -1.29 2.84E-05
ENSDARG00000058556 CR854881.1 -1.31 2.68E-09
ENSDARG00000073874 crygm2d6 -1.32 0.00E+00
ENSDARG00000086912 zgc:86723 -1.33 0.00E+00
ENSDARG00000076055 RPGRIP1 -1.33 1.26E-02
ENSDARG00000086360 RP1 (2 of 2) -1.33 1.94E-02
ENSDARG00000076192 ankrd1b -1.40 0.00E+00
ENSDARG00000035544 agxt2l1 -1.40 3.88E-02
ENSDARG00000069817 crygm2d8 -1.42 0.00E+00
ENSDARG00000027495 elovl4b -1.46 1.60E-11
ENSDARG00000028396 fkbp5 -1.50 0.00E+00
ENSDARG00000073750 crygm2d9 -1.50 9.85E-13
ENSDARG00000037656 C17H2orf71 -1.55 2.29E-02
ENSDARG00000007480 rpe65a -1.58 1.36E-13
ENSDARG00000019902 rcv1 -1.58 0.00E+00
ENSDARG00000069826 crygm2d15 -1.62 0.00E+00
ENSDARG00000012126 zgc:109965 -1.62 9.69E-12
ENSDARG00000038634 CCK (1 of 2) -1.68 1.58E-03
ENSDARG00000056511 arr3a -1.71 7.92E-08
ENSDARG00000069615 CKMT2 (1 of 2) -1.75 8.26E-05
ENSDARG00000075295 - -1.76 7.06E-06
ENSDARG00000088589 ponzr3 -1.78 1.83E-06
ENSDARG00000094990 si:dkey-91f15.1 -1.87 2.33E-02
ENSDARG00000052223 rcvrna -1.88 1.03E-07
ENSDARG00000038894 tmx3 -1.90 1.82E-07
ENSDARG00000087301 crygm2d14 -1.90 0.00E+00
ENSDARG00000076624 ptprb -1.93 1.22E-11
ENSDARG00000076790 si:ch211-55g3.6 -2.07 0.00E+00
ENSDARG00000088687 zgc:165347 -2.11 0.00E+00
ENSDARG00000027236 rs1 -2.11 4.49E-06
ENSDARG00000057427 SV2B (3 of 3) -2.21 7.83E-05
ENSDARG00000087440 ponzr4 -2.27 7.35E-10
ENSDARG00000045677 opn1sw1 -2.82 8.92E-03
ENSDARG00000094310 si:ch211-255g12.6 -2.86 0.00E+00
ENSDARG00000069827 crygm2d11 -3.05 5.59E-11
ENSDARG00000026855 cacna2d4a -3.10 3.22E-03
ENSDARG00000044861 opn1lw2 -3.43 2.06E-04

168

Chapter 5 - Discussion
Individual PAHs interact with biological systems in numerous ways, resultin g in complex responses in many organs and tissue types. To respond to exposures, organisms at all levels of biological complexity have the ability to detect, metabolize and excrete these multi - ringed structures that are ubiquitous in the environment. In ve rtebrates, the most well - known mediator of xenobiotic response is arguably the AHR, which can bind some PAH structures with high affinity. Metabolism and receptor affinity play large roles in the toxicity of AHR agonists. TCDD, one of the most toxic haloge nated hydrocarbons, binds the AHR with high affinity but is not readily metabolized. Compounds naturally present in plants, such as indole 3 -carbinol, are also AHR agonists but are more commonly thought of as beneficial components of the diet (Denison and Nagy 2003) . Research on PAHs both in aquatic systems and mammals has highlighted a need to better understand both the AHR - dependent and -independent mechanisms by which PAH structures exert toxicological effects, and the concentrations at which these biolo gical mechanisms become concerns for human and wildlife health.

Many studies have screened PAHs for their mutagenic activity, and PAHs are classified as carcinogenic or non-carcinogenic for human health risk assessment purposes. Because little is known abo ut other toxicity mechanisms, it is not currently possible to classify PAHs as “inflammatory” or “ion channel disrupters”, which would be necessary for predicting risk of these effects from exposure to complex PAH mixtures. The involvement of metabolism an d receptor binding in the toxicological mechanisms causes small difference in PAH structure to result in very different biological effects. The presence of a “bay” or “fjord” region in higher molecular weight PAHs affects DNA damage potential (Mattsson et al. 2009) . A recent study demonstrated that this is also true for non -genotoxic effects of methylanthracenes; 1 -methylanthracene (1 -MeA), which contains a bay -like region, was compared with 2-methylanthracene (2-MeA) for inflammatory related effects (Osgood et al. 2013) In an alveolar cell line, 1 -MeA induced inflammatory signaling associated with tumor promotion, including MAP Kinase signaling, p38 phosphorylation, and inhibited gap junctional intercellular communication, while 2 -MeA did not (Osgood et al. 2013).

Recognizing these differences is particularly important as we begin to assess the toxicity of 169

substituted PAHs. Associating molecular mechanisms with structural differences would greatly increase our predictive power for toxicological effects of thi s large family of compounds.

The objective of the research presented here was to create profiles of biological activities of diverse PAH structures during embryonic development. The zebrafish model can rapidly provide a rich in vivo dataset from which to t ease apart different biological mechanisms.

Morphology analysis can be used to assess toxicity potential with high sensitivity (Truong et al. 2011) . Whole genome analysis of mRNA expression via microarrays and RNA -seq provides a snapshot of the transcripti onal effects of exposure at a point in time. By comparing these snapshots between compounds, we identified common effects, biomarker genes, and proposed mechanisms for further investigation of PAH-mediated toxixity.

In chapter 2, we compared microarray mRN A profiles of three parent PAHs, dibenzothiophene (DBT), pyrene (PYR), and benz(a)anthracene (BAA). By analyzing early transcriptional changes at concentrations that eventually elicited biological effects, we identified genes that were differentially invol ved in toxicological effects of these PAHs. We also measured body burdens of PAHs at the exposure concentrations and time points employed in the microarray. This information was essential for discerning potential mechanism from body burden differences. PAH s are hydrophobic; they are readily absorbed by embryos but also adhere to plastic, complicating exposures. While the studies presented here were all conducted in glass vials, solubility likely played a substantial role in the biological effects observed f or the parent PAHs. Precipitation of PAH occurs for both PYR and BAA at the 25 μM concentration. PYR is slightly more soluble than BAA, and we expected some differences in body burden, but the extent to which this varied between compounds was surprising. The large difference in body burden between DBT and the 4 - ring parent PAHs was important for interpreting the expression data. By comparing log 2 FC values across all transcripts that were significantly induced by DBT or PYR, we found very similar patterns of expression. We found little evidence for distinct biological mechanisms between DBT and PYR, while BAA induced a very different profile of genes. BAA - induced genes were examined in further chapters. DBT and PYR, however, induced a large suite of inflammatory-related genes, which has been previously observed for PAHs in other model 170

systems. The exact mechanism of inflammation is not clear, and I observed no sign of a specific inflammatory site, using transgenic zebrafish with mpx -driven expression of GF P (neutrophils). I also observed a hyperactive phenotype elicited by PYR exposure. DBT also induced hyperactivity, but it decreased, while PYR -exposed embryos remained hyperactive for 5 hours (Appendix 2). Hyperactivity was observed at concentrations as lo w as 2.5 μM; future studies will investigate the mechanism that results in this behavioral phenotype.

The role of the AHR in mediating PAH -induced developmental toxicity is not yet well - defined. In Chapter 3, we characterized an ahr2 mutant zebrafish, whic h will be an important tool for defining the role of the AHR in both PAH -mediated toxicity as well as normal development. We determined that the TILLING-identified ahr2 mutant was indeed a functional AHR2 knockout, and was completely resistant to TCDD -induced toxicity. We additionally investigated the effects of an alternative ligand, leflunomide, in the ahr2hu3335 line. Interestingly, we found that leflunomide induced liver -specific Cyp1a in the mutant line, and found that the AHR isoform AHR1A was respons ible for mediating this induction.

While this study showed that AHR1A is a functional receptor, future work is needed to characterize the functions of this receptor and potential roles in mediating toxicity.

Building off of a screen of substituted PAHs, we investigated the role of AHR2 in mediating the developmental toxicity of two similarly -structured OPAHs, BEZO and 7,12 -B[a]AQ in Chapter 4. Intriguingly, we found that despite different CYP1A induction profiles, developmental effects of both BEZO and 7,1 2-B[a]AQ were mediated by AHR2. We conducted paired -end 50 bp RNA -seq to compare mRNA expression profiles of embryos exposed to the two compounds from 6 -48 hpf. Using the Tuxedo suite of software, we identified novel transcripts that are promising targets for future investigation of AHR - dependent toxicological pathways. By comparing BEZO and 7,12 -B[a]AQ transcriptional profiles, we found differences in groups of transcripts, such as eye -related genes, that suggest the AHR interacts with other transcription factors or coactivators to mediate the differential toxicological effects observed for these two compounds. Future studies using morpholino knockdown of other predicted interactors would be a promising way to determine their roles in the toxicity of BEZO and 7,12-B[a]AQ, as well as discover new AHR toxicity mechanisms.
171

We identified transcripts with common identifiers that could be compared across the microarray and RNA -seq platforms. By clustering transcripts by expression changes observed with all 5 PAHs , we identified genes that could be potential biomarkers of PAH - related toxicity. Ctsl.1, for example, was induced by BAA, BEZO, and 7,12 -B[a]AQ. Cathepsin L1 is a protease that influences blood pressure, and a recent study identified a polymorphic locus in its promoter, which was associated with high blood pressure in a human population (Mbewe-Campbell et al. 2012) . By examining the locus, the authors found it was located in an XRE, and determined that CTSL.1 is AHR -responsive. To our knowledge, the effec ts of xenobiotic exposure on CTSL.1 and blood pressure regulation have not yet been investigated. Chemokine receptor 4a ( cxcr4a) was induced mildly, but consistently, by all PAHs in these studies and would also be an interesting target for future investigation.

In the work presented here, we investigated global transcriptional responses to diverse PAH structures and began to identify expression patterns associated with PAH exposure.

Investigating PAH toxicity in AHR2 deficient zebrafish added additional ins ight into potential toxicity mechanisms. This work, along with other recent studies of both the AHR and PAHs, demonstrates that the AHR has functions well beyond xenobiotic -activated binding to the canonical AHR -responsive genes. Unraveling the pathways by which PAHs differentially interact with the receptor will be an engaging direction for future research.

References
Denison, M. S. and S. R. Nagy (2003). "Activation of the aryl hydrocarbon receptor by structurally diverse exogenous and endogenous chemicals." Annu Rev Pharmacol Toxicol 43: 309-334.

Jonsson, M. E., D. G. Franks, et al. (2009). "The tryptophan photoproduct 6-formylindolo[3,2- b]carbazole (FICZ) binds multiple AHRs and induces multiple CYP1 genes via AHR2 in zebrafish." Chem Biol Interact.

Mattsson, A., B. Jernstrom, et al. (2009). "H2AX phosphorylation in A549 cells induced by the bulky and stable DNA adducts of benzo[a]pyrene and dibenzo[a,l]pyrene diol epoxides." Chem Biol Interact 177(1): 40-47.

Mbewe-Campbell, N., Z. Wei, et al. (2012). "Genes and environment: novel, functional polymorphism in the human cathepsin L (CTSL1) promoter disrupts a xenobiotic response element (XRE) to alter transcription and blood pressure." J Hypertens.

Osgood, R. S., B. L. Upham, et al. (2013). "Polycyclic aromatic hydrocarbon-induced signaling events relevant to inflammation and tumorigenesis in lung cells are dependent on molecular structure." PLoS One 8(6): e65150.

Truong, L., S. L. Harper, et al. (2011). "Evaluation of embryotoxicity using the zebrafish model." Methods Mol Biol 691: 271-279.
172

Chapter 6 - Future Directions
In the studies present in this dissertation, we employed a comparative transcriptomics approach to identify global changes in mRNA expression that lead to developmental effects of PAH exposure in develop ing embryos. We identified clusters of genes and associated biological processes that are misexpressed in response to different PAH exposures, and further investigated the role of AHR2 in mediating responses. The rich datasets presented in this dissertatio n could lead to many future research directions. The ion channel disruption/inflammatory/NfKB-related processes disrupted by DBT and PYR exposure are complex, and attempted to characterize endpoints more specifically related to these processes in Appendix 2. Further studies of the PYR behavioral response, in comparison with other PAHs and in combination with potential antagonists are in progress in the laboratory.

Involvement of the AHR in toxicity pathways and normal development remains an exciting and complex area for research. With three AHR isoforms, zebrafish provide an opportunity to unravel differential functions that may be partitioned between the isoforms. We showed the AHR1A is a functional receptor, and can mediate CYP1A expression. Beyond ability to induce that downstream target, little is known about function of AHR1A or AHR1B.

Characterization of where they are expressed over the course of development, and their interactions with a variety of ligands could provide valuable information about diff erent functions of the AHR. These studies will be easier to carry out in the ahr2hu3335 line.

The need to outcross the ahr2hu3335 line has put many exciting studies of endogenous AHR2 functions on hold, but there is promise to further investigate how loss of AHR2 affects developmental processes with them in the future. Beyond this, determining global transcriptional responses to PAH exposure in the AHR2 mutant line would be an exciting direction for research. In Chapter 4, we investigated the dependence of select transcriptional changes on AHR2. This could be continued with additional genes, but provides only a piece-wise picture, and we have continually found the AHR to be involved in mechanisms in ways we have not predicted. Analysis across the genome woul d provide information about the entirety of processes the AHR is affecting. 173

From the RNA-seq analysis, we were able to identify misexpressed transcripts, and compare across PAHs. This dataset however, could be analyzed in many more interesting ways. As annotation of the zebrafish genome improves, promoter analysis and/or prediction of non - coding RNA targeting may be able to better predict expression regulation than our current analysis based on known interactions in mammals. Some novel transcripts were iden tified and discussed in this dissertation; many are present within the dataset that would be interesting to pursue for interaction with the AHR and involvement in zebrafish development processes. Annotation of the zebrafish genome has improved much over th e past 5 years, but there are many transcripts that remain un -annotated and could be important mediators of toxicity.

Comparison of all 5 PAH structures in these studies shed light on interesting patterns in expression. Better statistical comparisons acro ss multiple treatment groups may help to identify promising transcriptional differences between BEZO and 7,12 -B[a]AQ for further analysis. Comparing additional PAH structures, will of course add power to identify which structural differences may be respons ible for mediating groups of transcripts involved in different biological processes. Not discussed here are other samples collected in parallel with our BEZO and 7,12 -B[a]AQ samples, which included phenanthrene -quinone and environmental mixture exposed sam ples. Adding these to the analysis, and comparing patterns across the larger dataset will be an interesting future direction.

Finally, comparing the genes and responses identified here across model systems will be a valuable direction for future studies. Homologues of many genes identified here have similar functions in other species, and eventually could be developed into biomarkers of PAH effects relevant to human or wildlife populations. Determining first whether PAH exposure affects their expression in rodents or human cell lines remains to be determined, and an interesting area of future work.

174

Bibliography

(2010). R: A language and environment for statistical computing. R. D. C. Team. Vienna, Austria.

Abbott, B. D., J. E. Schmid, et al. (1999). "Adverse reproductive outcomes in the transgenic Ah receptor-deficient mouse." Toxicol Appl Pharmacol 155(1): 62-70.

Alexeyenko, A., D. M. Wassenberg, et al. (2010). "Dynamic zebrafish interactome reveals transcriptional mechanisms of dioxin toxicity." PLoS One 5(5): e10465.

Andreasen, E. A., M. E. Hahn, et al. (2002). "The zebrafish (Danio rerio) aryl hydrocarbon receptor type 1 is a novel vertebrate receptor." Mol Pharmacol 62(2): 234-249.

Andreasen, E. A., L. K. Mathew, et al. (2006). "Regenerative growth is impacted by TCDD:
gene expression analysis reveals extracellular matrix modulation." Toxicol Sci 92(1): 254-269.

Andreasen, E. A., J. M. Spitsbergen, et al. (2002). "Tissue-specific expression of AHR2, ARNT2, and CYP1A in zebrafish embryos and larvae: effects of developmental stage and 2,3,7,8-tetrachlorodibenzo-p-dioxin exposure." Toxicol Sci 68(2): 403-419.

Antkiewicz, D. S., R. E. Peterson, et al. (2006). "Blocking expression of AHR2 and ARNT1 in zebrafish larvae protects against cardiac toxicity of 2,3,7,8-tetrachlorodibenzo-p- dioxin." Toxicol Sci 94(1): 175-182.

Archuleta, M. M., G. L. Schieven, et al. (1993). "7,12-Dimethylbenz[a]anthracene activates protein-tyrosine kinases Fyn and Lck in the HPB-ALL human T-cell line and increases tyrosine phosphorylation of phospholipase C-gamma 1, formation of inositol 1,4,5-trisphosphate, and mobilization of intracellular calcium." Proc Natl Acad Sci U S A 90(13): 6105-6109.

Baba, T., J. Mimura, et al. (2001). "Structure and expression of the Ah receptor repressor gene." J Biol Chem 276(35): 33101-33110.

Baird, W. M., L. A. Hooven, et al. (2005). "Carcinogenic polycyclic aromatic hydrocarbon- DNA adducts and mechanism of action." Environ Mol Mutagen 45(2-3): 106-114.

Barker, C. W., J. B. Fagan, et al. (1994). "Down-regulation of P4501A1 and P4501A2 mRNA expression in isolated hepatocytes by oxidative stress." J Biol Chem 269(6): 3985- 3990.

Barron, M. G., M. G. Carls, et al. (2004). "Evaluation of fish early life-stage toxicity models of chronic embryonic exposures to complex polycyclic aromatic hydrocarbon mixtures." Toxicol Sci 78(1): 60-67.

Barron, M. G., R. Heintz, et al. (2004). "Relative potency of PAHs and heterocycles as aryl hydrocarbon receptor agonists in fish." Mar Environ Res 58(2-5): 95-100.

Beischlag, T. V., J. Luis Morales, et al. (2008). "The aryl hydrocarbon receptor complex and the control of gene expression." Crit Rev Eukaryot Gene Expr 18(3): 207-250.

Benjamini, Y. and Y. Hochberg (1995). "Controlling the False Discovery Rate - a Practical and Powerful Approach to Multiple Testing." Journal of the Royal Statistical Society Series B-Methodological 57(1): 289-300.

Billiard, S. M., M. E. Hahn, et al. (2002). "Binding of polycyclic aromatic hydrocarbons (PAHs) to teleost aryl hydrocarbon receptors (AHRs)." Comp Biochem Physiol B Biochem Mol Biol 133(1): 55-68.
175

Billiard, S. M., A. R. Timme-Laragy, et al. (2006). "The role of the aryl hydrocarbon receptor pathway in mediating synergistic developmental toxicity of polycyclic aromatic hydrocarbons to zebrafish." Toxicol Sci 92(2): 526-536.

Binder, R. L. and J. J. Stegeman (1984). "Microsomal electron transport and xenobiotic monooxygenase activities during the embryonic period of development in the killifish, Fundulus heteroclitus." Toxicol Appl Pharmacol 73(3): 432-443.

Bisson, W. H., D. C. Koch, et al. (2009). "Modeling of the aryl hydrocarbon receptor (AhR) ligand binding domain and its utility in virtual ligand screening to predict new AhR ligands." J Med Chem 52(18): 5635-5641.

Bock, K. W. (2012). "Ah receptor- and Nrf2-gene battery members: Modulators of quinone- mediated oxidative and endoplasmic reticulum stress." Biochem Pharmacol 83(7):
833-838.

Bostrom, C. E., P. Gerde, et al. (2002). "Cancer risk assessment, indicators, and guidelines for polycyclic aromatic hydrocarbons in the ambient air." Environ Health Perspect 110 Suppl 3: 451-488.

Brasier, A. R., M. Jamaluddin, et al. (2000). "Angiotensin II induces gene transcription through cell-type-dependent effects on the nuclear factor-kappaB (NF-kappaB) transcription factor." Mol Cell Biochem 212(1-2): 155-169.

Brasier, A. R. and J. Li (1996). "Mechanisms for inducible control of angiotensinogen gene transcription." Hypertension 27(3 Pt 2): 465-475.

Bugel, S. M., L. A. White, et al. (2010). "Impaired reproductive health of killifish (Fundulus heteroclitus) inhabiting Newark Bay, NJ, a chronically contaminated estuary." Aquat Toxicol 96(3): 182-193.

Burstyn, I., H. Kromhout, et al. (2005). "Polycyclic aromatic hydrocarbons and fatal ischemic heart disease." Epidemiology 16(6): 744-750.

Bussmann, J., S. A. Wolfe, et al. (2011). "Arterial-venous network formation during brain vascularization involves hemodynamic regulation of chemokine signaling." Development 138(9): 1717-1726.

Carney, S. A., J. Chen, et al. (2006). "Aryl hydrocarbon receptor activation produces heart- specific transcriptional and toxic responses in developing zebrafish." Mol Pharmacol 70(2): 549-561.

Carney, S. A., R. E. Peterson, et al. (2004). "2,3,7,8-Tetrachlorodibenzo-p-dioxin activation of the aryl hydrocarbon receptor/aryl hydrocarbon receptor nuclear translocator pathway causes developmental toxicity through a CYP1A-independent mechanism in zebrafish." Mol Pharmacol 66(3): 512-521.

Carney, S. A., A. L. Prasch, et al. (2006). "Understanding dioxin developmental toxicity using the zebrafish model." Birth Defects Res A Clin Mol Teratol 76(1): 7-18.

Cavalieri, E. L. and E. G. Rogan (1995). "Central role of radical cations in metabolic activation of polycyclic aromatic hydrocarbons." Xenobiotica 25(7): 677-688.

Chew, G. L., A. Pauli, et al. (2013). "Ribosome profiling reveals resemblance between long non-coding RNAs and 5' leaders of coding RNAs." Development 140(13): 2828- 2834.

Choi, H., W. Jedrychowski, et al. (2006). "International studies of prenatal exposure to polycyclic aromatic hydrocarbons and fetal growth." Environ Health Perspect 114(11): 1744-1750.

Ciganek, M., J. Neca, et al. (2004). "A combined chemical and bioassay analysis of traffic- emitted polycyclic aromatic hydrocarbons." Sci Total Environ 334-335: 141-148. 176

Collins, J. F., J. P. Brown, et al. (1998). "Potency equivalency factors for some polycyclic aromatic hydrocarbons and polycyclic aromatic hydrocarbon derivatives." Regul Toxicol Pharmacol 28(1): 45-54.

Couroucli, X. I., S. E. Welty, et al. (2002). "Regulation of pulmonary and hepatic cytochrome P4501A expression in the rat by hyperoxia: implications for hyperoxic lung injury." Mol Pharmacol 61(3): 507-515.

Cubbage, C. C. and P. M. Mabee (1996). "Development of the cranium and paired fins in the zebrafish Danio rerio (Ostariophysi, cyprinidae)." Journal of Morphology 229(2):
121-160.

Das, M., K. Garg, et al. (1994). "Attenuation of benzanthrone toxicity by ascorbic acid in guinea pigs." Fundam Appl Toxicol 22(3): 447-456.

Denison, M. S. and S. R. Nagy (2003). "Activation of the aryl hydrocarbon receptor by structurally diverse exogenous and endogenous chemicals." Annu Rev Pharmacol Toxicol 43: 309-334.

Detmar, J., M. Y. Rennie, et al. (2008). "Fetal growth restriction triggered by polycyclic aromatic hydrocarbons is associated with altered placental vasculature and AhR- dependent changes in cell death." Am J Physiol Endocrinol Metab 295(2): E519-530.

Dong, P. D., C. A. Munson, et al. (2007). "Fgf10 regulates hepatopancreatic ductal system patterning and differentiation." Nat Genet 39(3): 397-402.

Duarte-Salles, T., M. A. Mendez, et al. (2012). "Dietary benzo(a)pyrene and fetal growth:
effect modification by vitamin C intake and glutathione S-transferase P1 polymorphism." Environ Int 45: 1-8.

Durant, J. L., W. F. Busby, Jr., et al. (1996). "Human cell mutagenicity of oxygenated, nitrated and unsubstituted polycyclic aromatic hydrocarbons associated with urban aerosols." Mutat Res 371(3-4): 123-157.

Durante, W. (2013). "Role of arginase in vessel wall remodeling." Front Immunol 4: 111.

Dwivedi, N., M. Das, et al. (2001). "Role of biological antioxidants in benzanthrone toxicity." Archives of Toxicology 75(4): 221-226.

Ebert, A. M., G. L. Hume, et al. (2005). "Calcium extrusion is critical for cardiac morphogenesis and rhythm in embryonic zebrafish hearts." Proc Natl Acad Sci U S A 102(49): 17705-17710.

Eden, E., R. Navon, et al. (2009). "GOrilla: a tool for discovery and visualization of enriched GO terms in ranked gene lists." BMC Bioinformatics 10: 48.

Ema, M., N. Ohe, et al. (1994). "Dioxin binding activities of polymorphic forms of mouse and human arylhydrocarbon receptors." J Biol Chem 269(44): 27337-27343.

EPA, U. (2010). Development of a relative potency factor (RPF) approach for polycyclic aromatic hydrocarbon (PAH) mixtures (External review draft). U. S. E. P. Agency.

Washington, DC.

EPA, U. (2012). "Integrated Risk Information System." from http://www.epa.gov/iris/.

Evans, B. R., S. I. Karchner, et al. (2008). "Repression of aryl hydrocarbon receptor (AHR) signaling by AHR repressor: role of DNA binding and competition for AHR nuclear translocator." Mol Pharmacol 73(2): 387-398.

Fan, R., D. Wang, et al. (2012). "Preliminary study of children's exposure to PAHs and its association with 8-hydroxy-2'-deoxyguanosine in Guangzhou, China." Environ Int 42: 53-58.
177

Fernandez-Salguero, P., T. Pineau, et al. (1995). "Immune system impairment and hepatic fibrosis in mice lacking the dioxin-binding Ah receptor." Science 268(5211): 722- 726.

Fleming, C. R. and R. T. Di Giulio (2011). "The role of CYP1A inhibition in the embryotoxic interactions between hypoxia and polycyclic aromatic hydrocarbons (PAHs) and PAH mixtures in zebrafish (Danio rerio)." Ecotoxicology 20(6): 1300-1314.

Fukunaga, B. N. and O. Hankinson (1996). "Identification of a novel domain in the aryl hydrocarbon receptor required for DNA binding." J Biol Chem 271(7): 3743-3749.

Fukunaga, B. N., M. R. Probst, et al. (1995). "Identification of functional domains of the aryl hydrocarbon receptor." J Biol Chem 270(49): 29270-29278.

Gao, J., A. A. Voss, et al. (2005). "Ryanodine receptor-mediated rapid increase in intracellular calcium induced by 7,8-benzo(a)pyrene quinone in human and murine leukocytes." Toxicol Sci 87(2): 419-426.

Garner, L. V. and R. T. Di Giulio (2012). "Glutathione transferase pi class 2 (GSTp2) protects against the cardiac deformities caused by exposure to PAHs but not PCB-126 in zebrafish embryos." Comp Biochem Physiol C Toxicol Pharmacol 155(4): 573-579.

Gonzalez, F. J. and P. Fernandez-Salguero (1998). "The aryl hydrocarbon receptor: studies using the AHR-null mice." Drug Metab Dispos 26(12): 1194-1198.

Goodale, B. C., S. C. Tilton, et al. (2013). "Structurally distinct polycyclic aromatic hydrocarbons induce differential transcriptional responses in developing zebrafish." Toxicol Appl Pharmacol.

Gu, Y. Z., J. B. Hogenesch, et al. (2000). "The PAS superfamily: sensors of environmental and developmental signals." Annu Rev Pharmacol Toxicol 40: 519-561.

Guengerich, F. P. (2000). "Metabolism of chemical carcinogens." Carcinogenesis 21(3): 345- 351.

Gurbani, D., S. K. Bharti, et al. (2013). "Polycyclic aromatic hydrocarbons and their quinones modulate the metabolic profile and induce DNA damage in human alveolar and bronchiolar cells." Int J Hyg Environ Health.

Hahn, M. E. (1998). "The aryl hydrocarbon receptor: a comparative perspective." Comp Biochem Physiol C Pharmacol Toxicol Endocrinol 121(1-3): 23-53.

Hahn, M. E. (2002). "Aryl hydrocarbon receptors: diversity and evolution." Chem Biol Interact 141(1-2): 131-160.

Hahn, M. E. (2002). "Biomarkers and bioassays for detecting dioxin-like compounds in the marine environment." Sci Total Environ 289(1-3): 49-69.

Hahn, M. E., S. I. Karchner, et al. (2006). "Unexpected diversity of aryl hydrocarbon receptors in non-mammalian vertebrates: insights from comparative genomics." J Exp Zool A Comp Exp Biol 305(9): 693-706.

Hansch, C., Albert, L., Hoekman, D. (1995). Exploring QSAR: Volume 2: Hydrophobic, Electronic, and Steric Constants, American Chemical Society.

Hansen, A. M., L. Mathiesen, et al. (2008). "Urinary 1-hydroxypyrene (1-HP) in environmental and occupational studies--a review." Int J Hyg Environ Health 211(5- 6): 471-503.

Hecht, S. S., S. G. Carmella, et al. (2010). "Analysis of phenanthrene and benzo[a]pyrene tetraol enantiomers in human urine: relevance to the bay region diol epoxide hypothesis of benzo[a]pyrene carcinogenesis and to biomarker studies." Chem Res Toxicol 23(5): 900-908.
178

Hermann, A., R. Donato, et al. (2012). "S100 calcium binding proteins and ion channels." Front Pharmacol 3: 67.

Hernandez-Ochoa, I., B. N. Karman, et al. (2009). "The role of the aryl hydrocarbon receptor in the female reproductive system." Biochem Pharmacol 77(4): 547-559.

Hertz-Picciotto, I., H. Y. Park, et al. (2008). "Prenatal exposures to persistent and non- persistent organic compounds and effects on immune system development." Basic Clin Pharmacol Toxicol 102(2): 146-154.

Hofsteen, P., J. Plavicki, et al. (2013). "Sox9b is Required for Epicardium Formation and Plays a Role in TCDD-induced Heart Malformation in Zebrafish." Mol Pharmacol.

Hu, S. W., Y. J. Chan, et al. (2011). "Urinary levels of 1-hydroxypyrene in children residing near a coal-fired power plant." Environ Res 111(8): 1185-1191.

Huang da, W., B. T. Sherman, et al. (2009). "Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources." Nat Protoc 4(1): 44-57.

Huang, L., C. Wang, et al. (2012). "Benzo[a]pyrene exposure influences the cardiac development and the expression of cardiovascular relative genes in zebrafish (Daniorerio) embryos." Chemosphere 87(4): 369-375.

Hylland, K. (2006). "Polycyclic aromatic hydrocarbon (PAH) ecotoxicology in marine ecosystems." J Toxicol Environ Health A 69(1-2): 109-123.

Incardona, J. P., M. G. Carls, et al. (2009). "Cardiac arrhythmia is the primary response of embryonic Pacific herring (Clupea pallasi) exposed to crude oil during weathering." Environ Sci Technol 43(1): 201-207.

Incardona, J. P., M. G. Carls, et al. (2005). "Aryl hydrocarbon receptor-independent toxicity of weathered crude oil during fish development." Environ Health Perspect 113(12):
1755-1762.

Incardona, J. P., T. K. Collier, et al. (2004). "Defects in cardiac function precede morphological abnormalities in fish embryos exposed to polycyclic aromatic hydrocarbons." Toxicol Appl Pharmacol 196(2): 191-205.

Incardona, J. P., H. L. Day, et al. (2006). "Developmental toxicity of 4-ring polycyclic aromatic hydrocarbons in zebrafish is differentially dependent on AH receptor isoforms and hepatic cytochrome P4501A metabolism." Toxicol Appl Pharmacol 217(3): 308-321.

Incardona, J. P., T. L. Linbo, et al. (2011). "Cardiac toxicity of 5-ring polycyclic aromatic hydrocarbons is differentially dependent on the aryl hydrocarbon receptor 2 isoform during zebrafish development." Toxicol Appl Pharmacol 257(2): 242-249.

Isidoro Tavares, N., P. Philip-Couderc, et al. (2009). "Angiotensin II and tumour necrosis factor alpha as mediators of ATP-dependent potassium channel remodelling in post- infarction heart failure." Cardiovasc Res 83(4): 726-736.

Jeng, H. A., C. H. Pan, et al. (2011). "Polycyclic aromatic hydrocarbon-induced oxidative stress and lipid peroxidation in relation to immunological alteration." Occup Environ Med 68(9): 653-658.

Jennings, A. A. (2012). "Worldwide regulatory guidance values for surface soil exposure to carcinogenic or mutagenic polycyclic aromatic hydrocarbons." J Environ Manage 110: 82-102.

Jennings, A. A. (2012). "Worldwide regulatory guidance values for surface soil exposure to noncarcinogenic polycyclic aromatic hydrocarbons." J Environ Manage 101: 173- 190.

Jensen, B. A., C. M. Reddy, et al. (2010). "Developing tools for risk assessment in protected species: Relative potencies inferred from competitive binding of halogenated 179

aromatic hydrocarbons to aryl hydrocarbon receptors from beluga (Delphinapterus leucas) and mouse." Aquat Toxicol 100(3): 238-245.

Jia, Y., D. Stone, et al. (2011). "Estimated reduction in cancer risk due to PAH exposures if source control measures during the 2008 Beijing Olympics were sustained." Environ Health Perspect 119(6): 815-820.

Jonsson, M. E., M. J. Jenny, et al. (2007). "Role of AHR2 in the expression of novel cytochrome P450 1 family genes, cell cycle genes, and morphological defects in developing zebra fish exposed to 3,3',4,4',5-pentachlorobiphenyl or 2,3,7,8-tetrachlorodibenzo-p- dioxin." Toxicol Sci 100(1): 180-193.

Jules, G. E., S. Pratap, et al. (2012). "In utero exposure to benzo(a)pyrene predisposes offspring to cardiovascular dysfunction in later-life." Toxicology.

Jung, D., C. W. Matson, et al. (2011). "Genotoxicity in Atlantic killifish (Fundulus heteroclitus) from a PAH-contaminated Superfund site on the Elizabeth River, Virginia." Ecotoxicology 20(8): 1890-1899.

Jung, K. H., J. H. Noh, et al. (2011). "Molecular signature for early detection and prediction of polycyclic aromatic hydrocarbons in peripheral blood." Environ Sci Technol 45(1):
300-306.

Jyethi, D. S., P. S. Khillare, et al. (2013). "Risk assessment of inhalation exposure to polycyclic aromatic hydrocarbons in school children." Environ Sci Pollut Res Int.

Karchner, S. I., D. G. Franks, et al. (2005). "AHR1B, a new functional aryl hydrocarbon receptor in zebrafish: tandem arrangement of ahr1b and ahr2 genes." Biochem J 392(Pt 1): 153-161.

Kerkvliet, N. I. (2009). "AHR-mediated immunomodulation: the role of altered gene transcription." Biochem Pharmacol 77(4): 746-760.

Kerley-Hamilton, J. S., H. W. Trask, et al. (2012). "Inherent and Benzo[a]pyrene-Induced Differential Aryl Hydrocarbon Receptor Signaling Greatly Affects Life Span, Atherosclerosis, Cardiac Gene Expression, and Body and Heart Growth in Mice." Toxicol Sci 126(2): 391-404.

Kim, K. T., T. Zaikova, et al. (2013). "Gold nanoparticles disrupt zebrafish eye development and pigmentation." Toxicol Sci 133(2): 275-288.

Kimmel, C. B., W. W. Ballard, et al. (1995). "Stages of embryonic development of the zebrafish." Dev Dyn 203(3): 253-310.

Kluver, N., L. Yang, et al. (2011). "Transcriptional response of zebrafish embryos exposed to neurotoxic compounds reveals a muscle activity dependent hspb11 expression." PLoS One 6(12): e29063.

Knecht, A. L., B. C. Goodale, et al. (2013). "Comparative developmental toxicity of environmentally relevant oxygenated PAHs." Toxicol Appl Pharmacol.

Kojic, S., D. Radojkovic, et al. (2011). "Muscle ankyrin repeat proteins: their role in striated muscle function in health and disease." Crit Rev Clin Lab Sci 48(5-6): 269-294.

Kraus, U., S. Breitner, et al. (2011). "Particle-associated organic compounds and symptoms in myocardial infarction survivors." Inhalation Toxicology 23(7): 431-447.

Krieger, J. A., J. L. Born, et al. (1994). "Persistence of calcium elevation in the HPB-ALL human T cell line correlates with immunosuppressive properties of polycyclic aromatic hydrocarbons." Toxicol Appl Pharmacol 127(2): 268-274. 180

Krieger, J. A., D. R. Davila, et al. (1995). "Inhibition of sarcoplasmic/endoplasmic reticulum calcium ATPases (SERCA) by polycyclic aromatic hydrocarbons in HPB-ALL human T cells and other tissues." Toxicol Appl Pharmacol 133(1): 102-108.

Kriek, E., M. Rojas, et al. (1998). "Polycyclic aromatic hydrocarbon-DNA adducts in humans:
relevance as biomarkers for exposure and cancer risk." Mutat Res 400(1-2): 215- 231.

Lahvis, G. P., R. W. Pyzalski, et al. (2005). "The aryl hydrocarbon receptor is required for developmental closure of the ductus venosus in the neonatal mouse." Mol Pharmacol 67(3): 714-720.

Langrish, J. P., X. Li, et al. (2011). "Reducing personal exposure to particulate air pollution improves cardiovascular health in patients with coronary heart disease." Environ Health Perspect 120(3): 367-372.

Layshock, J. A., G. Wilson, et al. (2010). "Ketone and Quinone-Substituted Polycyclic Aromatic Hydrocarbons in Mussel Tissue, Sediment, Urban Dust, and Diesel Particulate Matrices." Environmental Toxicology and Chemistry 29(11): 2450-2460.

Le Mevel, J. C., F. Lancien, et al. (2008). "Central cardiovascular actions of angiotensin II in trout." Gen Comp Endocrinol 157(1): 27-34.

Lee, M. S., S. Magari, et al. (2011). "Cardiac autonomic dysfunction from occupational exposure to polycyclic aromatic hydrocarbons." Occup Environ Med 68(7): 474-478.

Li, X., Y. Feng, et al. (2012). "The dose-response decrease in heart rate variability: any association with the metabolites of polycyclic aromatic hydrocarbons in coke oven workers?" PLoS One 7(9): e44562.

Livak, K. J. and T. D. Schmittgen (2001). "Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method." Methods 25(4):
402-408.

Lundstedt, S., Y. Persson, et al. (2006). Transformation of PAHs during ethanol-Fenton treatment of an aged gasworks' soil. Chemosphere. England. 65: 1288-1294.

Lundstedt, S., P. A. White, et al. (2007). "Sources, fate, and toxic hazards of oxygenated polycyclic aromatic hydrocarbons (PAHs) at PAH-contaminated sites." Ambio 36(6):
475-485.

Mahler, B. J., P. C. Van Metre, et al. (2005). "Parking lot sealcoat: an unrecognized source of urban polycyclic aromatic hydrocarbons." Environ Sci Technol 39(15): 5560-5566.

Mandrell, D., L. Truong, et al. (2012). "Automated zebrafish chorion removal and single embryo placement: optimizing throughput of zebrafish developmental toxicity screens." J Lab Autom 17(1): 66-74.

Mao, C., L. Shi, et al. (2009). "Development of fetal brain renin-angiotensin system and hypertension programmed in fetal origins." Prog Neurobiol 87(4): 252-263.

Mathew, L. K., E. A. Andreasen, et al. (2006). "Aryl hydrocarbon receptor activation inhibits regenerative growth." Mol Pharmacol 69(1): 257-265.

Mathew, R., J. A. McGrath, et al. (2008). "Modeling polycyclic aromatic hydrocarbon bioaccumulation and metabolism in time-variable early life-stage exposures." Environ Toxicol Chem 27(7): 1515-1525.

Matsumura, F., A. Puga, et al. (2009). "Biological functions of the arylhydrocarbon receptor:
beyond induction of cytochrome P450s. Introduction to this special issue." Biochem Pharmacol 77(4): 473.
181

McIntosh, B. E., J. B. Hogenesch, et al. (2010). "Mammalian Per-Arnt-Sim proteins in environmental adaptation." Annu Rev Physiol 72: 625-645.

Meeker, N. D. and N. S. Trede (2008). "Immunology and zebrafish: spawning new models of human disease." Dev Comp Immunol 32(7): 745-757.

Menzie, C. A., B. B. Potocki, et al. (1992). "Exposure to Carcinogenic Pahs in the Environment." Environmental Science & Technology 26(7): 1278-1284.

Meylan, W. M. and P. H. Howard (1995). "Atom Fragment Contribution Method for Estimating Octanol-Water Partition-Coefficients." Journal of Pharmaceutical Sciences 84(1): 83-92.

Moens, C. B., T. M. Donn, et al. (2008). "Reverse genetics in zebrafish by TILLING." Brief Funct Genomic Proteomic 7(6): 454-459.

Morey, J. S., J. C. Ryan, et al. (2006). "Microarray validation: factors influencing correlation between oligonucleotide microarrays and real-time PCR." Biol Proced Online 8: 175- 193.

Mumtaz, M. and J. George (1995). Toxicological profile for polycyclic aromatic hydrocarbons (PAHs). Atlanta, GA, Agency for Toxic Substances and Disease Registry, U.S.

Department of Health and Human Services.

Murk, A. J., J. Legler, et al. (1996). "Chemical-activated luciferase gene expression (CALUX): a novel in vitro bioassay for Ah receptor active compounds in sediments and pore water." Fundam Appl Toxicol 33(1): 149-160.

Murray, I. A., C. A. Flaveny, et al. (2011). "Suppression of cytokine-mediated complement factor gene expression through selective activation of the Ah receptor with 3',4'- dimethoxy-alpha-naphthoflavone." Mol Pharmacol 79(3): 508-519.

Nacci, D., M. Huber, et al. (2009). "Evolution of tolerance to PCBs and susceptibility to a bacterial pathogen (Vibrio harveyi) in Atlantic killifish (Fundulus heteroclitus) from New Bedford (MA, USA) harbor." Environ Pollut 157(3): 857-864.

Narayanan, G. A., I. A. Murray, et al. (2012). "Selective aryl hydrocarbon receptor modulator- mediated repression of CD55 expression induced by cytokine exposure." J Pharmacol Exp Ther 342(2): 345-355.

Narayanan, S. P., J. Suwanpradid, et al. (2011). "Arginase 2 deletion reduces neuro-glial injury and improves retinal function in a model of retinopathy of prematurity." PLoS One 6(7): e22460.

Naumova, Y. Y., S. J. Eisenreich, et al. (2002). "Polycyclic aromatic hydrocarbons in the indoor and outdoor air of three cities in the U.S." Environ Sci Technol 36(12): 2552- 2559.

Nebert, D. W., T. P. Dalton, et al. (2004). "Role of aryl hydrocarbon receptor-mediated induction of the CYP1 enzymes in environmental toxicity and cancer." J Biol Chem 279(23): 23847-23850.

Nebert, D. W., J. R. Robinson, et al. (1975). "Genetic expression of aryl hydrocarbon hydroxylase activity in the mouse." J Cell Physiol 85(2 Pt 2 Suppl 1): 393-414. 182

Nebert, D. W., A. L. Roe, et al. (2000). "Role of the aromatic hydrocarbon receptor and [Ah] gene battery in the oxidative stress response, cell cycle control, and apoptosis." Biochem Pharmacol 59(1): 65-85.

Nielsen, T., A. Feilberg, et al. (1999). "The variation of street air levels of PAH and other mutagenic PAC in relation to regulations of traffic emissions and the impact of atmospheric processes." Environmental Science and Pollution Research 6(3): 133- 137.

Nikolsky, Y., E. Kirillov, et al. (2009). "Functional analysis of OMICs data and small molecule compounds in an integrated "knowledge-based" platform." Methods Mol Biol 563:
177-196.

O'Donnell, E. F., K. S. Saili, et al. (2010). "The anti-inflammatory drug leflunomide is an agonist of the aryl hydrocarbon receptor." PLoS One 5(10).

Ovrevik, J., V. M. Arlt, et al. (2010). "Differential effects of nitro-PAHs and amino-PAHs on cytokine and chemokine responses in human bronchial epithelial BEAS-2B cells." Toxicol Appl Pharmacol 242(3): 270-280.

Pandini, A., M. S. Denison, et al. (2007). "Structural and functional characterization of the aryl hydrocarbon receptor ligand binding domain by homology modeling and mutational analysis." Biochemistry 46(3): 696-708.

Pandini, A., A. A. Soshilov, et al. (2009). "Detection of the TCDD binding-fingerprint within the Ah receptor ligand binding domain by structurally driven mutagenesis and functional analysis." Biochemistry 48(25): 5972-5983.

Park, K. and A. L. Scott (2010). "Cholesterol 25-hydroxylase production by dendritic cells and macrophages is regulated by type I interferons." J Leukoc Biol 88(6): 1081- 1087.

Patel, R. D., I. A. Murray, et al. (2009). "Ah receptor represses acute-phase response gene expression without binding to its cognate response element." Lab Invest 89(6): 695- 707.

Perera, F., W. Y. Tang, et al. (2009). "Relation of DNA methylation of 5'-CpG island of ACSL3 to transplacental exposure to airborne polycyclic aromatic hydrocarbons and childhood asthma." PLoS One 4(2): e4488.

Perera, F. P., Z. Li, et al. (2009). "Prenatal Airborne Polycyclic Aromatic Hydrocarbon Exposure and Child IQ at Age 5 Years." Pediatrics.

Petersen, G. I. and P. Kristensen (1998). "Bioaccumulation of lipophilic substances in fish early life stages." Environmental Toxicology and Chemistry 17(7): 1385-1395.

Peterson, R. E., H. M. Theobald, et al. (1993). "Developmental and reproductive toxicity of dioxins and related compounds: cross-species comparisons." Crit Rev Toxicol 23(3):
283-335.

Pfaffl, M. W. (2001). "A new mathematical model for relative quantification in real-time RT- PCR." Nucleic Acids Res 29(9): e45.

Planchart, A. and C. J. Mattingly (2010). "2,3,7,8-Tetrachlorodibenzo-p-dioxin upregulates FoxQ1b in zebrafish jaw primordium." Chem Res Toxicol 23(3): 480-487.

Polidori, A., J. Kwon, et al. (2010). "Source proximity and residential outdoor concentrations of PM(2.5), OC, EC, and PAHs." J Expo Sci Environ Epidemiol 20(5): 457-468. 183

Postlethwait, J., A. Amores, et al. (2004). "Subfunction partitioning, the teleost radiation and the annotation of the human genome." Trends Genet 20(10): 481-490.

Prasch, A. L., H. Teraoka, et al. (2003). "Aryl hydrocarbon receptor 2 mediates 2,3,7,8- tetrachlorodibenzo-p-dioxin developmental toxicity in zebrafish." Toxicol Sci 76(1):
138-150.

Price, R. L., W. Carver, et al. (1997). "The effects of angiotensin II and specific angiotensin receptor blockers on embryonic cardiac development and looping patterns." Dev Biol 192(2): 572-584.

Puga, A., C. Ma, et al. (2009). "The aryl hydrocarbon receptor cross-talks with multiple signal transduction pathways." Biochem Pharmacol 77(4): 713-722.

Ramesh, A., S. A. Walker, et al. (2004). "Bioavailability and risk assessment of orally ingested polycyclic aromatic hydrocarbons." Int J Toxicol 23(5): 301-333.

Ramirez, N., A. Cuadras, et al. (2011). "Risk assessment related to atmospheric polycyclic aromatic hydrocarbons in gas and particle phases near industrial sites." Environ Health Perspect 119(8): 1110-1116.

Reimers, M. J., J. K. La Du, et al. (2006). "Ethanol-dependent toxicity in zebrafish is partially attenuated by antioxidants." Neurotoxicol Teratol 28(4): 497-508.

Ren, A., X. Qiu, et al. (2011). "Association of selected persistent organic pollutants in the placenta with the risk of neural tube defects." Proc Natl Acad Sci U S A 108(31):
12770-12775.

Rennie, M. Y., J. Detmar, et al. (2011). "Vessel tortuousity and reduced vascularization in the fetoplacental arterial tree after maternal exposure to polycyclic aromatic hydrocarbons." Am J Physiol Heart Circ Physiol 300(2): H675-684.

Reynaud, S. and P. Deschaux (2006). "The effects of polycyclic aromatic hydrocarbons on the immune system of fish: a review." Aquat Toxicol 77(2): 229-238.

Saeed, A. I., V. Sharov, et al. (2003). "TM4: a free, open-source system for microarray data management and analysis." Biotechniques 34(2): 374-378.

Safe, S. H. (1998). "Development validation and problems with the toxic equivalency factor approach for risk assessment of dioxins and related compounds." J Anim Sci 76(1):
134-141.

Saili, K. S. (2012). Developmental neurobehavioral toxicity of bisphenol A in zebrafish (Danio rerio) [electronic resource] / by Katerine Schletz Saili. Corvallis, Or. :, Oregon State University.

Sarkar, A., D. Ray, et al. (2006). "Molecular Biomarkers: their significance and application in marine pollution monitoring." Ecotoxicology 15(4): 333-340.

Sartor, M. A., M. Schnekenburger, et al. (2009). "Genomewide analysis of aryl hydrocarbon receptor binding targets reveals an extensive array of gene clusters that control morphogenetic and developmental programs." Environ Health Perspect 117(7):
1139-1146.

Schmidt, J. V. and C. A. Bradfield (1996). "Ah receptor signaling pathways." Annu Rev Cell Dev Biol 12: 55-89.

Schmidt, J. V., G. H. Su, et al. (1996). "Characterization of a murine Ahr null allele:
involvement of the Ah receptor in hepatic growth and development." Proc Natl Acad Sci U S A 93(13): 6731-6736.

Schoeny, R. a. K. P. (1993). Provisional guidance for quantitative risk assessment of polycyclic aromatic hydrocarbons. U. S. E. P. Agency. Washington, DC. 184

Scott, J. A., J. P. Incardona, et al. (2011). "AhR2-mediated, CYP1A-independent cardiovascular toxicity in zebrafish (Danio rerio) embryos exposed to retene." Aquat Toxicol 101(1): 165-174.

Shannon, P., A. Markiel, et al. (2003). "Cytoscape: a software environment for integrated models of biomolecular interaction networks." Genome Res 13(11): 2498-2504.

Shi, Z., N. Dragin, et al. (2010). "Organ-specific roles of CYP1A1 during detoxication of dietary benzo[a]pyrene." Mol Pharmacol 78(1): 46-57.

Shimizu, Y., Y. Nakatsuru, et al. (2000). "Benzo[a]pyrene carcinogenicity is lost in mice lacking the aryl hydrocarbon receptor." Proc Natl Acad Sci U S A 97(2): 779-782.

Singh, K. P., F. L. Casado, et al. (2009). "The aryl hydrocarbon receptor has a normal function in the regulation of hematopoietic and other stem/progenitor cell populations." Biochem Pharmacol 77(4): 577-587.

Singh, R. P., R. Khanna, et al. (2003). "Comparative effect of benzanthrone and 3- bromobenzanthrone on hepatic xenobiotic metabolism and anti-oxidative defense system in guinea pigs." Archives of Toxicology 77(2): 94-99.

Smith, B. W., S. S. Rozelle, et al. (2013). "The aryl hydrocarbon receptor directs hematopoietic progenitor cell expansion and differentiation." Blood.

Song, M. K., M. Song, et al. (2012). "Identification of molecular signatures predicting the carcinogenicity of polycyclic aromatic hydrocarbons (PAHs)." Toxicol Lett 212(1):
18-28.

Suresh, R., A. Shally, et al. (2009). "Assessment of association of exposure to polycyclic aromatic hydrocarbons with bronchial asthma and oxidative stress in children: A case control study." Indian J Occup Environ Med 13(1): 33-37.

Svoboda, K. R., A. E. Linares, et al. (2001). "Activity regulates programmed cell death of zebrafish Rohon-Beard neurons." Development 128(18): 3511-3520.

Tang, D., T. Y. Li, et al. (2006). "PAH-DNA adducts in cord blood and fetal and child development in a Chinese cohort." Environ Health Perspect 114(8): 1297-1300.

Tanguay, R. L., C. C. Abnet, et al. (1999). "Cloning and characterization of the zebrafish (Danio rerio) aryl hydrocarbon receptor." Biochim Biophys Acta 1444(1): 35-48.

Teraoka, H., W. Dong, et al. (2003). "Induction of cytochrome P450 1A is required for circulation failure and edema by 2,3,7,8-tetrachlorodibenzo-p-dioxin in zebrafish." Biochem Biophys Res Commun 304(2): 223-228.

Tian, B., D. E. Nowak, et al. (2005). "A TNF-induced gene expression program under oscillatory NF-kappaB control." BMC Genomics 6: 137.

Tian, Y., S. Ke, et al. (1999). "Ah receptor and NF-kappaB interactions, a potential mechanism for dioxin toxicity." J Biol Chem 274(1): 510-515.

Tilton, S. C., T. L. Tal, et al. (2012). "Bioinformatics resource manager v2.3: an integrated software environment for systems biology with microRNA and cross-species analysis tools." BMC Bioinformatics 13(1): 311.

Timme-Laragy, A. R., C. J. Cockman, et al. (2007). "Synergistic induction of AHR regulated genes in developmental toxicity from co-exposure to two model PAHs in zebrafish." Aquat Toxicol 85(4): 241-250.

Timme-Laragy, A. R., S. I. Karchner, et al. (2012). "Nrf2b, novel zebrafish paralog of oxidant- responsive transcription factor NF-E2-related factor 2 (NRF2)." J Biol Chem 287(7):
4609-4627.

Timme-Laragy, A. R., L. A. Van Tiem, et al. (2009). "Antioxidant responses and NRF2 in synergistic developmental toxicity of PAHs in zebrafish." Toxicol Sci. 185

Trapnell, C., L. Pachter, et al. (2009). "TopHat: discovering splice junctions with RNA-Seq." Bioinformatics 25(9): 1105-1111.

Trapnell, C., A. Roberts, et al. (2012). "Differential gene and transcript expression analysis of RNA-seq experiments with TopHat and Cufflinks." Nat Protoc 7(3): 562-578.

Truong, L., S. L. Harper, et al. (2011). "Evaluation of embryotoxicity using the zebrafish model." Methods Mol Biol 691: 271-279.
van Delft, J., S. Gaj, et al. (2012). "RNA-Seq provides new insights in the transcriptome responses induced by the carcinogen benzo[a]pyrene." Toxicol Sci 130(2): 427-439.

Van den Berg, M., L. S. Birnbaum, et al. (2006). "The 2005 World Health Organization reevaluation of human and Mammalian toxic equivalency factors for dioxins and dioxin-like compounds." Toxicol Sci 93(2): 223-241.

Van Metre, P. C. and B. J. Mahler (2005). "Trends in hydrophobic organic contaminants in urban and reference lake sediments across the United States, 1970-2001." Environ Sci Technol 39(15): 5567-5574.

Van Metre, P. C. and B. J. Mahler (2010). "Contribution of PAHs from coal-tar pavement sealcoat and other sources to 40 U.S. lakes." Sci Total Environ 409(2): 334-344.

Van Tiem, L. A. and R. T. Di Giulio (2011). "AHR2 knockdown prevents PAH-mediated cardiac toxicity and XRE- and ARE-associated gene induction in zebrafish (Danio rerio)." Toxicol Appl Pharmacol 254(3): 280-287.

Vanwezel, A. P. and A. Opperhuizen (1995). "Narcosis Due to Environmental-Pollutants in Aquatic Organisms - Residue-Based Toxicity, Mechanisms, and Membrane Burdens." Critical Reviews in Toxicology 25(3): 255-279.

Vlecken, D. H., J. Testerink, et al. (2009). "A critical role for myoglobin in zebrafish development." Int J Dev Biol 53(4): 517-524.

Vogel, C. F., E. Sciullo, et al. (2004). "Dioxin increases C/EBPbeta transcription by activating cAMP/protein kinase A." J Biol Chem 279(10): 8886-8894.

Walgraeve, C., K. Demeestere, et al. (2010). "Oxygenated polycyclic aromatic hydrocarbons in atmospheric particulate matter: Molecular characterization and occurrence." Atmospheric Environment 44(15): 1831-1846.

Wan, B., J. W. Yarbrough, et al. (2008). "Structure-related clustering of gene expression fingerprints of thp-1 cells exposed to smaller polycyclic aromatic hydrocarbons." SAR QSAR Environ Res 19(3-4): 351-373.

Wang, W., N. Jariyasopit, et al. (2011). "Concentration and photochemistry of PAHs, NPAHs, and OPAHs and toxicity of PM2.5 during the Beijing Olympic Games." Environ Sci Technol 45(16): 6887-6895.

Wassenberg, D. M., A. L. Nerlinger, et al. (2005). "Effects of the polycyclic aromatic hydrocarbon heterocycles, carbazole and dibenzothiophene, on in vivo and in vitro CYP1A activity and polycyclic aromatic hydrocarbon-derived embryonic deformities." Environ Toxicol Chem 24(10): 2526-2532.

Wei, S. L., B. Huang, et al. (2012). "Characterization of PM2.5-bound nitrated and oxygenated PAHs in two industrial sites of South China." Atmospheric Research 109: 76-83.

White, S. S. and L. S. Birnbaum (2009). "An overview of the effects of dioxins and dioxin-like compounds on vertebrates, as documented in human and ecological epidemiology." J Environ Sci Health C Environ Carcinog Ecotoxicol Rev 27(4): 197-211.

Wienholds, E., F. van Eeden, et al. (2003). "Efficient target-selected mutagenesis in zebrafish." Genome Res 13(12): 2700-2707.
186

Wiens, G. D. and G. W. Glenney (2011). "Origin and evolution of TNF and TNF receptor superfamilies." Dev Comp Immunol 35(12): 1324-1335.

Wilhelm, M., J. K. Ghosh, et al. (2012). "Traffic-related air toxics and term low birth weight in Los Angeles County, California." Environ Health Perspect 120(1): 132-138.

Willett, K. L., D. Wassenberg, et al. (2001). "In vivo and in vitro inhibition of CYP1A- dependent activity in Fundulus heteroclitus by the polynuclear aromatic hydrocarbon fluoranthene." Toxicol Appl Pharmacol 177(3): 264-271.

Wills, L. P., C. W. Matson, et al. (2010). "Characterization of the recalcitrant CYP1 phenotype found in Atlantic killifish (Fundulus heteroclitus) inhabiting a Superfund site on the Elizabeth River, VA." Aquat Toxicol 99(1): 33-41.

Wilson, S. R., A. D. Joshi, et al. (2013). "The tumor suppressor Kruppel-like factor 6 is a novel aryl hydrocarbon receptor DNA binding partner." J Pharmacol Exp Ther 345(3):
419-429.

Wirgin, I., N. K. Roy, et al. (2011). "Mechanistic basis of resistance to PCBs in Atlantic tomcod from the Hudson River." Science 331(6022): 1322-1325.

Wittkopp, N., E. Huntzinger, et al. (2009). "Nonsense-mediated mRNA decay effectors are essential for zebrafish embryonic development and survival." Mol Cell Biol 29(13):
3517-3528.

Wu, C., H. Lu, et al. (2011). "Molecular and Pathophysiological Features of Angiotensinogen:

A Mini Review." N Am J Med Sci (Boston) 4(4): 183-190.

Wu, M. T., T. C. Lee, et al. (2011). "Whole genome expression in peripheral-blood samples of workers professionally exposed to polycyclic aromatic hydrocarbons." Chem Res Toxicol 24(10): 1636-1643.

Xiong, K. M., R. E. Peterson, et al. (2008). "Aryl hydrocarbon receptor-mediated down- regulation of sox9b causes jaw malformation in zebrafish embryos." Mol Pharmacol 74(6): 1544-1553.

Xu, X., H. Hu, et al. (2013). "Studying the effects of polycyclic aromatic hydrocarbons on peripheral arterial disease in the United States." Sci Total Environ 461-462C: 341- 347.

Yu, H. (2002). "Environmental carcinogenic polycyclic aromatic hydrocarbons:
photochemistry and phototoxicity." J Environ Sci Health C Environ Carcinog Ecotoxicol Rev 20(2): 149-183.

187

Appendices
188

Appendix 1 - Developmental toxicity of DBT, BAA and PYR in ahr2hu3335 zebrafish
Objective
The developmental toxicity of BAA, DBT and PYR was compared between wild -type 5D embryos (ahr2+) and embryos with non-functional AHR2 (ahr2hu3335) in order to determine whether developmental toxicity of these three PAHs was mediated by ahr2.

Methods
Embryos were exposed to 25 uM benz(a)anthracene (BAA), dibenzothiophene (DBT), pyrene (PYR) or 1% DMSO vehicle control dissolved in embryo media from 6-48 hpf in glass vials. Exposures were conducted at 28C on a rocker, protected from light. At 48 hpf, exposure solutions were removed, embryos were rinsed 4 times and solution was replaced with 2 ml fresh embryo media. Embryos were incubated in the dark until 120 hpf, when they were evaluated for developmental malformations.

Results
As observed previously (Chapter 2), we observed significant malformation incidence in ahr2+embryos exposed to 25μM BAA, DBT and PYR (Figure 1A). In ahr2hu3335 embryos, 25 uM BAA did not cause a significant increase in any malformations. DBT and PYR, however, did cause increases in malformations, suggesting that they cause toxicity through ahr2- independent mechanisms.

Conclusions
With these s tudies, we confirmed the dependence of BAA -induced developmental toxicity on AHR2. We also demonstrate that absence of AHR2 does not protect from DBT and PYR - induced tocicity. This has been previously observed in AHR2 -deficient (AHR2 morpholino injected) e mbryos exposed to these compounds (Incardona, Day et al. 2006 ). Background malformations and low quantity of embryos precl uded the use of the ahr2hu33335 line for the remaining studies in this dissertation. It is generally accepted that 5 outcrosses are required to sufficiently dilute background mutations in zebrafish lines generated by TILLING. At the writing of this thesis, the fourth outcrossed generation is spawning more reliably and better 189

quality eggs. The ahr2hu3335 shows promise for future work investigating AHR related questions in zebrafish.

190

Figure A1-1
Profile plots of malformations observed at 5 dpf in PAH-treated wild-type and ahr2-null embryos indicate subtle differential responses between PAHs. PAH exposures did not induce significant mortality (Mort) in either fish line, but pericardial edema (PE), axis, yolk sac, jaw, and eye malformations were observed. All three PAH treatments caused a significant increase in the percent of wild-type embryos with at least one malformation (Any Mal.). BAA did not cause a significant increase in malformations in ahr2hu3335 embryos compared to control. Data represent 3 independent replicates analyzed by One-way Analysis of Variance with Tukey’s post-test for pairwise comparisons.

191

Appendix 2 - Characterization of behavioral and neutrophil responses to PAH exposure
Objective
During developmental toxicity studies, we observed hyperactivity at 3 and 4 days post fertilization in embryos exposed to pyrene (PYR). By 120 hpf, when toxicity evaluations are conducted, behavior was no longer noticeable. We hypothesized that embryos wer e sensing and responding to PYR in a more immediate manner , and not because of effects on development. In this preliminary study we investigated whether zebrafish respond to acute PAH exposure at 120 hpf, when organ systems are fully developed.

In our micr oarray study (Chapter 2) we observed a large number of inflammatory -related transcripts misexpressed in response to PAH exposure. PAHs are also known skin irritants.

We investigated whether localized inflammatory activity could be observed following PAH exposure using a transgenic zebrafish line (mpx:gfp) that expresses green fluorescent protein in neutrophils (Elks, Loynes et al. 2011).

Methods
Behavioral challenge assay
Zebrafish were dechorionated and placed individually into wells of a 96 -well plate at 6 hpf in 100 μl embryo media. They were incubated with normal lighting conditions until 120 dpf. 100 ul of exposure solution (DBT, PYR or BAA) was then added to each well, to a final concentration of 1% DMSO. Embryos were exposed to 0, 1, 2.5, 5, 10 or 25 μM final concentrations of PAH. Behavior was recorded under lit conditions using Viewpoint Zebraboxes, starting 10 min following exposure for 5 hours. Distance traveled was calculated per minute for each fish. Data represent the mean of 32 fish per exposure group.

Imaging neutrophil response
GFP:mpo zebrafish were exposed to 25 uM PYR or 1% DMSO control in a 96 -well plate at 72 hpf. 8 embryos from each group were anesthetized and imaged on a zeiss axiovert microscope using a 5X objective at 2, 4, 8 and 24 hou rs post exposure. Images were created from z-stack of 5 slices.
192

Results
Behavioral challenge assay
In this preliminary dataset, we observed distinct differen ces in behavioral patterns of fish exposed to DBT, PYR and BAA. Movement was heighted by exposure t o all PAHs for the first hour of the study. Zebrafish exposed to DBT and BAA returned to activity levels comparable to controls after ~ 1.5 hr (Figure 1A, 1B). The 25uM DBT group remained more active, at 80 millimeters per minute, for almost 3 hours. In co ntrast, PYR -exposed zebrafish were less active immediately following exposure, but activity increased to ~120 millimeters per minute 1 hr after exposure in the 10 and 25 uM exposure groups (Figure 1C). Activity remained elevated for the duration of the stu dy. We observed a concentration -response in hyperactivity, where zebrafish exposed to 2.5 uM PYR also had activity elevated above controls for the 5 hr exposure. This preliminary data demonstrates that zebrafish can immediately sense PAHs, and respond uniq uely to PYR. The sensory mechanism and cause of activity remains unknown.

Imaging neutrophil response
From the whole-embryo images collected, we were not able to discern any specific areas of neutrophil activity in the PYR -exposed embryos. We found that ne utrophil number and location was variable between fish. While a quantitative method may be able to detect small changes that are not readily observable, we were not able to identify a specific area in the fish that exhibited increased localization of these cells involved in the inflammatory response (Figure 2A). We note that we were only visualizing neutrophils; macrophages are also active at these developmental time points and could be more involved in response to PAH exposure. The waterborne exposure also would not necessarily be expected to result in a localized response.

Conclusions
In this preliminary data, we found that zebrafish can sense and immediately respond to PAH exposure with increased swimming behavior at 5 dpf. This was also observed anecdota lly at 3 and 4 dpf. PYR induced a unique response, wherein zebrafish exhibited increased swimming activity for the duration of the study (5 hours). We test lower concentrations, and found increased activity at 2.5 uM, which is 10 fold lower than the concen tration that 193

induced developmental malformations (in glass vial exposures). Fish have been shown to exhibit increased activity in response to water contaminants (Hellou 2011 ). The sensory mechanism, however, is not known. Given their structural similarity, the difference in response to PYR and BAA is particularly intriguing and under further investigation.

PAHs can cause skin irritation, and induce inflammatory activity in a number of model systems. We hypothesized that an inflammatory response to PYR might result in differential neutrophil localization. We did not observe any notable difference, however, in transgenic mpx:gfp zebrafish at 2, 4, 8 or 28 hours post exposure. Other cell types may be more involved in inflammatory response to PAHs, or this method may not be sufficient to discern differences.

194

Figure A2-1
Activity of zebrafish (120 hpf) exposed to BAA (A), DBT (B), or PYR (C) was recorded for 5 hours following exposure.

A
B
C
195

Figure A2-2 No treatment-related differences in neutrophil localization following PYR exposure
Representative images of 3 dpf zebrafish exposed to 1% DMSO at 2 and 4 hours post exposure (A, B). Treatment effects were not observed in embryos exposed to 25 μM PYR at 2 and 4 hpe (C,D).

References
Elks, P. M., C. A. Loynes, et al. (2011). "Measuring inflammatory cell migration in the zebrafish." Methods Mol Biol 769: 261-275.

Hellou, J. (2011). "Behavioural ecotoxicology, an "early warning" signal to assess environmental quality." Environ Sci Pollut Res Int 18(1): 1-11.

Management and Conservation Article
Greater Sage-Grouse Winter Habitat Selection and
Energy Development
KEVIN E. DOHERTY, 1,2 Wildlife Biology Program, College of Forestry and Conservation, University of Montana, Missoula, MT 59812, USA DAVID E. NAUGLE, Wildlife Biology Program, College of Forestry and Conservation, University of Montana, Missoula, MT 59812, USA BRETT L. WALKER, Wildlife Biology Program, College of Forestry and Conservation, University of Montana, Missoula, MT 59812, USA JON M. GRAHAM, Department of Mathematical Sciences, University of Montana, Missoula, MT 59812, USA ABSTRACT Recent energy development has resulted in rapid and large-scale changes to western shrub-steppe ecosystems without a complete understanding of its potential impacts on wildlife populations. We modeled winter habitat use by female greater sage-grouse (Centrocercus urophasianus) in the Powder River Basin (PRB) of Wyoming and Montana, USA, to 1) identify landscape features that influenced sage-grouse habitat selection, 2) assess the scale at which selection occurred, 3) spatially depict winter habitat quality in a Geographic Information System, and 4) assess the effect of coal-bed natural gas (CBNG) development on winter habitat selection. We developed a model of winter habitat selection based on 435 aerial relocations of 200 radiomarked female sage-grouse obtained during the winters of 2005 and 2006. Percent sagebrush ( Artemisia spp.) cover on the landscape was an important predictor of use by sage-grouse in winter. The strength of habitat selection between sage-grouse and sagebrush was strongest at a 4-km 2 scale. Sage-grouse avoided coniferous habitats at a 0.65-km2 scale and riparian areas at a 4-km 2 scale. A roughness index showed that sage-grouse selected gentle topography in winter. After controlling for vegetation and topography, the addition of a variable that quantified the density of CBNG wells within 4 km 2 improved model fit by 6.66 Akaike’s Information Criterion points (Akaike wt¼ 0.965). The odds ratio for each additional well in a 4-km 2 area (0.877; 95% CI ¼ 0.834–

0.923) indicated that sage-grouse avoid CBNG development in otherwise suitable winter habitat. Sage-grouse were 1.3 times more likely to occupy sagebrush habitats that lacked CBNG wells within a 4-km 2 area, compared to those that had the maximum density of 12.3 wells per 4 km2 allowed on federal lands. We validated the model with 74 locations from 74 radiomarked individuals obtained during the winters of 2004 and 2007. This winter habitat model based on vegetation, topography, and CBNG avoidance was highly predictive (validationR2 ¼ 0.984). Our spatially explicit model can be used to identify areas that provide the best remaining habitat for wintering sage-grouse in the PRB to mitigate impacts of energy development. (JOURNAL OF WILDLIFE MANAGEMENT 72(1):187–195; 2008) DOI: 10.2193/2006-454
KEY WORDS Centrocercus urophasianus, coal-bed natural gas, energy development, greater sage-grouse, habitat, land-use change, resource selection function, sagebrush, scale, winter.

Understanding landscape-scale habitat selection during
critical life stages is essential for developing conservation plans for sensitive species. Studies of habitat selection at small scales further our ecological understanding of species– habitat relationships but do not convey spatially explicit information about habitat quality at a scale useful for prioritizing landscapes for conservation. Recent advances in modeling habitat selection from high-resolution satellite imagery using resource selection functions (RSF) offers the ability to rank speciﬁc areas by their relative probability of use (Manly et al. 2002). Resulting probability layers can then be mapped in a Geographic Information System (GIS) to identify regions where high-quality habitat is available.

Further, these models allow cross-validation and testing against independent datasets to ensure that inferences
regarding habitat selection are robust (Boyce et al. 2002, Johnson et al. 2006). The relative inﬂuence of variables thought to be important in habitat selection can also be assessed in a competing-model framework (Burnham and
Anderson 2002).

Previously widespread, greater sage-grouse ( Centrocercus urophasianus; hereafter sage-grouse) have been extirpated from approximately 50 % of their original range in western North America (Schroeder et al. 2004), with an estimated range-wide population decline of 45–80% and local declines of 17–92 % (Connelly and Braun 1997, Connelly et al.
2000, Aldridge and Brigham 2003). Despite increased
concern for their populations, little effort has gone into measuring landscape-scale winter habitat selection by
greater sage-grouse. Previous winter habitat studies have focused on the importance of micro-site vegetation features such as height, canopy cover, or crude protein levels of sagebrush (e.g., Eng and Schladweiler 1972, Beck 1977,
Connelly et al. 2000, Crawford et al. 2004, Sauls 2006). In winter, sage-grouse inhabit areas with moderate to dense sagebrush (Eng and Schladweiler 1972, Homer et al. 1993, Connelly et al. 2000) and typically prefer areas with gentle (,10%), south- or west-facing slopes (Beck 1977, Hupp
and Braun 1989). Previous demographic studies have
documented high rates of winter survival (reviewed in
Connelly et al. 2004). However, Moynahan et al. (2006)
demonstrated that severe winters can have substantial
population-level impacts. Birds also must often move long distances to ﬁnd suitable winter habitat (Patterson 1952 in Connelly et al. 2004; Connelly et al. 1988; Robertson 1991).

Impacts to wintering habitat may have disproportionate
effects on regional population size and persistence. For example, Beck (1977) found that 80 % of use sites occurred in ,7% of the area of sagebrush available in northern
Colorado, USA, suggesting that winter habitat may be
1 Present address: Wyoming Audubon, 358 N 5th Street, Unit A, Laramie, WY 82072, USA
2 E-mail: [email protected]
Doherty et al. /C15 Sage-Grouse Habitat and Energy Development 187 limited. The relationship between sagebrush and sage-
grouse is arguably the closest during winter when birds switch from a diet of insects, forbs, and sagebrush to one composed of .96% sagebrush (Remington and Braun
1985, Welch et al. 1991, Connelly et al. 2000, Crawford et al. 2004). Heavy snowfall may even further reduce the
amount of suitable habitat by limiting the abundance of sagebrush above the snow (Hupp and Braun 1989; Connelly et al. 2000, 2004).

Coal-bed natural gas (CBNG) development in the Powder
River Basin (PRB) has caused rapid, large-scale changes to sagebrush habitats in Montana and Wyoming, USA. The
sage-grouse sub-population in the PRB is a critical
component of the larger Wyoming Basin population, which represents 25 % of sage-grouse in the species’ range
(Connelly et al. 2004). The population in the PRB has a high density of active leks and serves as a link to populations in eastern Wyoming and western South Dakota, USA, and
between the Wyoming Basin and central Montana, USA
(Connelly et al. 2004). The CBNG ﬁeld in the PRB is one of the largest developed energy ﬁelds in North America. In this region, approximately 29,000 CBNG wells have been
drilled on public and private lands, and another approx- imately 37,000 wells are expected within a 2.4-million ha area, roughly the size of the state of New Hampshire
(Bureau of Land Management 2003 a, b). Drilling is
typically authorized at a maximum density of 1 well per 32 ha on lands where federally owned gas reserves are
extracted; however, there are no well-density restrictions placed on private or state-owned gas reserves. Wells, power lines, roads, vehicle trafﬁc, pipelines, compressor stations, and water storage ponds within a gas ﬁeld this size
contribute to fragmentation of sagebrush habitats and may impact sagebrush obligates (Knick et al. 2003).

We investigated sage-grouse winter habitat use in the
PRB as part of a larger study of the potential impacts of CBNG development on sage-grouse populations. Our
objectives were to 1) create a robust habitat selection model for sage-grouse in winter, 2) evaluate the appropriate scale at which females select winter habitat, 3) spatially depict habitat suitability in a GIS to identify areas with a high probability of use, and 4) assess the inﬂuence of CBNG
development on winter habitat selection .

STUDY AREA
Our study area in the PRB covered portions of Johnson,
Sheridan, and Campbell counties in Wyoming, and Big-
horn, Rosebud, and Powder River counties in Montana.

Shrub-steppe habitat in the PRB was dominated by
Wyoming big sagebrush ( Artemisia tridentata wyomingensis ) with an understory of native and nonnative grasses such as bluebunch wheatgrass ( Pseudoroegneria spicata ), western wheatgrass ( Agropyron smithii ), prairie junegrass ( Koeleria macrantha), blue grama ( Bouteloua gracilis ), Japanese brome (Bromus japonicus ), cheatgrass ( B. tectorum ), and crested wheatgrass ( Agropyron cristatum ). Plains silver sagebrush (Artemisia cana cana ) was also present in drainages but at much lower abundance. Rocky mountain juniper ( Juniperus scopulorum) and ponderosa pine ( Pinus ponderosa ) were located in wooded draws and formed forests across the
extreme northern extent of the study area. Conifers were largely absent from the southern half of the study area. Land use was dominated by cattle ranching; only 4 % of the
landscape consisted of dry land or irrigated agriculture. The PRB typically was cold and dry in January with average
temperatures of /C0 6.08 C and 16.3 cm of snowfall. Winter weather conditions in 2004 and 2005 were almost identical to historical averages. The winter of 2006 was mild; in January, temperatures were 6.5 8 C above normal and
snowfall was 15 cm below average. The January 2007
average temperature of /C0 5.58 C was near historical norms; however, snowfall was 60 % above normal.

METHODS
Marking and Monitoring Protocols
We captured sage-grouse by rocket-netting (Giesen et al. 1982) and spotlighting (Wakkinen et al. 1992) on and
around leks in 3 study areas: 1) Bighorn County, Montana, 2) Campbell County, Wyoming, and 3) Johnson County,
Wyoming during March–April and August of 2003–2006.

We aged and sexed grouse and ﬁtted females with a 21.6-g necklace-style radiocollar with a 4-hour mortality switch (model A4060; Advanced Telemetry Systems, Isanti, MN).

Sage-grouse in the Bighorn and Campbell county study
areas were nonmigratory. In contrast, many birds in the Johnson Country study area were migratory, with distinct breeding, summer, and winter ranges. In all study sites, we obtained winter locations after birds in our migratory
population had moved to wintering areas but before they had moved back to the breeding grounds. We monitored
sage-grouse via aerial radiotracking during the winters of 2005–2007. We used a ﬁxed-wing airplane with aerial
telemetry antennas mounted on both wings struts and
connected to a switch box. We used a Global Positioning System (GPS) receiver to record locations of used sites as we circled sage-grouse at approximately 100–200-m elevation above the ground. We radiotracked sage-grouse on foot
during the winter of 2004, and recorded their positions with a GPS receiver when we obtained visual sightings of
radiomarked birds. We estimated the 95 % error ellipse of aerial locations by relocating a transmitter placed in rolling sagebrush cover 40 times from the air in a blind trial. We then calculated a bivariate normal home range estimator (Jennrich and Turner 1969) using these relocations to
quantify our maximum resolution to estimate the location of an unknown collar (78.2-m radius). The ability of our plane to tightly circle sage-grouse was not constrained by rugged areas or conifer-dominated landscapes in the PRB because birds were not located in these habitat features; thus, our test was representative of the maximum precision of our aerial telemetry locations in rolling sagebrush habitats. We did not quantify error for ground-based locations, but we assumed error estimates were smaller than aerial-based methods.

Since we treated our aerial telemetry error test as a
188 The Journal of Wildlife Management /C15 72(1)
maximum precision estimate, we conducted all analyses at scales /C21 100 m to ensure that our inference was not
confounded by location error.

Designation of Used and Available Sites
We employed a used–available design to evaluate sage-
grouse habitat relationships in winter (Boyce et al. 2002, Manly et al. 2002, Johnson et al. 2006). We deﬁned used points as the sites where we located radiomarked sage-
grouse during radiotracking. We split sage-grouse used
locations into those we analyzed to build a statistical model to quantify large-scale habitat relationships and those we analyzed to test the predictive ability of our spatially explicit winter habitat model. We located birds we used to build the model during 3 ﬂights from 2 January to 25 January 2005 ( n ¼ 292 locations on 106 individuals) and on 3 ﬂights from 24 December 2005 to 1 February 2006 (n ¼ 241 locations on 94 individuals). To test the model, we used 87 locations
collected on the ground from 15 to 18 January 2004 ( n ¼ 30 locations on 28 individuals) and on 2 ﬂights on 18 and 26 January 2007 (n ¼ 57 locations on 57 individuals). Of the 85 individuals used to test the model, 57 were not included among birds marked during 2005 or 2006. We found some
radiomarked birds together in ﬂocks. To avoid the
possibility of dependency in our data, we retained only one used location per ﬂock. The ﬁnal data set contained 435 locations for building the model and 74 locations for testing the model.

We selected available points within circles that had a
radius to the farthest winter used point and were centered on either the lek of capture or on the lek closest to where birds were captured via spotlighting. We merged circles that overlapped within each study area to create 3 nonoverlap- ping polygons that corresponded with our 3 study areas. We randomly selected available points from a spatial Poisson distribution (Beyer 2004) proportional to twice the number of used points within a polygon and year to ensure a
representative sample of available habitats.

GIS Habitat Classification
We acquired SPOT-5 satellite imagery (Terra Image USA,
Santa Barbara, CA) for the northern portion of the study area in August 2003 and for the southern portion in August 2004 when the project expanded to encompass a larger
geographic area. We ortho-rectiﬁed SPOT-5 imagery to
existing digital ortho-quads of the study area. The SPOT-5 panchromatic and multi-spectral images were combined into a single panchromatic, multi-spectral ﬁle. We then used the panchromatic 25-m
2-pixel image to perform pan-sharpen-
ing to reduce the multi-spectral image pixel size from 100 m
2 to 25 m 2, greatly increasing the resolution of our
analysis. We used eCognition e 4.0 software (Deﬁniens
Imaging, Munich, Germany) to cluster the pixels into
regions representing spectrally similar ground features. We exported clusters into ArcGIS 9.2 software to create a
polygon database. We collected ﬁeld training points ( n ¼ 7,092) that were stratiﬁed by space and landowner access to classify 5 habitat cover classes as sagebrush, conifer, grassland, riparian, and barren. Classiﬁcation accuracy assessed by withholding subsamples of data (i.e., k-fold cross-validation with 10 folds; Boyce et al. 2002) was 83 % for sagebrush, 77% for conifer, 76% for grassland, 70% for riparian, and 80 % for barren with an overall accuracy of 78%. We removed urban areas and strip mines from
analyses.

Vegetation, Topography, and Energy Development
Variables
We quantiﬁed characteristics of vegetation, topography
(e.g., Beck 1977, Remington and Braun 1985, Hupp and
Braun 1989, Sauls 2006), and energy development around
used and available points using a GIS to evaluate landscape predictors of sage-grouse winter habitat selection. We
utilized used and available points to select individual 5 3 5-m raster pixels, which we then buffered by 100 m, 400 m, and 1,000 m. We quantiﬁed variables within a square
centered on each used and available pixel at 3 spatial scales:
205 3 205-m (0.04-km
2), 805 3 805-m (0.65-km 2), and
2,005 3 2,005-m (4-km 2). We calculated the percent of
total area covered by each of the 5 vegetation cover classes to quantify vegetation. To quantify topography, we processed a 900-m2 resolution digital elevation model (DEM) using
Spatial Analyst in ArcGIS 9.2 and used it to estimate slope and solar radiation for each pixel in the landscape. Solar radiation calculates how much sun a particular pixel receives dependent on slope and aspect. We estimated solar radiation using the hillshade command in Spatial Analyst using the angle and aspect of the sun during 15 January 2007 at 1300 hours (U.S. Navy Astronomical Applications Department
2007). We used the standard deviation of the DEM
elevations within each buffer size to calculate an index to describe the roughness of the landscape. Elevation was not included as a predictor variable for GIS habitat modeling because elevational migration of sage-grouse does not occur in the PRB, and minor differences in elevation at used and available locations were biol ogically irrelevant. In the northern PRB, mean elevation was 1,210 m (SE ¼ 3.8)
for available locations and 1,248 m (SE ¼ 3.9) for used locations. In the southern PRB, mean elevation was 1,363 m (SE ¼ 4.1) for available locations and 1,378 m (SE ¼ 3.4) for used locations. We used the density of CBNG wells as a
measure of the extent of energy development. Wells are the only segment of the energy footprint accurately mapped and publicly available for the entire PRB from the Wyoming Oil and Gas Conservation Commission and Montana Board of
Oil and Gas Conservation, and well density within a buffer is strongly correlated with other features of CBNG
development such as roads, ponds, and power lines (D. E.

Naugle, University of Montana, unpublished data).

Statistical Analyses
We employed logistic regression with used and available points for model selection and RSF model parameter
estimates (Boyce et al. 2002, Manly et al. 2002, Johnson et al. 2006). We pooled used locations of individual animals Doherty et al. /C15 Sage-Grouse Habitat and Energy Development 189 and made inferences at the population level (Design I;
Erickson et al. 2001, Manly et al. 2002).

We ﬁrst assigned variables into one of 3 model categories:
vegetation, topography, or energy development. Because no published landscape-scale studies existed upon which to base a priori models (Burnham and Anderson 2002), we tested all variables individually and removed variables with odds ratios overlapping 1.0. We tested all buffer distances for each variable and identiﬁed the scale that best represented sage- grouse habitat selection for each variable using log-like- lihood values. We then allowed the best scale for each
variable to compete with all possible combinations of other variables within the same category to identify the most parsimonious model. We used information-theoretic meth- ods (Burnham and Anderson 2002) to choose between
competing models by converting log-likelihood values
computed in logistic regression to Akaike’s Information Criterion (AIC) values. We brought models within 2 AIC
points to the next hierarchy of model selection. After
identifying the top model(s) within vegetation, topography, and energy development, we allowed models to compete
across categories to see if the additional information
increased model ﬁt.

We did not allow correlated predictors ( r /C21j 0.7j) in the same model at any level of model selection. If variables were correlated (r /C21j 0.7j), we chose the variable we felt had the greatest biological meaning according to known character- istics of winter sage-grouse habitat from published studies.

When variables were moderately correlated (i.e., j0.3j/C20 r , j0.7j), we checked for stability and consistency of regression coefﬁcient estimates as we added predictor variables to models. If a regression coefﬁcient switched signs or standard errors increased substantially when correlated variables were in the same model, we removed one variable from analysis if the other was an important predictor.

We evaluated whether sage-grouse avoided energy devel-
opment in winter by using AIC values to determine if the addition of CBNG wells/km
2 to the top habitat model
explained more information than habitat alone. We then
examined the resulting corresponding model coefﬁcient for CBNG wells to determine if sage-grouse avoided or were
attracted to energy development and to what degree. We
performed a bootstrap analysis to quantify the change in odds of use with the introduction of CBNG wells in the
form of 95% conﬁdence intervals around the odds ratios for differences in the number of wells. Because the best
approximating model had a high AIC weight ( w
i ¼

0.965), we used beta coefﬁcients from the best approximat- ing model for all computations (see results; Burnham and Anderson 2002). For each bootstrap data set ( n ¼ 5,000) we calculated and stored model coefﬁcients and the mean value for all used locations for each variable. We then repeated this bootstrap analysis, varying the number of CBNG wells in a 4-km
2 area from 0 wells to 22 wells, the full range of
well density we observed in our original data set. For each of the 5,000 simulations we computed the odds of use with the logistic equation. We then ordered these ratios and used a rankit adjustment (Chambers et al. 1983) to compute 2.5 % and 97.5 % percentiles for the upper and lower 95 %
conﬁdence interval bounds.

We then used the same bootstrap technique to quantify
how the amount of sagebrush within a 4-km
2 area affected
the odds of use in winter with and without CBNG
development (12.3 wells/4 km 2 and 0.0 wells/4 km 2,
respectively). We used the logistic equation to generate odds of use for each bootstrap dataset ( n ¼ 5,000) by
applying stored model coefﬁcients to mean values of
parameters at used locations while systematically varying percent sagebrush within 4 km
2 from 0 % to 100 % at 0.0
and 12.3 wells per 4 km 2. To test if the odds of use were signiﬁcantly different with the addition of CBNG we
computed the difference in odds generated from each
bootstrap data set with and without CBNG. Again, we
ordered odds ratios with and without CBNG and their
differences and used a rankit adjustment (Chambers et al. 1983) to compute 2.5% and 97.5% percentiles for the upper and lower 95 % conﬁdence interval bounds.

To turn our statistical model into a spatially explicit GIS habitat model, we employed a RSF model that had the
form:
wðxÞ¼ expðb
1x1 þ b2x2 þ/C1/C1/C1þ bkxkÞ; ð1Þ
where w(x) is the raw RSF value for each pixel in the
landscape, and x1, x2, ... xk represent values for vegetation, topography, and energy development generated from a
moving-window analysis for each pixel, and b
1, ... , bk are
the model parameters estimated with logistic regression (Boyce et al. 2002, Manly et al. 2002, Johnson et al. 2006).

We applied b-coefﬁcients from equation 1 to GIS layers in ArcView Spatial Analyst. The output was a new GIS layer that represents the RSF values generated from equation 1 for each individual 25-m
2 pixel for the entire landscape. We
created each component GIS layer by moving-window
analyses for key vegetation, topographic, and energy
development variables identiﬁed in model selection. These analyses resulted in summary statistics for each pixel in the GIS layer at the desired scale. We resampled sagebrush to a 900-m
2 pixel size because the time required to process a 4-
km2 area for 625 million pixels exceeded our computational capacity. Sagebrush resampled well and little information was lost when evaluating the 900-m 2 resampled sagebrush layer versus the original 25-m 2 resolution sagebrush layer ( r ¼ 0.934). Conifer resampled poorly ( r ¼ 0.793) so we kept this variable at the original pixel size.

We categorized RSF values into 5 ordinal 20 % quantile
bins representing progressively selected habitats. We
validated our spatial model with the test data set of sage- grouse locations collected during the winters of 2004 and 2007. We regressed the observed proportion of the test data set in each RSF bin against the expected proportion of use from the original RSF model to evaluate model ﬁt (Johnson et al. 2006). A good model ﬁt should have a high validation R
2 value, a slope not different from 1.0, and an intercept not different from zero (Johnson et al. 2006).
190 The Journal of Wildlife Management /C15 72(1)

RESULTS
Sagebrush at the 4-km 2 scale was the dominant variable in univariate space (Table 1). Sagebrush and grassland
accounted for .95% of the total vegetation cover at used locations, which explains their strong negative correlation ( r ¼/C0 0.78). Within a 4-km
2 area, used sites contained .75%
sagebrush cover intermixed with grassland. There was 14.5% more sagebrush at used (76.0%,S E ¼ 0.55) than at available sites (61.5 %,S E ¼ 0.61). Sage-grouse used sites averaged

19.1% (SE ¼ 0.53) grassland cover within a 4-km 2 area.

The best model for sage-grouse vegetation use consisted of sagebrush and riparian (4-km 2 scale), as well as conifer and barren (0.65-km2 scale; Table 2). The roughness index at a

0.65-km2 scale and slope were both important topographic predictors of sage-grouse use (Table 2). The number of
CBNG wells within a 4-km
2 area was the best model to
represent energy development (Table 1).

Model ﬁt increased when the best approximating models
from vegetation, topography, and energy development were combined (Table 3). We removed barren ground from the
ﬁnal vegetation model because it lacked stability and
Table 1. Vegetation, topographic, and energy development variables that we evaluated as potential landscape predictors of sage-grouse winter habitat selection, Powder River Basin, Montana and Wyoming, USA, 2005 and 2006. We used log-likelihoods to identify best scale at which selection occurred fo r individual variables along with confidence intervals of odds ratios that did not overlap 1 to choose variables for model selection.

Model category Variable Buffer area (km 2) Log-likelihood Odds ratio Upper 95 % CI Lower 95 % CI Vegetationa Sagebrush 4 /C0 799.550 1.052 1.060 1.044
Sagebrush 0.65 /C0 814.010 1.048 1.043 1.034
Sagebrush 0.04 /C0 825.694 1.030 1.035 1.024
Grass 4 /C0 877.583 0.972 0.980 0.964
Grass 0.04 /C0 878.044 0.982 0.987 0.976
Grass 0.65 /C0 884.551 0.980 0.987 0.973
Conifer 0.65 /C0 813.051 0.765 0.822 0.712
Conifer 0.04 /C0 833.587 0.793 0.859 0.732
Conifer 4 /C0 818.951 0.810 0.850 0.772
Riparian 4 /C0 851.246 0.843 0.882 0.805
Riparian 0.65 /C0 860.729 0.870 0.909 0.833
Riparian 0.04 /C0 889.368 0.958 0.979 0.938
Barren 0.65 /C0 890.643 0.897 0.940 0.856
Barren 4 /C0 890.197 0.866 0.919 0.816
Barren 0.04 /C0 898.349 0.960 0.987 0.934
Topography Roughness 0.65 b /C0 838.257 0.888 0.909 0.868 Roughness 0.04 /C0 844.885 0.815 0.850 0.782
Roughness 4 /C0 848.668 0.921 0.936 0.905
Solar radiation 0.0009 /C0 902.677 0.997 1.002 0.992
Slope 0.0009 /C0 863.384 0.879 0.907 0.852
Energy Development Distance to nearest well /C0 865.638 1.000 1.002 0.997 No. wells 4 /C0 857.717 0.961 0.985 0.939
No. wells 0.65 /C0 859.699 0.833 0.943 0.736
No. wells 0.04 /C0 863.083 0.434 1.102 0.171
a We excluded grass from further habitat models because of its correlation with sagebrush ( r ¼/C0 0.78). b Roughness ¼ index calculated using the SD of a digital elevation model.

Table 2. Log-likelihood (LL), number of parameters (K), Akaike’s Information Criterion value (AIC), change in AIC value from the top model (DAIC), and Akaike weight ( wi) of sage-grouse winter habitat selection for vegetation and topography models, Powder River Basin, Montana and Wyoming, USA, winters of 2005 and 2006.

Model LL K AIC DAIC wi
Vegetation models
Sagebrusha þ Conifer þ Riparian þ Barren /C0 716.337 5 1,442.674 0.000 0.998 Sagebrush þ Conifer þ Riparian /C0 723.772 4 1,455.544 12.870 0.002 Sagebrush þ Conifer þ Barren /C0 744.539 4 1,497.078 54.404 0.000 Sagebrush þ Conifer /C0 749.355 3 1,504.710 62.036 0.000 Sagebrush þ Riparian þ Barren /C0 780.350 4 1,568.700 126.026 0.000 Sagebrush þ Riparian /C0 787.762 3 1,581.524 138.850 0.000 Sagebrush þ Barren /C0 799.877 3 1,605.754 163.080 0.000 Topography models
Roughnessb þ Slopec /C0 835.881 3 1,677.762 0.000 0.798 Roughness /C0 838.257 2 1,680.514 2.752 0.202
Slope /C0 863.384 2 1,730.768 53.006 0.000
a Vegetation variables ¼ % cover of each Geographic Information System vegetation category within a selected buffer distance ( % sagebrush, barren and riparian within 4 km 2 þ % conifer and roughness within 0.65 km 2). b Roughness ¼ index calculated using the SD of a digital elevation model (DEM). c Slope ¼ slope of pixel calculated using a DEM.

Doherty et al. /C15 Sage-Grouse Habitat and Energy Development 191 consistency due to its correlation with roughness ( r ¼ 0.32).

When roughness and barren ground were in the same
model, the coefﬁcient for barren ground switched from a negative to a positive effect and its standard error increased, causing the odds ratio interval to overlap 1.0 (odds 0.96–

1.06). Roughness was a more stable predictor and was
unaffected by the inclusion of barren ground. The ﬁnal
combined model was 1.96 AIC points better when barren
ground was removed.

Sage-grouse selected large expanses of sagebrush with
gentle topography and avoided conifer, riparian, and energy development (Table 4). The addition of the average number of wells per 4-km
2 improved model ﬁt by 6.66 AIC points
(Table 3). An Akaike weight (wi ¼ 0.965) indicated that the model with both habitat and energy variables had over-
whelming support (Table 3). The resulting model coefﬁcients from the habitat and energy model indicate that after
adjusting for sage-grouse habitat preference, birds avoid CBNG development in otherwise suitable habitat (Table 4).

Our bootstrap analysis demonstrated that current legal
maximum well density on federal lands (approx. 12.3 wells/4 km
2, or 32-ha spacing) decreased the odds of sage-grouse
use by 0.30 compared to the average landscape selected by our radiomarked sage-grouse (odds 0.57 vs. 0.87; Fig. 1).

Sage-grouse were 1.3 times more likely to use winter habitat if CBNG development was not present. The odds of sage-
grouse winter habitat use increased with greater percentage sagebrush cover within 4 km 2 (Fig. 2a). The difference in odds of use with and without CBNG development was
statistically signiﬁcant at all levels of sagebrush ( P , 0.05); however, these differences were more pronounced in high- quality winter habitats dominated by sagebrush cover (Fig. 2b). Avoidance of CBNG was not relevant to winter habitat selection at low levels of sagebrush cover because sage- grouse showed strong avoidance of those areas prior to
development (Fig. 2a).

The best approximating model including vegetation,
topography, and energy variables accurately predicted an independent data set of 74 winter locations (validation R 2 ¼

0.98; Fig. 3). Using 6-, 7-, or 8-bin ordinal RSF models with quantile breaks did not change the strength or pattern of model validation. The slope of observed versus expected values did not differ from 1.0 (slope ¼ 1.14, 95% CI ¼ 0.87–

1.41) and the intercept did not differ from zero (/C0 2.85, 95% CI ¼/C0 1.06–4.9). The top 2 RSF classes accounted for

86.6% of the 435 locations used to build the RSF model
and 90.5% of the 74 locations used to test the winter habitat model (Fig. 3).

DISCUSSION
Our study is the ﬁrst to show that abundance of sagebrush at a landscape scale inﬂuences sage-grouse habitat selection in Table 3. Log-likelihood (LL), number of parameters ( K), Akaike’s Information Criterion value (AIC), change in AIC value from the top model ( DAIC), and Akaike weight ( wi) of sage-grouse winter habitat model selection, Powder River Basin, Montana and Wyoming, USA, winters of 2005 and 2006.

Modela LL K AIC DAIC wi
Vegetationb þ Topographyc þ CBNGd /C0 683.644 7 1,381.288 0.000 0.965 Vegetation þ Topography /C0 687.974 6 1,387.948 6.660 0.035 Vegetation þ CBNG /C0 718.083 5 1,446.166 64.878 0.000
Vegetation /C0 723.772 4 1,455.544 74.256 0.000
Topography þ CBNG /C0 826.657 3 1,659.314 278.026 0.000 Topography /C0 835.881 3 1,677.762 296.474 0.000
CBNG /C0 857.717 2 1,719.434 338.146 0.000
a Models represent the AIC best combination of variables within each model category. b Vegetation ¼ % sagebrush and riparian within 4 km 2 þ % conifer within 0.65 km 2. c Topography ¼ roughness of land within 0.65 km 2 þ slope. d CBNG ¼ no. of coal-bed natural gas wells/4 km 2.

Table 4. Logistic regression b-coefficients (SE) and odds ratios from the best model (Akaike wt ¼ 0.965) describing winter habitat selection and energy avoidance for sage-grouse, Powder River Basin, Montana and Wyoming, USA, 2005 and 2006.

Parameters Estimate SE Odds ratio Upper 95 % CI Lower 95 % CI Constant /C0 1.106 0.369
Roughnessa /C0 0.039 0.017 0.962 0.994 0.931
Slopeb /C0 0.102 0.022 0.903 0.943 0.865
Coniferc /C0 0.203 0.033 0.816 0.871 0.765
Sagebrushd 0.028 0.004 1.028 1.037 1.020
Ripariane /C0 0.131 0.026 0.877 0.923 0.834
CBNG wellsf /C0 0.035 0.014 0.966 0.992 0.940
a Roughness ¼ topographic index calculated as the SD of a digital elevation model (DEM) within 0.65 km 2. b Slope ¼ slope of pixel calculated from DEM.
c Conifer ¼ % conifer cover within 0.65 km 2.
d Sagebrush ¼ % sagebrush cover within 4 km 2.
e Riparian ¼ % riparian cover within 4 km 2.
f CBNG ¼ no. of coal-bed natural gas wells within 4 km 2. 192 The Journal of Wildlife Management /C15 72(1)
winter. Recent advances in RSF modeling and habitat
mapping using satellite imagery enabled us to document
what all major reviews on sage-grouse habitat requirements have suggested (Schroeder et al. 1999; Connelly et al. 2000, 2004; Crawford et al. 2004). At the largest scale evaluated (4 km2), sage-grouse selected for sagebrush and grassland
landscapes (.95% area) that were dominated by sagebrush (.75%) with little tolerance for other cover types.

Conversion of sagebrush negatively inﬂuences sage-grouse populations (Leonard et al. 2000, Smith et al. 2005). Sage- grouse avoided riparian areas at the 4-km 2 scale and conifer h a b i t a t sa n dr u g g e dl a n d s c a p e sa ta0 . 6 5 - k m2 scale, relationships that would have been less discernible at
broader spatial scales. Our roughness index was a much
stronger predictor than the rest of our suite of topographic variables, but slope further increased model ﬁt. Roughness is readily calculated from available DEMs and may be
applicable to other life stages for sage-grouse. In the only other sage-grouse landscape study that has evaluated habitat selection at multiple scales, birds selected large expanses (.1 km2) of sagebrush and avoided anthropogenic edge during the breeding season (Aldridge and Boyce 2007). Our
ﬁndings from winter in conjunction with those of Aldridge and Boyce (2007) highlight the need for landscape-scale research to gain further insight into sage-grouse ecology.

Our habitat model was highly predictive. We built our
model using sage-grouse locations collected during mild to average winter conditions and validated it in years with average temperatures or above-average snowfall. We do not know whether we deﬁned winter habitat broadly enough to include refugia necessary for birds to survive a 50- or 100- year winter storm event (Moynahan et al. 2006), but we
believe the model is useful to identify habitat available in most winters. Extreme events may move birds into rugged landscapes as they search for exposed sagebrush, thermal Figure 1. Reduction in the odds of sage-grouse winter habitat use versus available habitat with increasing coal-bed natural gas (CBNG) well density, Powder River Basin, Montana and Wyoming, USA, 2005–2006. Odds (solid line) and 95 % confidence intervals (dashed line) are based on 5,000 bootstrap samples with densities varying between 0–22 wells per 4 km 2, the range of CBNG development we observed in our sample of used and available points.

Figure 2. Odds of sage-grouse winter habitat use in relation to percent sagebrush cover per 4 km 2, Powder River Basin, Montana and Wyoming, USA, 2005–2006. Odds and 95 % confidence intervals are based on 5,000 bootstrap samples with sagebrush varying from 0 % to 100 %, with and without coal-bed natural gas (CBNG) development. a) The gray line represents CBNG development (12.3 wells/4 km 2,9 5 % CI small dashed line) and the black line represents no CBNG development (0.0 wells/4 km2, 95% CI large dashed line). b) The difference of means for odds of use with and without CBNG (black line minus grey line from part [a] above) is plotted against varying amounts of sagebrush cover per 4 km 2 (95% CI dashed line).

Doherty et al. /C15 Sage-Grouse Habitat and Energy Development 193 cover, and protection from high winds (Beck 1977, Hupp
and Braun 1989, Robertson 1991, Connelly et al. 2004).

A multi-scale approach is needed to understand the
relative importance of local and landscape factors inﬂuenc- ing sage-grouse habitat selection. Local vegetation measures have been the primary focus of sage-grouse habitat research to date (Eng and Schladweiler 1972, Beck 1977, Connelly et al. 2000, Crawford et al. 2004, Sauls 2006). Ideally, local variables should compete against landscape factors in an AIC framework to predict sage-grouse habitat use. Exami- nation of ecological processes at the landscape scale does not eliminate the need to understand habitat relationships at local scales; rather, it will likely require a combination of scales to completely understand how sage-grouse respond to their environment.

Our spatially explicit habitat model provides resource
managers with a practical tool to guide conservation
planning. Effective planning requires that we know which habitats are selected at landscape scales, where those habitats are located, and how species respond to disturbances. Recent advances in wildlife ecology enable biologists to develop RSF models that link resource use with changes in habitat quality and potential stressors (Manly et al. 2002, Johnson et al. 2004). Moreover, RSFs estimate the strength of selection and enable predictive equations to be linked in a GIS to depict spatial relationships across a planning region (Manly et al. 2002, Johnson et al. 2004). Spatially explicit planning tools should be used to prioritize landscapes with the
highest probability of supporting populations. Once iden- tiﬁed, local biologists provide on-site recommendations for how to best deliver on-the-ground conservation.

After adjusting for sage-grouse habitat preference, sage- grouse avoided energy development in otherwise suitable habitats in winter. Previous research has shown that
breeding sage-grouse in oil and gas ﬁelds avoid develop- ment, experience higher rates of mortality, or both
(Holloran 2005, Kaiser 2006, Aldridge and Boyce 2007).

Accumulating evidence of the impacts of energy develop- ment in sagebrush-steppe ecosystems extends beyond that of sage-grouse. Mule deer ( Odocoileus hemionus )a v o i d e d otherwise suitable habitats within 2.7–3.7 km of gas wells (Sawyer et al. 2006) and densities of Brewer’s sparrow
(Spizella breweri ) and sage sparrow ( Amphispiza belli ) declined 36–57 % within 100-m of dirt roads in gas ﬁelds (Ingelﬁnger and Anderson 2004). Some suitable winter
habitat remains undeveloped for sage-grouse in the PRB
(RSF bins 4 and 5; Fig. 3), but the anticipated addition of another 37,000 CBNG wells at 32-ha spacing has the
potential to affect .1.18 million ha of land. As remaining winter habitats are developed, and sage-grouse can no longer avoid CBNG, it is unclear whether birds will be able to adapt to a disturbance of this magnitude.

MANAGEMENT IMPLICATIONS
Sage-grouse avoidance of energy development in winter
shows that a comprehensive strategy is needed to maintain suitable habitats in all seasons. Identifying and setting aside areas of undeveloped, high-quality habitat within the project area should be top priority. Currently, only 0.5-km 2
(quarter-mile buffer) of land surrounding a lek is excluded from development, an area that is 8 times smaller than the scale at which individual sage-grouse selected winter
habitats (i.e., 4 km 2). Timing stipulations that restrict CBNG development within 3.2 km of a lek during the
breeding season (15 Mar–15 Jun) are insufﬁcient because they do not prevent infrastructure from displacing sage- grouse in winter. An additional stipulation in Montana that restricts new drilling activities within crucial winter range (1 Dec–31 Mar) only protects sage-grouse habitat during the winter in which the drilling is scheduled. Current
stipulations leave only a small fraction of the land
undeveloped, place no restrictions on the location of wells in winter habitat, and allow human access to all areas
throughout the life of the producing gas ﬁeld. Our spatially explicit winter habitat model can be used to identify areas in the PRB that provide the best remaining habitat for sage- grouse in winter.

ACKNOWLEDGMENTS
Comments by J. W. Hupp, J. W. Connelly, C. L. Aldridge
and one anonymous reviewer greatly improved this manu-
script. We thank landowners in the Powder River Basin that granted access to private lands. We thank J. Hess for her outstanding efforts in the ﬁeld. Funding for this work came from Bureau of Land Management ofﬁces in Montana and
Wyoming. Additional project support came from the
Bureau of Land Management (Washington, D.C.), United
States Department of Energy, Montana Department of
Fish, Wildlife and Parks, Wyoming Game and Fish
Department, National Fish and Wildlife Foundation,
National Science Foundat ion (EPS–CORE program),
Montana Cooperative Wildlife Research Unit, Petroleum
Association of Wyoming, Western Gas Resources Inc.,
Wolf Creek Charitable Foundation, Bighorn Environ-
mental Consulting, Anheuser-Busch Companies, Inc., and
the University of Montana.

Figure 3. Percent of sage-grouse use locations in each of 5 ordinal resource- selection function bins we used to build (black bars, n ¼ 436 locations from 2005 to 2006) and test (gray bars, n ¼ 74 locations from 2004 and 2007) the winter habitat model, Powder River Basin, Montana and Wyoming, USA. 194 The Journal of Wildlife Management /C15 72(1)

LITERATURE CITED
Aldridge, C. L., and M. S. Boyce. 2007. Linking occurrence and fitness to persistence: a habitat-based approach for endangered greater sage-grouse.

Ecological Applications 17:508–526.

Aldridge, C. L., and R. M. Brigham. 2003. Distribution, abundance, and status of the greater sage-grouse, Centrocercus urophasianus , in Canada.

Canadian Field-Naturalist 117:25–34.

Beck, T. D. I. 1977. Sage grouse flock characteristics and habitat selection in winter. Journal of Wildlife Management 41:18–26.

Beyer, H. L. 2004. Hawth’s analysis tools for ArcGIS. ,http://www. spatialecology.com/htools.. Accessed 15 May 2006.

Boyce, M. S., P. R. Vernier, S. E. Nielsen, and F. K. A. Schmiegelow. 2002. Evaluating resource selection functions. Ecological Modeling 157:
281–300.

Bureau of Land Management. 2003 a. Final environmental impact state- ment and proposed plan amendment for the Powder River Basin Oil and Gas Project. ,http://www.wy.blm.gov/nepa/prb-feis/index.htm.. Ac- cessed 18 Mar 2007.

Bureau of Land Management. 2003 b. Montana statewide final oil and gas EIS and amendment of the Powder River and Billings Resource Management Plans. ,http://www.mt.blm.gov/mcfo/CBNG/eis/index. html.. Accessed 18 Mar 2007.

Burnham, K. P., and D. R. Anderson. 2002. Model selection and inference:
a practical information-theoretic approach. Second edition. Springer- Verlag, New York, New York, USA.

Chambers, J. M., W. S. Cleveland, B. Kleiner, and P. A. Tukey. 1983.

Graphical methods for data analysis. Duxbury Press, Boston, Massachu- setts, USA.

Connelly, J. W., and C. E. Braun. 1997. Long-term changes in sage grouse Centrocercus urophasianus populations in western North America. Wildlife Biology 3:229–234.

Connelly, J. W., H. W. Browers, and R. J. Gates. 1988. Seasonal move- ments of sage grouse in southeastern Idaho. Journal of Wildlife Management 52:116–122.

Connelly, J. W., S. T. Knick, M. A. Schroeder, and S. J. Stiver. 2004.

Conservation assessment of greater sage-grouse and sagebrush habitats.

Western Association of Fish and Wildlife Agencies, Cheyenne, Wyoming, USA.

Connelly, J. W., M. A. Schroeder, A. R. Sands, and C. E. Braun. 2000.

Guidelines to manage sage grouse populations and their habitats.

Wildlife Society Bulletin 28:967–985.

Crawford, J. A., R. A. Olson, N. E. West, J. C. Moseley, M. A. Schroeder, T. D. Whitson, R. F. Miller, M. A. Gregg, and C. S. Boyd. 2004.

Ecology and management of sage-grouse and sage-grouse habitat. Journal of Range Management 57:2–19.

Eng, R. L., and P. Schladweiler. 1972. Sage grouse winter movements and habitat use in central Montana. Journal of Wildlife Management 36:141– 146.

Erickson, W. P., T. L. McDonald, K. G. Gerow, S. Howlin, and J. W.

Kern. 2001. Statistical issues in resource selection studies with radio- marked animals. Pages 209–242 in J. J. Millspaugh and J. M. Marzluff, editors. Radiotracking and animal populations. Academic Press, San Diego, California, USA.

Giesen, K. M., T. J. Schoenberg, and C. E. Braun. 1982. Methods for trapping sage grouse in Colorado. Wildlife Society Bulletin 10:224–231.

Holloran, M. J. 2005. Greater sage-grouse ( Centrocercus urophasianus ) population response to natural gas field development in western Wyoming. Dissertation, University of Wyoming, Laramie, USA.

Homer, C. G., T. C. Edwards, Jr., R. D. Ramsey, and K. P. Price. 1993.

Use of remote sensing methods in modeling sage grouse winter habitat.

Journal of Wildlife Management 57:78–84.

Hupp, J. W., and C. E. Braun. 1989. Topographic distribution of sage grouse foraging in winter. Journal of Wildlife Management 53:823–829.

Ingelfinger, F., and S. Anderson. 2004. Passerine response to roads associated with natural gas extraction in a sagebrush steppe habitat.

Western North American Naturalist 64:385–395.

Jennrich, R. I., and F. B. Turner. 1969. Measurement of non-circular home range. Journal of Theoretical Biology 22:227–237.

Johnson, C. J., S. E. Nielsen, E. H. Merrill, T. L. McDonald, and M. S.

Boyce. 2006. Resource selection functions based on use–availability data:
theoretical motivation and evaluation methods. Journal of Wildlife Management 70:347–357.

Johnson, C. J., D. R. Seip, and M. S. Boyce. 2004. A quantitative approach to conservation planning: using resource selection functions to map the distribution of mountain caribou at multiple spatial scales. Journal of Applied Ecology 41:238–251.

Kaiser, R. C. 2006. Recruitment by greater sage-grouse in association with natural gas development in western Wyoming. Thesis, University of Wyoming, Laramie, USA.

Knick, S. T., D. S. Dobkin, J. T. Rotenberry, M. A. Schroeder, W. M.

Vander Haegen, and C. Van Riper, III. 2003. Teetering on the edge or too late? Conservation and research issues for avifauna of sagebrush habitats. Condor 105:611–634.

Leonard, K. M., K. P. Reese, and J. W. Connelly. 2000. Distribution, movements and habitats of sage grouse, Centrocercus urophasianus, on the Upper Snake River Plain of Idaho: changes from the 1950s to the 1990s.

Wildlife Biology 6:265–270.

Manly, B. F. J., L. L. McDonald, D. L. Thomas, T. L. McDonald, and W.

P. Erickson. 2002. Resource selection by animals: statistical design and analysis for field studies. Second edition. Kluwer Academic, Dordrecht, The Netherlands.

Moynahan, B. J., M. S. Lindberg, and J. W. Thomas. 2006. Factors contributing to process variance in annual survival of female greater sage- grouse in north-central Montana. Ecological Applications 16:1529–1538.

Patterson, R. L. 1952. The sage grouse in Wyoming. Wyoming Game and Fish Commission and Sage, Denver, Colorado, USA.

Remington, T. E., and C. E. Braun. 1985. Sage grouse food selection in winter, North Park, Colorado. Journal of Wildlife Management 49:1055– 1061.

Robertson, M. D. 1991. Winter ecology of migratory sage grouse and associated effects of prescribed fire in southern Idaho. Thesis, University of Idaho, Moscow, USA.

Sauls, H. S. 2006. The role of selective foraging and cecal microflora in sage-grouse nutritional ecology. Thesis, University of Montana, Mis- soula, USA.

Sawyer, H., R. M. Nielson, F. Lindzey, and L. L. McDonald. 2006. Winter habitat selection of mule deer before and during development of a natural gas field. Journal of Wildlife Management 70:396–403.

Schroeder, M. A., C. L. Aldridge, A. D. Apa, J. R. Bohne, C. E. Braun, S.

D. Bunnell, J. W. Connelly, P. A. Diebert, S. C. Gardner, M. A.

Hilliard, G. D. Kobriger, and C. W. McCarthy. 2004. Distribution of sage-grouse in North America. Condor 106:363–376.

Schroeder, M. A., J. R. Young, and C. E. Braun. 1999. Sage grouse (Centrocercus urophasianus). Account 425 in A. Poole and F. Gill, editors.

The birds of North America. The Academy of Natural Sciences, Philadelphia, Pennsylvania, and The American Ornithologists’ Union, Washington, D.C., USA.

Smith, J. T., L. D. Flake, K. F. Higgins, G. D. Kobriger, and C. G.

Homer. 2005. Evaluating lek occupancy of greater sage-grouse in relation to landscape cultivation in the Dakotas. Western North American Naturalist 65:310–320.

U.S. Navy Astronomical Applications Department. 2007. Altitude and azimuth of the sun during one day. ,http://aa.usno.navy.mil/ ..

Accessed 7 Mar 2007.

Wakkinen, W. L., K. P. Reese, J. W. Connelly, and R. A. Fischer. 1992.

An improved spotlighting technique for capturing sage grouse. Wildlife Society Bulletin 20:425–426.

Welch, B. L., F. J. Wagstaff, and J. A. Roberson. 1991. Preference of wintering sage grouse for big sagebrush. Journal of Range Management 44:462–465.

Associate Editor: Hupp.

Doherty et al. /C15 Sage-Grouse Habitat and Energy Development 195 Prioritizing Conservation of Ungulate Calving Resources in Multiple-Use Landscapes
Matthew R. Dzialak *, Seth M. Harju, Robert G. Osborn ¤, John J. Wondzell, Larry D. Hayden-Wing, Jeffrey B. Winstead, Stephen L. Webb
Hayden-Wing Associates, LLC, Natural Resource Consultants, Laramie, Wyoming, United States of America Abstract
Background: Conserving animal populations in places where human activity is increasing is an ongoing challenge in many parts of the world. We investigated how human activity interacted with maternal status and individual variation in behavior to affect reliability of spatially-explicit models intended to guide conservation of critical ungulate calving resources. We studied Rocky Mountain elk ( Cervus elaphus ) that occupy a region where 2900 natural gas wells have been drilled.

Methodology/Principal Findings: We present novel applications of generalized additive modeling to predict maternal status based on movement, and of random-effects resource selection models to provide population and individual-based inference on the effects of maternal status and human activity. We used a 2 62 factorial design (treatment vs. control) that included elk that were either parturient or non-parturient and in areas either with or without industrial development.

Generalized additive models predicted maternal status (parturiency) correctly 93% of the time based on movement. Human activity played a larger role than maternal status in shaping resource use; elk showed strong spatiotemporal patterns of selection or avoidance and marked individual variation in developed areas, but no such pattern in undeveloped areas. This difference had direct consequences for landscape-level conservation planning. When relative probability of use was calculated across the study area, there was disparity throughout 72–88% of the landscape in terms of where conservation intervention should be prioritized depending on whether models were based on behavior in developed areas or undeveloped areas. Model validation showed that models based on behavior in developed areas had poor predictive accuracy, whereas the model based on behavior in undeveloped areas had high predictive accuracy.

Conclusions/Significance: By directly testing for differences between developed and undeveloped areas, and by modeling resource selection in a random-effects framework that provided individual-based inference, we conclude that: 1) amplified selection or avoidance behavior and individual variation, as responses to increasing human activity, complicate conservation planning in multiple-use landscapes, and 2) resource selection behavior in places where human activity is predictable or less dynamic may provide a more reliable basis from which to prioritize conservation action.

Citation: Dzialak MR, Harju SM, Osborn RG, Wondzell JJ, Hayden-Wing LD, et al. (2011) Prioritizing Conservation of Ungulate Calving Resources in Multiple-Use Landscapes. PLoS ONE 6(1): e14597. doi:10.1371/journal.pone.0014597 Editor: Michael Somers, University of Pretoria, South Africa Received June 23, 2010; Accepted December 15, 2010; Published January 26, 2011 Copyright: /C2232011 Dzialak et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Funding: Pioneer Natural Resources Company funded this work. Hayden-Wing Associates, LLC provided in-kind contributions including travel costs associate d with dissemination of this work, and materials associated with data collection and analysis including radio-telemetry equipment, GPS devices, and statistical analysis software. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Competing Interests: This work was funded by a commercial source, Pioneer Natural Resources Company. Hayden-Wing Associates, LLC provided in-kind contributions including travel costs associated with dissemination of this work, and materials associated with data collection and analysis inclu ding radio- telemetry equipment, GPS devices, and statistical analysis software. Having received funding from this source, the consultancy could reasonably b e perceived as a financial competing interest. This does not alter the authors’ adherence to all the PLoS ONE policies on sharing data and materials. * E-mail: [email protected]
¤ Current address: New Mexico Game and Fish Department, Santa Fe, New Mexico, United States of America Introduction
Identifying resources associated with critical life-history phases in ungulates is a conservation priority. Winter range, parturition areas, and migration routes are important seasonal habitats in North America that provide resources necessary for survival and reproduction such as high-quality forage, reduced exposure to inclement conditions, and reduced risk of predation [1], [2]. In the Intermountain West USA, these habitats have become increas- ingly fragmented. Here, energy development is of broad
conservation interest because its prominence has increased in recent decades along with concern about its potential impact on wildlife and their habitats [3]. Understanding how human activity such as resource extraction interacts with wildlife, and developing tools to guide conservation planning in areas where human activity is widespread or increasing are ongoing challenges in conservation science. In this paper, we investigated interactions between human activity associated with energy development and resource selection by female Rocky Mountain elk ( Cervus elaphus ) during calving season with the larger goal of informing conservation planning for ungulates in places where human activity is widespread or increasing. We asked: 1) what are the relative influences of maternal status (parturiency) and human activity on resource selection, 2) to what extent does behavior vary among individuals, PLoS ONE | www.plosone.org 1 January 2011 | Volume 6 | Issue 1 | e14597 3) how does individual variation interact with human activity, and 4) how can our findings be applied in conservation planning and decision making? First, we describe a novel application of generalized additive models for designating maternal status using movement data. Model-based methods to designate maternal status were necessary because data bearing directly on maternal status were unavailable for most elk. Second, we estimated group- dependent random-effects resource selection functions (RSFs) to identify how resource selection patterns differed relative to maternal status and human activity, and how behavior varied among individuals. We applied results by developing and validating group-dependent predictive maps of critical calving resources, and quantifying discrepancies among maps relative to predictive accuracy.

A key approach in studying wildlife-human interaction is resource selection modeling [4]. Resource selection is a funda- mental ecological process that structures animal movement and distribution [5]. The choices animals make as they move throughout the landscape reflect trade-offs between selecting resources that meet their needs for survival and reproduction, and minimizing perceived risk of harm – such risk often is a function of interaction with predators or humans [6], [7], [8]. A set of analytical methods used with increasing frequency for investigating resource selection in animals is the estimation of RSFs [9]. RSFs describe the relative probability of occurrence of animals as a function of behavioral responses to features of the environment.

The probability of occurrence is described as relative because RSFs are estimated in a use-versus-availability framework in which selection is quantified relative to available but presumed non-used features. Environmental features can include a wide range of variables such as vegetation, terrain, group/herd size, risk of predation, or human-modifications of the landscape [10], [11].

RSFs have strong application in conservation planning where wildlife-human interaction is a concern; specifically, in establishing a spatially explicit basis from which to prioritize conservation action such as reclamation, mitigation, or minimizing human activity in particular habitats.

Areas associated with parturition are important because female ungulates make resource-related choices that affect offspring development during gestation and provisioning of recently born calves that are susceptible to malnutrition and predation [12], [13].

While there is little evidence of consistent parturition site fidelity in many ungulates, strong fidelity among females to seasonal ranges, particularly around calving time, has been demonstrated [2], [14].

New light has been shed on the adaptive significance of resource selection during the period that encompasses reproductive activity in many vertebrates through the study of maternal effects – developmental mechanisms by which parents translate their environmental experience into adaptive variation in their offspring [15], [16], [17]. The adaptive significance of resource selection during reproductive periods suggests that conservation strategies designed around ungulate parturition areas might be most reliable when based on parturient females rather than samples including both parturient and non-parturient individuals.

Another key feature of animal ecology that warrants further attention as part of conservation planning is variation in behavior among individuals. Individual variation is widespread and well- known in many animal species [18] and can reflect long-term selection for a given trait or learned behavior [19], [20]. Fitness is influenced by the choices individuals make in terms of resource selection because each resource type has particular costs and benefits to the individual [21]. In risky or rapidly changing environments such as those in which human activity is increasing, optimal behavioral strategies may vary among individuals on a situation-specific basis [22] making it difficult to generalize behavior across the population and thus effectively guide conservation planning.

Methods
Study area
The 1845 km 2 study area encompassed northern portions of the Raton basin in south-central Colorado, USA. Topography is rugged with steep slopes, rocky outcrops, ridges, and valleys ranging in elevation from 2000–3000 m. Mean annual precipita- tion is about 40–53 cm depending on elevation [23]. Vegetation includes conifer forest, montane shrub, and grassland. Dominant species include ponderosa pine ( Pinus ponderosa ), one-seed juniper (Juniperus monosperma), two-needle pinyon ( Pinus edulis), Gambel oak (Quercus gambelii ) which commonly forms shrub-thickets on southern aspects, antelope bitterbrush ( Purshia tridentata), skunkbush sumac ( Rhus trilobata ), and willow ( Salix spp.) in riparian areas.

Predators of elk (including neonates) include black bear ( Ursus americanus), mountain lion ( Felis concolor ), and coyote ( Canis latrans ); no wolf ( Canis lupus ) pack occurred in the study area. The study area encompassed historic and ongoing energy development.

Bituminous coal mining was a dominant land use during 1873– 1970. Coal-bed methane development was initiated in Raton basin in 1982 and accelerated in the late 1990s [24], [25]. In 2009, there were about 2900 wells associated with methane development in the Basin (Figure 1).

Capturing elk
In February and March 2006–2009, helicopter net-gunning was used to capture yearling (1.5 years) and adult ( $2.5 years) female elk throughout and adjacent to the gas field. Elk were fitted with Global Positioning Systems (GPS) collars (TGW-3590, Telonics, Inc., Mesa, AZ 85204) configured with store-on-board and Very High Frequency (VHF) beacon options. Twenty-five female elk were fitted with GPS collars in 2006, 40 in 2007, 50 in 2008, and 50 in 2009. GPS collars attempted to record location information every 3 h resulting in a maximum of 8 locations/elk/day. Age of elk was estimated based on dental eruption and wear patterns [26].

Blood samples were collected from captured elk in 2008 and 2009 to determine pregnancy, but not from elk captured in previous years. Animal capture and handling protocols were approved by the Colorado Division of Wildlife (Permit #s 06TR1083,
07TR1083, 08TR1083 and 09TR1083A001).

Grouping Elk Relative to Maternal Status and Human
Activity
We predicted maternal status (parturient versus non-parturient) by using generalized additive models (GAMs) to parameterize response curves depicting daily movement of elk during calving season. GAMs are semi-parametric extensions of generalized linear models [27]. The central concept is that the function of a covariate is estimated nonparametrically from the data by means of scatterplot smoothers. The functional form of the relationship between the response and covariate(s) is therefore determined by the data rather than being restricted to a parametric form [28].

Formally, the linear regression model
y~az
X
I
i~1
bi (Xi )ze
is generalized by modeling y as being related to covariates additively by
Ungulate Conservation
PLoS ONE | www.plosone.org 2 January 2011 | Volume 6 | Issue 1 | e14597 g(y)~az
XI
i~1
fi (Xi )ze
where g is the link function and e is a random error term.

Functions fi may be linear or nonparametric functions defined by smoothers such as smoothing splines or locally estimated scatterplot smoothers (loess; also referred to as locally estimated, or weighted, polynomial regression). Smoothers provide a series of data summaries of the response that are specific to regions of the covariates; a well known smoother is the moving average [29]. The amount of smoothing is calibrated by the size of the neighborhood, or percentage of the data points, over which averaging is done; a quantity known as span. A larger span yields a smoother data summary (less curvature) whereas a smaller span yields a less smooth data summary (more curvature). This data summary ( i.e., the effective number of parameters of a smoother) is described by the quantity equivalent degrees of freedom ( df). Span is related inversely to df so as span increases df decreases [27]. Data that are best described by a straight line (a single parameter comprises the smoother) correspond to a span of 100% and thus 1 df. Conversely, data that are best described by gradients or turning points (several or many parameters comprise the smoother) correspond to a smaller span and thus to a larger number of df [27], [30]. The analyst may set the span by specifying df based on visual examination of the data, or implement generalized cross validation methods in which appropriate df are identified automatically given the data. From a GAM perspective, movement of female ungulates during the calving season likely contains information on maternal status. Restlessness and the seeking of solitude characterize imminent parturition [31], [32], [33] and establish a general pattern of increased daily movement pre-partum and decreased movement post-partum relative to barren females (Figure 2a; [34], Figure 1. The Raton Basin gas field and adjacent areas in south-central Colorado, USA. Each natural gas well is encircled by a 1-km buffer (shaded region). We designated elk locations occurring within the buffered region to be ‘‘inside of the gas field’’ whereas locations adjacent but external to the buffered region were designated ‘‘outside of the gas field’’. doi:10.1371/journal.pone.0014597.g001
Ungulate Conservation
PLoS ONE | www.plosone.org 3 January 2011 | Volume 6 | Issue 1 | e14597 [35], [36]). We hypothesized that during calving time, movement associated with restlessness and solitude seeking in parturient elk provides a general pattern of complexity that is not observed in non-parturient females, and we predicted that smoothers associ- ated with movement data on parturient elk consistently would be comprised of more parameters and thus more df than smoothers associated with non-parturient elk.

We used GAMs to regress distance traveled within a 24 h period (using locations recorded at 1200 h on consecutive days) against date. Date encompassed the 15 May – 1 July calving season [33] in each of the 4 years comprising the study period.

We used the generalized cross validation option to assign df to response curves depicting daily changes in the distance moved by elk between successive locations. We used PROC GAM in SAS H (SAS Institute, Inc., Cary, North Carolina, USA), specifying automated calculation of df using the generalized cross validation method and loess smoothing, to assign df to response curves depicting calving season movement patterns in each elk. We established the following prediction: movement described by ,3d f in GAMs would depict non-parturiency whereas movement
described by $3 df would depict parturiency. This prediction was based on two observations. First, the shape of the response curve depicting relatively simple movement, as in non-parturient individuals, appears to correspond with #2 regions that differ in Figure 2. Movement in female ungulates during calving time. Generalized from the literature [34], [35], [36], parturient females often will make long-distance movements associated with pre-parturient ‘‘restlessness’’ within days of parturition and then exhibit reduced movement associated with provisioning the neonate (a). This pattern may provide a quantifiable distinction between parturient and non-parturient females. Also shown (b) is parameterization (equivalent degrees of freedom) as a function of span based on generalized additive models of distance moved between consecutive locations in 103 female Rocky Mountain elk. Loess smoothing and automated calculation of degrees of freedom using the generalized cross validation method were specified.
doi:10.1371/journal.pone.0014597.g002
Ungulate Conservation
PLoS ONE | www.plosone.org 4 January 2011 | Volume 6 | Issue 1 | e14597 slope, each of which being calibrated over a span of $50% of the data (Figure 2a; [34], [35]). Second, by plotting df as a function of span, our data show that a span of 50% corresponds to ,3 df (Figure 2b).

We tested the GAM approach using blood samples obtained from elk during capture in 2008 and 2009, and field observation of females and calves in those same years. Blood sera were tested for presence of pregnancy-specific protein-B (PSPB; BioTracking, LLC, Moscow, Idaho, USA; [37]). We conducted field observation from dawn to 0900 and from 1800 until dusk using binoculars and spotting scopes to watch for behavior that suggested a maternal bond between a female and calf. This behavior included nursing and licking bouts, traveling as a female/calf unit, and heightened attentiveness between a female and calf [13], [38]. Females for which PSPB testing indicated pregnancy and for which field observation suggested a strong female-calf bond were designated as parturient. Determining non-parturiency is never definitive; however, we designated females as non-parturient with negative PSPB results and for which field observation was unable to associate the female with a calf [13]. We thus established two groups of elk relative to maternal status by which we analyzed resource selection.

Similarly, we established two groups of elk relative to human activity: elk occupying developed areas versus elk occupying undeveloped areas. Elk locations occurring within 1 km of a gas well were considered to be in the developed area (Figure 1).

Human activity was apparent in the areas we called undeveloped including some ranching and residences; however, no industrial development occurred in undeveloped areas and human activity was limited relative to developed areas. Given the short temporal window within which we conducted these analyses, calving season use areas (100% minimum convex polygons) generally occurred wholly within developed areas or wholly within undeveloped areas. Only a small number of elk occupied both areas during the calving season – we discuss these elk separately. We estimated minimum convex polygons because the temporal window of the study was relatively short, calving season use areas comprised a portion of the annual use areas, and we wanted to err on the side of inclusiveness rather than potentially omitting a portion of critical calving range from the analysis. In comparing resource selection between elk occupying developed versus undeveloped areas, a key assumption is that human activity associated with energy development is the primary difference between areas and that other factors to which elk respond were similar between areas.

We compared landscape and habitat covariate values between areas to inform this assumption.

Random-effects Resource Selection Modeling
Modeling variables as random effects can improve our
understanding of resource selection, which will enhance the practical application of RSFs in management decisions [4], [39].

Random effects models assume that sample units are drawn at random from a larger population and that the data are structured hierarchically ( i.e., within subject responses are more similar than between-subject responses). Mechanistically, assumptions of ran- dom-effects models are: (1) the random effects are distributed normally with mean equal to 0 and unknown variance, (2) within- group correlation is constant through time, and (3) the analyst has correctly specified the variance-covariance structure (see [39] for a review of the application of random-effects models in resource selection analysis). In RSF models, it is appropriate to model intercepts and/or covariates as random effects when variance among sample units is of interest, animal response to gradients in a resource is suspected, or population-level inference is of interest.

Random intercepts can account for unbalanced data, correlation among observations, and provide improved model fit and
parameter estimation [39] ( but see [11]). Random-effects (or mixed-effects) RSF models provide information on individual behavior, how individuals contribute to population-level observa- tions, and how their responses to a resource may change as a function of its availability – a process known as a functional response [40], [41]. Analytical approaches to model functional responses in resource selection are particularly important when there is a trade-off in selection for a particular resource [4], [40], which would be the case if human activity is perceived as a risk of harm [7].

We incorporated random effects into the use versus availability design [9] in which covariates representing important resources are compared at used and available (but presumed non-used) locations using
^ww(x)~exp(^bb
0z^bb1x1z^bb2x2z:::z^bbnxn)
where ^ww(x) is the relative probability of use as a function of covariates xn with coefficients ^bbn estimated from logistic regression.

Availability was defined for each elk by including random locations within 100% minimum convex polygon seasonal use area estimates; the number of random locations generated was 3 times the number of used locations for each elk. We examined resource selection within seasonal use areas ( i.e.,3
rd order
selection). We modeled resource selection separately during day and night because we expected behavior of elk to differ between day and night. We assigned time of day at random to available locations for day versus night comparisons; times assigned to available locations corresponded to times associated with used locations (e.g., every 3 hours on the sampled hour). Day models included the times 0900, 1200, 1500, and 1800 h whereas night models included 0000 and 0300 h. Using a Geographic Informa- tion System (GIS; ArcGIS 9.2), we calculated 7 covariates at used and random locations (Table 1). Four of these covariates including cover type, slope, elevation, and habitat edge density were calculated at locations both within and outside of the gas field.

Three covariates including road density, distance to a human-built structure, and industrial development footprint (Table 1) were calculated only within the gas field because human activity in areas adjacent but outside of the gas field was less intense and not associated with industrial development. Raster data for cover type were developed from annual aerial photography of the study area, terrain covariates were calculated from a 30-m resolution digital elevation model, and human activity covariates including roads, structures, and industrial development footprint were heads-up- digitized from aerial photography and analyzed as year-specific covariates (Table 1). We used Spatial Analyst in ArcGIS to extract values from raster data for all covariates.

We estimated a three-level random-effects model in which locations i =1 … I occurred within strata representing individual elk j =1 … J [39]. Considering a random intercept and random coefficients, the RSF is estimated by
g(x)~b
0zb1x1ij z:::zbk xkij zckj xkj zc0j
where covariates k (k =1 … K) have values x, c0j is the random intercept and ckj is the random coefficient of xk for elk j, which is the difference in the intercept and coefficient for elk j from the mean population-level intercept b0 and coefficient bkj, respectively.

We estimated models using the GLIMMIX procedure in SAS H.

We specified the conditional probability distribution of the data as Ungulate Conservation
PLoS ONE | www.plosone.org 5 January 2011 | Volume 6 | Issue 1 | e14597 binomial and used a logit link function. We included ordinal date as a class variable and specified a variance components covariance structure in which random intercepts for ordinal date were nested within each individual to address within-day autocorrelation among locations. R-side, or marginal, random-effects models estimated using GLIMMIX provide conditional coefficient estimates for each individual, that is, estimates for individual animals that are conditional on the distribution of coefficient estimates across all individuals in the population. To estimate marginal ( i.e., population-level) coefficients we assumed that conditional coefficients for each elk represented a random sample from a normal distribution with the mean of that distribution representing the population-level effects of covariates on the probability of use [42]. We estimated marginal coefficients using ^/C22bb/C22bb
k ~ 1
J
XJ
j~1
^bbkj
where ^bbkj was the estimate of coefficient k for elk j [43], [44], and we estimated variance using
var ^bbk
/C16/C17
~ 1
J{1
XJ
j~1
(^bbkj {^/C22bb/C22bbk )2
Population-level coefficient estimates (^/C22bb/C22bbk) are similar to the average of the conditional estimates because the conditional coefficient estimates are constrained to have means ^bbk [44]. In all, five groups of elk were available for comparison: 1) all sample elk inside of the gas field, 2) elk predicted to be parturient based on GAMs inside the gas field, 3) field-observed parturient elk inside of the gas field, 4) field-observed non-parturient elk inside of the gas field, and 5) all sample elk outside of the gas field (sample size was too low to develop maternal groups; see Results).

Mapping Responses
We applied the results from random-effects models to map relative probability of use across the study area which encom- passed developed and undeveloped areas. We used marginal coefficients from logistic regression to derive an RSF at a resolution of 30 m using
w(x)~exp
XK
k~1
b/C3
k xk
!
where covariates k (k =1 … K) have values x [9]. We were interested in differences in resource selection patterns depending on maternal status and between elk that occupied developed versus undevel- oped areas. It is important to note that, although we analyzed an independent group of field-observed parturient elk to facilitate comparison of marginal estimates, these elk were withheld from final RSF development as a validation sample. Thus, we examined differences in how the relative probability of use was assigned throughout the landscape depending on whether this probability was based on models of: 1) all sample elk inside the gas field, 2) elk predicted to be parturient based on GAMs inside the gas field, and 3) all sample elk outside of the gas field (groups 1,2, and 5 above).

We conducted this examination using the following methods separately for day and night. We developed annual RSF maps for each of the 3 groups identified above. We estimated quantiles in SASH (PROC RANK, PROC MEANS) by which pixels
comprising the raster surface were partitioned into 5 equal-sized subsets based on pixel value. In GIS we reclassified RSF values based on quantiles establishing 5 ranks of the relative probability of use (1 = low probability, 5 = high probability). We summed within-year maps across all years and ranked relative probability of use as described above yielding 6 multi-year predictive maps (day map and night map for each group) with relative probability of use ranging from 1 (low) to 5 (high). We validated predictive maps using locations from 24 elk that were observed in the field to be Table 1. Covariates used in random-effects resource selection models and their descriptions.

Covariate Description
Slope Digital elevation model (DEM) provided at a resolution of 1.52 m and re-sampled for covariate calculation to a resolution of 30 m and measured in degrees. Values for slope calculated at point locations in our study ranged from 0.02 (flat ground) to 41.8 degrees (very steep).

Elevation DEM re-sampled for covariate calculation to a resolution of 30 m and measured in m. Elevation at sample points in our study ranged from 1918.7–2970.0 m.

Cover type Raster dataset compiled from 1-foot resolution aerial imagery of the study area using Image Analysis TM and re-sampled to 30 m
resolution for covariate calculation. From an elk-centric perspective, habitat in the Raton basin functioned in one of two ways:
as security cover or as forage resources. Raster cells were assigned one of two values (binary covariate) representing habitat that functioned to provide cover versus habitat that did not function to provide cover. All tree or oak-thicket dominated habitats were considered security cover, whereas all shrub and grassland dominated habitats were considered non-cover.

Edge density Density of line features depicting the interface of cover and non-cover habitat calculated for the central grid cell within a 990 m
2 moving window. Values in our study were 22.2–136.0 km/km 2.

Road density Density of line features depicting roads calculated for grid cells at the center of a 990 m 2 moving window. Road density was calculated as a year-specific variable – that is, new road features were added to the data set as annual aerial imagery became available. Values in our study, calculated within the gas field only, were 0.0–7.6 km/km 2.

Distance to structure Linear distance from a sample location to a human-built structure including houses, agricultural facilities, and industrial facilities. Distance to structure was a year-specific variable and was analyzed as natural log transformed distance +0.1 to allow its magnitude to decrease with increasing distance. Values in our study, calculated within the gas field only, were 0.0–5854.9 m.

Industrial development footprint Area (density) of physically modified ground calculated from aerial photography as a year-specific variable. Ph ysically modified ground primarily reflected industrial development including well pads, pipe yards, pipelines, construction areas, or clearing for facilities development. Industrial development footprint was calculated for the central grid cell within a 990 m 2 moving window. This covariate excludes physical disturbance associated with roads. Values in our study, calculated only within the gas field, were 0.0–0.62 km 2/km2.
doi:10.1371/journal.pone.0014597.t001
Ungulate Conservation
PLoS ONE | www.plosone.org 6 January 2011 | Volume 6 | Issue 1 | e14597 parturient (see below); these elk were withheld from final RSF development so they represent an independent validation sample.

Locations from these elk were plotted on multi-year predictive maps. We tested whether the number of locations that occurred within each predicted probability of use rank (1–5) differed from expectation using a chi-square test for specified proportions (PROC FREQ, SAS H). To provide a measure of the amount to which a map that validated well differed from a map that validated poorly, we calculated the number of pixels comprising the raster surface that differed between maps in terms of relative probability rank.

Results
Capturing Elk and the Environment Inside versus Outside of the Gas Field
We fitted 25 female elk with GPS collars in 2006, 40 in 2007, 50 in 2008, and 50 in 2009. The entire within-year sample was unavailable for these analyses because the analyses spanned a relatively short period during calving (Table 2), and in each year several elk moved from the study area to alpine habitat for parturition. Inadequate sample size ( i.e., #2 elk comprising maternal groups) in undeveloped areas restricted a more comprehensive assessment. Differences in elevation, slope, edge density, and cover type between developed and undeveloped areas were minimal. Based on 25,290 GIS-generated random sample points (12,645 inside and 12,645 outside of the gas field, respectively),
x 6 SD values inside versus outside of the gas field
respectively were 2,347.7 6 146.4 m and 2,382.6 6 243.4 m for elevation, 10.6 6 6.2 degrees and 9.4 6 7.0 degrees for slope, and

82.0 6 10.2 km/km 2 and 77.5 6 12.7 km/km 2 for edge density.

The proportion of sample points that occurred within security habitat was 0.84 and 0.80 inside versus outside of the gas field, respectively. Topographic covariates (elevation and slope) were not ground truthed; however, vegetation attributes were ground- truthed based on field-established polygons ( n = 1177) of known vegetation type.

Generalized Additive Modeling to Designate Maternal
Status
We applied results from PSPB and field observation of female elk in 2008 and 2009 to test the GAM approach, in which movement patterns parameterized by ,3df were predicted to depict non-parturiency whereas movement patterns parameterized by $3 df were predicted to depict parturiency. Based on PSPB results and 788 hours of field observation, we assigned maternal status to 34 elk; 24 were identified as parturient and 10 as non- parturient. Movement during calving time was highly variable, especially among parturient elk, with parameterization ranging from 1.24–16.9 df. However, among elk observed to be parturient in the field, 93% (13/14) were parameterized by $3 df (Table 3).

Random-effects Resource Selection Modeling – Marginal
Inference
We found little evidence of lack of fit or overdispersion in RSF models; generalized chi-square/degrees of freedom was 0.79–0.92 for night time models and 0.87–0.97 for day time models. The most notable response by elk inside of the gas field was strong avoidance of the industrial development footprint during the day among all groups (at the population-level; Figure 3a). Other notable behavior in field-observed parturient elk included strong selection for cover and avoidance of high road density during the day relative to other groups (Figure 3b,c). While field-observed parturient elk showed some day-time preference for flatter areas and lower elevation relative to the larger sample, differences among maternal groups in selection for these and other resources were small, inconsistent, or not apparent.

Table 2. Resource selection by Rocky Mountain elk during calving time; sample size and number of locations used in analyses.
#Elk #Locations
Year Group a Day Night Day Night
2006 Inside – all sample elk 11 10 1,759 879
Inside – GAM-predicted parturient 5 5 917 458
Outside – all sample elk 7 7 122 61
2007 Inside – all sample elk 26 24 3,281 1,641
Inside – GAM-predicted parturient 9 9 1,174 587
Outside – all sample elk 16 14 900 451
2008 Inside – all sample elk 19 19 2,056 1,455
Inside – GAM-predicted parturient 7 7 776 438
Inside – field-observed parturient 12 12 1,598 802
Inside – field-observed non-parturient 5 5 425 227
Outside – all sample elk 19 16 417 230
2009 Inside – all sample elk 15 15 2,735 1,482
Inside – GAM-predicted parturient 8 8 1,381 757
Inside – field-observed parturient 12 12 1,976 1,062
Inside – field-observed non-parturient 5 5 685 496
Outside – all sample elk 16 14 411 215
aInside refers to areas within the gas field and outside refers to areas adjacent but external to the gas field. GAM-predicted parturient refers to female elk that we predicted to be parturient based on results of generalized additive modeling (GAM) of movement data. Field-observed refers to elk for which we observed behavior in the field to assess parturition status.

Total sample fitted with Global Positioning Systems collars was 25, 40, 50, and 50 in 2006, 2007, 2008, and 2009, respectively. Within-year deviation from the available sample arises from 3 sources: 1) some collared elk moved out of the study area during calving, 2) the GAM-predicted sample was nested within the total sample, and 3) some individuals occupied areas both inside and outside of the gas field.
doi:10.1371/journal.pone.0014597.t002
Table 3. Parameterizing movement using generalized
additive models.

Year/Group df,2a 2,df,3 df$3
2008 Field-observed parturient 2 1 9
2009 Field-observed parturient 5 3 4
2008 Field-observed non-parturient 2 2 1
2009 Field-observed non-parturient 3 2 0
aAs part of generalized additive modeling we specified the generalized cross validation option to assign parameters ( df) to polynomials depicting daily changes in the distance moved by elk between successive locations.

Maternal status was assigned to 34 female Rocky Mountain elk (10 non- parturient and 24 parturient) using information from blood samples (pregnancy-specific protein-B) and field observation (788 h). We assessed how generalized additive models of location data parameterized movement, and how parameterization corresponded with calf status as determined in the field.

Results show that among 14 elk for which movement was parameterized by $3 df, 13 were observed in the field to be with calf.
doi:10.1371/journal.pone.0014597.t003
Ungulate Conservation
PLoS ONE | www.plosone.org 7 January 2011 | Volume 6 | Issue 1 | e14597 We estimated selection for 4 landscape/habitat covariates in both developed and undeveloped areas including security cover, slope, elevation, and habitat edge density. Considering population- level coefficient estimates, selection for resources that functioned to provide security cover differed markedly between elk that occupied developed areas versus undeveloped areas (Figure 4). Elk inside of the gas field consistently showed strong selection for security cover during the day (Figure 4a) and strong selection for non-cover or forage habitats during the night (Figure 4b) throughout the study period whereas elk occupying areas outside of the gas field generally selected randomly for security cover with coefficient estimates near zero. Another difference between elk inside versus outside of the gas field involved selection for elevation. Elk outside of the gas field generally selected randomly for elevation with coefficient estimates always near zero. Elk inside of the gas field generally selected for higher elevation during the day (Figure 4c; however, 95% CL overlapped zero), and for lower elevation at night (Figure 4d). Diurnal patterns in selection for slope and edge density were apparent but differences in selection for these features between elk inside versus outside of the gas field generally were small (Table 4).

Random-effects Resource Selection Modeling –
Conditional Inference
Conditional estimates revealed a complex association between human activity and individual variation in response to environmental and anthropogenic features (Figure 5). Elk inside the gas field showed greater heterogeneity among individuals relative to elk outside of the Figure 3. Group-dependent population-level responses: inside of the gas field. Marginal coefficient estimates 6 95% CL of selection for human disturbance (a), security cover (b), and road density (c) by female elk during 2006-2009 in Raton Basin, Colorado, USA. Day time results are displayed among maternal statuses for elk inside of the gas field. doi:10.1371/journal.pone.0014597.g003
Ungulate Conservation
PLoS ONE | www.plosone.org 8 January 2011 | Volume 6 | Issue 1 | e14597 gas field in their responses to edge density and elevation. Selection for slope (Figure 5a,b) and edge density (Figure 5c, d) was estimated more precisely among elk inside of the gas field; however, elk outside the gas field selected (randomly) for elevation with high precision (Figure 5e, f). Also apparent inside the gas field but not outside was a temporal trend of increasing avoidance of high edge density (Figure 5c).

Within-year marginal estimates of day time selection for road density were positive (i.e., elk selected for higher road density) in 3 of 4 years.

However, conditional estimates revealed that, across all years, only 37% of sampled elk selected for relatively high road density; 32% selected neither for nor against high road density, and 31% avoided high road density (Figure 5g).

Seasonal use areas of 15 elk overlapped both developed and undeveloped areas. Individuals showed notable behavioral differ- ences, including diurnal variation, relative to selection for security cover and elevation depending on whether they were in developed versus undeveloped areas (Figure 6). During the day, 13 of 15 elk showed stronger selection for security cover when inside the gas field, but at night this pattern broke down with no apparent consistent behavioral response. Selection for elevation was variable inside the gas field during day and night with some elk showing relatively strong selection for higher or lower elevations. When elk were outside of the gas field they generally used elevation at random (Figure 6).

Figure 4. Population-level responses: inside versus outside of the gas field. Marginal coefficient estimates 695% CL of day-time (a) and night-time (b) selection for security cover, and day-time (c) and night-time (d) selection for elevation by female elk during 2006-2009 in Raton Basi n, Colorado, USA. Inside refers to elk occupying an active gas field whereas outside refers to elk occupying adjacent undeveloped areas; maternal statu s is not considered ( i.e., all sample elk are grouped depending on whether they occurred inside versus outside of the gas field). doi:10.1371/journal.pone.0014597.g004
Ungulate Conservation
PLoS ONE | www.plosone.org 9 January 2011 | Volume 6 | Issue 1 | e14597 Table 4. Marginal coefficient estimates ( ^/C22bb/C22bbn[SE]) for two-level random-effects models of resource selection by Rocky Mountain elk during calving time (15 May - 1 July), 2006–2009.

Day Night
Effect/Year Inside Outside Inside Outside
Slope 2006 20.001 (0.002) 20.003 (0.034) 20.064 (0.013) - Slope 2007 0.011 (0.006) 0.005 (0.013) 20.030 (0.010) 20.040 (0.016) Slope 2008 20.003 (0.007) 0.003 (0.003) 20.035 (0.007) 20.045 (0.008) Slope 2009 0.012 (0.005) 0.024 (0.014) 20.023 (0.011) 20.027 (0.013) Edge Density 2006 0.015 (0.016) 20.022 (0.004) 20.007 (0.021) - Edge Density 2007 0.001 (0.013) 20.003 (0.012) 0.006 (0.011) 20.013 (0.004) Edge Density 2008 20.026 (0.009) 20.008 (0.003) 20.026 (0.008) 20.010 (0.002) Edge Density 2009 20.027 (0.009) 20.006 (0.010) 20.036 (0.012) 20.044 (0.021) - Model failed to converge.
doi:10.1371/journal.pone.0014597.t004
Figure 5. Population-level and individual responses. Marginal (dash) and conditional (circles) coefficient estimates (SE) of day-time selection for slope (a, b) edge density (c, d), elevation (e, f), and road density (g) by female elk during 2006-2009 in Raton basin, Colorado, USA. Results are displayed by year for elk that occupied developed (panels a, c, e, and g) and undeveloped (panels b, d, and f) areas. Orange depicts elk that were predicted to be parturient using generalized additive modeling; open circles ( #) indicate conditional estimates for field-observed parturient elk in 2008 and 2009.
doi:10.1371/journal.pone.0014597.g005
Ungulate Conservation
PLoS ONE | www.plosone.org 10 January 2011 | Volume 6 | Issue 1 | e14597 Functional Responses
Consistent with a functional response in resource selection, elk selected randomly for disturbed areas when disturbance was minimal. As human activity increased ( i.e., across individual elk calving season areas, or through time), elk showed stronger avoidance of the industrial development footprint during the day but not at night; this spatiotemporal pattern of avoidance revealed that elk continued to use physically disturbed areas but modified their behavior to avoid human activity (Figure 7a). One way elk avoided human activity in the day was by modifying selection for security cover. In developed areas elk were constrained to select resources near physical disturbance and, during the day, showed a stronger functional response to the proportion of cover than did elk in undeveloped areas (Figure 7b, c).

Mapping Responses
We developed six models (day and night separately for each of three groups): the first was based on all sample elk that occurred inside of the gas field; the second was a subset of the first group for which GAMs predicted females to be parturient; and the third was based on all sample elk that occurred outside of the gas field. We validated day and night models separately (1830 night locations Figure 6. Movement from within the gas field to areas adjacent to the gas field. A subsample of female elk ( n = 15) occupied areas within the gas field as well as areas outside but adjacent to the gas field. This figure shows differences in day-time (a) and night-time (b) selection for habi tat that provides security cover, and day-time (c) and night-time (d) selection for elevation among these elk depending on whether they were inside versus outside of the gas field. Symbols represent conditional coefficient estimates (y-axes) for each elk in 2007 (squares; n = 6), 2008 (circles; n = 4), and 2009 (triangles; n = 5); lines show how selection changed within each elk depending on whether it was inside versus outside of the gas field. Solid lines depict larger selection coefficients inside the gas field whereas dashed lines depict larger coefficients outside of the gas field. Coefficie nts are informed by 36-293 location per elk inside the gas field, and 20-248 locations per elk outside the gas field. doi:10.1371/journal.pone.0014597.g006
Ungulate Conservation
PLoS ONE | www.plosone.org 11 January 2011 | Volume 6 | Issue 1 | e14597 and 3505 day locations) and then summed day and night values within the three groups of interest to estimate overall within-group model performance. For both models based on elk behavior inside of the gas field (Figure 8a), habitat with low predicted probability of use was modeled accurately; however, validation elk used habitat predicted to have high probability of use only as much as would be expected if elk used all predictive classes equally (Figure 8c). The model based on elk behavior outside of the gas field (Figure 8b) validated well with few locations occurring in lowest ranked areas and many locations occurring in highest ranked areas (Figure 8d). The proportion of pixels ranked differently between the model that validated most poorly and the model that validated best was 0.72–0.88 depending on year.

Discussion
We successfully integrated GAMs, field observation, and random-effects RSF modeling to designate parturiency, describe individual and population-level resource selection, and determine the relative influence of maternal status and human activity on the reliability of spatially explicit models intended to guide conserva- tion of critical ungulate calving resources. We assumed that planning tools such as spatially explicit models would be more reliable if they account for adaptive behavior which, among parturient elk, should be reflected in movement and resource selection [45]. This was the basis for using GAMs to designate parturiency based on movement; our thinking was that ungulate Figure 7. Functional responses in resource selection by female elk. Selection for the human disturbance footprint (a), security cover inside of the gas field (b), and security cover outside of the gas field (c) change as a function of availability. Availability was calculated at the seasonal use area level and conditional coefficients were estimated from generalized linear random-effects resource selection models. Dashes (-) and dashed lines symbolize night-time resource selection, and circles ( N) and solid lines symbolize day-time resource selection. Dashes and circles outlined by open circles ( #) or boxes ( %) indicate conditional estimates for field-observed parturient elk in 2008 and 2009. doi:10.1371/journal.pone.0014597.g007
Ungulate Conservation
PLoS ONE | www.plosone.org 12 January 2011 | Volume 6 | Issue 1 | e14597 behavior contains information on reproductive status and GAMs would reveal a difference in the shape and relative complexity of the response curve between parturient and non-parturient females.

Maternal status was evident in elk movement with our application of GAMs correctly predicting parturiency 93% of the time. We note, however, the GAM approach as presented would require further development if improved sensitivity (reduced false negative rate) is desirable. GAMs have been a standard tool in
epidemiologic analyses and have found broad application in ecology for modeling tolerance thresholds and spatial distributions [46], [47], [48], [49]. Using movement data to designate functional groups, seasonality, or behavior in animals has increased in prevalence as GPS-based research has become more frequent [50], [51]. Although we found some potentially important differences in resource selection patterns between parturient and non-parturient elk (see below), behavior in these groups was more similar than expected. Had maternal status been more apparent in shaping resource selection, methods to designate status might have found better application in landscape-level mapping of the relative predicted probability of use. Here, both parturient and non- parturient elk were examined inside the gas field where any behavior distinctly associated with parturiency was obscured by apparent risk-aversive behavior related to human activity during the day time.

At the population-level parturient elk, while avoiding roads and selecting for security cover more strongly than other elk, conformed to a general pattern of avoiding human activity during the day by occupying upland forest or Q. gambelii thickets, and selecting forage resources in valley bottoms at night regardless of proximity to infrastructure. Trade-offs that structure resource selection in many ungulates, most notably between forage requirements and risk avoidance [6], [52], often are amplified in Figure 8. Assigning the relative probability of use throughout the landscape. Comparison of how the relative predicted probability of use was assigned throughout the landscape depending on whether RSF mapping was based on resource selection behavior among female elk that occupied a natural gas field (a) or occupied areas adjacent to but outside of a natural gas field (b). Maps depict day-time probability of use. Wells are depicted as black dots. Borders of wildlife areas managed by the state of Colorado are depicted in red. Probability of use is scaled from low to high with each of 5 ordinal bins representing quantiles of the total number of pixels (30-m resolution) comprising the area. Charts (c, d) display map validation with columns depicting the number of locations from an independent sample of elk that, when plotted on RSF maps (a, b), occurred within each ordinal bin. The black horizontal line depicts expectation if resources were selected at random. The map based on behavior inside of the gas field validated poorly (c; x
2 = 577.46, df =4 , p,0.001) relative to the map based on behavior outside of the gas field (d; x2 = 849.18, df =4 , p,0.001) with the distribution of validation locations differing significantly from random in both cases. RSF maps (a, b) are based on all sample elk inside and outside of the gas field, respectively; not shown is the RSF map based on elk predicted to be parturient using generalized additive models. doi:10.1371/journal.pone.0014597.g008
Ungulate Conservation
PLoS ONE | www.plosone.org 13 January 2011 | Volume 6 | Issue 1 | e14597 parturient females through nutritional demands to support body condition, lactation, and neonatal defense [53]. Here, responses to human activity were more apparent than differences in resource selection as a function of parturiency with elk exhibiting a clear spatiotemporal avoidance of human activity at the population- level and modified patterns of selection for environmental features, notably security cover and elevation, in developed areas relative to undeveloped areas. Research has shown that ungulates exhibit avoidance behavior relative to human development, recreation, hunting, and other activities [54]. Modeling resource selection separately during day and night offered insight into whether the physical presence of infrastructure versus the operation and maintenance of such infrastructure (only occurring during the day) was more important in avoidance behavior. Elk inside the gas field used cover and elevation to modulate avoidance of human activity during the day [55] whereas cover and elevation were less influential outside of the gas field suggesting that human activity during the day was the factor to which elk were responding [3]. At night during calving time female elk showed no aversion to infrastructure and selected for areas characterized by valley- bottoms and foraging resources that in many instances were in close proximity to, or directly associated with, disturbances such as well pads and roads [56]. Day time refugia characterized by minimal human disturbance, security cover, and upper slope positions will be critical for maintenance of calving and perhaps other seasonal habitats in multiple-use landscapes.

To our knowledge, fully random-effects RSF models do not appear in the published literature. The random-effects framework provided insight into individual behavior and how elk modified patterns of calving season resource selection relative to develop- ment. Individuals that moved between developed and undevel- oped areas (Figure 6) were spatially aware and showed marked changes in their behavior that were consistent with the hypothesis that human activity during the day was perceived as a source of risk [6]. An observation that warrants further attention in wildlife- human interaction studies is that heterogeneity among individuals in response to their environment was apparent, and this variation was amplified in developed areas relative to undeveloped areas.

Associated with this observation was generally more precise estimation of selection within individuals in developed areas (except selection for elevation). Among-individual variation can comprise the majority of a population’s niche width, and only when the within-individual component of total niche width is constrained does between-individual variation become prominent.

Of particular relevance here is that trade-offs remain among the most plausible mechanisms for the observation of limited within- individual variation [18]. If ungulates in developed areas must make trade-offs associated with avoiding human activity, partic- ularly during daylight hours, we might expect constrained within- individual variation and thus more heterogeneity among individ- uals. Human activity functioning to constrain decision-making in ungulates is consistent with the notion that risky environments impose pressures that disallow animals to respond to other features as they otherwise would [57]. This could make it difficult to establish or predict general patterns of resource selection during periods in which ungulates show fidelity to historic ranges, yet human activity rapidly modifies the landscape. In fact, we showed that RSF maps based on day-time resource selection behavior in developed areas had poor predictive accuracy.

The study of behavioral syndromes offers a relevant framework within which to discuss the conservation implications of individual variation in human-wildlife interaction studies [58]. A behavioral syndrome is a suite of correlated behaviors reflecting among- individual consistency across multiple contexts. Within a syn- drome, individuals have a behavioral type such as risk-aversion (i.e., more risk-averse versus less risk-averse types; [58]). The notion that individuals can be more or less risk aversive implies a limit to their range of behavioral plasticity. From a conservation perspective, animals exhibiting limited plasticity in environments undergoing rapid change, such as those affected by industrial development, may be less able to adapt. Elk clearly show some ability to adapt to human activity [59]. Nonetheless, if the effect of human activity is a threshold phenomenon [60], we might expect there to exist a limit of physical disturbance corresponding to a limit in the range of behavioral plasticity in ungulates, beyond which redistribution, social, or demographic effects may be observed [61], [62].

Examining only average responses across populations obscures variability among individuals that may have important implica- tions for management or provide new ecological insight. For example, management strategies designed to conserve a resource that is important, on average, to the population may overlook resources that are critical to individuals that comprise a smaller demographic segment that functions disproportionately in popu- lation persistence. It has been stated ‘‘information on individual resource use is necessary if we are to make the transition from phenomenological models of population dynamics to mechanistic models in which the dynamics of a population are predicted from the properties of its components’’ [18]. Conditional estimates also provide information on why marginal estimates may be counter- intuitive, or how individuals assemble to comprise the marginal estimate. For example, the marginal estimate of selection for road density in our study indicated that elk selected for higher road density in 3 of the 4 years comprising the study period (Figure 5).

This observation is counter to most research on roads and ungulates [63], [64]. Conditional estimates revealed that ,1/3 of elk showed a positive association with higher road density but just as many avoided high road density; this observation sets the stage for examining potential links between a particular behavioral strategy such as road avoidance and demographic responses. The increase in the within-year proportion of elk occupying areas with higher road density from 2007 to 2009 (Figure 5d), concurrent with an increase in new road development, suggests that elk maintained relatively consistent calving season use areas during the study period, but modified their behavior as they became increasingly constrained to select resources in proximity to development. Such modification was consistent with a functional response in resource selection; elk response to human activity changed as a function of availability and with time of day. During the day when humans are active elk showed avoidance behavior that strengthened with increasing disturbance. This response spilled over to elk selection for security cover which appeared to have high importance to elk inside of the gas field as a day time refugium. Outside of the gas field elk showed relatively weak day time selection for cover and, as might be expected, selection weakened as the proportion of cover within the seasonal use area increased ( sensu [40], Figure 7). These observations are consistent with the hypothesis that ungulates face a trade-off that is mediated by human activity in multiple-use landscapes – that is, the strength of the trade-off varies in direct relation to resource availability which is driven by human activity [41].

Spatially explicit models of relative predicted probability of use validated poorly when they were based on resource selection behavior inside of the gas field. The model based on behavior outside of the gas field validated well. To our knowledge, this treatment/control concept is a novel approach in RSF-based conservation planning. Typically, quantifying relative predicted probability of use in human-modified areas is based on animal Ungulate Conservation
PLoS ONE | www.plosone.org 14 January 2011 | Volume 6 | Issue 1 | e14597 behavior in those areas. This approach has been used effectively; it has been shown that predictive maps based on mule deer
(Odocoileus hemionus ) behavior in human-modified areas validated well [3]. In their study [3], variation among individuals was present but limited relative to our findings. We suggest that ungulates in developed areas often respond to human activity in situation-specific ways. In our study, the local attributes of human activity varied in space and time within the seasonal use area of each elk making it difficult to generalize across the population.

Situation-specific responses, including individual variation in the strength of selection or avoidance response, induced heterogeneity which complicated the application of models based on such behavior for conservation planning.

Protecting unmodified habitat in multiple-use landscapes typically is not an option [65]; therefore, measures to account for or reconcile changes in animal behavior across gradients of human-modification might be considered. If conservation objec- tives include establishing zones within such landscapes intended to function as refugia from human activity and promote long-term population persistence, information on resource selection patterns in existing refugia would be expected to provide valuable guidance in prioritizing the creation of new zones in modified landscapes.

Establishing refugia based on resource selection patterns that reflect responses to human activity risks uncertainty in the performance of such refugia once human disturbance pressure is released allowing animals to respond to features as they otherwise would have [57]. In our study, models based on resource selection patterns that reflected responses to human activity classified the relative probability of use differently from the models based on behavior in the absence of industrial development throughout extensive portions of the landscape, potentially resulting in large errors in where conservation action would have its greatest impact.

In population response research it is always desirable to include spatial and temporal controls (before-after, control-impact design), and demographic responses. Availability of such components is uncommon in large-scale wildlife-human interaction studies. In such situations the approach we describe, including efforts to account for possible adaptive behavior among reproductive groups, attention to the treatment/control concept, and a random-effects modeling framework, should have general appli- cation in human-wildlife interaction research particularly among species that inhabit places where human activity is intense, or among special status species for which little information on resource needs exists.

Acknowledgments
We thank Pioneer Natural Resources Company for supporting this work.

We thank C.R. Anderson, C.J. Bishop, B.P. Dreher, V.J. Dreitz, P.M.

Lukacs, and M.B. Rice, of the Colorado Division of Wildlife for providing insightful comments on an earlier version of this paper. We thank J.J. Cox and B. Augustine of the University of Kentucky Department of Forestry for providing comments on earlier versions. We thank landowners in Raton Basin who participated in this effort for their patience and ongoing enthusiasm for this work.

Author Contributions
Conceived and designed the experiments: MRD SMH RGO LDHW JBW. Performed the experiments: MRD SMH. Analyzed the data: MRD SMH SLW. Wrote the paper: MRD SMH SLW. Field observation: MRD SMH RGO JJW. Data collection: SMH RGO JJW.

References

1. Bowyer RT, Van Ballenberghe V, Kie JG, Maier JK (1999) Birth-site selection by Alaskan moose: maternal strategies for coping with a risky environment.

Journal of Mammalogy 80: 1070–1083.

2. White PJ, Davis TL, Barnowe-Meyer KK, Crabtree RL, Garrott RA (2006) Partial migration and philopatry of Yellowstone pronghorn. Biological Conservation 135: 502–510.

3. Sawyer H, Kauffman MJ, Nielsen RM (2009) Influence of well pad activity on winter habitat selection patterns of mule deer. Journal of Wildlife Management 73: 1052–1061.

4. Hebblewhite M, Merrill E (2008) Modelling wildlife-human relationships for social species with mixed-effects resource selection models. Journal of Applied Ecology 45: 834–844.

5. Fortin D, Courtois R, Etcheverry P, Dussault C, Gingras A (2008) Winter selection of landscapes by woodland caribou: behavioral response to geograph- ical gradients in habitat attributes. Journal of Applied Ecology 45: 1392–1400.

6. Frid A, Dill L (2002) Human-caused disturbance stimuli as a form of predation risk. Conservation Ecology 6: 11–25.

7. Beale CM, Monaghan P (2004) Human disturbance: people as predation-free predators? Journal of Applied Ecology 41: 335–343.

8. Creel S, Christianson D, Liley S, Winnie JA (2007) Predation risk affects reproductive physiology and demography of elk. Science 315: 960.

9. Manly BFJ, McDonald LL, Thomas DL, McDonald TL, Erickson WP (2002) Resource selection by animals: statistical design and analysis for field studies.

Norwell: Kluwer Academic Publishers. 240 p.

10. Kittle AM, Fryxell JM, Desy GE, Hamr J (2008) The scale-dependent impact of wolf predation risk on resource selection by three sympatric ungulates.

Oecologia 157: 163–175.

11. Fortin D, Fortin ME, Beyer HL, Duchesne T, Courant S, et al. (2009) Group- size-mediated habitat selection and group fusion-fission dynamics of bison under predation risk. Ecology 90: 2480–2490.

12. Singer FJ, Harting A, Symonds KK, Coughenour MB (1997) Density dependence, compensation, and environmental effects on calf mortality in Yellowstone National Park. Journal of Wildlife Management 61: 12–25.

13. Phillips GE, Alldredge AW (2000) Reproductive success of elk following disturbance by humans during calving season. Journal of Wildlife Management 64: 521–530.

14. Schaefer JA, Bergman CM, Luttich SN (2000) Site fidelity of female caribou at multiple spatial scales. Landscape Ecology 15: 731–739.

15. Bernardo J (1996) Maternal effects in animal ecology. Am Zool 36: 83–105.

16. Mousseau TA, Fox CW (1998) The adaptive significance of maternal effects.

Trends in Ecology and Evolution 13: 403–407.

17. Blount JD, Surai PF, Nager RG, Houston DC, Møller AP, et al. (2002) Carotenoids and egg quality in the lesser black-backed gull Larus fuscus :a supplemental feeding study of maternal effects. Proceedings of the Royal Society B 269: 29–36.

18. Bolnick DI, Svanba ¨ ck R, Fordyce JA, Yang LH, Davis JM, et al. (2003) The ecology of individuals: incidence and implications of individual specialization.

The American Naturalist 161: 1–28.

19. Austin D, Bowen WD, McMillan JI (2004) Intraspecific variation in movement patterns: modeling individual behavior in a large marine predator. Oikos 105:
15–30.

20. Thompson MJ, Henderson RE (1998) Elk habituation as a credibility challenge for wildlife professionals. Wildlife Society Bulletin 26: 477–483.

21. Estes JA, Riedman ML, Staedler MM, Tinker MT, Lyon BE (2003) Individual variation in prey selection by sea otters: patterns, causes, and implications.

Journal of Animal Ecology 72: 144–155.

22. Gillespie RG, Caraco T (1987) Risk-sensitive foraging strategies of two spider populations. Ecology 68: 887–899.

23. Western Regional Climate Center (2010) Available: http://www.wrcc.dri.edu/ coopmap/. Accessed 2010 Jan 14.

24. Hemborg HT (1998) Spanish Peak Field, Las Animas County, Colorado:

Geologic setting and early development of a coalbed methane reservoir in the central Raton basin. Colorado Geological Survey Department of Natural Resources, Resources Series 33.

25. Vitt A (2007) Trinchera Data Analysis Unit E-33, Game Management Units 83, 85, 140, 851, Elk Management Plan. Colorado Division of Wildlife, Pueblo.

26. Quimby DC, Gaab JE (1957) Mandibular dentition as an age indicator in Rocky Mountain elk. Journal of Wildlife Management 21: 435–451.

27. Hastie TJ, Tibshirani RJ (1990) Generalized additive models. London:

Chapman and Hall. 335 p.

28. Guisan A, Edwards Jr. TC, Hastie T (2002) Generalized linear and generalized additive models in studies of species distributions: setting the scene. Ecological Modelling 157: 89–100.

29. Beck N, Jackman S (1997) Getting the mean right is a good thing: generalized additive models. Society for Political Methodology Working Papers.

30. Fewster RM, Buckland ST, Siriwardena GM, Baillie SR, Wilson JD (2000) Analysis of population trends for farmland birds using generalized additive models. Ecology 81: 1970–1984.

31. Clutton-Brock TH, Guinness FE (1975) Behaviour of red deer ( Cervus elaphus L.) at calving time. Behaviour 55: 287–299.

32. Cowie GM, Moore GH, Fisher MW, Taylor MJ (1985) Calving behavior of farmed red deer. Proceedings of a Deer Course for Veterinarians 2:1 43-154.

Ungulate Conservation
PLoS ONE | www.plosone.org 15 January 2011 | Volume 6 | Issue 1 | e14597

33. Hudson RJ, Haigh JC (2002) Physical and physiological adaptations. In:

Toweill DE, Thomas JW, eds. North American elk: ecology and management.

Washington DC: Smithsonian Institution Press. pp 199–257.

34. Vore JM, Schmidt EM (2001) Movements of female elk during the calving season in northwest Montana. Wildlife Society Bulletin 29: 720–725.

35. Poole KG, Serrouya R, Stuart-Smith K (2007) Moose calving strategies in interior montane ecosystems. Journal of Mammalogy 88: 139–150.

36. Long RA, Kie JG, Bowyer RT, Hurley MA (2009) Resource selection and movements by female mule deer Odocoileus hemionus: effects of reproductive stage.

Wildlife Biology 15: 288–298.

37. Noyes JH, Sasser RG, Johnson BK, Byrant LD, Alexander B (1997) Accuracy of pregnancy detection by serum protein (PSPB) in elk. Wildlife Society Bulletin 25:
695–698.

38. Johnson DE (1951) Biology of the elk calf, Cervus elaphus nelsoni . Journal of Wildlife Management 15: 396–410.

39. Gillies CS, Hebblewhite M, Nielsen SE, Krawchuk MA, Aldridge C, et al. (2006) Application of random effects to the study of resource selection by animals.

Journal of Animal Ecology 75: 887–898.

40. Mysterud A, Ims RA (1998) Functional responses in habitat use: availability influences relative use in trade-off situations. Ecology 79: 1435–1441.

41. Godvik IMR, Loe LE, Vik JO, Veiberg V, Langvatn R, et al. (2009) Temporal scales, trade-offs, and functional responses in red deer habitat selection. Ecology 90: 699–710.

42. Sawyer H, Nielson RM, Lindzey F, McDonald LL (2006) Winter habitat selection of mule deer before and during development of a natural gas field.

Journal of Wildlife Management 70: 396–403.

43. Marzluff JM, Millspaugh JJ, Hurvitz P, Handcock MS (2004) Relating resources to a probabilistic measure of space use: forest fragments and Steller’s jays.

Ecology 85: 1411–1427.

44. Thomas DL, Johnson D, Griffith B (2006) A Bayesian random effects discrete- choice model for resource selection: population-level selection inference. Journal of Wildlife Management 70: 404–412.

45. Geist V (2002) Adaptive behavioral strategies. In: Toweill DE, Thomas JW, eds.

North American elk: ecology and management. Washington DC: Smithsonian Institution Press. pp 389–433.

46. Dominici F, McDermott A, Zeger SL, Samet JM (2002) On the use of generalized additive models in time-series studies of air pollution and health.

American Journal of Epidemiology 156: 193–203.

47. Yuan LL (2004) Assigning macroinvertebrate tolerance classifications using generalized additive models. Freshwater Biology 49: 662–677.

48. Potvin MJ, Drummer TD, Vucetich JA, Beyer Jr. DE, Peterson RO, et al. (2005) Monitoring and habitat analysis for wolves in upper Michigan. Journal of Wildlife Management 69: 1660–1669.

49. Winter AG, Swartzman GL (2006) Interannual changes in distribution of age-0 walleye Pollock near the Pribilof Islands, Alaska, with reference to the prediction of Pollock year-class strength. Journal of Marine Science 63: 1118–1135.

50. Franke A, Caelli T, Kuzyk G, Hudson RJ (2006) Prediction of wolf ( Canis lupus ) kill-sites using hidden Markov models. Ecological Modelling 197: 237–246.

51. Vander Wal E, Rodgers AR (2009) Designating seasonality using rate of movement. Journal of Wildlife Management 73: 1189–1196.

52. Bowyer RT, Kie JG, Van Ballenberghe V (1998) Habitat selection by neonatal black-tailed deer: climate, forage, or risk of predation? Journal of Mammalogy 79: 415–425.

53. Keech MA, Bowyer RT, Ver Hoef JM, Boertje RD, Dale BW, et al. (2000) Life- history consequences of maternal condition in Alaskan moose. Journal of Wildlife Management 64: 450–462.

54. Morrison JR, de Vergie WJ, Alldredge AW, Byrne AE, Andree WW (1995) The effects of ski area expansion on elk. Wildlife Society Bulletin 23: 481–489.

55. Edge WD, Marcum CL (1991) Topography ameliorates the effects of roads and human disturbance on elk. In: Christensen AG, Lyon LJ, Lonner TN, editors.

Proceedings of a symposium on elk vulnerability. Bozeman: Montana State University. pp 132-137.

56. Anderson DP, Turner MG, Forester JD, Zhu J, Boyce MS, et al. (2005) Scale- dependent summer resource selection by reintroduced elk in Wisconsin, USA.

Journal of Wildlife Management 69: 298–310.

57. Winnie Jr. J, Christianson D, Creel S, Maxwell B (2006) Elk decision-making is simplified in the presence of wolves. Behavioral Ecology and Sociobiology 61:
277–289.

58. Sih A, Bell A, Johnson JC (2004) Behavioral syndromes: and ecological and evolutionary overview. Trends in Ecology and Evolution 19: 372–378.

59. Edge WD, Marcum CL (1985) Movements of elk in relation to logging disturbances. Journal of Wildlife Management 49: 926–930.

60. Harju SM, Dzialak MR, Taylor RC, Hayden-Wing LD, Winstead JB (2010) Thresholds and time lags in the effects of energy development on greater sage- grouse populations. Journal of Wildlife Management 74: 437–448.

61. Cameron RD, Reed DJ, Dau JR, Smith WT (1992) Redistribution of calving caribou in response to oil field development on the Arctic slope of Alaska. Arctic 45: 338–342.

62. Manor R, Saltz D (2003) Impact of human nuisance disturbance on vigilance and group size of a social ungulate. Ecological Applications 13: 1830–1834.

63. Rowland MM, Wisdom MJ, Johnson BK, Kie JG (2000) Elk distribution and modeling in relation to roads. Journal of Wildlife Management 64: 672–684.

64. Friar JL, Merrill EH, Beyer HL, Morales JM (2008) Thresholds in landscape connectivity and mortality risks in response to growing road networks. Journal of Applied Ecology 45: 1504–1513.

65. Moilanen A, Franco AMA, Early RI, Fox R, Wintle B, Thomas CD (2005) Prioritizing multiple-use landscapes for conservation: methods for large multi- species planning problems. Proceedings of The Royal Society B 272: 1885–1891.

Ungulate Conservation
PLoS ONE | www.plosone.org 16 January 2011 | Volume 6 | Issue 1 | e14597 Available online at www.sciencedirect.com

Energy Procedia 00 (2008) 000–000

Energy
Procedia

www.elsevier.com/locate/XXX

GHGT-9
Environmental impact of amines
Ingvild Eide-Haugmoa *, Odd Gunnar Brakstadb, Karl Anders Hoffb, Kristin Rist Sørheimb, Eirik Falck da Silvab, Hallvard F. Svendsena

aNorwegian University of Science and Technology, 7491 Trondheim, Norway bSINTEF Materials and Chemistry, 7465 Trondheim, Norway Elsevier use only: Received date here; revised date here; accepted date here Abstract
The process of post combustion CO 2-capture by chemical absorption relies on large scale use of chemicals, typically amines in aqueous solution. In such operations, emissions of amines may occur through the cleaned exhaust gas, as degraded solvent and as accidental spills. It is thus important that the chemicals used have low or no environmental effects. To check this, standar d ecotoxicity and biodegradability tests for a marine environment were performed on more than 40 amines, including both solvents already in use for CO
2-removal and new promising chemicals. The results form a database for environmental risk assessment of common absorption solvents and will be used to correlate chemic al structure of the solvents to degradation and toxicity data fo r use in solvent screening. Some of the solvents used for carbon capture, have been shown to have low biodegradability. The tertiary amines which have been tested do not degrade easily, while the amino acids tested both have low toxicity and degrade easily. The fate estimation model EPI Suite™ [US EPA, Wa shington DC, USA] has also been used to estimate the biodegradation and toxicity of the chemicals. It was compared to the experimental results to investigate if this tool could be used for future solvent screening. For the biodegradability the predictions showed agreement with 48% of the tested compounds, while for ecotoxicity the predictions showed agreement with 66%.

© 2008 Elsevier Ltd. All rights reserved
Keywords: amine; marine environment; biodegradability; degradation; toxicity; EPI Suite™; solvent;

* Corresponding author. Tel.:+47-7355-0346; fax: +47-7359-4080.

E-mail address: [email protected].
c⃝ 2009 Elsevier Ltd.

Energy Procedia 1 (2009) 1297–1304
www.elsevier.com/locate/procedia
doi:10.1016/j.egypro.2009.01.170
Open access under CC BY-NC-ND license.
2 Author name / Energy Procedia 00 (2008) 000–000

1. Introduction
Amine based processes have been used commercially for removal of acid gas impurities from process gas stream, and is currently the most popular way to remove CO2 in industry[1]. For natural gas sweetening operations, typically alkanolamines like monoethanolamine (MEA), diethanolam ine (DEA), and N-methyldiethanolamine (MDEA), as well as mixtures of alkanolamines are used. Capture by absorption relies on large scale use of chemicals and emissions of the solvent may occur through the cleaned exhaust gas, as degraded solvent and as accidental spills. It is thus important that the chemicals used have low or no environmental effect. In this work, the main focus is on emissions to the marine environment.

An environmental assessment of the alkanolamines have be en done by Davis and Carpenter [2], reviewing their potential for degradation as well as aquatic toxicology and bi oconcentration potential. The alkanolamines were said to be highly susceptible to biodegradation and it was further stated that the toxicity was low for the majority of species. However, the assessment was based on soil a nd freshwater, not on the marine environment. Calamari et al.[3] conducted a series of biodegradation and ecotoxicity tests on 8 amines using three different inocula in aqueous solution, wherein two amines, tert-but ylamine and morpholine, showed no degradation at all. The rest of the compounds - aniline, dimethylamine, diethylamine, di-n-butylamine, diisopropan olamine and cyclohexylamine – underwent degradation to some extent. Price et al. [4] performed biodegradation tests both in seawater and freshwater, however the bacterial content used in the tests was higher than that expected in the ocean. The amines tested were diethanolamine, diethyle netriamine, ethylenediamine and trieth anolamine. In general the rate of degradation in the seawater biodegradation tests was found to be slower than in the freshwater system, and only the most degradable compounds from the freshwater system were tested. Dieth ylenetriamine was the only amine showing no degradation in freshwater, while ethylenediamine showed poor degradation in seawater.

2. Experimental
Two tests, a biodegradation and an ecotoxicity test, have been performed to determine the environmental fate of the chemicals in the marine enviro nment. Biodegradation was determined by a marine biodegradation test, conducted according to the OECD guideline 306, “Biodegradability in seawater” [5]. The chemicals were diluted in normal seawater which had been aged for 3-5 days, and essential nutrients for the bacteria in the seawater were added. Seawater without chemicals was used for blank samp les, and as a reference so lution, aniline diluted in seawater was used. The solutions were distributed as dupl icates in bottles with stoppers , and incubated in the dark for 28 days at 20 ± 2°C. A dissolved oxygen meter was us ed to measure the concentration of dissolved oxygen in solution: At the start of the experiment and after 5, 15 and 28 days bottles in each series with blank, test and reference solutions were removed for measurements. After measurment the bottles were discharged. The biological oxygen demand (BOD) was used as a measure for biodeg radation. The BOD values were calculated as the difference in dissolved oxygen between solution with blank and test solutions, and the percentage of the theoretical oxygen demand (ThOD) was calculated. The method for esti mating the theoretical oxyge n demand is given in equation 1. Here c, h, cl, n, s, p, na and o are the num bers of carbon, hydrogen, chlorine, nitrogen, sulfur, phosphorous, sodium and oxygen atoms present, and MW is the molecular weight of the substance.

16 1 5 1[2 ( 3 ) 3 ]22 2ThOD c h cl n s p na oMW=+ − − + + + + E q 1

The ecotoxicity was determined by a marine phytop lankton test, conducted according to ISO/DIS guideline 10253, using the alga Skeletonema Costatum [6]. The algae were inoculated in a concentration series of the chemicals, with triplicates of each concentration. The choice of test concentrations was based on a preliminary test with ten-fold dilutions of the test subs tance. As controls, algae were also inoculated to algal medium without test substance. All test tubes were incubated with ag itation under constant light intensity (60-120 ȝE/sec/m2) and a temperature of 20°C for 72 hours. In vivo chlorophyll fluorescence was measured daily in a fluorometer, as the algae performs photosynthesis and contains ch lorophyll. Growth rate were calculate d by linear regression analyses of the logarithmic of algal growth curves for each concentration of chemical and for controls. 1298 I. Eide-Haugmo et al. / Energy Procedia 1 (2009) 1297–1304 Author name / Energy Procedia 00 (2008) 000–000 3
From these parameters the effective concentrations (mg/ l) of test substance inhibiting algal growth by 50 % (EC50) relative to the control cultures, were calculated by 95 % confidence interval using the program TOXEDO [7]. Based on the Convention for the Protection of the Marine Environment of the North-East Atlantic (OSPAR Convention), the Norwegian Po llution Control Authority (SFT) has a minimum recommended value for the marine biodegradation test of 20%, while for the marine phytoplankton test the minimum value is 10 mg/l [8].

EPI Suite™ is an environmental fate estimation model de veloped by the EPA’s Office of Pollution Prevention Toxics and Syracuse Research Corporation (SRC)[9]. In Epi suite TM the aerobic linear prediction model in BIOWIN
was used to estimate the probability of a chemical biodegrading in the environment, while the EC50 value for the 96 hour duration test with green algae in ECOSAR was us ed for predictions on ecotoxicity. BIOWIN uses a group- contribution model to asess the biode gradability of a compound; the methodo logy and description of the groups are discussed by Howard et al.[10]. In BIOWIN a chemical is classified as biodegrading fast, if the value calculated for the compound by the linear model is above 0,5. ECOSAR is based upon the octanol/water partition coefficient (Kow) of the compound, used together with constants based on the class of the compound (including aliphatic amine).

3. Results and discussion
Experimental results are shown in figure 1 for the biodegr adation and figure 2 for ecotoxiciy. The results varied considerably in both biodegradation an d ecotoxicity tests, even though almost all of the species tested are amines and alkanolamines. In the biodegradati on test the results varied from < 1% to 100 % biodegradability (median 21.3 % biodegradability), while the Skeletonema t e s t s ho w e d a n E C50 range from 1.84 to > 10000 mg/L (median 198 mg/L). Risk is a combination of persistence and toxicity after discharge of a chemical to the environment. In terms of environmental risks, low biodegradation (persistance) contributed more than ecot oxicity. The ecotoxicity was mostly above the preferred value of 10 mg/l. From figure 1 it can be seen that some of the amines commonly used for capture (MDEA, AMP, Piperazine), would have long pe rsistence in the marine environment due to their low degradability.

To check which functional groups of the chemical that influence the results, the chemicals were sorted according to amine type, and also according to the number of hydroxyl groups. Results are shown in figure 3 and 4, with biodegradability plotted against ecotoxi city, and significant variations in resu lts were determined. All amino acids tested had a high bi odegradability in combination with low toxi city, showing good environmental characteristics, and suggesting that this class of compounds were intere sting from an environmental point of view. However, it should be taken into consideration that the tested amino acids are all known to be abundant in nature. Most of the tertiary amines tested had a low biode gradability, with the ex ception of N,N-dimethylethanolamine (DMMEA).

Alkanolamines were shown to have toxicity between 10 and 1000 mg/l. It was also observed that all the amines having a quarternary carbon had low biodegradability, while the compounds with carbon chains of four were highly biodegradable. Compounds with ring struc tures typically showed the highest toxi city of the tested chemicals, with the exception of piperazine and 1-(2-hydroxyethyl)-piperazine.

The group contribution model discussed by Howard et al.[10] suggests that compou nds containing quarternary carbons and tertiary amines in general are poorly biodegradable, while compounds with carbonchains of four are more easily degradable. It is also expected that pr imary and secondary amino and hydroxyl groups are more degradable than tertiary amines and compounds containing quarternary carbon. Calamari et al. [3] found that the amino group and the linear alky l chains were usually degraded by bacteria, while branched alkyl chains and etheric linkage gave some resistance to degradation. Aniline, contain ing a benzene ring with no other substitutes than an amino group was easily degraded. Tert-butylamine, with a branched chain, and morpholine, with an epoxidic group, showed no degradation. Diisopropanolamine, which is less branched than tertbutylamine, was partially degraded by bacteria adapted to the compound. In co mparison, tertbutylamine and morpholin e were not biodegradable in our study, however, neither was diisopropanolamine. Dimethylamine was found to be biodegradable in our study as well as by Calamari et al. [3].

I. Eide-Haugmo et al. / Energy Procedia 1 (2009) 1297–1304 1299 4 Author name / Energy Procedia 00 (2008) 000–000
In the study by Price et al. [4] it was found that diethylentriamine was not degradable in freshwater, ethylenediamine was poorly degradabl e in seawater, while diethanolamine and triethanolamine were degradable both in freshwater and in seawater. W ith an exception fo r triethanolamine, which in ou r study is not found to degrade, the results are consistent with our study.

0 20 40 60 80 100
Triethylamine
Triethanolamine
Tetra-N-methyl-propanediyldiamine
Tetrahydrothiophenedioxide
Spermine
Spermidine
Sarcosine
Pyrrolidine

1.3-Propanediamine
Piperidine
1-(2-Hydroxyethyl)-Piperazine
Piperazine
N-tertbutylethanolamine
N-methyldiethanolamine(MDEA)
Neopentandiamine
NN-dimethylethanolamine
NN'-bis(2-hydroxyethyl)ethylenediamine
N-(2-hydroxyethyl)-ethylenediamine
Morpholine
2-Methylaminoethanol
Glycine
Ethylenediamine
2-Ethylaminoethanol
Ethanolamine(MEA)
Dimethylpropanolamin
Dimethylaminopropylamin
Dimethylamin
Diisopropanolamine
Diglycolamine(DGA)
Diethylenetriamine
Diethylaminoethanol
Diethanolamine(DEA)
4-Amino-1-butanol
3-Aminopropanol
3-Amino-1-methylaminopropane
3-Amino-1-cyclohexylaminopropan
3-(2-Aminoetyl)aminopropylamine
2-Amino-2-methylpropanol(AMP)
2-Amino-2-methyl-13-propanediol
2-Amino-2-ethyl-13-propanediol
1-Amino-2-propanol
Alanine
BOD (% of ThOD)

Figure 1: The biodegradability of all the chemicals tested, results shown as percent degraded with regards to the theoretical oxygen demand (ThOD). The red line shows the lowest acceptable value for a chemical to be released in the marine environment, while the green line is the lower limit for a chemical to be released independent of the ecotoxicity [8]. 1300 I. Eide-Haugmo et al. / Energy Procedia 1 (2009) 1297–1304 Author name / Energy Procedia 00 (2008) 000–000 5

Figure 2: Results of the ecotoxicity testing, shown as concentration where compounds inhibited algal growth by 50% (EC-50). The blue line shows the lowest acceptable value (10 mg/l) for a chemical to be released in the marine environment [8]. 10 0
10 1
10 2
10 3
10 4
Triethylamine
Tetra-N-methyl-propanediyldiamine
Tetrahydrothiophenedioxide
Spermine
Spermidine
Sarcosine
Pyrrolidine

Figure 3: Biodegradability plotted against toxicity for the monoamines, sorted according to amine groups.

1 10 100 1000 10000
0 20
40 60
80 100
No hydroxylgroup
1 hydroxylgroup
2 hydroxylgroups
Aminoacids
EC-50 (mg/l)
BOD (% of ThOD)

Figure 4: Biodegradability plotted against toxicity, sorted according to amino acid-group and the number of hydroxylgroups.

The predictions from EPI Suite
TM
compared with the test results are shown in figure 5 and 6. The linear prediction model for aerobic degradability in BIOWIN fails in classifying the non-biodegradable compounds. If the term “biodegrades fast” in BIOWIN is defined as an experimental BOD-value of 20% or more, 20 out of the 42 tested compounds were classified correctly. While the groups em ployed in the group-contribution model agree with the results obtained in this study, the data base which the model is base d on does not include biodegradation tests with pure seawater. As the model is now, it is not suitable as a screening tool for marine biodegradation studies. For ecotoxicity the results were better than for the biodegradation, but there were still large deviations. 1302 I. Eide-Haugmo et al. / Energy Procedia 1 (2009) 1297–1304 Author name / Energy Procedia 00 (2008) 000–000 7
The reproducability of th e results from the ecotoxicity test was within the limits 0- 10, 10-100, 100-1000 and 1000-10000 mg/l. However, ECOSAR only predicted 27 out of 41 tested compound into the right category when using these limits. In general it seemed that the toxi city was over-predicted by the model. ECOSAR uses the octanol-water partition coefficient together with consta nts based on the class of ch emical to calculate the predictions. Altough there seems to be a correlation between the octanol-water coefficient and toxicity, additional descriptors, or a different type of cl assification of the compounds, are requi red. It should also be taken into consideration that the test used in our studies are based on 72 hours exposure, not on 96 hours, and the type of algae used is different.

0 10 20 30 40 50 60 70 80 90 100

0.0 0.5

1.0 1.5
Experimental BOD (% of ThOD)
Predicted aerobic biodegradability

Figure 5: Predicted aerobic biodegradability by the linear model in BIOWIN, as a function of experimentally determined biodegradability (BOD 28). A chemical is classified as “biodegrades fast” by BIOWIN if the predicted value is above 0,5. For experimental BOD, a value above 60% means the chemical is defined as easily degraded, while a value above 20% is the minimum preferred value [8]. 1 10 100 1000 10000
1 10
100
1000
10000
Experimental EC-50 (mg/l)
Predicted EC-50 (mg/l)

Figure 6: Predictet EC-50 values for green algae by ECOSAR as a function of experimentally determined EC-50 values for the algae Skeletonema Costatum.

I. Eide-Haugmo et al. / Energy Procedia 1 (2009) 1297–1304 1303 8 Author name / Energy Procedia 00 (2008) 000–000

4. Conclusion
Ecotoxicity and biodegradability in the marine environment have been tested for 41 and 42 compounds respectively. Some of the solvents commonly used for carbon capture, such as N-methyldiethanolamine (MDEA), piperazine and 2-amino-2-me thylpropanol (AMP), were sh own to have low biodegra dability. The tested tertiary amines and compounds containing quar ternary carbon did not degrade easily, while the aminoacids both show low toxicity and high biodegradation potential. EPI Suite
TM was used to predict the biodegradability and ecotoxicity for the chemicals tested. For biodegradab ility the predictions only showed ag reement with 48 %, while for the ecotoxicity the predictions showed agreement with 66 % of the tested compounds.

Acknowledgement
This publication forms a part of the REACT project, perf ormed under the strategic Norwegian research program Petromaks. The authors acknowledge the partners: Stat oilHydro, Shell and the Research Council of Norway (175918/S30) for their support.

References

1. S. Wong and Bioletti, R., Carbon Dioxide Separation Technologies. 2002, Carbon & Energy Managment, Alberta Research Council Inc.:

Edmonton, Alberta.

2. J.W. Davis and Carpenter, C.L. , Rev Environ Contam Toxicol. No.149 (1997) 87-137.

3. D. Calamari, et al. , Chemosphere. No.9 (1980) 753-762.

4. K.S. Price, Waggy, G.T., and Conway, R.A., Journal (Water Pollution Control Federation). No.46 (1974) 63-77.

5. OECD guidelines for testing of chemicals no. 306, adopted by the Council 17.7.92: Biodegradability in Seawater.

6. ISO Guideline ISO/DIS 10253 "Water quality - Marine algae growth inhibition test with Skeletonema costatum and Phaeodactylum tricornutum".

7. VKI., Program for statistical estimation of EC-values based on experimental data from ecotoxicological assays in TOXEDO ver. 1.2 . 1992.

8. Novatech, Supplementary Guidance for the completing of Harmonised Offshore Chemical Notification Format (HOCNF) 2000 for Norwegian sector. Harmonised Offshore Chemical Notification Format OSPAR Recommendations 2000/5. 2005.

9. US EPA Estimation Programs Interface Suite™ for Microsoft® Windows, v[3.20] . 2008; United States Environmental Protection Agency, Washington DC, USA. Available from: http://www.epa.gov/opptintr/exposure/pubs/episuite.htm [cited 2008 10. march]

10. P.H. Howard, et al. , Environmental Toxicology and Chemistry. No.11 (1992) 593-603.

1304 I. Eide-Haugmo et al. / Energy Procedia 1 (2009) 1297–1304 SDMS DocID 463456 Superfimd Records Center
SITE: >^ne-, JVc^_ _
BREAK: ^^ ' ^
Biological Report 85(1.11) Contaminant Hazard Reviews
May 1987 Report No. 11

POLYCYCLIC AROMATIC HYDROCARBON HAZARDS
TO FISH, WILDLIFE, AND INVERTEBRATES: A

SYNOPTIC REVIEW
by
Ronald Eisler
U.S. Fish and Wildlife Service
Patuxent Wildlife Research Center
Laurel, MD 20708
Fish and Wildlife Service
U.S. Department of the Interior

DISCLAIMER
Mention of trade names or commercial products does not constitute U.S. Government endorsement or recommendation for use.

This report should be cited as follows;
Eisler, R. 1987. Polycyclic aromatic hydrocarbon hazards to fish, wildlife, and invertebrates: a synoptic review. U.S. Fish Wildl. Serv. Biol.

Rep. 85(1.11). 81 pp.

SUMMARY
This account synthesizes available technical literature on ecological and toxicological aspects of polycyclic aromatic hydrocarbons (PAHs) in the environment, with special reference to natural resources. Subtopics include:
chemical properties, sources, and
fate; background concentrations in
biological and nonbiological samples; toxic and sublethal effects to flora and fauna; and proposed criteria and research needs for the protection of sensitive species.

PAHs consist of hydrogen and carbon arranged in the form of two or more fused benzene
rings. There are thousands of PAH compounds, each differing in the number and position of aromatic rings, and in the position of substituents on the basic ring system. Environmental concern has focused on PAHs that range in molecular weight from 128.16 (naphthalene,
2-ring structure) to
300.36 (coronene, 7-ring structure). Unsubstituted lower molecular weight PAH compounds, containing 2 or 3 rings, exhibit significant acute toxicity and other adverse effects to some organisms, but are noncarcinogenic; the higher molecular weight PAHs, containing 4 to 7 rings, are significantly less toxic, but many of the 4- to 7-ring compounds are demonstrably carcinogenic, mutagenic, or teratogenic to a wide variety of organisms, including fish and other aquatic life, amphibians, birds, and mammals. In general, PAHs show little tendency to biomagnify in food chains, despite their high lipid solubility, probably because most PAHs are rapidly metabolized. Inter- and intraspecies responses to individual PAHs are quite variable, and are significantly modified by many inorganic and organic compounds, including other
PAHs. Until these interaction effects are clarified, the results of single substance laboratory tests may be extremely difficult to apply to field situations of suspected PAH contamination.

PAHs are ubiquitous in nature--as evidenced by their detection in sediments,
soils, air, surface waters, and plant and animal tissues--primarily as a result of natural processes such as forest fires, microbial synthesis, and volcanic activities. Anthropogenic activities associated with significant production of PAHs--leading, in some cases, to localized areas of high contamination--include high-temperature (>700 C) pyrolysis of organic materials typical of some processes used in the iron and steel industry, heating and power generation, and petroleum refining. Aquatic environments may receive PAHs from accidental releases of petroleum and its products, from sewage effluents, and from other sources. Sediments heavily contaminated m

with industrial PAH wastes have directly caused elevated PAH body burdens and increased frequency of liver neoplasia in fishes.

At present, no criteria or standards have been promulgated for PAHs by any regulatory agency for the protection of sensitive species of aquatic organisms or wildlife. This observation was not unexpected in view of the paucity of data on PAH background concentrations in wildlife and other natural resources, the absence of information on results of chronic oral feeding studies of PAH mixtures, the lack of a representative PAH mixture for test purposes, and the demonstrable--and, as yet, poorly understood--effects of biological and nonbiological modifiers on PAH toxicity and metabolism. By contrast, criteria for human health protection and total PAHs, carcinogenic PAHs, and benzo(a)pyrene have been proposed for drinking water and air, and for total PAHs and benzo(a)pyrene in food: drinking water, 0.01 to <0.2 ug total PAHs/1, <0.002 ug carcinogenic PAHs/1, and <0.0006 ug benzo(a)pyrene/l; air, <0.01 ug -total PAHs/m , <0,002 ug carcinogenic PAHs/m , and <0.0005 ug benzo(a)pyrene/m ; food, 1.6 to <16.0 ug total PAHs daily, and 0.16 to <1.6 ug benzo(a)pyrene daily. In view of the carcinogenic characteristics of many PAH compounds and their increasing concentrations in the environment, it now seems prudent to reduce or eliminate them wherever possible, pending acquisition of more definitive ecotoxicological data.

CONTENTS
Page
SUMMARY. iii
TABLES -vi
FIGURES vii
ACKNOWLEDGMENTS viii
INTRODUCTION 1
ENVIRONMENTAL CHEMISTRY, SOURCES, AND FATE 3
Properties , 3
Sources 7
Fate 10
BACKGROUND CONCENTRATIONS 15
General 15
Nonbiological Samples 15
Biological Samples 25
TOXIC AND SUBLETHAL EFFECTS 34
General 34
Fungi 34
Terrestrial Plants ..... 34
Aquatic Biota 36
Amphibians and Reptiles 49
Birds 50
Mammals 50
RECOMMENDATIONS 58
LITERATURE CITED 64

TABLES
Number Page
1 Some physical and chemical properties of
selected PAHs 7
2 Major sources of PAHs in atmospheric and
aquatic environments 8
3 PAH concentrations in selected nonbiological
materials 16
4 PAH concentrations in field collections of
selected biota. Values are shown in ug/kg
fresh weight (FW), or dry weight (DW) 26
5 Toxicities of selected PAHs to aquatic
organisms 37
6 PAH bloconcentratlon factors (BCF) for
selected species of aquatic organisms 41
7 Some effects of PAHs on selected laboratory
animals 52
8 Proposed PAH criteria for human health
protection (modified from EPA 1980; Lee
and Grant 1981; Pucknat 1981) 60
VI

FIGURES
Number Page
1 Nomenclature of PAHs (modified from Lee and
Grant 1981, and Grimmer 1983) 4
2 Ring structures of representative
noncarcinogenic PAHs (modified from Lee and
Grant 1981, and Neff 1985) 5
3 Ring structures of representative tumorogenic,
cocarcinogenic, and carcinogenic PAHs (modified
from Lee and Grant 1981) 6
4 The bay region dihydrodiol epoxide route
of benzo(a)pyrene (modified from
Dipple 1985) 14
vn

ACKNOWLEDGMENTS
I thank Jean Higgis and Julia Armstrong for secretarial help; Lynda J. Garrett, Nancy A. Bushby, and Joyce E. Haber for literature retrieval; Matthew C. Perry, Martha L. Gay, David J. Hoffman, and Paul C.

Baumann for constructively reviewing the manuscript; and Curtis H. Halvorson and James R. Zuboy for editorial services.
vm

x&
rm-^
'W!
'i%-^^-"
m^
m

INTRODUCTION
Several polycyclic aromatic hydrocarbons (PAHs) are among the most potent carcinogens known to exist, producing tumors in some organisms through single exposures to microgram quantities. PAHs act at both the site of application and at organs distant to the site of absorption; their effects have been demonstrated in nearly every tissue and species tested, regardless of the route of administration (Lee and Grant 1981). The evidence implicating PAHs as an inducer of cancerous and precancerous lesions is becoming overwhelming, and this class of substances is probably a major contributor to the recent increase in cancer rates reported for industrialized nations (Cooke and Dennis 1984). PAHs were the first compounds known to be associated with carcinogenesis (Lee and Grant 1981). Occupational skin cancer was first documented in London chimney sweeps in 1775 and in German coal tar workers in the late 1800's. By the early 1900's, soot, coal tar, and pitch were all found to be carcinogenic to humans. By 1918, it was shown that topical applications of coal tar produced skin tumors in
mice and rabbits;
benzo(a)pyrene, a PAH, was identified as one of
the most carcinogenic
compounds in coal tar (Dipple 1985). The carcinogenic activity to man of ¥:' soots, tars, and oils is beyond dispute. In addition to the skin cancers
noted initially, higher incidences of respiratory tract and upper
tract tumors were associated with occupational
gastrointestinal exposures to
(Dipple 1985). PAH-induced cancers in laboratory animals is these carcinogens
Benzo(a)pyrene, for example, has produced tumors in mice, well documented,
guinea pigs, rabbits, ducks, and monkeys following
rats, hamsters,
oral, dermal, and intraperitoneal routes (Pucknat 1981). administration by
Teratogenic or carcinogenic responses have been induced in sponges, planarians, echinoderm larvae, teleosts, amphibians, and plants by exposure to carcinogenic PAHs (Neff 1979, 1982b). An unusually high prevalence of oral, dermal, and hepatic neoplasms have been observed in bottom-dwelling fish from polluted sediments containing grossly-elevated PAH levels (Couch and Harshbarger 1985). PAH compounds have damaged chromosomes in cytogenetic tests, have produced mutations in mammalian cell culture systems, and have induced DNA repair synthesis in human fibroblast cultures (EPA 1980). While some PAHs are potent mutagens and carcinogens, others are less active or suspected carcinogens. Some, especially those of biological origin, are probably not carcinogens (Jackim and Lake 1978). Certain lower molecular weight, noncarcinogenic PAHs, at environmentally realistic levels, were acutely toxic to aquatic organisms, or produced deleterious sublethal t
responses (Neff 1985). However, few generalizations can be made about the class of PAH compounds because of the extreme variability in toxicity and physicochemical properties of PAHs and their various effects on individual species (Lee and Grant 1981).

PAHs are widely-distributed in the environment, almost ubiquitous, and have been detected in animal and plant tissues, sediments, soils, air, surface water, drinking water, industrial effluents, ambient river water, well water, and groundwater (EPA 1980). Man probably has always been exposed to PAHs from the natural background level in soils and plants ,(Harrison et al. 1975); avoiding exposure to nanogram quantities of these substances on a daily basis is now considered essentially impossible for all living resources (Dipple 1985). Ever since benzo(a)pyrene was recognized as a carcinogen at the beginning of this century, the presence of it and of other PAHs in the environment has received continuous attention. As one consequence, many reviews have been published on ecological and toxicological aspects of PAH in the environment, with special reference to their carcinogenic properties.

In this report, I summarize selected data on environmental aspects of PAHs, emphasizing PAH effects to aquatic and wildlife resources. This brief review is part of a continuing series prepared in response to informational requests from environmental specialists of the U. S. Fish and Wildlife Service.
^ Harrison et al. (1975); Barnett (1976); Suess (1976); Gelboin and Ts'o (1978a, 1978b, 1981); Jackim and Lake (1978); Jones and Freudenthal (1978); Lo and Sandi (1978); Jones and Leber (1979); Neff (1979, 1982a, 1982b, 1985); Tsang and Griffin (1979); Bjorseth and Dennis (1980); EPA (1980); Cooke and Dennis (1981, 1983, 1984); Futoma et al. (1981); Lee and Grant (1981); Pucknat (1981); Sims and Grover (1981); Stegemen (1981); Cooke et al. (1982); Richards and Jackson (1982); Couch et al. (1983); Edwards (1983); Grimmer (1983); Quaghebeur et al. (1983); Sims and Overcash (1983); Couch and Harshbarger (1985); Harvey (1985); Johnson et al. (1985); Sugimura (1986).

ENVIRONMENTAL CHEMISTRY, SOURCES, AND FATE

PROPERTIES
Polycyclic aromatic hydrocarbons (PAHs), also known as polynuclear aromatic hydrocarbons (PNAs) and polycyclic organic matter (POM), are composed of hydrogen and carbon arranged in the form of two or more fused benzene rings in linear, angular, or cluster arrangements, which may or may not have substituted groups attached to one or more rings (Sims and Overcash 1983). In some cases, the newly defined substituted PAH has strikingly greater toxicological effects than does the parent compound (Cooke and Dennis 1984).

The nomenclature of PAH compounds has been ambiguous in the past due to different peripheral numbering systems. The currently accepted nomenclature is shown in Figure 1.

Of major environmental concern are mobile PAHs that vary in molecular weight from 128.16 (naphthalene, CJQHQ) to 300.36 (coronene, C^^H^p)- Higher molecular weight PAHs are relatively immobile because of their Targe molecular volumes and their extremely low volatility and solubility. Among the mobile forms are thousands of compounds that differ in the number and position of aromatic rings, and in the position of substituents on the basic ring system.

The lower molecular weight unsubstituted PAH compounds, containing 2 to 3 rings, such as naphthalenes, fluorenes, phenanthrenes, and anthracenes (Figure 2), have significant acute toxicity to some organisms, whereas the higher molecular weight 4- to 7-ring aromatics do not. However, all known PAH carcinogens, cocarcinogens, and tumor producers are in the high molecular weight PAH group (Figure 3).

Physical and chemical characteristics of PAHs generally vary with molecular weight. With increasing molecular weight, aqueous solubility decreases, and melting point, boiling point, and the log Kow (octanol/water partition coefficient) increases (Table 1), suggesting increased solubility in fats, a decrease in resistance to oxidation and reduction, and a decrease in vapor pressure. Accordingly, PAHs of different molecular weight vary substantially in their behavior and distribution in the environment and in

Figure I. Nomenclature of PAHs (modified from Lee and Grant 1981, and Grimmer 1983). The PAH formula is oriented so that the greatest number of rings are in a horizontal row and a maximum number of rings are above and to the right of the horizontal row. The first carbon atom that belongs to the uppermost ring and is not engaged in ring fusion with another ring is given the number C-1; numbering continues in a clockwise direction omitting those carbon atoms which do not carry a hydrogen atom. The bond between C-1 and C-2 is designated as side "a"; other peripheral sides continue in clockwise direction in alphabetical order. Examples are: (1) pyrene (correctly oriented, numbered, and
lettered), (2) benzo(a)pyrene (not oriented correctly), (3) benzo(a)pyrene (correctly oriented, numbered, and lettered).

CxX)
03 02
'^y&-'^
c& C&9
OS 06
CX^ oxo
08 0? 07
to <
co5)
_>f?"^Y'''
oxo
SSa-^'. •
Wit' ®,
^•".•'
y}:%'••• n
.'ifi'-':.
•If-' 65?
16 18
Figure 2. Ring structures of representative noncarcinogenic PAHs (modified from Lee and Grant 1981, and Neff 1985). The numbering and lettering system for several PAHs Is also given. Compounds are: (1) naphthalene, (2) fluorene, (3) anthracene, (4) phenanthrene, (5) aceanthrylene, (6) benzo(a)fluorene, (7) i)enzo(b)fluorene, (8) benzo(c)fluorene, (9) fluoranthene, (10) naphthacene, (.11) pyrene, (12) benzo(k)fluoranthene, (13) benzo(g,h,i)fluoranthene, (14) perylene, (15) benzo(e)pyrene, (16) benzo(g,h,i)perylene, (17) anthanthrene, (18) coronene.

« ' «
03 02
01 01. 05 07
06 H, H.
c—c'
09 08
oK" .W?
H, H,
16 13
O ^
19 20
Figure 3. Ring structures of representative tumorogenic, cocarc inogenic, and carcinogenic PAHs (modified from Lee and Grant 1981). The numbering and lettering system for several PAHs is also given. Compounds are: (1) chrysene, (2) benz(a)anthracene, (3) dibenzo(a,h)fluorene, (4) dibenzo(a,g )fluorene, (5) dibenzo(a,c)fluorene, (6) dibenz(a,c)anthracene, (7) dibenzja , j)anthracene, (8) indeno (1,2,3-cd) pyrene, (9) dibenzo(a,l)pyrene, (10) cholanthrene, (11) benzo(j)fluoranthene, (12) benzo(b)fluoranthene, (13) diben zo(a,e)pyrene, (14) dimethylbenz(a)anthracene, (15) benzo(c)phenanthrene, (16) 3-methylchol anthrene, (17) dibenz(a,h)anthracene, (18) benzo(a)pyrene, (19) dibenzo(a,h) pyrene, (20) dibenzo(a, ijpyrene. Compounds 1 to 9 are weakly carcinogenic, cocarcinogenic, or tumorogenic; compounds 10 to 13 are carcinoge nic; compounds 14 to 20 are strongly carcinogenic.

Table 1. Some physical and chemical properties of selected PAHs.

Approximate Melting Solubility
Number of molecular point in water log
rings weight ( C) (mg/l) Kow
Naphtha!ene 2 128 80 30.0 3.37
6S4i©iWithracene 3 178 216 0.07 4.45
•^^^^J^tBenz(a)anthracene 4 228 158 0.014 5.61
Benzo(a)pyrene 5 252 179 0.0038 6.04
f^?te~Benzo(g,h,i) perylene 6 276 222 0.00026 7.23
their biological effects. Additional and more comprehensive data on the physical and chemical properties of PAHs are given in Barnett (1976), Lo and Sandi (1978), Neff (1979, 1985), EPA (1980), Futoma et al. (1981), Lee and Grant (1981), Pucknat (1981), Edwards (1983), Grimmer (1983), Sims and Overcash (1983), and Whitehouse (1985).

SOURCES
About 43,000 metric tons of PAHs are discharged into the atmosphere each year, and another 230,000 tons enter aquatic environments (Table 2). PAHs are ubiquitous in nature as a consequence of synthesis in terrestrial vegetation, microbial synthesis, and volcanic activity, but quantities formed by these natural processes are small in comparison with those produced from forest and prairie fires and anthropogenic sources (Barnett 1976; Suess 1976; Lo and Sandi 1978; Neff 1979, 1985; EPA 1980; Lee and Grant 1981; Pucknat 1981; Edwards 1983; Grimmer 1983; Sims and Overcash 1983). Anthropogenic activities associated with significant production of PAHs include: coke production in the iron and steel industry; catalytic cracking in the petroleum Industry; the manufacture of carbon black, coal tar pitch, and asphalt; heating and power generation; controlled refuse incineration; open burning; and emissions from Internal combustion engines used in transportation. Thus, the formation of PAHs in the environment is due to an endogenous synthesis by microorganisms. i!B«.;

Table 2. Major sources of PAHs in atmospheric and aquatic environments (modified from Lo and Sandi 1978; Neff 1979; Edwards 1983; Sims and Overcash 1983).

Ecosystem and sources Annual input, in metric tons

ATMOSPHERE
Total PAHs
Forest and prairie fires
Agricultural burning
Refuse burning
Enclosed incineration
Heating and power
• •
19,513
13,009
. 4,769
3,902
2,168
!nzo(a)pyrene
Heating and power
Worldwide
USA only
Industrial processes (mostly coke
Worldwide
USA only
Refuse and open burning
Worldwide
USA only
Motor vehicles
Worldwide
USA only
production)
2,604
475
1,045
198
1,350
588
45 22

AQUATIC ENVIRONMENTS
Total PAHs
Petroleum spillage
Atmospheric deposition
Wastewaters
Surface land runoff
Biosynthesis
170,000
50,000
4,400
2,940
2,700
Total benzo(a)pyrene 700
algae, and macrophytes which provide natural background, and to a second process which is connected to man-controlled high-temperature (>700 C) pyrolysis of organic materials, to open burning, and to natural volcanic activities. The discovery in fossil fuels of complex mixtures of PAHs spanning a wide range of molecular weights has led to the conclusion that, given sufficient time (i.e., millions of years), pyrolysis of organic materials at temperatures as low as 100 to 150 C can also lead to production of PAHs (Neff 1985).

Forest and prairie fires release much greater amounts of PAHs to the atmosphere than does fossil fuel burning. Nearly all of the airborne PAHs produced by flame pyrolysis are associated with the particulate fraction produced during combustion, and these are significantly modified by the chemical composition of the
fuel, the pyrolysis temperature, the duration of
exposure to elevated temperatures, and to other factors (Neff 1979; Edwards 1983). In one study, a PAH profile was established for a series of laboratory fires simulating the prescribed burning of pine needle litter (McMahon and Tsoukalas 1978). Heading fires (moving with wind) produced more total particulate matter than backing fires (moving against wind), but backing fires produced significantly higher amounts of PAHs, with the actual amounts formed dependent on fuel loading and the residence time of combustible gases in the burning zone. Emission factors for benzo(a)pyrene varied from 238 to 3,454 ug/kg in backing fires and 38 to 97 ug/kg in heading fires.

PAHs present in the atmosphere enter rain as a result of in-cloud and below-cloud scavenging (van Noort and Wondergem 1985). Total PAHs deposited on land and water are almost equivalent to PAH content in rainfall; significant quantities of PAHs are found in presumed pollution-free areas, indicating the importance of rain in transport and distribution of PAHs (Quaghebeur et al. 1983).

PAHs may reach aquatic environments in domestic and industrial sewage effluents, in surface runoff from land, from deposition of airborne particulates, and especially from spillage of petroleum and petroleum products into water bodies (Jackim and Lake 1978; Lake et al. 1979; Neff 1979; EPA 1980; Martens 1982; Boehm and Farrington 1984; Hoffman et al. 1984; Prahl et al. 1984). The majority of PAHs entering aquatic environments remains close to sites of deposition, suggesting that lakes, rivers, estuaries, and coastal marine environments near centers of human populations are the primary repositories of aquatic PAHs (Neff 1979). Large variations in aquatic PAH contents were evident due to localized source inputs and physicochemical conditions. For example, urban runoff from stormwater and highways to Narragansett Bay, Rhode Island, accounted for 71% of the total inputs for higher molecular weight
PAHs, and 36% of the total PAHs (Hoffman et al.
1984). More than 30% of all combustion-derived PAHs in coastal sediments of Washington State is supplied by riverine transport of suspended particulate materials, while direct atmospheric input accounts for a maximum of 10% (Prahl et al. 1984). In contrast, concentrations of PAHs in sediments from the

vicinity of Georges Bank, off the US northeastern coast, varied from 1 to 100 ug/kg dry weight, and were directly related to total organic carbon, silt, and clay contents in sediments; combustion-derived PAHs dominated at the higher concentrations, while lower levels were often associated with a fossil fuel origin (Boehm and Farrington
1984).

Discharge water from hydrostatic testing of natural gas pipelines is a significant source of PAH loading into aquatic environments, contributing as much as 32,000 ug PAHs/1 of discharge water, mostly as naphthalenes (Eiceman et al. 1984). More than 25 PAHs, primarily anthracenes and pyrenes, were detected in pipeline residues on inner walls of natural gas pipelines at concentrations up to
2,400 ug/m of inner surface; the same compounds may be
reasonably expected in aqueous wastes from pipeline maintenance (Eiceman et al. 1985). Release of these, or similar, discharge waters directly into aquatic environments will result in contamination similar to that caused by oil spills; however, these sites for pollution may occur in locations far distant from oil production and refinery activities (Eiceman et al. 1984).

PAHs are also present in tap water at concentrations of 0.1 to 1.0 ng/1, primarily as mono- and dichlorinated derivitives of naphthalene, phenanthrene, fluorene, and fluoranthene (Shiraishi et al. 1985). The presence of PAHs and chlorinated PAHs in tap water indicates the reaction of PAHs with chlorine; however, their significance to human health and to aquatic biota is unknown.

FATE
Concern about PAHs in the environment is due to their persistence and to the fact that some are known to be potent mammalian carcinogens, although environmental effects of most noncarcinogenic PAHs are poorly understood (Neff 1985). Prior to 1900, a natural balance existed between the production and the degradation of PAHs. Synthesis of PAHs by microorganisms and volcanic activity and production by man-made high temperature pyrolytic reactions and open burning seemed to be balanced by PAH destruction via photodegradation and microbial transformation, With increased industrial development and increased emphasis of fossil fuels as energy sources, the balance has been disturbed to the extent that PAH production and introduction into the environment greatly exceeds known PAH removal processes (Suess 1976; Sims and Overcash 1983).

When released into the atmosphere, PAH compounds will become associated with particulate materials. Their residence time in the atmosphere and transport to different geographic locations are governed by particle size, meteorological conditions, and atmospheric physics. The highly reactive PAHs photodecompose readily in the atmosphere by reaction with ozone and various oxidants; degradation times range from several days to six weeks for PAHs adsorbed onto particulates <1 um in diameter (in the absence of rainfall) to <1 day to several days for those adsorbed to larger particles (Suess 1976).
10 Smaller atmospheric particulates containing PAHs are easily inhaled (Lee and Grant 1981), and may pose special problems, as yet unevaluated, for airborne organisms such as birds, insects, and bats. Photooxidation, one of the most important processes in the removal of PAHs from the atmosphere, can also produce reaction products that are carcinogenic or mutagenic, although little is known of their persistence (Edwards 1983). One of the more common photooxidation reactions of PAHs is the formation of endoperoxides that ultimately undergo a series of reactions to form quinones (Edwards 1983).

Various parameters may modify chemical and photochemical transformation of PAHs in the atmosphere, including light intensity, concentration of gaseous pollutants (0,, NO , SO ), and chemicophysical characteristics of particulates or substrates into whicn the PAHs are adsorbed; depending on these variables, the half-life of benzo(a)pyrene in the atmosphere varies from 10 minutes to 72 days (Valerio et al.
1984). Atmospheric PAHs are transported over relatively long distances from industrial areas and from natural forest and prairie fires (Edwards 1983); however, sites nearer urban centers have much higher PAH deposition rates than more rural areas (Hites and Gschwend 1982).

Much of the PAHs released into the atmosphere eventually reaches the soil by direct deposition or by deposition on vegetation. The PAHs may be adsorbed or assimilated by plant leaves before entering the animal food chain, although some adsorbed PAHs may be washed off by rain, chemically oxidized to other products, or returned to the soil as the plants decay. PAHs assimilated by vegetation may be translocated, metabolized, and possibly photodegraded within the plant. In some plants growing in highly contaminated areas, assimilation
may exceed metabolism and degradation, resulting in an accumulation in plant tissues (Edwards 1983).

In water, PAHs may either evaporate, disperse into the water column, become incorporated into bottom sediments, concentrate in aquatic biota, or experience chemical oxidation and biodegradation (Suess 1976). The most important degradative processes for PAHs in aquatic systems are photooxidation, chemical oxidation, and biological transformation by bacteria and animals (Neff 1979). Most PAHs in aquatic environments are associated with particulate materials; only about 33% are present in dissolved form (Lee and Grant 1981). PAHs dissolved in the water column will probably degrade rapidly through photooxidation (EPA 1980), and degrade most rapidly at higher concentrations, at elevated temperatures, at elevated oxygen levels, and at higher incidences of solar radiation (McGinnis and Snoeyink 1974; Suess 1976; Bauer and Capone 1985). The ultimate fate of those PAHs that accumulate in sediments is believed to be biotransformation and biodegradation by benthic organisms (EPA 1980). PAHs in aquatic sediments, however, degrade very slowly in the absence of penetrating radiation and oxygen (Suess 1976), and may persist indefinitely in oxygen-poor basins or in anoxic sediments (Neff 1979). PAH degradation in aquatic environments occurs at a slower rate than that in the atmosphere (Suess 1976), and the cycling of PAHs in aquatic environments, as is true for other ecological systems, is poorly understood (Neff 1979).
11 Animals and microorganisms can metabolize PAHs to products that may ultimately experience complete degradation. The degradation of most PAHs is not completely understood. Those in the soil may be assimilated by plants, degraded by soil microorganisms, or accumulated to relatively high levels in the soil. High PAH concentrations in soil can lead to increased populations of microorganisms capable of degrad.ing the compounds. Of equal importance to PAH cycling dynamics is the physical state of the PAH, i.e., whether in vapor phase or associated with particles such as flyash. Particles may increase or decrease the susceptibility of PAHs to degradation, depending on the PAH and particles involved (Edwards 1983).

PAHs can be taken into the mammalian body by inhalation, skin contact, or ingestion, although they are poorly absorbed from the gastrointestinal tract.

The main routes of elimination of PAHs and their metabolites include the hepatobiliary system and the gastrointestinal tract (Sims and Overcash 1983).

In mammals, an enzyme system variously known as the cytochrome P-450-dependent mixed-function oxidase, mixed-function oxidase, mixed-function oxygenase, aryl hydrocarbon hydroxylase, or drug metabolizing system, is responsible for initiating the metabolism of various lipophilic organic compounds, including PAHs. The primary function of this system is to render poorly water soluble lipophilic materials more water soluble, and therefore more available for excretion. Some PAHs are transformed to intermediates, which are highly toxic, mutagenic, or carcinogenic to the host. Oxidative metabolism of PAHs in this system proceeds via high electrophilic intermediate arene oxides, some of which bind covalently to cellular macromolecules such as DNA, RNA, and protein. Most authorities agree that metabolic activation by the mixed-function oxidase system is a necessary prerequisite for PAH-induced carcinogenesis and mutagenesis (Neff 1979). This enzyme system is known to be present in rodent tissues, and human liver, skin, placenta, fetal liver, macrophages, lymphocytes, and monocytes (Lo and Sandi 1978). Studies with rodents have shown that the mixed-function oxidase system can convert PAHs to various hydroxylated derivatives including phenols, quinones, and epoxides, and can also activate PAHs to produce carcinogenic metabolites (Lo and Sandi 1978). Fish and most crustaceans tested to date possess the enzymes necessary for activation (Statham et al. 1976; Varanasi et al. 1980; Fabacher and Baumann 1985), but some molluscs and other invertebrates are unable to efficiently metabolize PAHs (Jackim and Lake 1978; Varanasi et al. 1985).

Although many aquatic organisms possess the requisite enzyme systems for metabolic activation of
PAHs, it is not certain in most cases whether these
enzymes produce the same metabolites as those produced by mammalian enzymes (Neff 1979).

PAHs are metabolized by liver mixed-function oxidases to epoxides, dihydrodiols, phenols, and quinones. The intermediate metabolites have been identified as the mutagenic, carcinogenic, and teratogenic agents (Sims and Overcash 1983). The activation mechanisms occur by hydroxylation or 12 production of unstable epoxides of PAHs which damage DNA, initiating the carcinogenic process (Jackim and Lake 1978). Metabolic formation of bay region diol epoxides represents an important pathway by which PAHs are activated to carcinogens (Figure 4). Such metabolic activation proceeds via initial formation of the dihydrodiol with the bay region double bond, followed by subsequent oxidation of the dihydrodiol to the bay region diol epoxide (Sims and Overcash
1983). Active epoxides may be converted to less toxic
products by various enzymatic and other reactions (Neff 1979). In the case of benzo(a)pyrene, the "ultimate carcinogen" (7 beta, 8 alpha-dihydroxy-,7,8,9,10 tetrahydrobenzo(a)pyrene- 9 alpha, 10 alpha-epoxide) reacts with the guanine of RNA and DNA, the linkage taking place between the C-10 atom of benzo(a)pyrene and the C-2 amino group of guanine (Grimmer 1983; Dipple 1985; Figure 4). Additional information on actual and theoretical mechanisms involved in the metabolic activation of PAHs are given in Cavalieri et al. (1978, 1980), Bjorseth and Dennis (1980), Herd and Greene (1980), Cooke and Dennis (1981), Sims and Grover (1981), Grimmer (1983), Szentpaly (1984), Harvey (1985), and Yan (1985).
13 Bay Raglon 12 1
HO-
7,8-dlhydrodlol 9,10-epoxide 7,8-dihydrodiol
Figure 4. The bay region dihydrodiol epoxide route of
benzo(a)pyrene (modified from Dipple 1985).
14 BACKGROUND CONCENTRATIONS

GENERAL
PAHs are ubiquitous in the environment. In nonbiological materials, concentrations are elevated in the vicinity of urban industrialized locales, and from areas of significant wood burning activities such as forest fires and residential home heating. Terrestrial vegetation and aquatic invertebrates can accumulate significant concentrations of PAHs, possibly due to inefficient or missing mixed-function oxidase systems. Fish do not appear to contain grossly elevated PAH residues; this may be related to their efficient degradation system. At present, data are lacking or unavailable on PAH background concentrations in natural populations of birds and other wildlife --although it seems unlikely that significant accumulations will occur. Some investigators have shown that aquatic invertebrates, fish, and amphibians collected from areas of high sediment PAH content show elevated frequencies of hyperplasia and neoplasia (Rose 1977; Mix 1982; Black 1983; Mai ins et al. 1984, 1985a, 1985b; Black et al. 1985; Baumann et al., in press), and, recently, that hepatic carcinoma has been induced in rainbow trout (Salmo qairdneri) by benzo(a)pyrene through dietary and intraperitoneal injection routes (Hendricks et al. 1985).

More comprehensive information on PAH background levels in various biological and nonbiological compartments is given in Lo and Sandi (1978), Neff (1979, 1985), Pucknat (1981), Edwards (1983), Grimmer (1983), and Sims and Overcash (1983).

NONBIOLOGICAL SAMPLES
Total PAH levels in air are usually much higher in winter than in summer, higher in urban communities than in rural areas (Table 3; Grimmer 1983), and appear to be related primarily to the weight of total suspended particulates in the atmosphere (Hites and Gschwend 1982; Greenberg et al. 1985; Srivastava et al. 1985; Ang et al. 1986). PAH levels in precipitation are significantly higher in winter than in summer, primarily due to emissions from household heating (Quaghebeur et al. 1983; van Noort and Wondergem 1985). Among 15 Table 3. PAH concentrations in selected nonbiological materials, Material (units),
and other variables Concentration Reference*
AIR (ng/m^)
USA cities, 1959, total PAHs
Detroit 95.1 EPA 1980
Birmingham 63.4
Nashville 60.6
New Orleans 33.6
Los Angeles 31.8
Atlanta 26.3
San Francisco 13.7
Sydney, Australia
Winter 8.2 Barnett 1976
Summer 0.6
USA cities, 1971-1977
Benzo(a)perylene = BaPER 0.2-9.2 EPA 1980
Benzo(e)pyrene = BeP 0.9-4.6
Benzo(k)fluoranthene = BkFL 0.03-1.3
Pyrene = PYR 0.18-5.2
Coronene = COR 0.2-6.4
Perylene = PER 0.01-1.2
Anthracene = A 0.07-0.3
Naphthalene = NA Max. 0.4
Benz(a)anthracene = BaA Max. 4.6
Indeno(l,2,3-cd)pyrene = IP • Max. 1.3
Steel mill, Ontario, Canada, 1971-1979
Station 0.8 km distant
Benzo(a)pyrene = BaP 9.4 (Max. n o.0) Potvin
BkFL 8.9 (Max. 142 .0) et al. 1981
Fluoranthene = FL 7.0 (Max. 43 .3)
PER 9.1 (Max. 106 .0)
Benzo(g,h,i)perylene = BghiPER 13.7 (Max. 90 .0)
Station 2.8 km distant
BaP 0.4 (Max. 7 9)
BkFL
FL

0.7 (Max.

1.1 (Max.
5 4
1)
8)
PER
BghiPER

0.7 (Max.

1.4 (Max.
9 8
1)
5)
Benzo(a)pyrene = BaP
Urban areas 0.1-61.0 Edwards 1983
Downwind from coal gasification
plant, Yugoslavia Max. 80.0
16 Table 3. (Continued)
Material (units),
and other variables Concentration
Urban areas
1966 3.2
1970 2.1
1976 0.5
Rural areas 0.01-1.9
Rural areas
1966 0.4
1976 0.1
SOILS
Near M6 Motorway, Lancaster, UK
(maximum deposition rate, ng/m /wee Jk)
Distance from roadway

3.8 meters
A 2,300
FL 15,200
BaA 5,800
Benzo(b)fluoranthene = BbFL 7,300
BkFL 2,800
BaP 4,900

9.0 - 47 meters
A 420
FL 1,700
BaA 260
BbFL 690
BkFL 470
BaP 290
Vicinity slash burn site, Oregon (g/ha)
0-2 cm depth
Preburn
Phenanthrene = PHEN 0,
FL 0,
103 days postburn
PHEN 9.

FL 3.
365 days postburn

PHEN ND
FL 0.
2-5 cm depth
* !
Reference
EPA 1980
Edwards 1983
EPA 1980
Johnston and
Harrison 1984
Sullivan
and Mix 1985
17 Table 3.

Material (units),
and other variables
105 days postburn
PHEN
FL
365 days postburn
PHEN
FL
BaP (ug/kg)
Rural areas
Industrial areas
Nonpolluted areas
Near known sources
Near coal-tar pitch
disposal site, Germany
Near recreation area, USSR
Forest soil
LITTER
Forest, Oregon (g/ha)
3 days postburn
PHEN
FL
32 days postburn
PHEN
FL
Coniferous trees (ug/kg)
BghiPER
BaP
IP
FL
DIMENTS (ug/kg)
Buffalo River, near Buffalo, NY
Sediments
BaA
Chrysene = CHRY
BbFL
(Continued)
Concentration

1.3 0.3
ND
ND

0.4 400.0
up to 1,000
>100,000
650,000

0.4 1.5-4.0
603
245
ND
ND
42 51
47 164
7,300
4,300
3,500
Reference'
Barnett 1976
Edwards 1983
Lee and
Grant 1981
Harrison
et al. 1975
Sullivan and
Mix 1985
Thomas
et al. 1984
Black 1983
18 Table 3. (Continued)
Material (units). J J
and other variables Concentration Reference
BaP 4,500
Dibenz(a,h)anthracene = DBA 1,000
IP 4,400
Sediment extracts
BaA 16,000
CHRY 14,000
BbFL 13,900
BaP 15,400
DBA 3,300
IP 12,300
Cayuga Lake, Ithaca, NY,, 1978
Total PAHs
Within marinas 4,600-13,900 Heit 1985
Deepwater 1,260-2,500
Near power plant 104-6,800
FL
Within marinas 1,700
Deepwater 285
Near power plant 8-1,000
Penobscot Bay, Maine
Total PAHs 286-8,794 Johnson
PHEN 17-252 et al. 1985
A ND-49
FL 156-3,700
Pyrene = PYR 16-539
BaA 14-540
CHRY 9-578
BbFL 17-1,000
BkFL 14-696
BaP 10-540
DBA 2-120
BghiPER 23-641
IP 9-228
Casco Bay, Maine, total PAH:5 215-14,425
Charles River, Mass., total PAHs 87,000-120,000
Boston Harbor, Mass., total PAHs 8,500
New Bedford Harbor, Mass;., 1total PAHs 63,000
Lake Erie, total PAHs 530-3,750
Adirondack Lakes, total PAH:5 4,070-12,807
19 Table 3. ( Continued.)
Material (units),
and other variables Concentration Reference
Alaska, total PAHs
Tamar estuary, UK, total PAHs
Southampton estuary, UK, total PAHs
Severn estuary, UK, total PAHs
Monaco, total PAHs
Gulf of Finland, total PAHs
Norway, total PAHs
Walvis Bay, Africa, total PAHs
Amazon River system, total PAHs
5-113
4,900
91,000-1,791,000
1,600-25,700
5,200-12,100
437
284-99,452
68 ND-544
SEWAGE
Waters, worldwide, total PAHs(ug/l)
Sludge, total PAHs
United Kingdom, 12 sites, (ug/kg)
Fresh weight
Dry weight
Texas, Reese Air Force Base
Effluent lagoon (ug/kg fresh weight)
PER
PYR
FL
BaA
CHRY
BaP
BeP
A
100-500
80-1,760
200-50,300
300.0

5.8 5.7

1.4 1.3

0.5 0.2

0.2 Lee and
Grant 1981
Mclntyre
etal. 1981
Rose 1977
MOTOR OILS (ug/1)
Unused
BaP
CHRY
PER
Used
BaP
CHRY
PER
115
56 11
1,382
10,170
1,024
Pasquini and
Monarca 1983
20 Table 3. (Continued)
Material (units).
and other variables
GROUNDWATER (ug/1)
Worldwide
Total PAHs
Total PAHs
Carcinogenic PAHs
Germany
Total PAHs
Carcinogenic PAHs
Champaign, Illinois
Total PAHs
Carcinogenic PAHs
Elkhart, Indiana
Total PAHs
Carcinogenic PAHs
DRINKING WATER (ug/1)
USA, total PAHs
Europe, total PAHs
Monongehala River, Pittst )urgh, PA
Untreated
Total PAHs
Carcinogenic PAHs
Treated
Total PAHs
Carcinogenic PAHs
Ohio River, Wheeling, WV
Untreated
Total PAHs
Carcinogenic PAHs
Treated
Total PAHs
Carcinogenic PAHs
Lake Winnebago, Appleton, WI
Untreated
Total PAHs
Carcinogenic PAHs
Concentration

0.01-0.05

0.045-0.51

0.00-0.081

0.04 0.003

0.007

0.003

0.02 0.004

0.015

0.04-0.06

0.6 0.14

0.003

0.002

1.59 0.57

0.14 0.011

0.007

0.002
» \
Reference
Lee and
Grant 1981
Harrison
et al. 1975
EPA 1980
Lee and
Grant 1981
- EPA 1980
21 Table 3. (Concluded)
Material (units),
and other variables Concentration Reference
Treated
Total PAHs 0.006
Carcinogenic PAHs 0.002
SURFACE WATER (ug/1)
Worldwide
Low level contamination 0.05-0.25 Lee and
Medium polluted 0.2-1.0 Grant 1981
Germany, Rhine River
Total PAHs 1.12 EPA 1980
Carcinogenic PAHs 0.49
Thames River, UK
Total PAHs 0.5-1.33
Carcinogenic PAHs 0.18-0.56
^Each reference applies to data in the same row and in the rows that immediately follow for which no reference is indicated. 22 industrial sources, the production of metallurgical coke is the single most significant source of atmospheric PAHs in Ontario, Canada. Coke production in 1977 represented about 52% of all PAH emissions from Ontario sources versus about 46% formed as a result of forest fires (Potvin et al. 1981). Beyond 2
km distant from the coke point source, PAH concentrations in air were typical of those measured in major urban nonindustrialized areas (Table 3; Potvin et al. 1981). A variety of PAHs have been detected in ambient air in the USA and elsewhere. Benzo(a)pyrene, because of its carcinogenic properties, has been monitored extensively, and has frequently been used as an indicator of PAHs (EPA 1980). In general, total PAHs in air is about lOX higher than benzo(a)pyrene levels, although this relation is extremely variable (Lee and Grant 1981). Benzo(a)pyrene levels, like total PAHs, were higher in winter than summer, probably due to residential and industrial heating; air levels in urban areas with coke ovens were 40% to 70% higher than in cities without coke ovens, but this may be related to higher industrial emissions in those cities (Lee and Grant 1981). In one case, benzo(a)pyrene levels in air from the center of a remote mountain community in Colorado were several times higher than what, is usually found in U.S. metropolitan areas, and was attributed to extensive residential wood burning (Murphy et al.
1982). Average
concentrations of benzoXa)pyrene in urban air Nationwide declined from 3.2 ng/m in 1966 to 0.5 ng/m in 1978, an 80% decrease (Lee and Grant 1981).

These decreases are believed to be due primarily to decreases in coal consumption for commercial and residential heating, improved disposal of solid wastes, and restrictions on open burning (EPA 1980).

A major source of PAHs in soils and soil litter is from emissions and deposition from forest fires. In a controlled burn study, Sullivan and Mix (1985) showed that lower molecular weight PAHs, such as phenanthrene and fluorene, which had been deposited in soil litter, degraded to nondetectable levels within 2 years after burning. Higher molecular weight PAHs such as benzo(k)fluorene, benzo(a)pyrene, benzo(g,h,i)perylene, perylene, and indeno(l,2,3-cd)pyrene, were more persistent in litter, decreasing after 5 years to about 20% of initial deposition. Although movement into the top 2 cm of the soil profile was initially more pronounced for lower molecular weight PAHs, all compounds appeared to reach equilibrium between litter and soil on the basis of organic content within one year postburn. Differential persistence and fate of PAHs on slash burn sites is explained by solubility, Kow, and other physicochemical properties (Sullivan and Mix 1985). PAHs from vehicle emissions constitute a minor, but measurable, source of soil PAHs (Table 3). The majority of highway-derived PAHs appears to be deposited within

3.8 m of the road, but the influence of the highway may extend to nearly 70 m (Johnston and Harrison 1984). The use of composted municipal wastes for conditioning of agricultural soils is not recommended, as these contain at least nine identified carcinogenic PAHs (Martens
1982).

Some sediments were found to be highly contaminated with PAHs. Sediments and sediment extracts from the Buffalo River, New York, contained elevated levels of carcinogenic PAHs (1,000-16,000
ug/kg). Brown bullheads (Ictalurus
23 nebulosus), in response to repeated applications of Buffalo River sediment extracts, showed epidermal hyperplasia and neoplasia when compared to controls (Black 1983). PAH concentrations in sediments from the Great Barrier Reef, Australia, were always <0.8 ug/kg dry weight, except in small areas close to sites frequently visited by power boats; in those instances, total PAH levels exceeded 13,400 ug/kg (Smith et al. 1985). Highest PAH levels measured in sediments of Cayuga Lake, New York, were found in marinas or areas of the lake receiving urban runoff, and were apparently not related to stack emissions from a nearby coal-fired power plant; Heit (1985) believed that stack emissions were either masked by other sources or were atmospherically transported and deposited elsewhere. Coastal and offshore sediments are subject to highly elevated PAH levels from a variety of sources, mostly unknown, relative to preindustrial times (Johnson et al. 1985). For example, PAH levels in sediments of Penobscot Bay, Maine, fell within the range found in sediments near industrialized regions, and were significantly higher than expected for an area previously considered to be uncontaminated (Table 3; Johnson et al. 1985).

Sewage effluents usually contained measurable levels of PAHs, although extreme variability between and among sites is common. For example, during a heavy storm, individual PAH levels in a sewage works may increase more than lOOX over a dry weather period (Harrison et al.
1975). Conventional sewage
treatment plant processes remove up to 90% of carcinogenic PAHs, and this may be increased to 99% using percolating filters and activated sludge processes (Harrison et al. 1975). Tiger salamanders (Ambystoma tigrinum), collected in 1975 from a 13-ha sewage effluent lagoon at Reese Air Force Base, Texas, showed a remarkably high incidence (53%) of neoplastic and other lesions (Rose 1977). Analysis of sludge composites showed elevated PAH levels, especially perylene; levels of organochlorine and organophosphorus pesticides, nitrosamines, and heavy metals were judged to be nonelevated (Rose 1977).

Careful disposal of used motor oils is warranted, as these contain high quantities of mutagenic and carcinogenic PAHs (Table 3; Pasquini and Monarca 1983).

All but the most heavily contaminated fresh and marine waters contain total PAH concentrations in the part-per-trillion or low part-per-billion range (Table 3; Neff 1982b). A large proportion of the PAH content in water is probably adsorbed onto suspended solids (Harrison et al. 1975). In Lake Michigan, concentrations of total PAHs in the surface microlayer gVaried from

0.15 to 0.45 ug/1, representing on a relative scale 10 times the concentration in air, suggesting that aerosols are a major source of these compounds and that the microlayer is a repository until the PAHs are removed by adsorption and sedimentation (Strand and Andren
1980).
24 BIOLOGICAL SAMPLES
Carcinogenic PAHs have been extracted from a large variety of fresh plants, including root and leaf vegetables, fruits, grains, and edible mushrooms, as well as from various marine bacteria and phytoplankton under circumstances suggesting that PAHs were present due to local biosynthesis (Suess 1976). Vegetation and soil near known PAH sources are more highly contaminated with PAHs than those collected at greater distances (Edwards 1983). PAH levels in lettuce (Lactuca sativa) grown in Sweden seemed to be directly related to its proximity to local recognized point sources of PAH emitters (Table 4; Larsson and Sahlberg 1982). Washing lettuce with water had little effect on phenanthrene levels, but significantly reduced other PAHs, such as benzo(a)pyrene, benz(a)anthracene, and benzo(g,h,i)perylene by 68% to 87% (Larsson and Sahlberg 1982). Fruits and vegetables grown in polluted atmospheres may contain up to lOOX higher levels of total PAHs than those grown in unpolluted environments (EPA 1980; Lee and Grant 1981). PAH concentrations for plants are generally greater on plant surfaces than internal tissues, greater in above ground plant parts than those below ground, and greater in plants with broad leaves (greater surface area) than those with narrow leaves (Edwards 1983). Plants can become contaminated with PAHs through environmental pollution, particularly through deposition from the atmosphere, and also through food processing. For example, the bran portion of milled wheat, as well as finished bran cereal, had a considerably higher PAH content than other fractions or finished products (Lawrence and Weber 1984b), Enrichment of PAHs in plants is associated with deposition of atmospheric particulate matter with relatively small particle sizes; thus, PAH content is usually in the order of humus > mosses > lichens (Thomas et al. 1984). Mosses appear to be good indicators of regional PAH air pollution and have been recommended for this purpose (Herrmann and Hubner 1984).

Concentrations of total PAHs in soils, usually the sum of 5 to 20 PAHs, typically exceeded benzo(a)pyrene levels by at least one order of magnitude; however, concentrations of benzo(a)pyrene in vegetation were generally less than those in soil where plants were growing (Edwards 1983).

PAH accumulations in marine molluscs have been reported (Table 4); however, some of these data may be misleadingly low. For example, lengthy cold storage of 10 months can result in loss of volatile PAHs, such as anthracene, in tissues of mussels (Smith et al. 1984); accordingly, background concentrations in these organisms may be underreported. Bivalve molluscs tend to accumulate high PAH levels due to their inability to metabolize and excrete them (Lawrence and Weber 1984a), presumably due to inefficient or missing mixed-function oxidase systems (Sirota and Uthe 1981). Cellular proliferative disorders, resembling neoplastic conditions in vertebrates, were found in mussels with the greatest PAH concentrations: 9.5% vs. 0.7% in control site (Mix 1982). Baseline levels of PAHs in indigenous bivalve molluscs reflected the degree of human onshore activity at the various sample sites, and presumably the level of water contamination; however, little relation was 25 Table 4. PAH concentrations in field collections of selected biota.

Values are shown in ug/kg (ppb) fresh weight (FW), or dry weight (DW).

Taxonomic group, compound.
and other variables

ALGAE AND OTHER PLANTS
Marine algae, Greenland
Total PAHs
Marine algae, Benzo(a)pyrene=BaP
Freshwater alga.

Chlorella vulgaris, BaP
Bacteria, BaP
Moss, Hypnum cuoressiforme
Southern Finland, 1982
Near center of
industrial town
BaP
Fluoranthene=FL
Benzo(g,h,i)perylene=BghiPER
Indeno(l,2,3 cd)pyrene=IP
Vegetation
Total PAHs
Nonpolluted areas
Near known source
BaP
Lettuce,
Lactuca sativa, total PAHs
Sweden, summer 1980
Grown near highway
8-15 m distant
15-50 m distant
Near airport, 150-800 m
Aluminum smelter

0.5-1.5 km distant

2.0-6.5 km
Industrial areas
Residential areas
Urban
Rural
Seedlings, wheat
and rye, BaP
Concentration
60 FW
Up to 60
DW
10-50 DW
2-6 DW
110 DW
250 DW
90 DW
41 DW
20-1,000 DW
25,000 DW

0.1-150.0 DW
50 FW
26 FW
24 FW
654 FW
128 FW
13 FW
13 FW
12 FW
10-20 DW
Reference^
I
Harrison et al. 1975
Lee and Grant 1981
Suess 1976
Herrmann
and Hubner 1984
Edwards 1983
Larsson and
Sahlberg 1982
Suess 1976
26 Table 4. (Continued)
Taxonomic group, compound.
and other variables Concentration Reference^

INVERTEBRATES
Rock crab. Cancer irroral :us
Edible portions, 1980
New York Bight
Total PAHs 1,600 FW Humason and
BaP 1 FW Gadbois 1982
Long Island Sound
Total PAHs 1,290 FW
BaP ND
American oyster,
Crassostrea virginica, soft parts
South Carolina, 1983,
residential resorts
Total PAHs
Spring months
Palmetto Bay 520 FW Marcus and Stokes 1985
Outdoor Resorts 247 FW
Fripp Island 55 FW
Summer months
Palmetto Bay 269 FW
Outdoor Resorts 134 FW
Fripp Island 21 FW
American lobster, Homarus ; americanus
Edible portions, 1980
New York Bight
Total PAHs 367 FW Humason and
BaP 15 FW Gadbois 1982
Long Island Sound
Total PAHs 328 FW
BaP 15 FW
Softshell clam, Mva arenaria
Coos Bay, Oregon, 19781979
Soft parts
Contaminated site
Total PAHs 555 FW Mix 1982
Phenanthrene = PHEN 155 FW
FL 111 FW
Pyrene = PYR 62 FW
BaP 55 FW
Benz(a)anthracene = BaA 42 FW
27 Tabl 2 4. (Continued)
Taxonomic group, compound.
and other variables Concentration
Chrysene = CHRY 27 FW
• Benzo(b)fluoranthene = BbFL 12 FW
Others <10 FW
Uncontaminated site
Total PAHs 76 FW
PHEN 12 FW
FL 10 FW
Others <10 FW
Bav mussel, Mytilus edulis
Oregon, 1979-1980
Soft parts, total PAHs
Near industrialized area 106-986 FW
Remote site 27-274 FW
Sea scallop, Placopectin maqell anicus
Baltimore Canyon, east coast USA
Muscle
BaA 1 FW
BaP <1 FW
PYR 4 FW
New York Bight, 1980
Edible portions
Total PAHs 127 FW
BaP 3 FW
Clam, Tridacna maxima
Australia, 1980-1982, Great Barr ier Reef
Soft parts, total PAHs
Pristine areas <0.07 FW
Power boat areas Up to 5 FW
BaP
Marine plankton
Greenland 5 FW
Italy 6-21 FW
France 400 FW
Worldwide Up to 400 DW
Mussel, Mytilus sp.

Greenland
Shell 60 FW
Soft parts 18 FW
Reference^
'
Mix and Schaffer 1983b
Brown and
Panelrov 1979
Humason and
Gadbois 1.982
Smith et al. 1984
Harrison et al. 1975
Lee and Grant 1981
Harrison et al. 1975
28 Table 4. (Continued)
Taxonomic group, compound,
and other variables Concentration Reference'
Italy
Shell 11 FW
Soft parts 130-540 FW
Bivalve molluscs, 5 spp.

Edible portions 6 (Max. 36) FW Stegeman 1981
Decapod crustaceans, 4 spp.

Edible portions 2 (Max. 8) FW
Softshell clam, Mya arenaria. soft parts
Coos Bay, Oregon
1976-1978
Near industrialized areas 6-20 FW Mix and Schaffer 1983a Remote areas 1-2 FW
1978-1979
Near industrialized areas 9 FW
Remote areas 4 FW

VERTEBRATES
Fish, muscle
Lake Ontario, 6 spp., total PAH 3-8 FW Lawrence and
Baltimore Canyon, Weber 1984a
east coast, USA, 5 spp.

BaA Max. 0.3 FW Brown and
BaP Max. <5 FW Pancirov 1979
PYR Max. <5 FW
Smoked
FL 3 FW EPA 1980
PYR 2 FW
Nonsmoked
FL Max. 1.8 FW
PYR Max. 1.4 FW
Winter flounder, Pseudopleuronectes americanus
amencanus
Edible portions, 1980
New York Bight
Total PAHs 315 FW
Humason and
BaP 21 FW Gadbois 1982
Long Island Sound
Total PAHs 103 FW
BaP ND
29 Table 4. (Continued)
Taxonomic group, compound,
and other variables
Windowpane, Scopthalmus aauosus
Edible portions, 1980
New York Bight
Total PAHs
BaP
Long Island Sound
Total PAHs
BaP
Red hake, Uroohvcus chuss
Edible portions, 1980
New York Bight
Total PAHs
BaP
Long Island Sound
Total PAHs
BaP
BaP
Fish
Marine, edible portions
9 spp.

Greenland
Italy
Steak, charcoal broiled
Ribs, barbecued

INTEGRATED STUDIES
Michigan, 1978, Hersey River
Near wastewater treatment plant
PHEN
Sediments
Insects, whole
Crustaceans, muscle
Fish, muscle
BaA
Sediments
Insects
Crustaceans
Fish
Concentration Reference'
536 FW
4 FW
86 FW
ND
412 FW
22 FW
124 FW
5 FW
Max. 3 FW Stegeman 1981
15 FW Harrison et al. 1975
65 FW
5-8 DW Barnett 1976
11 DW
4,097 FW Black et al. 1981
5,488 FW
447 FW
28-15,313 FW
3,504 FW
2,893 FW
40 FW

0.2-19 FW
30 Table 4. (Concluded)
Taxonomic group, compound,
and other variables Concentration Reference^
BaP
Sediments 1,194 FW
Insects 725 FW
Crustaceans 8 FW
Fish 0.07-1 FW
Control location
Sediments and biota
PHEN - 2-42 FW
BaA ND-6.7 FW
BaP 0.04-1.2 FW
Nova Scotia, 1980, total PAHs
Near coking facility
Sediments 2,830,000 DW Sirota et al. 1983
American lobster.

Homarus americanus
Hepatopancreas 57,300-88,100 FW
Tail muscle 1,910-2,670 FW
Control area
Sediments <8,220 DW
American lobster
Hepatopancreas 1,185 FW
Tail muscle 216 FW
Black River, Ohio, contaminated
area, total PAHs
Sediments 6,700 DW West et al. 1984
Brown bullhead,
Ictalurus nebulosus 660 FW
Water 153 FW
Each reference applies to data in the same row and in the rows that immediately follow for which no reference is indicated. 31 evident between accumulated levels of individual PAHs and total PAHs (Mix 1982). Elevated PAH concentrations, especially benz(a)anthracene, chrysene, fluorene, phenanthraene, and pyrene in oyster tissues and sediments were measured in samples from the vicinity of marinas, and were higher in oysters in cooler months, when lipids and glycogen were being stored preparatory to spawning (Marcus and Stokes 1985). In general, PAH concentrations in marine clams were highest in areas adjacent to industrialized bayfronts and lowest in clams inhabiting more remote areas; concentrations were lowest in autumn-winter, and highest during spring-summer (Mix and Schaffer 1983a). A similar pattern was observed in mussels, Mytilus edulis, with the more water soluble, lower molecular weight, PAHs bioconcentrated lOX to lOOX above that of the higher molecular weight, less water soluble PAHs (Mix and Schaffer 1983b); PAH levels in mussels seemed to be independent of water salinity (Mix and Schaffer 1979). Clams contaminated with PAHs and removed to clean seawater for 24 hours showed significant depuration of unsubstituted 3- and 4-ring PAHs; in contrast, concentrations of all 5-, 6-, and 7-ring compounds, which includes most of the carcinogenic PAHs, were not significantly depurated (Mix 1982). A positive relation exists between PAH isomers in sediments, soft tissues of the mussel Mytilus edulis, and a seaweed (Fucus sp.) collected at Vancouver, British Columbia (Dunn 1980). For mussels, the general trend towards lower levels of higher molecular weight PAHs relative to levels in sediments suggests an uptake mechanism which involves the solution of PAHs in water; superimposed on this pattern is the more rapid turnover and shorter half-life of lower molecular weight PAHs in mussels (Dunn 1980).

PAH residues were higher than expected in American lobsters (Homarus americanus) collected offshore (mean weight 3.6 kg) when compared to smaller (0.6 kg) lobsters collected inshore (Sirota and Uthe 1981), suggesting that age or body size are important modifiers in PAH accumulation dynamics. PAH concentrations in sediments collected near a coking facility in Nova Scotia in 1980 contained up to 2,830 mg/kg dry weight, or more than 20X the levels recorded in Boston (Mass.) Harbor; concentrations in excess of 100 mg/kg dry weight sediment were recorded for phenanthrene, fluorene, pyrene, benz(a)anthracene, chrysene, benzo(e)pyrene, benzo(b)fluoranthene, and benzo(a)pyrene, and these seemed to reflect the elevated tissue levels in American lobsters collected from that locale (Sirota et al. 1983). PAH residues in digestive glands of American lobsters collected in 1979 in Nova Scotia from the vicinity of a major oil spill were higher than those from coastal control sites; however, PAH contents of edible muscle from control and oiled lobsters were similar (Sirota and Uthe 1981).

PAH levels in fish are usually, low because this group rapidly metabolizes PAHs (Lawrence and Weber 1984a); furthermore, higher molecular weight PAHs, which include the largest class of chemical carcinogens, do not seem to accumulate in fish (West et al. 1984). Raw fish from unpolluted waters usually do not contain detectable amounts.of PAHs, but smoked or cooked fish contain varying levels. The concentration of benzo(a)pyrene in skin of cooked 32 fish was much higher than in other tissues, suggesting that skin may serve as a barrier to the migration of PAHs in body tissues (EPA 1980).

Sediments and biota collected from the Hersey River, Michigan, in 1978, were heavily contaminated with phenanthrene, benz(a)anthracene, and benzo(a)pyrene when compared to a control site. Elevated PAH concentrations were recorded in sediments, whole insect larvae, crayfish muscle, and flesh of lampreys (family Petromyzontidae), brown trout (Salmo trutta), and white suckers (Catostomus commersoni), in that general order (Black et al. 1981).

The polluted collection locale was the former site of a creosote wood preservation facility between 1902 and 1949, and, at the time of the study, received Reed City wastewater treatment plant effluent, described as an oily material with a naphthalene-like odor (Black et al. 1981). In many cases, aquatic organisms from PAH-contaminated environments have a higher incidence of tumors ahd hyperplastic diseases than those from nonpolluted environments.

Carcinogenic PAHs have not been unequivocally identified as the causative agent for an increased incidence of cancer in any natural population of aquatic organisms, according to Neff
(1982b). However, a growing body of
evidence, mostly circumstantial, links PAHs to cancer in feral fish populations, especially bottom dwelling fish from areas with, sediments heavily contaminated with PAHs (Baumann and Lech, in press).
33 TOXIC AND SUBLETHAL EFFECTS

GENERAL
A'wide variety of PAH-caused adverse biological effects have been reported in numerous species of organisms under laboratory conditions, including effects on survival, growth, metabolism, and especially tumor formation. Inter- and intraspecies responses to carcinogenic PAHs were quite variable, and were significantly modified by many chemicals including other PAHs that are weakly carcinogenic or noncarcinogenic. Until these interaction effects are clarified, the results of single substance laboratory tests may be extremely difficult to apply to field situations of suspected PAH contamination.

FUNGI
Fungal degradation of PAHs may be important in the detoxification and elimination of PAHs in the environment. The fungus Cunninghamella elegans, for example, inhibited the mutagenic activity of benzo(a)pyrene, 3-methyl cholanthrene, benz(a)anthracene, and 7,12-dimethylbenz(a)anthracene, as judged by results of the Ames test using Salmonella typhimurium (Cerniglia et al. 1985). The rate of decrease in mutagenic activity in bacterial cultures incubated with PAHs was coincident with the rate of increase in fungal metabolism. C.elegans metabolized PAHs to dihydrodiols, phenols, quinones, and dihydrodiol epoxides, and to sulfate, glucuronide, and glucoside conjugates of these primary metabolites in a manner similar to that reported for mammalian enzyme systems, suggesting that this organism (and perhaps other fungi) is important in PAH metabolism and inactivation (Cerniglia et al. 1985).
34 TERRESTRIAL PLANTS
Biological effects of PAHs on terrestrial vegetation have been reviewed by EPA (1980), Lee and Grant (1981), Wang and Meresz (1982), Edwards (1983), and Sims and Overcash (1983). In general, these authorities agreed on several points. First, plants and vegetables can absorb PAHs from soils through their roots, and translocate them to other plant parts such as developing shoots.

Uptake rates were governed, in part, by PAH concentration, PAH water solubility, soil type, and PAH physicochemical state (vapor or particulate).

Lower molecular weight PAHs were absorbed by plants more readily than higher molecular weight PAHs. Under laboratory conditions, some plants concentrated selected PAHs above that of their immediate geophysical surroundings, but this has not been conclusively demonstrated in field-grown cultivated crops or other vegetation. Second, above-ground parts of vegetables, especially the outer shell or skin, contained more PAHs than underground parts, and this was attributed to airborne deposition and subsequent adsorption. Externally deposited PAHs in vegetables were difficult to remove with cold water washings; not more than 25% were removed from lettuce, kale, spinach, leeks, and tomatoes using these procedures. Third, PAH-induced phytotoxic effects were rare; however, the data base on this subject is small. Fourth, most higher plants can catabolize benzo(a)pyrene, and possibly other PAHs, but metabolic pathways have not been clearly defined. Finally, the biomagnification potential of vegetation in terrestrial and aquatic food chains needs to be measured; this work should be conducted with a variety of PAHs in both field and laboratory experiments.

Some plants contain chemicals known to protect against PAH effects.

Certain green plants contain ellagic acid, a substance that can destroy the diol epoxide form of benzo(a)pyrene, inactivating its carcinogenic and mutagenic potential (Edwards 1983). PAHs synthesized by plants may act as plant growth hormones (Edwards 1983). Some vegetables, such as cabbage, brussel sprouts, and cauliflower, contain naturally occurring antineoplastic compounds including benzyl isothiocyanate and phenethyl isothiocyanate; these compounds are known to inhibit mammary cancers, stomach tumors, and pulmonary edemas induced in rats by benzo(a)pyrene and 7,12-dimethylbenz(a)anthracene (EPA
1980). Decreased activation of carcinogens has also been demonstrated in animals fed diets that were high in protein, low in carbohydrate, and containing adequate choline; the reverse was observed in diets high in carbohydrate, low in protein, or containing certain organophosphorus insecticides, piperonyl butoxide, carbon tetrachloride, nickel carbonyl, or tin (EPA 1980). In cases where dietary constituents can alter the metabolism of foreign agents, such as PAHs, the anticarcinogenic effect may result from an alteration of steady state levels of activated versus detoxified metabolites (EPA 1980). The implications of these observations to herbivorous wildlife are unknown at present.
35 AQUATIC BIOTA
PAHs vary substantially in their toxicity to aquatic organisms (Table 5). In general, toxicity increases as molecular weight increases (although high molecular weight PAHs have low acute toxicity, perhaps due to their low solubility in water) and with increasing alkyl substitution on the aromatic ring. Toxicity is most pronounced among crustaceans and least among teleosts (Neff 1979; Table 5). In all but a few cases, PAH concentrations that are acutely toxic to aquatic organisms are several orders of magnitude higher than concentrations found in even the most heavily polluted waters (Neff 1979).

Sediments from polluted regions, however, may contain PAH concentrations similar to those which are acutely toxic, but their limited bioavailability would probably render them substantially less toxic than PAHs in solution (Neff 1979).

A growing literature exists on uptake, retention, and translocation of PAHs by aquatic plants and animals. Authorities generally agree that: most species of aquatic organisms studied to date rapidly accumulate (i.e., bioconcentrate) PAHs from low concentrations in the ambient medium; uptake of PAHs is highly species specific, being higher in algae, molluscs, and other species which are incapable of metabolizing
PAHs; bloconcentratlon factors
(BCF) tend to increase as the molecular weight of the PAH increases, with increasing octanol/water partition coefficient values, with time until approaching an apparent equilibrium level (sometimes within 24 hours), and with increases in dissolved organic matter in the medium, lipid content of organism, and a variety of endogenous and exogenous factors (Jackim and Lake 1978; Southworth et al. 1978; Lee and Grant 1981; Neff 1982a). BCF values have been determined for selected PAHs and aquatic organisms (Table 6); additional BCF data for aquatic biota are available for plants (Dobroski and Epifanio 1980; Boyle et al.
1984), crustaceans (Southworth 1979; Sirota and
Uthe 1981; Fox and Rao 1982; Neff 1982a; Williams et al. 1985), tunicates (Baird et al. 1982), molluscs (Jackim and Wilson 1979; Dobroski and Epifanio 1980; Neff 1982a), and fishes (Southworth 1979; Neff 1982a; Stoker et al. 1984). Algal accumulation of benzo(a)pyrene increased linearly in a 24-hour exposure period, and correlated positively with surface area (Leversee et al. 1981), suggesting adsorption rather than absorption. Algae readily transform benzo(a)pyrene to oxides, peroxides (Kirso et al. 1983), and dihydrodiols (Warshawsky et al. 1983). Photosynthetic rates of algae, and presumably PAH accumulations, were significantly modified by light regimens. For reasons still unexplained, algae grown in "white" light (major energy in blue-green portion of the spectrum) were more sensitive to benzo(a)pyrene than were cultures grown in "gold" light (Warshawsky et al. 1983; Schoeny et al. 1984).

Accumulation by oysters (Crassostrea virginica) and clams (Rangia cuneata) of naphthalene, phenanthrene, fluorene, and their methylated derivatives increased with increasing methylation and PAH molecular weight; uptake was more rapid under conditions of aontinuous flow than in static tests (Neff et al. 1976). When returned to PAH-free seawater, molluscs released PAHs to non 36 Table 5. Toxicities of selected PAHs to aquatic organisms, PAH compound, Concentration
organism, and in medium
other variables (ug/I)
BENZ(a)ANTHRACENE
Bluegill, Lepomis
macrochirus 1,000
BENZ0(a)PYRENE
Sandworm, Neanthes
arenceodentata >1,000

CHRYSENE
Sandworm >1,000
7,12-DIMETHYLBENZ(a)ANTHRACENE
Minnows, Poeiciliopsis spp.

Juveniles 250
Juveniles 500
DIBENZ(a,h)ANTHRACENE
Sandworm >1,000

FLUORANTHENE
Sandworm 500

FLUORENE
Grass shrimp, Palaemonetes
puqio 320
Bluegill 500
Amphipod, Gammarus
pseudoliminaeus 600
Rainbow trout, Salmo
qairdneri 820
Bluegill 910
Sandworm 1,000
Effect'
LC-87 (6 m)
LC-50 (96 h)
LC-50 (96 h)
LC-0 (20 h)
LC-lOO (20 h)
LC-50 (96 h)
LC-50 (96 h)
LC-50 (96 h)
LC-12 (30 d)
LC-50 (96 h)
LC-50 (96 h)
LC-50 (95 h)
LC-50 (96 h)
Reference
EPA 1980
Neff 1979
Schultz
and Schultz 1982
Neff 1979
Finger et al. 1985
Neff 1979
37 Table 5. (Continued)
PAH compound,
organism, and
other variables
Concentration
in medium
(ug/1) Effect' Reference
Sheepshead minnow,
Cyprinodon variegatus 1,680
Snail, Mudalia
potosensis 5,600
Mayfly, Hexagenia
bilineata 5,800
Fathead minnow, Pimephales
promelas >100,000
LC-50 (96 h)
LC-50 (96 h)
LC-50 (120 h)
LC-0 (96 h)
Finger et al. 1985

NAPHTHALENE
Copepod, Eurytemora
affinis 50
Pink salmon, Oncorhynchus
gorbuscha. fry 920
Dungeness crab. Cancer
maqister 2,000
Grass shrimp 2,400
Sheepshead minnow 2,400
Brown shrimp, Penaeus
aztecus 2,500
2,680
Amphipod, Elasmopus
pectenicrus
Coho salmon, Oncorhyncus
kisutch. fry 3,200
Sandworm 3,800
Mosquitofish, Gambusia
affinis 150,000 LC-50 (96 h)
LC-30 (10 d)
LC-50 (24 h)
LC-50 (96 h)
LC-50 (96 h)
LC-50 (24 h)
LC-50 (24 h)
LC-50 (96 h)
LC-50 (96 h)
LC-50 (96 h)
Neff 1979
Neff 1985
Neff 1979
Neff 1985
Neff 1979
1-METHYLNAPHTHALENE
Dungeness crab. Cancer
magister
Sheepshead minnow
1,900
3,400
LC-50 (96 h)
LC-50 (24 h)
2-METHYLNAPHTHALENE
Grass shrimp
Dungeness crab
Sheepshead minnow
1,100
1,300
2,000
LC-50 (96 h)
LC-50 (96 h)
LC-50 (24 h)
Neff 1985
Neff 1979
38 miiLom
Table 5. (Concluded)
PAH compound, Concentration
organism, and in medium
other variables (ug/1) Effect^ Reference

TRIMETHYLNAPHTHALENES
Copepod, Eurytemora
affinis 320 LC-50 (24 h)
Sandworm 2,000 LC-50 (96 h)

PHENANTHRENE
Grass shrimp 370 LC-50 (24 h)
Sandworm 600 LC-50 (96 h) EPA 1980
1-METHYLPHENANTHRENE
Sandworm 300 LC-50 (96 h)
m = months, d = days, h = hours.
•*Each reference applies to data in the same row and in the rows that immediately follow for which no reference is indicated.
39 detectable levels in about 60 days, with high molecular weight PAHs depurated more slowly than low molecular weight compounds; brown shrimp (Penaeus aztecus) and longnose killifish (Fundulus similis), which can metabolize PAHs; lost PAHs more quickly than clams and oysters, which apparently lack the detoxifying enzymes (Neff et al. 1976). Pink shrimp (Penaeus duorarum) exposed to 1.0 ug chrysene/1 for 28 days and then transferred to unpolluted seawater for an additional 28 days contained concentrations of chrysene (91 ug/kg fresh weight in abdomen, 48 ug/kg in cephalothorax) that were considered potentially hazardous to human consumers over extended periods (Miller et al. 1982). Eggs of the sand sole (Psettichthys melanostictus) exposed to 0.1 ug benzo(a)pyrene /I for 5 days showed reduced and delayed hatch and, when compared to controls, produced larvae with high accumulations (2.1 mg/kg fresh weight) and gross abnormalities, such as twinning and tissue overgrowths, in 50% of the test larvae (Hose et al. 1982). Naphthalene and benzo(a)pyrene were rapidly accumulated from the medium by three species of California marine teleosts; loss was rapid, being >90% for naphthalene in 24 hours, and 20% (muscle) to 90% (gill) for benzo(a)pyrene in a similar period (Lee et al. 1972). Phenanthrene is metabolized by many species of aquatic organisms, including fish. A marine flounder, Platichthys flesus, given a single oral dose of 0.7 mg phenanthrene/kg body weight, contained elevated phenanthrene concentrations in lipids, melanin-rich tissues (such as skin), and the eye
lens; most was eliminated within 2 weeks (Solbakken et al. 1984). Different rates of accumulation and depuration of benzo(a)pyrene and naphthalene in bluegill (Lepomis macrochirus) and Daphnia magna have been documented by McCarthy and Jimenez (1985) and McCarthy et al.
(1985). Benzo(a)pyrene
accumulations in bluegill, for example, were lOX greater than naphthalene, but benzo(a)pyrene is extensively metabolized, whereas naphthalene is not.

Consequently, postexposure accumulations of naphthalene greatly exceeded that of the parent benzo(a)pyrene. Because the more hydrophobic PAHs, such as benzo(a)pyrene, show a high affinity for binding to dissolved humic materials and have comparatively rapid biotransformation rates, these interactions may lessen or negate bloaccumulation and food chain transfer of hydrophobic PAHs (McCarthy and Jimenez 1985; McCarthy et al.
1985).

Time to depurate or biotransform 50% of accumulated PAHs (Tb 1/2) varied widely. Tb 1/2 values for Daphnia pulex and all PAH compounds studied ranged between 0.4 and 0.5 hours (Southworth et al. 1978). For marine copepods and naphthalene, a Tb 1/2 of about 36 hours was recorded (Neff 1982a). For most marine bivalve molluscs, Tb 1/2 values ranged from 2 to 16 days. Some species, such as the hardshell clam (Mercenaria mercenaria), showed little or no depuration, while others, such as oysters, eliminated up to 90% of accumulated PAHs in 2 weeks--although the remaining 10% was released slowly, and traces may remain indefinitely (Jackim and Lake 1978). Percent loss of various PAHs in oysters (Crassostrea virginica), 7 days postexposure, ranged from no loss for benzo(a)pyrene to 98% for methyl naphthalene; intermediate were benz(a)anthracene (32%), fluoranthene (66%), anthracene (79%), dimethyl- naphthalene (90%), and naphthalene (97%) (Neff 1982a). Teleosts and 40 Table 6. PAH bloconcentratlon factors (BCF)
for selected species of aquatic organisms.

PAH compound, organism, Exposure
and other variables period

ANTHRACENE
Cladoceran, Daphnia
magna 60 m
Fathead minnow.

Pimephales promelas 2 to 3 d
Cladoceran, Daphnia
pulex 24 h
Mayfly, Hexagenia sp. 28 h
Rainbow trout, Salmo
gairdneri 72 h
9-METHYLANTHRACENE
Cladoceran, Daphnia
pulex 24 h
BENZ(a)ANTHRACENE
Cladoceran, Daphnia
pulex 24 h
BENZ0(a)PYRENE
Teleosts, 3 spp.,
Muscle 1 h
Clam, Rangia
cuneata 24 h
Bluegill, Lepomis
macrochirus 4 h
Atlantic salmon, Salmo salar
Egg 168 h
Midge, Chironomus
riparius, larvae 8 h
Rainbow trout, liver 10 d
BCF
200
485
760 to 1200
3,500
4,400 to
9,200
4,583
10,109

0.02 to 0.1
9 to 236
12 71
166
182 to 920
Reference
EPA 1980
Southworth 1979
Southworth et al. 1978;
Southworth 1979;
EPA 1980; Neff 1985
EPA 1980
Linder et al. 1985
Neff 1985
Southworth et al. 1978
EPA 1980
Neff 1979; EPA 1980
Leversee et al. 1981
Kuhnhold and
Busch 1978
Leversee et al. 1981
Gerhart and
Carlson 1978
41 Table 6. (Continued)
PAH compound, organism. Exposure
and other variables period BCF Reference
Oyster, Crassostrea
virginica
Northern pike Esox lucius
Bile and gallbladder
II
II
II
Liver
II
II
II
Gills
II
II
II
Kidney
II
II
Other tissues
Mosquitofish, Gambusia
affinis
Bluegill
No dissolved humic
material (DHM)
20 mg/l DHM
Cladoceran, Daphnia
magna
Alga,
Oedogonium cardiacum
Periphyton, mostly
diatoms
Mosquito, Culex oipiens
quinouefasciatus
Sand sole, Psettichthys
melanostictus
Egg
Snail, Physa sp.

Cladoceran, Daphnia
pulex
14 d

3.3 h

19.2 h

8.5 d
23 d

3.3 h

19.2 h

8.5 d
23 d

3.3 h

19.2 h

8.5 d
23 d

3.3 h

19.2 h

8.5 d

3.3 h to
3 d
48 h
48 h
6 h
3 d
24 h
3 d
6 d
3 d
3 d
242
3,974
36,656
82,916
53,014
259
578
1,376
619
283
382
372
213
192
872
1,603
23 d <55
930
2,657
225
2,837
5,258
9,600
11,536
21,000
82,231
134,248
EPA 1980
Balk et al. 1984
Lu et al. 1977
McCarthy and
Jimenez 1985
Leversee et al. 1981
Lu et al. 1977
Leversee et al. 1981
Lu et al. 1977
Hose et al. 1982
Lu et al. 1977
42 Table 6. (Continued)
PAH compound, organism. Exposure
and other variables period BCF

CHRYSENE
Clam, Rangia
cuneata 24 h 8
Mangrove snapper,
Lutjanus griseus
Liver 4 d 83 to 104
Liver 20 d 258 to 367
Pink shrimp, Penaeus duorarum
Cephalothorax 28 d 248 to 361
Cephalothorax 28 d + 28 d
postexposure 21 to 48
Abdomen 28 d 84 to 199
Abdomen 28 d + 28 d
postexposure 22 to 91

FLUORANTHENE
Rainbow trout,
liver 21 d 379

FLUORENE
Bluegill 30 d 200 to 1,800

NAPHTHALENE
Clam, Rangia
cuneata 24 h 6
Sandworm, Neanthes
arenaceodenta 3 to 24 h 40
Sandworm 24 h + 300 h not
post- detectable
treatment
Atlantic salmon, egg 168 h 44 to 83
Cladoceran, Daphnia
pulex 24 h 131
Crustaceans, 3 spp. 72 h 195 to 404
Bluegill, whole 24 h 310
Reference
Neff 1979
Miller et al. 1982
Gerhart and
Carlson 1978
Finger et al. 1985
Neff 1979
Neff 1982a
Kuhnhold and
Busch 1978
Neff 1985
Neff 1979
McCarthy and
Jimenez 1985
43 Table 6. (Concluded)
PAH compound, organism,
and other variables

DIMETHYLNAPHTHALENES
Crustaceans, 3 spp.

PERYLENE
Cladoceran, Daphnia
pulex

PHENANTHRENE
Clam, Rangia
cuneata
Cladoceran, Daphnia
pulex
PYRENE
Cladoceran Daphnia
pulex
Rainbow trout, liver
Exposure
period BCF
72 h 967 to
1,625
24 h 7,191
24 h 32
24 h 325
24 h 2,702
21 d 69
Reference
Neff 1979
Neff 1985
Neff 1979
Neff 1985
Gerhart and
Carlson 1978
m = minutes, h = hours, d = days.
'Each reference applies to the values in the same row and in the rows that follow for which no other reference is indicated. 44 arthropods usually had low Tb 1/2 values. In bluegill, 89% loss of benzo(a)pyrene was recorded 4 hours postexposure; for midge larvae it was 72% in 8 hours, and for daphnids it was 21% in 18 hours (Leversee et al. 1981).

The role of sediments in PAH uptake kinetics should not be discounted.

Sediment-associated anthracene contributed about 77% of the steady state body burden of this compound in the amphipod Hyalella azteca (Landrum and Scavia 1983),. For benzo(a)pyrene and the amphipod Pontoporeia hoyi, the sediment source (including interstitial water) accounted for 53% in amphipods collected at 60 m, but only 9% at 23 to 45 m (Landrum et al. 1984). Benthos from the Great Lakes, such as oligochaete worms (Limnodrilus sp., Stylodrilus sp.) and amphipods (Pontoporeia hoyi), obtain a substantial fraction of their PAH body content from the water when sediment PAH concentrations are low. However, when sediment PAH concentrations are elevated, benthos obtain a majority of their PAHs from that source through their ability to mobilize PAHs from the sediment/pore water matrix; the high concentrations of phenanthrene, fluorene, benzo(a)pyrene, and other PAHs measured in these organisms could provide a significant source of PAHs to predator fish (Eadie et al. 1983). Great Lakes benthos appear to contain as much PAHs as the fine grain fraction of the sediment which serves as their food, although overlying water or pore water appears to contribute a larger proportion of PAHs to the organism's body burden than does sediments (Eadie et al.
1984). Marine mussels (Mytilus
edulis) and polychaete annelid worms (Nereis virens) exposed for 28 days to sediments heavily contaminated with various PAH compounds accumulated significant concentrations (up to
1,000X control levels) during the first 14
days of exposure, and little thereafter; during a 5-week postexposure period, depuration was rapid, with the more water soluble PAHs excreted most rapidly; PAH levels usually remained above control values to the end of the postexposure period (Lake et al. 1985). English sole (Parophrys vetulus), during exposure for 11 to 51 days to PAH-contaminated sediments, showed significant accumulations of naphthalenes in liver (up to 3.1 mg/kg dry weight) after 11 days, with concentrations declining markedly thereafter; uptake of phenanthrene, chrysene, and benzo(a)pyrene was negligible during the first 7 days (Neff 1982a).

Fluorene effects in freshwater pond ecosystems have recently been evaluated (Boyle et al. 1984, 1985; Finger et al. 1985). In ponds exposed to initial fluorene concentrations of 0.12 to 2.0 mg/l, Tb 1/2 values in water ranged from 6 to 11 days. Ten weeks after fluorene introduction, little degradation had occurred in the organic bottom sediments; fluorene residues were present in fish, invertebrates, and rooted submerged macrophytes.

Studies with fingerling bluegills showed that

0.062 mg fluorene/1 adversely
affected their ability to capture chironomid prey, 0.12 mg/l reduced growth, and 1.0 mg fluorene/1 increased their vulnerability to predation by largemouth bass (Micropterus
salmoides). The authors concluded that fluorene, at
concentrations well below its solubility and at levels that could realistically occur, in the environment, represents a potential hazard to aquatic organisms.
45 Large interspecies differences in ability to absorb and assimilate PAHs from food have been reported. For example, crustaceans (Neff 1982a) and fish (Maccubbin et al. 1985; Mai ins et al. 1985a, 1985b) readily assimilated PAHs from contaminated food, whereas molluscs and polychaete annelids were limited (Neff 1982a). In all cases where assimilation of ingested PAHs was demonstrated, metabolism and excretion of PAHs were rapid (Neff 1982a). Thus, little potential exists for food chain biomagnification of PAHs (Southworth 1979; Dobroski and Epifanio 1980; Neff 1982a). . In laboratory aquatic ecosystem studies, Lu et al. (1977) found that benzo(a)pyrene can be accumulated to high, and potentially hazardous, levels in fish and invertebrates. In the case of mosquitofish (Gambusia affinis), almost all of the accumulated benzo(a)pyrene was from its diet, with negligible accumulations from the medium. However, mosquitofish degraded benzo(a)pyrene about as rapidly as it was absorbed, in contrast to organisms such as snails (Physa sp.) which retained most (88%) of the accumulated benzo(a)pyrene for at least 3 days postexposure, presumably due to deficiencies in their mixed- function oxidase detoxication system (Lu et al.
1977). Benzo(a)pyrene, when
administered to northern pike (Esox lucius) through the diet or the medium, followed similar pathways: entry via the gills or gastrointestinal system, metabolism in the liver, and excretion in the urine and bile (Balk et al. 1984). Benthic marine fishes exposed to naphthalene or benzo(a)pyrene, either in diet or through contaminated sediments, accumulated substantial concentrations in tissues and body fluids (Varanasi and Gmur 1981). The tendency of fish to metabolize PAHs extensively and rapidly may explain why benzo(a)pyrene, for example, is frequently undetected, or only detected in low concentrations in livers of fish from environments heavily contaminated with PAHs (Varanasi and Gmur 1980,
1981). Extensive metabolism of benzo(a)pyrene
plus the presence of large proportions of polyhydroxy metabolites in liver of English sole indicates the formation of reactive intermediates such as diol epoxides and phenol epoxides of benzo(a)pyrene, both of which are implicated in mammalian mutagenesis and carcinogenesis (Varanasi and Gmur 1981).

Cytotoxic, mutagenic, and carcinogenic effects of many PAHs are generally believed to be mediated through active epoxides formed by interaction with microsomal monooxygenases. These highly active arene oxides can interact with macromolecular tissue components and can further be metabolized or rearranged to phenols or various conjugates. They can also be affected by epoxide hydrolase to form dihydrodiols, which are precursors of biologically active diol epoxides--a group that has been implicated as ultimate carcinogens.

Investigators generally agree that marine and freshwater fishes are as well equipped as mammals with liver PAH-metabolizing enzymes; rapidly metabolize PAHs by liver mixed-function oxidases, with little evidence of accumulation; translocate conjugated PAH metabolites to the gall bladder prior to excretion in feces and urine; and have mixed-function oxidase degradation rates that are significantly modified by sex, age, diet, water temperature, dose-time relationships, and other variables. In addition, many species of fishes can convert
PAHs, benzo(a)pyrene for example, to potent mutagenic metabolites, but because detection of the 7,8-dihyclrodiol, 9,10-epoxide by analytical 46 methods is extremely difficult, most investigators must use biological assays, such as the Ames test, to detect mutagenic agents. At present, the interaction effects of PAHs with inorganic and other organic compounds are poorly understood. Specific examples of the above listed phenomena for PAH compounds and teleosts are documented for benzo(a)pyrene (Ahokas et al. 1975; Lu et al. 1977; Gerhart and Carlson 1978; Melius et al. 1980; Varanasi et al. 1980, 1984; Stegeman et al. 1982; Couch et al. 1983; Hendricks 1984; Melius 1984; Schoor 1984; Schoor and Srivastava 1984; Hendricks et al. 1985; Neff 1985; Fair 1985; von Hofe and Puffer 1985), 3-methylcholanthrene (Gerhart and Carlson 1978; Melius et al. 1980; Melius and Elam 1983; Schoor and Srivastava 1984; Neff 1985), benz(a)anthracene, chrysene, and pyrene (Gerhard and Carlson 1978), and 7,12-climethylbenz(a)anthracene (Stegeman et al. 1982).

Baumann et al. (1982) summarized reports on increasing frequencies of liver tumors in wild populations of fish during the past decade, especially in brown bullhead (Ictalurus nebulosus) from the Fox River, Illinois (12% tumor frequency), in Atlantic hagfish (Myxine qlutinosa) from Swedish estuaries (5%), in English sole from the Duwamish estuary, Washington (32%), and in tomcod (Microgadus tomcod) from the Hudson River, New York (25%). In all of these instances, significant levels of contaminants were present in the sediments, including PAHs. PAHs have been identified as genotoxic pollutants in sediments from the Black River, Ohio, where a high incidence of hepatoma and other tumors has been observed in ictalurid fishes (West et al. 1984, 1985). Reports of tumors in Great Lakes fish populations have been increasing. Tumors of thyroid, gonad, skin, and liver are reported, with tumor frequency greatest near areas contaminated by industrial effluents such as PAHs; liver tumors were common among brown bullhead populations at sites with large amounts of PAHs in sediments (Baumann 1984). A positive relationship was finally established between sediment PAH levels and prevalence of liver lesions in English sole in Puget Sound, Washington (Maiins et al. 1984; Varanasi et al. 1984), and sediment levels and liver tumor frequency in brown bullheads from the Black River, Ohio (Baumann and Harshbarger 1985; Black et al. 1985). Sediment PAH levels in the Black River, Ohio, from the vicinity of a coke plant outfall, were up to 10,000 times greater than those from a control location: concentrations were greater than 100 mg/kg for pyrene, fluoranthene, and phenanthrene; between 50 and 100 mg/kg for benz(a)anthracene, chrysene, and benzofluoranthenes; and between 10 and 50 mg/kg for individual naphthalenes, benzo(e)pyrene, benzo(a)pyrene, perylene, indeno(l,2,3-cd)pyrene, benzo(g,h,i)perylene, and anthanthrene (Baumann et al. 1982). Brown bullheads from this location contained >1.0 mg/kg of acenapthalene (2.4), phenanthrene (5.7), fluoranthene (1.9), and pyrene (1.1), and lower concentrations of heavier molecular weight PAHs; bullheads also exhibited a high (33%) liver tumor frequency, which seemed to correspond to their PAH body burdens. Investigators concluded that the elevated frequency of liver neoplasia in Black River bullheads was chemically induced, and was the result of exposure to PAHs (Baumann et al. 1982; Baumann and Harshbarger 1985).
47 Neoplasms in several species of fishes have been produced experimentally with 3-methylcholanthrene, acetylaminofluorene, benzo(a)pyrene, and 7,12 dimethylbenz(a)anthracene, with tumors evident 3 to 12 months postexposure (Couch and Harshbarger 1985; Hendricks et al. 1985). Under laboratory conditions, liver neoplasms were induced in two species of minnows (Poeciliopsis spp.) by repeated short-term exposures (5 hours once a week, for 5 weeks) to an aqueous suspension of 5 mg/l of 7,12-dimethyl benz(a)anthracene. About 44% of the fish surviving this treatment developed hepatocellular neoplasms 6 to 9 months postexposure (Schultz and Schultz 1982). Eastern mudminnows (Umbra pygmaea) kept in water containing up to 700 ug PAHs/1 for 11 days showed increased frequencies of chromosomal aberrations in gills: 30% vs. 8% in controls (Prein et al. 1978). High dietary benzo(a)pyrene levels of 500 mg/kg produced significant elevations in hepatic mixed-function oxidase levels in rainbow trout after 9 weeks (Hendricks et al. 1985). Rainbow trout fed diets containing 1,000 mg benzo(a)pyrene/kg for 12 months developed liver tumors (Couch et al. 1983). About 25% of rainbow trout kept on diets containing 1,000 mg benzo(a)pyrene/kg for 18 months had histologically confirmed liver neoplasms as compared to 15% after 12 months, with no evidence of neoplasia in controls (Hendricks et al. 1985). Young English sole may activate and degrade carcinogenic PAHs, such as benzo(a)pyrene, to a greater extent than adults, but additional research is needed to determine if younger fish are at greater risk than older sole to PAH-induced toxicity (Varanasi et al. 1984). In English sole, a high significant positive correlation between PAH metabolites (1-and 3-hydroxy benzo(a)pyrene, hydroxy and dihydrodiol metabolites of pyrene and fluoranthene) in bile, and idiopathic liver lesions, prevalance of neoplasms, megalocytic hepatosis, and total number of hepatic lesions (Krahn et al. 1986) suggests that selected PAH metabolites and key organs or tissues may be the most effective monitors of PAH contamination in aquatic organisms.

In addition to those effects of PAHs emphasizing survival, uptake, depuration, and carcinogenesis previously listed, a wide variety of additional effects have been documented for aquatic organisms. These include: inhibited reproduction of daphnids and delayed emergence of larval midges by fluorene (Finger et al. 1985); decreased respiration and heart rate in mussels (Mytilus californianus) by benzo(a)pyrene (Sabourin and Tullis 1981); increased weight of liver, kidney, gall bladder, and spleen of sea catfish (Arius felis) by 3-methylcholanthrene, which was dose-related (Melius and Elam 1983); photosynthetic inhibition of algae and macrophytes by anthracene, naphthalene, phenanthrene, pyrene (Neff 1985), and fluorene (Finger et al. 1985); immobilization of the protozoan, Paramecium caudatum, by anthracene, with an EC-50 (60 min) of 0.1 ug/1 (EPA 1980); perylene accumulation by algae (Stegeman 1981); accumulation without activation of benzo(a)pyrene and benzo(a)anthracene by a marine protozoan (Parauronema acutum), and biotransformation of various fluorenes by P. acutum to mutagenic metabolites (Lindmark 1981); interference by toluene and anthracene with benzo(a)pyrene uptake by freshwater amphipods (Landrum 1983); abnormal blood chemistry in oysters (Crassostrea virginica) exposed for one year to 5 ug 3-methyl 48 •nfr'"iMi
cholanthrene/1 (Couch et al. 1983); and enlarged livers in brown bullheads from a PAH-contaminated river (Fabacher and Baumann 1985).

AMPHIBIANS AND REPTILES
Limited data were available on biological effects of benzo(a)pyrene, 3-methylcholanthrene, and perylene to reptiles and amphibians (Balls 1964; Stegeman 1981; Anderson et al. 1982; Schwen and Mannering 1982a, 1982b; Couch et al.
1983).

Implantation of 1.5 mg of benzo(a)pyrene crystals into the abdominal cavity of adult South African clawed toads (Xenopus laevis) produced lymphosarcomas in 11 of the 13 toads (85%) after 86 to 288 days (Balls 1964).

Immature toads were more resistant, with only 45% bearing lymphoid tumors of liver, kidney, spleen, or abdominal muscle 272 to 310 days after implantation of 1.5 mg of benzo(a)pyrene crystals in the dorsal lymph sac or abdominal cavity. Implantation of 3-methylcholanthrene crystals into X. laevis provokes development of lymphoid tumors similar to those occurring naturally in this species; moreover, these tumors are readily transplantable into other Xenopus or into the urodele species Triturus cristatus (Balls 1964). Intraperitoneal injection of perylene into tiger salamanders can result in hepatic tumors (Couch et al. 1983).

A critical point of interaction between PAHs and reptiles/amphibians involves the transformation of these compounds by cytochrome P-450-dependent monooxygenase systems (Stegeman 1981); in general, reaction rates in this group are considerably slower than those observed in hepatic microsomes from mammals (Schwen and Mannering 1982a). Mixed-function oxidation systems can be induced in liver and skin of tiger salamanders by perylene (Couch et al. 1983) and 3-methylcholanthrene (Anderson et al. 1982), and in liver of the leopard frog (Rana pipiens) and garter snake (Thamnophis sp.) by benzo(a)pyrene and 3-methylcholanthrene (Stegeman 1981; Schwen and Mannering 1982a, 1982b). A single dose of 40 mg/kg body weight of 3-methylcholanthrene was sufficient to induce mixed-function oxidase activity for several weeks in the leopard frog (Schwen and Mannering 1982b). Amphibians, including tiger salamanders, are quite resistant to PAH carcinogenesis when compared to mammals, according to Anderson et al. (1982). This conclusion was based on studies with Ambystoma hepatic microsomes and their inability to produce mutagenic metabolites of benzo(a)pyrene and perylene (as measured by bacterial Salmonella typhimurium strains used in the Ames test); however, rat liver preparations did produce mutagenic metabolites under these procedures (Anderson et al. 1982). 49 BIRDS
Only two articles were available on PAHs and avian wildlife, and both concerned mallards (Anas platyrhynchos). In one study, Patton and Dieter (1980) fed mallards diets that contained 4,000 mg PAHs/kg (mostly as naphthalenes, naphthenes, and phenanthrene) for a period of 7 months. No mortality or visible signs of toxicity were evident during exposure; however, liver weight increased 25% and blood flow to liver increased 30%, when compared to controls. In the second study, Hoffman and Gay (1981) measured embryotoxicity of various PAHs applied externally, in a comparatively innocuous synthetic petroleum mixture, to the surface of mallard eggs. The most embryotoxic PAH tested was 7,12-dimethylbenz(a)anthracene: approximately

0.002 ug/egg (equivalent to about 0.036 ug/kg fresh weight, based on an average weight of 55 g per egg) caused 26% mortality in 18 days, and, among survivors, produced significant reduction in embryonic growth and a significant increase in the percent of anomalies, e.g., incomplete skeletal ossification, defects in eye, brain, liver, feathers, and bill. At 0.01 ug 7,12-dimethylbenz(a)anthracene/egg, only 10% survived to day 18. Similar results were obtained with 0.015 ug (and higher) chrysene/egg. For benzo(a)pyrene, 0.002 ug/egg did not affect mallard survival, but did cause embryonic growth reduction and an increased incidence of abnormal survivors.

At 0.01 ug benzo(a)pyrene/egg, 60% died in 18 days; at 0.05 ug/egg, 75% were dead within 3 days of treatment. Embryos may contain microsomal enzymes that can metabolize PAHs to more highly toxic intermediates than can adults, and avian embryos may have a greater capacity to metabolize PAHs in this manner than do mammalian embryos and fetuses (as quoted in Hoffman and Gay 1981); this observation warrants additional research. Several investigators have suggested that the presence of PAHs in petroleum, including benzo(a)pyrene, chrysene, and 7,12-dimethylbenz(a)anthracene, significantly enhances the overall embryotoxicity in avian species, and that the relatively small percent of the aromatic hydrocarbons contributed by PAHs in petroleum may confer much of the adverse biological effects reported after eggs have been exposed to .-microliter quantities of polluting oils (Hoffman and Gay 1981; Albers 1983).

MAMMALS
Numerous PAH compounds are distinct in their ability to produce tumors in skin and in most epithelial tissues of practically all animal species tested; malignancies were often induced by acute exposures to microgram quantities. In some cases, the latency period can be as short as 4 to 8 weeks, with the tumors resembling human carcinomas (EPA 1980). Certain carcinogenic PAHs are capable of passage across skin, lungs, and intestine, and can enter the rat fetus, for example, following intragastric or intravenous administration to pregnant dams (EPA 1980). In most cases, the process of carcinogenesis occurs 50 over a period of many months in experimental animals, and many years in man.

The tissue affected is determined by the route of administration and species under investigation. Thus, 7,12-dimethylbenz(a)anthracene is a potent carcinogen for the mammary gland of young female rats after oral or intravenous administration; dietary benzo(a)pyrene leads to leukemia, lung adenoma, and stomach tumors in mice; and both PAH compounds can induce hepatomas in skin of male mice when injected shortly after birth (Dipple 1985). Acute and chronic exposure to various carcinogenic PAHs have resulted in destruction of hematopoietic and lymphoid tissues, ovotoxicity, antispermatogonic effects, adrenal necrosis, changes in the intestinal and respiratory epithelia, and other effects (Table 7; EPA 1980; Lee and Grant 1981). For the most part, however, tissue damage occurs at dose levels that would also be expected to induce carcinomas, and thus the threat of malignancy predominates in evaluating PAH toxicity. There is a scarcity of data available on the toxicological properties of PAHs which are not demonstrably carcinogenic to mammals (EPA 1980; Lee and Grant 1981).

Target organs for PAH toxic action are diverse, due partly to extensive distribution in the body and also to selective attack by these chemicals on proliferating cells (EPA
1980). Damage to the hematopoietic and lymphoid
system in experimental animals is a particularly common observation (EPA 1980). In rats, the target organs for 7,12-dimethylbenz(a)anthracene are skin, small intestine, kidney, and mammary gland, whereas in fish the primary target organ is liver (Schultz and Schultz 1982). Application of carcinogenic PAHs to mouse skin leads to destruction of sebaceous glands and to hyperplasia, hyperkeratosis, and ulceration (EPA 1980). Tumors are induced in mouse skin by the repeated application of small doses of PAHs, by a single application of a large dose, or by the single application of a subcarcinogenic dose (initiation) followed by repeated application of certain noncarcinogenic agents (promotion) (Dipple 1985). Newborn mice were highly susceptible to 3-methylcholanthrene, with many mice dying from acute or chronic wasting disease following treatment; some strains of mice eventually developed thynomas, but other strains showed no evidence despite serious damage to the thymus (EPA 1980).

In general, PAH carcinogens transform cells through genetic injury involving metabolism of the parent compound to a reactive diol epoxide. This, in turn, can then form adducts with cellular molecules, such as DNA, RNA, and proteins, resulting in cell transformation (Dipple 1985; Ward et al. 1985).

In the case of benzo(a)pyrene, one isomer of the 7,8-diol, 9,10-epoxide is an exceptionally potent carcinogen to newborn mice and is believed to be the ultimate carcinogenic metabolite of this PAH (Slaga et al. 1978). One of the
most toxicologically significant processes involved in the response to PAH absorption is the interaction with drug metabolizing enzyme systems (Lee and Grant 1981). Increased production of mixed-function oxidase enzymes in various small mammals has been induced by halogenated napthalenes (Campbell et al. 1983), 3-methylcholanthrene (Miranda and Chhabra 1980), and numerous other PAHs (EPA 1980). PAH metabolites produced by microsomal enzymes in mammals 51 Table 7. Some effects of PAHs on selected laboratory animals Effect (units),
organism,
PAH compound
LD-50, ACUTE ORAL
(mg/kg body weight)
Rodents (Rattus spp., Mus spp.)
Benzo(a)pyrene
Phenanthrene
Naphthalene
Fluoranthene

CARCINOGENICITY, CHRONIC ORAL
(mg/kg body weight)
>dents
7,12-dimethylbenz(a)anthracene
Benzo(a)pyrene
Dibenz(a,h)anthracene
Benz(a)anthracene
Benzo(b)fluoranthene
Benzo(k)f1uoranthene
Indeno(1,2,3-cd)pyrene
Chrysene
Anthracene

CARCINOGENICITY, APPLIED EXTERNALLY
AS TOPICAL (mg)
Mice, Mui spp.

Benzo(a)pyrene
Dibenz(a,c)anthracene
7,12-dimethylbenz(a)anthracene
Dibenz(a,j)anthracene
Anthracene
Benzo(g,h,i)perylene
Benz(a)anthracene
CARCINOGENICITY, SUBCUTANEOUS (mg)
Mice
Dibenz(a,h)anthracene
Adults
Concentration
50 700
1,780
2,000

0.00004-0. 00025

0.002

0.006

2.0 40.0

72.0 72.0

99.0 3, 300.0

0.001

0.02 0.039

0.08 0.8

1.0 >0.0002
Reference
Sims and
Overcash 1978
Lo and Sandi 1978
Sims and
Overcash 1983
Lo and Sandi 1978
52 uaa
Table 7. (Continued)
Effect (units),
organism,
PAH compound Concentration Reference
Newborn >0.00008
Dibenzo(a,i)pyrene
In sesame oil 0.05
In peanut oil 0.6
Benzo(a)pyrene 0.06
Dibenzo(a,e)pyrene >0.6
Benzo(b)fluoranthene 1.8
Benz(a)anthracene 5.0
Dibenzo(a,h)pyrene 6.0
TESTICULAR DAMAGE (mg)
Rat, Rattus spp.

Benzo(a)pyrene, oral 100.0 (no effect) EPA 1980
7,12-dimethylbenz(a)anthracene
Intravenous
Young rats 0.5 - 2.0
Older rats 5.0
Oral 20.0

OOCYTE AND FOLLICLE DESTRUCTION,

SINGLE INTRAPERITONEAL INJECTION
(mg/kg body weight)
Mice
Benzo(a)pyrene 80.0 Mattison 1980
3-methylcholanthrene 80.0
7,12-dimethylbenz(a)anthracene 80.0

ALTERED BLOOD SERUM CHEMISTRY AND

NEPHROTOXICITY, SINGLE INTRAPERITONEAL
INJECTION (mg/kg body weight)
Rat
Phenanthrene 150.0 Yoshikawa
Pyrene 150.0 et al. 1985
53 Table 7. (Concluded)
Effect (units),
organism,
PAH compound Concentration Reference
FOOD CONSUMPTION, DAILY FOR 5
(mg/kg body weight)
DAYS
Deer mice, Peromvscus maniculatus
2-methoxynaphthalene
30% reduction
2-ethoxynaphthalene
3% reduction
House mice, Mui musculus
2-methoxynaphthalene
50% reduction
2-ethyoxynaphthalene
50% reduction
825
1,213
^
825
1,213
Schafer and
Bowles 1985
Each reference applies to the values in the same row, and in the rows that follow for which no other reference is indicated. 54 can be arbitrarily divided into water soluble groups, and organosoluble groups such as phenols, dihydrodiols, hydroxymethyl derivatives, quinones, and epoxides (EPA 1980). In the case of benzo(a)pyrene, the diol epoxides are usually considered as the ultimate carcinogens. Other microsomal enzymes convert epoxide metabolites to easily excretable water soluble compounds, with excretion primarily through feces and the hepatobiliary system (EPA 1980).

Interspecies differences in sensivity to PAH-induced carcinogenesis are due largely to differences in levels of mixed function oxidase activities, and these will directly affect rates at which active metabolites are converted to less active products (Neff 1979).

Investigators agree that unsubstituted aromatic PAHs with less than 4 condensed rings have not shown tumorgenic activity; that many, but not all, 4-, 5-, and 6-ring PAH compounds are carcinogenic; and that only a few unsubstituted hydrocarbons with 7 rings or greater are tumorogenic or carcinogenic (Neff 1979; EPA 1980; Dipple 1985). Many PAH compounds containing 4 and 5 rings, and some containing 6 or more rings, provoke local tumors after repeated application to the dorsal skin of mice; the tumor incidence exhibited a significant dose-reponse relationship (Grimmer et al. 1985). Among unsubstituted PAHs containing a nonaromatic ring, e.g., cholanthrene and acenapthanthracene, all active carcinogens retained an intact phenanthrene segment (EPA
1980). The addition of alkyl substitutents in
certain positions in the ring system of a fully aromatic PAH will often confer carcinogenic activity or dramatically enhance existing carcinogenic potency.

For example, monomethyl substitution of benz(a)anthracene can lead to strong carcinogenicity in mice, with potency depending on the position of substitution in the order 7 > 6 > 8 = 12 > 9; a further enhancement of carcinogenic activity is produced by appropriate dimethyl substitution, with 7,12-dimethylbenz(a)anthracene among the most potent PAH carcinogens known.

Alkyl substitution of partially aromatic condensed ring systems may also add considerable carcinogenic activity, as is the case with 3-methylcholanthrene.

With alkyl substitutes longer than meiihyl, carcinogenicity levels decrease, possibly due to a decrease in transport through cell membranes (EPA 1980).

A good correlation exists between skin tumor initiating activities of various benzo(a)pyrene metabolites and their mutagenic activity in mammalian cell mutagenesis systems (Slaga et al.
1978), although variations in
chromosome number and structure may accompany tumors induced by various carcinogenic PAHs in
rats, mice, and hamsters (Bayer 1978; EPA 1980). Active PAH metabolites, e.g., dihydrodiols or diol epoxides, can produce sister chromatid exchanges in Chinese hamster ovary cell (Bayer 1978; EPA 1980; Pal 1984). When exchanges were induced by the diol epoxide, a close relationship exists between the frequency of sister chromatid exchanges and the levels of deoxyribnucleoside-diol-epoxide adduct formation (Pal 1984). In general, noncarcinogenic PAHs were not mutagenic (EPA 1980).

Laboratory studies with mice have shown that many carcinogenic PAHs adversely affect the immune system, thus directly impacting an organism's 55 general health, although noncarcinogenic analogues had no immunosuppressive effect; further, the more carcinogenic the PAH, the greater the immuno suppression (Ward et al. 1985).

Destruction of oocytes and follicles in mice ovary is documented following intraperitoneal injection of benzo(a)pyrene; 3-methylcholanthrene, and 7,12-dimethylbenz(a)anthracene; the rate of destruction was proportional to the activity of the ovarian cytochrome P-450 dependent monooxygenase, as well as the carcinogenicity of the PAH (Mattison 1980). However, no information is presently available to indicate whether PAHs present a hazard to reproductive success. In those cases where teratogenic effects are clearly evident, e.g., 7, 12-dimethylbenz(a)anthracene, the required doses were far in excess of realistic environmental exposures (Lee and Grant 1981).

Numerous studies show that unsubstituted PAHs do not accumulate in mammalian adipose tissues despite their high lipid solubility, probably because they tend to be rapidly and extensively metabolized (EPA 1980; Lee and Grant 1981).

Biological half-life (Tb 1/2) of PAHs is limited, as judged by rodent studies. In the case of benzo(a)pyrene and rat blood and liver, Tb 1/2 values of 5 to 10 minutes were recorded; the initial rapid elimination phase was followed by a slower disappearance phase lasting 6 hours or more (EPA 1980).

Tb 1/2 values from the site of subcutaneous injection in mice were 1.75 weeks for benzo(a)pyrene, 3.5 weeks for 3-methylcholanthrene, and 12 weeks for dibenz(a,h)anthracene; the relative carcinogenicity of each compound was directly proportional to the time of retention at the injection site (Pucknat 1981).

Many chemicals are known to modify the action of carcinogenic PAHs in experimental animals, including other PAHs that are weakly carcinogenic or noncarcinogenic. The effects of these modifiers on PAH metabolism appear to fall into three major categories: those which alter the metabolism of the carcinogen, causing decreased activation or increased detoxification; those which scavenge active molecular species of carcinogens to prevent their reaching critical target sites in the cell; and those which exhibit competitive antagonism (DiGiovanni and Slaga 1981b). For example, benz(a)anthracene, a weak carcinogen, when applied simultaneously with dibenz(a,h)anthracene, inhibited the carcinogenic action of the latter in mouse skin; a similar case is made for benzo(e)pyrene or dibenz(a,c)anthracene applied to mouse skin shortly prior to initiation with 7,12 dimethylbenz(a)anthracene, or 3-methylcholanthrene (DiGiovanni and Slaga 1981a). Benzo(a)pyrene, a known carcinogen, interacts synergistically with cyclopenta(cd)pyrene, a moderately strong carcinogen found in automobile exhausts, according to results of mouse skin carcinogenicity studies (Rogan et al. 1983). Other PAH combinations were cocarcinogenic, such as benzo(e)pyrene, pyrene, and fluoranthene applied repeatedly with benzo(a)pyrene to the skins of mice (DiGiovanni and Slaga 1981a). Effective 56 inhibitors of PAH-induced tumor development include selenium, vitamin E, ascorbic acid, butylated hydroxytoluene, and hydroxyanisole (EPA 1980). In addition, protective effects against PAH-induced tumor formation have been reported for various naturally occurring compounds such as flavones, retenoids, and vitamin A (EPA 1980). Until these interaction effects are clarified, the results of single substance laboratory studies may be extremely difficult to apply to field situations of suspected PAH contamination.

Additional work is also needed on PAH dose-response relationships, testing relevant environmental PAHs for carcinogenicity, and elucidating effects of PAH mixtures on tumor formation (Grimmer
1983).
57 RECOMMENDATIONS
At present, no criteria or standards have been promulgated for PAHs by any regulatory agency for the protection of sensitive species of aquatic organisms or wildlife. This observation is not unexpected in view of several factors: (1) the paucity of data on PAH background concentrations in wildlife and other natural resources; (2) the absence of information on results of chronic oral feeding studies of PAH mixtures and the lack of a representative PAH mixture for test purposes; and (3) the demonstrable--and as yet, poorly understood—effects of biological modifiers, such as sex, age, and diet, and interaction effects of PAHs with inorganic and other organic compounds, including other PAHs.

Nevertheless, the growing data base for aquatic life indicates a number of generalizations: (1) many PAHs are acutely toxic at concentrations between 50 and 1,000 ug/1; (2) deleterious sublethal responses are sometimes observed at concentrations in the range of 0.1 to 5.0 ug/1; (3) uptake can be substantial, but depuration is usually rapid except in some species of invertebrates; and (4) whole body burdens in excess of 300 ug benzo(a)pyrene/kg (and presumably other PAHs) in certain teleosts would be accompanied by a rise in the activity of detoxifying enzymes.

Current aquatic research has focused on PAHs because of their known relationship with carcinogenesis and mutagenesis. Many reports exist of high incidences of cancer-like growths and developmental anomalies in natural populations of aquatic animals and plants, but none conclusively demonstrate the induction of cancer by exposure of aquatic animals to environmentally realistic levels of carcinogenic PAHs in the water column, diet, or sediments (Neff 1982b,
1985). However, recent studies by Baumann, Mai ins, Black, Varanasi and their coworkers, among others, have now established that sediments heavily contaminated with PAHs from industrial sources were the direct cause of elevated PAH body burdens and elevated frequencies of liver neoplasia in fishes from these locales. At present, only a few sites containing high PAH concentrations in sediments have been identified (Couch and Harshbarger
1985), suggesting an urgent need to identify and to evaluate other PAH-contaminated aquatic sites. Most fishery products consumed by upper trophic levels, including man, contain PAH concentrations similar to those in green vegetables and smoked and charcoal-broiled meats, and would probably represent a minor source of PAH toxicity; however, consumption of aquatic 58 organisms, especially filter-feeding bivalve molluscs, from regions severely contaminated with petroleum or PAH-containing industrial wastes, should be avoided (Jackim and Lake 1978; Neff 1982b). Neff (1982b) suggested that repeated consumption of PAH-contaminated shellfish may pose a cancer risk to humans. If true, this needs to be evaluated using seabirds, pinnipeds, and other wildlife groups which feed extensively on molluscs that are capable of accumulating high burdens of carcinogenic PAHs, in order to determine if similar risks exist.

For avian wildlife, data are missing on PAH background concentrations and on acute and chronic toxicity; these data should be collected posthaste.

Studies with mallard embryos and PAHs applied to the egg surface showed toxic and adverse sublethal effects at concentrations between

0.035 and 0.18 ug
PAH/kg whole egg (Hoffman and Gay 1981). Additional research is needed on petroleum-derived PAHs and their effects on developing embryos of seabirds and other waterfowl.

PAH criteria for human health protection (Table 8) were derived from tests with small laboratory mammals, primarily rodents. Accordingly, these proposed criteria should become interim guidelines for protection of nonhuman mammalian resources pending acquisition of more definitive data. The proposed PAH criteria are controversial. Pucknat (1981) states that there is no way at present to quantify the potential human health risks incurred by the' interaction of any PAH with other PAHs or with other agents in the environment, including tumor initiators, promotors, and inhibitors. The problem arises primarily from the diversity of test'systems and bioassay conditions used for determining carcinogenic potential of individual PAHs in experimental animals, and is confounded by the lack of a representative PAH mixture for test purposes, the absence of data for animal and human chronic oral, exposures to PAH mixtures, and the reliance on data derived from studies with benzo(a)pyrene to produce generalizations concerning environmental effects of PAHs--generalizations which may not be scientifically sound--according to Pucknat (1981). EPA (1980) emphasizes that only a small percentage of PAH compounds are known to be carcinogenic, and that measurements of total PAHs (i.e., the sum of all multiple fused-ring hydrocarbons having no heteroatoms) can not be equated with carcinogenic potential; furthermore, when the term "total
PAHs" is used, the compounds
being considered should be specified for each case. Lee and Grant (1981) state that an analysis of dose-response relationships for PAH-induced tumors in animals shows, in some cases, deviation from linearity in dose-response curves, especially at low doses, suggesting a two-stage model consistent with a linear nonthreshold pattern. Because overt tumor induction follows a dose-response relationship consistent with a multihit promotion process, the multihit component of carcinogenesis may be supplied by environmental stimuli not necessarily linked or related to PAH exposure.

The well-documented existence of carcinogenic and anticarcinogenic agents strongly suggests that a time assessment of carcinogenic risk for a particular 59 Table 8. Proposed PAH criteria for human health protection (modified from EPA 1980; Lee and Grant 1981; Pucknat 1981).

Criterion, PAH group, and units

DRINKING WATER
Total PAHs
ug/r
Daily intake, ug
Yearly intake, ug
Benzo(a)pyrene
ug/1
Daily intake, ug
Carcinogenic PAHs
ug/1 1
Daily intake, ug
ug/r
-5
Cancer risk 10
-6
Cancer risk 10
-7
Cancer risk 10
c
Daily intake, ug
-5
Cancer risk 10
-6
Cancer risk 10
-7
Cancer risk 10
FOOD
Total PAHs .

Daily intake, ug
Yearly intake, ug
Benzo(a)pyrene
Daily intake, ug
AIR
Total PAHs
ug/m r,
Daily intake, ug
Cyclohexane extractable fractions
Coke oven emissions, coal tar
products, ug/m , 8 to 10
hour-weighted average
Concentration

0.0135-0.2

0.027-0.4

4.0 0.00055
o.oon

0.0021

0.0042

0.028

0.0028

0.00028

0.056

0.0056

0.00056

1.5-16.0
4,150.0

0.16-1.6

0.0109

0.164-0.251
100.0-150.0
60 Table 8. (Concluded)
Criterion, PAH group, and units Concentration
Benzene soluble fractions
Coal tar pitch volatiles,
ug/m , 8-hour, time-weighted
average
Benzo(a)pyrene
ug/m f
Daily intake, ug
Carcinogenic PAHs
ug/m"^ r
Daily intake, ug
'
'
200.0

0.0005

0.005-0.0115

0.002

0.03-0.046

ALL SOURCES
Total PAHs
Daily, ug
Benzo(a)pyrene
Daily intake, ug
Daily allowable limit, ug^
Carcinogenic PAHs /except diet)
Daily intake, ug

1.79-16.6

0.166-1.61

0.048

0.086-0.102
^Total of 6 PAHs: fluoranthene, benzo(a)pyrene,
benzo(g,h,i)perylene, benzo(b)fluoranthene, benzo(k)fluoranthene, and indeno(l,2,3-cd)pyrene.

Total of 3 PAHs: benzo(a)pyrene, benzo(j)fluoranthene, and indeno(l,2,3-cd)pyrene.
^Based on all carcinogenic PAHs,
d
Assuming 1,600 g food daily, 70 kg adult, 1 to 10 ug total PAHs/diet.
^As above, except 0.1 to 1.0 ug benzo(a)pyrene/diet.
f 3
Assuming average of 15 to 23 m of air inhaled daily.
^From Wang and Meresz (1982).
51 PAH can be evaluated only through a multifactorial analysis (Lee and Grant 1981). One of the most toxicologically significant processes involved in the response to PAH absorption is the interaction with drug metabolizing enzyme systems. The induction of this enzyme activity in various body tissues by PAHs and other xenobiotics is probably critical to the generation of reactive PAH metabolites at the target site for tumor induction. At present, wide variations occur in human and animal carcinogen-metabolizing capacity.

Moreover, it has not yet been possible to definitely correlate enzyme activity with susceptibility to carcinogenesis. The obligatory coupling of metabolic activation with PAH-induced neoplasia in animals indicates that the-modulation of drug metabolizing enzymes is central to carcinogenesis (Lee and Grant 1981).

PAHs from drinking water contribute only a small proportion of the average total human intake (Harrison et al. 1975). The drinking water quality criterion for carcinogenic PAH compounds is based on the assumption that each compound is as potent as benzo(a)pyrene, and that the carcinogenic effect of the compounds is proportional to the sum of their concentrations (EPA 1980).

Based on an oral feeding study of benzo(a)pyrene in mice, the concentration of this compound estimated to result in additiopal risk of one additional case for every 100,000 individuals exposed (i.e, 10 ) is

0.028 ug/1. Therefore,
with this assumption, the sum of the concentrations of all carcinogenic PAH compounds shou]d be less than 0.028 ug/1 in order to keep the lifetime cancer risk below 10' (EPA 1980). The corresponding recommended criteria which may result in an incremental cancer risk of 10' and 10' over the lifetime are

0.0028 and 0.00028 ug/1, respectively (Table 8). If the above estimatesare made for consumption of aquatic organisms, only, the levels are 0.311 (10" ),

0.031(10'°), and 0.003(10'^) ug/kg, respectively (EPA 1980). The use of contaminated water for irrigation can also spread PAHs into other vegetable foodstuffs (EPA 1980). When vegetables grown in a PAH-polluted area are thoroughly washed and peeled, their contribution to total PAH intake in humans is not significant (Wang and Meresz 1982). Herbivorous wildlife, however, may ingest significant quantities of various PAHs from contaminated vegetables- but no data were available on this subject.

PAHs are widely distributed in the environment as evidenced by their detection in sediments, soils, air, surface waters, and plant and animal tissues. However, the ecological impact of PAHs is uncertain. PAHs show little tendency for bloconcentratlon, despite their high lipid solubility (Pucknat 1981), probably because most PAHs are rapidly metabolized. Sims and Overcash (1983) list a variety of research needs regarding PAHs in soil-plant systems. Specifically, research is needed to establish: the rates of PAH decomposition in soils; the soil PAH levels above which PAH constituents adversely affects the food chain; and enhancement factors that increase degradation rates of PAHs, especially PAHs with more than 3 rings. Once these factors have been determined, PAH disposal into soils may become feasible at environmentally nonhazardous levels.
62 Diet is the major source of PAHs to humans. Authorities agree that most foods contain 1 to 10 ug total PAHs/kg fresh weight, that smoking or barbecuing fish and meats increases total PAH content up to lOOX, that contaminated molluscs and crustaceans may contribute significantly to PAH intake, and that PAH carcinogenic risk to humans has existed at least since man began to cook his food (Barnett 1976; EPA 1980; Lee and Grant 1981; Lawrence and Weber 1984a). A total of 22 PAHs has been identified in foods, of which 11 have been found to be carcinogenic in experimental animals. Of these, only 5 (benzo(a)pyrene, benz(a)anthracene, 3-methylcholanthrene, dibenz(a,h)anthracene, and 7,12-dimethylbenz(a)anthracene) have been demonstrated to Induce tumors following oral administration to rats and mice, and only 3 of the 11 exhibited positive dose-response relationships in chronic studies with mice (Lo and Sandi 1978). At the present time, there is no evidence that any of the 11 known carcinogenic PAHs or their combinations can cause cancer in human beings via the oral route, especially in quantities likely to be present in foods (Lo and Sandi 1978).

In view of the carcinogenic characteristics of many PAH compounds, their increasing concentrations in the environment should be considered alarming, and efforts should be made to reduce or eliminate them wherever possible (Suess 1976).
63 LITERATURE CITED
Ahokas, J.T., 0. Pelkonen, and N.T. Karki. 1975. Metabolism of polycyclic hydrocarbons by a highly active aryl hydrocarbon hydroxylase system in the liver of a trout species. Biochem. Biophys. Res. Comm. 63:635-641.

Albers, P.H. 1983. Effects of oil on avian reproduction: a review and discussion, Pages 78-96 vn The effects of oil on birds. A multidiscipline symposium. Tri-State Bird Rescue and Research, Inc., Wilmington, Delaware.

Anderson, R.S., J.E. Doos, and F.L. Rose. 1982. Differential ability of Ambystoma tigrinum hepatic microsomes to produce mutagenic metabolites from polycyclic aromatic hydrocarbons and aromatic amines. Cancer Lett. 16:33-41.

Ang, K.P., H. Gunasingham, and B.T, Tay. 1986. The distribution of polynuclear aromatic hydrocarbons in ambient air particulates in Singapore. Environ. Monitor. Assess. 6:171-180.

Baird, W.M., R.A. Chemerys, L. Diamond, T.H. Meedel, and J.R. Whittaker. 1982. Metabolism of benzo(a)pyrene by Ciona intestinal is. Pages 191-200 Ul N.L. Richards and B.L. Jackson (eds.). Symposium: carcinogenic polynuclear aromatic hydrocarbons in the marine environment. U.S. Environ.

Protection Agency Rep. 600/9-82-013.

Balk, L., J. Meijer, J.W. DePierre, and L.E. Appelgren. 1984. The uptake and distribution of ( H) benzo(a)pyrene in the northern pike (Esox lucius). Examination by whole-body autoradiography and scintillation counting. Toxicol. Appl. Pharmacol. 74:430-449.

Balls, M.
laevis.
1964. Benzpyrene-induced
Experientia 20:143-145.
tumours in the clawed toad, Xenopus
Barnett, D. 1976, Polycyclic aromatic hydrocarbons and foods. CSIRO Food Res. Quart. 36: 8-12.

Bauer, J.E., and D.G. Capone. 1985. Degradation and mineralization of the polycylic aromatic hydrocarbons anthracene and naphthalene in intertidal marine sediments. Appl. Environ. Microbiol. 50:81-90.
64 Baumann, P.C. 1984. Cancer in wild freshwater fish populations with emphasis on the Great Lakes. J. Great Lakes Res. 10:251-253.

Baumann, P.C, and J.C. Harshbarger. 1985. Frequencies of liver neoplasia in a feral fish population and associated carcinogens. Mar. Environ. Res. • 17:324-327.

Baumann, P.C, and J.J. Lech. In press. PAH's, metabolites, and neoplasia in feral fish populations, ui U. Varanasi (ed.). PAH metabolism in the aquatic environment. CRC Press, Boca Raton, Florida.

Baumann, P.C, W.D. Smith, and W.K. Parland. In press. Contaminant concentrations and tumor frequencies in brown bullhead from an industrialized river and a recreational lake. Trans. Am. Fish. Soc.

Baumann, P.C, W.D. Smith, and M. Ribick. 1982. Hepatic tumor rates and polynuclear aromatic hydrocarbon levels in two populations of brown bullhead (Ictalurus nebulosus). Pages 93-102 in M.W. Cooke, A.J. Dennis, and G.L. Fisher (eds.). Polynuclear aromatic hydrocarbons: physical and biological chemistry. Battelle Press, Columbus, Ohio.

Bayer, U. 1978. In^ vivo induction of sister chromatid exchanges by three polyaromatic hydrocarbons. Pages 423-428 ui P.W. Jones and R.I. Freudenthal (eds.). Carcinogenesis - a comprehensive survey. Vol. 3. Polynuclear aromatic hydrocarbons: second international symposium on analysis, chemistry, and biology. Raven Press, New York.

Bjorseth, A., and A.J. Dennis (eds.). 1980. Polynuclear aromatic hydrocarbons: chemistry and biological effects. Battelle Press, Columbus, Ohio. 1,097 pp.

Black, J., H. Fox, P. Black, and F. Bock. 1985. Carcinogenic effects of river sediment extracts in fish and mice. Pages 415-427 jji R.L. Jolley, R.J. Bull, W.P. Davis, S. Katz,- M.H. Roberts, Jr., and V.A. Jacobs (eds.). Water chlorination. Volume 5. Lewis Publ., Chelsea, Michigan.

Black, J.J. 1983. Epidermal hyperplasia and neoplasia in brown bullheads (Ictalurus nebulosus) in response to repeated applications of a PAH containing extract of polluted river sediment. Pages 99-111 ui M. Cooke and A.J. Dennis (eds.). Polynuclear aromatic hydrocarbons: formation, metabolism and measurement. Battelle Press, Columbus, Ohio.

Black, J.J., T.F. Hart, Jr., and E. Evans. 1981. HPLC studies of PAH pollution in a Michigan trout stream. Pages 343-355 in M. Cooke and A.J.

Dennis
(eds.). Chemical analysis and biological fate: polynuclear aromatic hydrocarbons. Fifth international symposium. Battelle Press, Columbus, Ohio.
65 Boehm, P.D., and J.W. Farrington. 1984. Aspects of the polycyclic aromatic hydrocarbon geochemistry of recent sediments in the Georges Bank region.

Environ. Sci. Technol. 18:840-845.

Boyle, T.P., S.E. Finger, R.L. Paulson, and CF. Rabeni. 1985. Comparison of laboratory and field assessment of fluorene - Part II: effects on the ecological structure and function of experimental pond ecosystems. Pages 134-151 in T.P. Boyle (ed.). Validation and predictability of laboratory methods for assessing the fate and effects of contaminants in aquatic ecosystems. ASTM STP 855. American Society for Testing and Materials, Philadelphia, Pennsylvania.

Boyle, T.P., S.E. Finger, J.D. Petty, L.M. Smith, and J.N. Huckins. 1984.

Distrubution and rate of disappearance of fluorene in pond ecosystems.

Chemosphere 13:997-1008.

Brown, R.A., andR.J. Pancirov. 1979. Polynuclear aromatic hydrocarbons in Baltimore Canyon fish. Environ. Sci. Technol. 13:878-879.

Campbell, M.A., S. Bandiera, L. Robertson, A. Parkinson, and S. Safe. 1983.

Hepta-, hexa-, tetra- and dichloronaphthalene congeners as inducers of hepatic microsomal drug-metabolizing enzymes. Toxicology 26:193-205.

Cavalieri, E., R. Roth, E. Rogan, C Grandjean, and J. Althoff. 1978.

Mechanisms of tumor initiation by polycyclic aromatic hydrocarbons. Pages 273-284 in P.W. Jones and R.I. Freudenthal (eds.). Carcinogenesis - a comprehensive survey. Vol. 3. Polynuclear aromatic hydrocarbons: second international symposium on analysis, chemistry, and biology. Raven Press, New York.

Cavalieri, E., D. Sinha, and E. Rogan. 1980. Rat mammary gland versus mouse skin: different mechanisms of activation of aromatic hydrocarbons. Pages 215-231 in A. Bjorseth and A.J. Dennis (eds.). Polynuclear aromatic hydrocarbons: chemistry and biological effects. Battelle Press, Columbus, Ohio.

Cerniglia, C.E., G.L. White, and R.H. Heflich. 1985. Fungal metabolism and detoxification of polycyclic aromatic hydrocarbons. Arch. Microbiol. 143:105-110.

Cooke, M., and A.J. Dennis (eds.). 1981, Chemical analysis and biological fate: polynuclear aromatic hydrocarbons, Fifth international symposium.

Battelle Press, Columbus, Ohio. 770 pp.

Cooke, M., and A.J. Dennis (eds.).. 1983. Polynuclear aromatic hydrocarbons; formation, metabolism and measurement. Battelle Press, Columbus, Ohio, 1,301 pp.
66 T'
Cooke, M., and A.J. Dennis (eds.). 1984. Polynuclear aromatic hydrocarbons:
mechanisms, methods and metabolism. Battelle Press, Columbus, Ohio. 1,464 pp.

Cooke, M., A.J. Dennis, and G.L. Fisher (eds.). 1982. Polynuclear aromatic hydrocarbons: physical and biological chemistry. Battelle Press, Columbus, Ohio. 947 pp.

Couch, J.A., and J.C Harshbarger. 1985. Effects of carcinogenic agents on aquatic animals: an environmental and experimental overview. J. Environ, Sci. Health, Part C, Environ. Carcin. Rev. 3:63-105.

Couch, J.A., W.P. Schoor, W. Davis, and L. Courtney. 1983. Effects of carcinogens, mutagens, and teratogens on nonhuman species (aquatic animals). U.S. Environ. Protection Agency Rep. 600/9-83-005. 46 pp.

DiGiovanni, J., and T.J. Slaga. 1981a. Effects of benzo(e)pyrene [B(e)P] and dibenz(a,c)anthracene [DB(a,c)A] on the skin tumor-initiating activity of polycyclic aromatic hydrocarbons. Pages 17-31 in M. Cooke and A.J. Dennis (eds.). Chemical analysis and biological fate: polynuclear aromatic hydrocarbons. Fifth international symposium. Battelle Press, Columbus, Ohio.

DiGiovanni, J., and T.J. Slaga. 1981b. Modification of polycyclic aromatic hydrocarbon carcinogensis. Pages 259-292 in H.V. Gelboin and P.O. Ts'o (eds.). Polycyclic hydrocarbons and cancer. Vol. 3. Academic Press, New York.

Dipple, A. 1985. Polycyclic aromatic hydrocarbon carcinogensis: an introduction. Pages 1-17 in R.D. Harvey (ed.). Polycyclic hydrocarbons and carcinogensis. ACS Symp. Ser. 283. Amer. Chem. Soc, Washington, D.C.

Dobroski, C.J., Jr., and CE. Epifanio. 1980. Accumulation of benzo(a)pyrene in a larval bivalve via trophic transfer. Canad. J. Fish. Aquatic Sci. 37:2318-2322.

Dunn, B.P. 1980. Polycyclic aromatic hydrocarbons in marine sediments, bivalves, and seaweeds: analysis by high-pressure liquid chromatography.

Pages 367-377 in A. Bjorseth and A.J. Dennis (eds.). Polynuclear aromatic hydrocarbons: chemistry and biological effects. Battelle Press, Columbus, Ohio.

Eadie, B.J., W.R. Faust, P.F. Landrum, and N.R. Morehead. 1984. Factors affecting bloconcentratlon of PAH by the dominant benthic organisms of the Great Lakes. Pages 363-377 in M. Cooke and A.J. Dennis (eds.).

Polynuclear aromatic hydrocarbons: mechanisms, methods and metabolism. 57 Battelle Press, Columbus, Ohio.

Eadie, B.J., W.R. Faust, P.F. Landrum, N.R. Morehead, W.S. Gardner, and T.

Nalepa. 1983. Bioconcentrations of PAH by some benthic organisms of the Great Lakes. Pages 437-449 in M. Cooke and A.J. Dennis (eds.).

Polynuclear aromatic hydrocarbons: formation, metabolism and measurement.

Battelle Press, Columbus, Ohio.

Edwards, N.T. 1983. Polycyclic aromatic hydrocarbons (PAH's) in the terrestrial environment - a review. J. Environ. Qual. 12:427-441.

Eiceman, G.A., B. Davani, and J.A. Dodson. 1984. Discharge water from hydrostatic testing of natural gas pipelines as source of PAH loading into aquatic environments. Int. J. Environ. Anal. Chem. 19:27-39.

Eiceman, G.A., B. Davani, M.E. Wilcox, J.L. Gardea, and J.A. Dodson. 1985.

High molecular weight hydrocarbons including polycyclic aromatic hydrocarbons in natural gas from consumer distribution pipelines and in pipeline residue. Environ. Sci. Technol. 19:603-508.

EPA. 1980. Ambient water quality criteria for polynuclear aromatic hydrocarbons. U.S. Environ. Protection Agency. Rep. 440/5-80-069. 193 pp.

Fabacher, D.L., and P.C. Baumann. 1985. Enlarged livers and hepatic microsomal mixed-function oxidase components in tumor-bearing brown bullheads from a chemically contaminated river. Environ. Toxicol. Chem. 4:703-710.

Fair, P.H. 1986. Interaction of benzo(a) pyrene and cadmium on GSH-S-trsnsferase and benzo(a)pyrene hydroxylase in the black sea bass Centropristis striata. Arch. Environ. Contam. Toxicol. 15:257-253.

Finger, S.E., E.F. Little, M.G. Henry, J.F. Fairchild, and T.P. Boyle. 1985.

Comparison of laboratory and field assessment of fluorene - Part I:
effects of fluorene on the survival, growth, reproduction, and behavior of aquatic organisms in laboratory
tests. Pages 120-133 in T.P. Boyle (ed.).

Validation and predictability of laboratory methods for assessing the fate and effects of contaminants in aquatic ecosystems. ASTM STP 855. American Society for Testing and Materials, Philadelphia, Pennsylvania.

Fox, F.R., and K.R. Rao. 1982. Accumulation, tissue distribution, and depuration of benzo(a)pyrene and benz(a)anthracene in the grass shrimp, Palaemonetes pugio. Pages 336-349 in N.L. Richards and B.L. Jackson (eds.). Symposium: carcinogenic polynuclear aromatic hydrocarbons in the marine environment. U.S. Environ. Protection Agency Rep. 600/9-82-013.

Futoma, D.J., S.R. Smith, T.E. Smith, and J. Tanaka. 1981. Polycyclic 68 iT
aromatic hydrocarbons in water systems. CRC Press, Boca Raton, Florida. 190 pp.

Gelboin, H.V., and P.O. Ts'o (eds.). 1978a. Polycyclic hydrocarbons and cancer. Vol. 1. Environment, chemistry, and metabolism. Academic Press, New York. 408 pp.

Gelboin, H.V., and P.O. Ts'o (eds.). 1978b. Polycyclic hydrocarbons and cancer. Vol 2. Molecular and cell biology. Academic Press, New York. 452 pp.

Gelboin, H.V., and P.O. Ts'o (eds.). 1981. Polycyclic hydrocarbons and cancer. Vol 3. Academic Press, New York. 351 pp.

Gerhart, E.H., and R.M. Carlson. 1978. Hepatic mixed-function oxidase activity in rainbow trout exposed to several polycyclic aromatic compounds.

Environ. Res. 17:284-295.

Greenberg, A., F. Darack, R. Harkov, P. Lioy, and J. Daisey. 1985. Polycyclic aromatic hydrocarbons in New Jersey: a comparison of winter and summer concentrations over a two-year period. Atmospher. Environ. 19:1325-1339.

Grimmer, G. (ed.). 1983. Environmental carcinogens: polycyclic aromatic hydrocarbons. Chemistry,- occurrence, biochemistry, carcinogenicity. CRC Press, Boca Raton, Florida. 261 pp.

Grimmer, G., H. Brune, R. Deutsch-Wenzel, G. Dettbarn, J. Misfeld, U. Able, and J. Timm. 1985. The contribution of polycyclic aromatic hydrocarbon fractions with different boiling ranges to the carcinogenic impact of emission condensate from coal fired residential furnaces as evaluated by topical application to the skin of mice. Cancer Lett. 28:203-211.

Harrison, R.M., R. Perry, and R.A. Wellings. 1975. Polynuclear aromatic hydrocarbons in raw, potable and waste waters. Water Res. 9:331-346.

Harvey, R.G. (ed.). 1985. Polycyclic hydrocarbons and carcinogensis. ACS Symp. Ser. 283. Amer. Chem. Soc, Washington, D.C. 406 pp.

Heit, M. 1985. The relationship of a coal fired power plant to the levels of polycyclic aromatic hydrocarbons (PAH) in the sediment of Cayuga Lake.

Water, Air, Soil Pollut. 24:41-61.

Hendricks, J.D. 1984. Use of small fish species in carcinogenicity testing.

Pages 397-404 in K.L. Hoover (ed.). National Cancer Institute Monograph 65, Washington, D.C.

Hendricks, J.D., T.R. Meyers, D.W. Shelton, J.L. Casteel, and G.S. Bailey. 1985. Hepatocarcinogenicity of benzo(a)pyrene to rainbow trout by dietary 69 exposure and intraperitoneal injection. J. Natl Cancer Inst. 74:839-851.

Herd, J.E., and F.E. Greene. 1980. Effects of perinatal exposure to benzo(a)pyrene on the aryl hydrocarbon hydroxylase system of adult rat liver. Biol. Neonate 38:291-299.

Herrmann, R., and D. Hubner. 1984. Concentrations of micropollutants (PAH, chlorinated hydrocarbons and trace metals) in the moss Hypnum cupressiforme in and around a small industrial town in southern Finland. Ann. Bot.

Fennici 21:337-342.

Hites, R.A., and P.M. Gschwend. 1982. The ultimate fates of polycyclic aromatic hydrocarbons in marine and lacustrine sediments. Pages 357-365 in M. Cooke, A.J. Dennis, and F.L. Fisher (eds.). Polynuclear aromatic hydrocarbons: physical and biological chemistry. Battelle Press, Columbus, Ohio.
loffman, D.J., and M.L. Gay. 1981. , Embryotoxic effects of benzo(a)pyrene, chrysene, and 7,12-dimethylbenz(a)anthracene in petroleum hydrocarbon mixtures in mallard ducks. J. Toxicol. Environ. Health 7:775-787.

Hoffman, E.J., G.L. Mills, J.S. Latimer, and J.G. Qiiinn, 1984. Urban runoff as a source of polycyclic aromatic hydrocarbons to coastal waters.

Environ. Sci. Technol. 18:580-587.

Hose, J.E., J.B. Hannah, D. DiJulio, M.L. Landolt, B.S. Miller, W.T. Iwaoka, and S.P. Felton 1982. Effects of benzo(a)pyrene on early development of flatfish. Arch . Environ. Contam. Toxicol. 11:167-171.

Humason, A.W., and D.F. Gadbois. 1982. Determination of polynuclear aromatic hydrocarbons in the New York Bight area. Mimeo MS6-506. Natl. Mar. Fish.

Serv., Gloucester, Massachusetts.

Jackim, E., and C Lake. 1978. Polynuclear aromatic hydrocarbons in estuarine and nearshore environments. Pages 415-428 in M.L. Wiley (ed.).

Estuarine interactions. Academic Press, New York.

Jackim, E., and L. Wilson. 1979. Benzo(a)pyrene accumulation and depuration in the soft-shell clam (Mya arenaria). Pages 91-94 ui D.S. Wilt (ed.).

Proceedings tenth National Shellfish S.anitation workshop. Hunt Valley, Maryland.

Johnson, A.C, P.F. Larsen, D.F. Gadbois, and A.W. Humason. 1985. The distribution of polycyclic aromatic hydrocarbons in the surficial sediments of Penobscot Bay (Maine, USA) in relation to possible sources and to other sites worldwide. Mar. Environ. Res. 5:1-15.
70 Johnston, W.R., and R.M. Harrison. 1984. Deposition of metallic and organic pollutants alongside the M5 motorway. Sci. Total Environ. 33:119-127.

Jones, P.W., and R.I. Freudenthal (eds.). 1978. Carcinogensis - a comprehensive survey. Vol. 3. Polynuclear aromatic hydrocarbons: second international symposium on analysis, chemistry, and biology, Raven Press, New York. 487 pp.

Jones, P.W., and P. Leber (eds.). 1979. Polynuclear aromatic hydrocarbons.

Third international symposium on chemistry and biology - carcinogenesis and mutagenesis. Ann Arbor Science Publ., Ann Arbor, Michigan. 892 pp.

Kirso, U., L. Belykh, D. Stom, N. Irha, and E. Urbas. 1983. Oxidation of benzo(a)pyrene by plant enzymes. Pages 679-687 in M. Cooke and A.J. Dennis (eds.). Polynuclear aromatic hydrocarbons: formation, metabolism and measurement. Battelle Press, Columbus, Ohio.

Krahn, M.M., L.D. Rhodes, M.S. Myers, L.K. Moore, W.D. MacLeod, Jr., and D.C Mai ins. 1986. Associations between metabolites of aromatic compounds in bile and the occurrence of hepatic lesions in English sole (Parophrys vetulus) from Puget Sound, Washington. Arch. Environ. Contam. Toxicol. 15:51-67
Kuhnhold, W.W., and F. Busch. 1978. On the uptake of three different types of hydrocarbons by salmon eggs (Salmo salar L.). Meeresforsch. 26:50-59.

Lake, J., G.L. Hoffman, and S.C Schimmel. 1985. Bloaccumulation of contaminants from Black Rock Harbor dredged material by mussels and polychaetes. U.S. Environ. Protection Agency Tech. Rep. D-85-2. 150 pp.

Avail, from Natl. Tech. Inform. Serv., Springfield, Virginia.

Lake, J.L., C Norwood, C Dimock, and R. Bowen. 1979. Origins of polycyclic aromatic hydrocarbons in estuarine sediments. Geochim. Cosmochim. Acta 43:1847-1854.

Landrum, P.F. 1983. The effect of co-contaminants on the bioavailability of polycyclic aromatic hydrocarbons to Pontoporeia hoyi. Pages 731-743 in M.

Cooke and A.J. Dennis (eds.). Polynuclear aromatic hydrocarbons:
formation, metabolism and measurement. Battelle Press, Columbus, Ohio.

Landrum, P.F., B.J. Eadie, W.R. Faust, N.R. Morehead, and M.J. McCormick. 1984. Role of sediment in the bloaccumulation of benzo(a)pyrene by the amphipod, Pontoporeia hoyi. Pages 799-812 in M. Cooke and A.J. Dennis (eds.). Polynuclear aromatic hydrocarbons: mechanisms, methods and metabolism. Battelle Press, Columbus, Ohio.

Landrum, P.F., and D. Scavia. 1983. Influence of sediment on anthracene uptake, depuration, and biotransformation by the amphipod Hyalella azteca. 71 Canad. J. Fish. Aquatic Sci. 40:298-305.

Larsson, B., and G. Sahlberg. 1982. Polycyclic aromatic hydrocarbons in lettuce. Influence of a highway and an aluminium smelter. Pages 417-426 in M. Cooke, A.J. Dennis, and G.L. Fisher (eds.). Polynuclear aromatic hydrocarbons: physical and biological chemistry. Battelle Press, Columbus, Ohio.

Lawrence J.F., and D.F. Weber. 1984a. Determination of polycyclic aromatic hydrocarbons in some Canadian commercial fish, shellfish, and meat products by liquid chromatography with confirmation by capillary gas chromatography-mass spectrometry. J. Agric. Food Chem. 32:789-794.

Lawrence, J.F., and D.F. Weber. 1984b. Determination of polycyclic aromatic hydrocarbons in Canadian samples of processed vegetable and dairy products by liquid chromatography with fluorescence detection. J. Agric. Food Chem. 32:794-797.

Lee, R.F., R. Sauerheber, and G.D. Dobbs. 1972. Uptake, metabolism and discharge of polycycllc aromatic hydrocarbons by marine fish. Mar. Biol. 17:201-208.

Lee, S.D., and L. Grant (eds.). 1981. Health and ecological assessment of polynuclear aromatic hydrocarbons. Pathotex Publ., Park Forest South, Illinois. 364 pp.

Leversee, 6.J., J.P. Geisy, P.F. Landrum, S. Bartell, S. Gerould, M. Bruno, A.

Spacie, J. Bowling, J. Haddock, and T. Fannin. 1981. Disposition of benzo(a)pyrene In aquatic systems components: periphyton, chironomids, daphnia, fish. Pages 357-366 in M. Cooke and A.J. Dennis (eds.). Chemical analysis and biological fate: polynuclear aromatic hydrocarbons. Fifth international symposium. Battelle Press, Columbus, Ohio.

Linder, G., H.L. Bergman, and J.S. Meyer. 1985. Anthracene bloconcentratlon in rainbow trout during single-compound and complex-mixture exposures.

Environ. Toxicol. Chem. 4:549-558.

Lindmark, D.G. 1981. Activation of polynuclear aromatic hydrocarbons to mutagens by the marine ciliate Parauronema acutum. Appl. Environ.

Microbiol. 41:1238-1242.

Lo, M-T., and E. Sandi. 1978. Polycyclic aromatic
(polynuclears) in foods. Residue Rev. 69:35-86.
hydrocarbons
Lu, P-Y., R.L. Metcalf, N. Plummer, and D. Mandrel. 1977. The environmental fate of three carcinogens: benzo-(a)-pyrene, benzidine, and vinyl chloride evaluated in laboratory model ecosystems. Arch. Environ. Contam. Toxicol. 6:129-142.
72 Maccubbin, A.E., P. Black, L. Trzeciak, and J.J. Black. 1985. Evidence for polynuclear aromatic hydrocarbons in the diet of bottom-feeding fish.

Bull. Environ. Contam. Toxicol. 34:876-882.

Mai ins, D.C, M.M. Krahn, D.W. Brown, L.D. Rhodes, M.S. Myers, B.B. McCain, and S.-L. Chan. 1985a. Toxic chemicals in marine sediment and biota from Mukilteo, Washington: relationships with hepatic neoplasms and other hepatic lesions in English sole (Parophrys vetulus). J. Natl. Cancer Inst. 74:487-494.

Malins, D.C, M.M. Krahn, M.S. Myers, L.D. Rhodes, D.W. Brown, CA. Krone, B.B. McCain, and S.-L. Chan. 1985b. Toxic chemicals in sediments and biota from a creosote-polluted harbor: relationships with hepatic neoplasms and other hepatic lesions in English sole (Parophrys vetulus).

Carcinogenesis 6:1463-1469.

Malins, D.C, B.B. McCain, D.W. Brown, S.-L. Chan, M.S. Myers, J.T. Landahl, P.G. Prohaska, A.J. Friedman, L.D. Rhodes, D.G. Burrows, W.D. Gronlund, and H.O. Hodgins. 1984. Chemical pollutants in sediments and diseases of bottom-dwelling fish in Puget Sound, Washington. Environ. Sci. Technol. 18:705-713.

Marcus, J.M., and T.P. Stokes. 1985. Polynuclear aromatic hydrocarbons in oyster tissue around three coastal marinas. Bull. Environ. Contam.

Toxicol. 35:835-844.

Martens, R. 1982. Concentrations and microbial mineralization of four to six ring polycyclic aromatic
Chemosphere 11:761-770.
hydrocarbons in composted municipal waste.

Mattison, D.R
polycycllc
53:249-259.
1980.
aromatic
Morphology
hydrocarbons
of
in
oocyte
mice.
and folli
Toxicol.
cle destruction by
Appl. Pharmacol.

McCarthy, J.F., and B.D. Jimenez. 1985. Reduction in bioavailability to bluegills of polycycllc aromatic hydrocarbons bound to dissolved humic material. Environ. Toxicol. Chem. 4:511-521.

McCarthy, J.F., B.D. Jimenez, and T. Barbee. 1985. Effect of dissolved humic material on accumulation of polycycllc aromatic hydrocarbons:
structure-activity relationships. Aquatic Toxicol. 7:15-24.

McGinnes, P.R., and V.L. Snoeyink. 1974. Determination of the fate of polynuclear aromatic hydrocarbons in natural water systems. Univ. Illinois at Urbana-Champaign, Water Res. Center, UILU-WRC--74-0080, Res. Rep. No.

80. 56 pp.
73 Mclntyre, A.E., R. Perry, and J.N. Lester. 1981. Analysis of polynuclear aromatic hydrocarbons in sewage sludges. Anal. Lett. 14(A4):291-309. .'icMahon, C.K., and S.N. Tsoukalas. 1978. Polynuclear aromatic hydrocarbons in forest fire smoke. Pages 61-73 ui P.W. Jones and R.I. Freudenthal (eds.). Carcinogenesis - a comprehensive survey. Vol. 3. Polynuclear aromatic hydrocarbons: second international symposium on analysis, chemistry, and biology. Raven Press, New York.

Melius, P. 1984. Comparative benzo(a)pyrene metabolite patterns in fish and rodents. Pages 387-390 ui K.L. Hoover (ed.).
carcinogenicity testing. Natl. Cancer In
Maryland.

Use of small
st. Mono. 65.
fish in
Bethesda,
Melius, P., and D.L. Elam. 1983. Mixed function oxidase in sea catfish.

Pages 877-895 in M. Cooke and A.J. Dennis (eds.). Polynuclear aromatic hydrocarbons: formation, metabolism and measurement. Battelle Press, Columbus, Ohio.

Melius, P., D. Elam, M. Kilgore, B. Tan, and W.P. Schoor. 1980. Mixed function oxidase inducibility and polyaromatic hydrocarbon metabolism in the mullet, sea catfish, and Gulf killifish. Pages 1059-1075 in A.

Bjorseth and A.J. Dennis (eds.). Polynuclear aromatic hydrocarbons:
chemistry and biological effects. Battelle Press, Columbus, Ohio.

Miller, D.L., J.P. Corliss, R.N. Farragut, and H.C Thompson, Jr. 1982. Some aspects of the uptake and elimination of the polynuclear aromatic hydrocarbon chrysene by mangrove snapper, Lutjanus griseus, and pink shrimp, Penaeus duorarum. Pages 321-335 in N.L. Richards and B.L. Jackson (eds.). Symposium: carcinogenic polynuclear aromatic hydrocarbons in the marine environment. U.S. Environ. Protection Agency Rep. 500/9-82-013.

Miranda, CL., and R.S. Chhabra. 1980. Species differences in stimulation of intestinal and hepatic microsomal mixed-function oxidase enzymes. Biochem.

Pharmacol. 29:1151-1165.

Mix, M.C. 1982. Polynuclear aromatic hydrocarbons and cellular proliferative disorders in bivalve molluscs from Oregon estuaries. U.S. Environ.

Protection Agency Rep. 500/4-82-025. 49 pp.

Mix, M.C, and R.L. Schaffer. 1979. Benzo(a)pyrene concentrations in mussels (Mytilus edulis) from Yaquina Bay, Oregon during June 1975 - May 1978.

Bull. Environ. Contam. Toxicol. 23:677-584.

Mix, M.C, and R.L. Schaffer. 1983a. Concentrations of unsubstituted polycyclic aromatic hydrocarbons in softshell clams from Coos Bay, Oregon, USA. Mar. Pollut. Bull. 14:94-97.
74 Mix, M.C, and R.L. Schaffer. 1983b. Concentrations of unsubstituted polynuclear aromatic hydrocarbons in bay mussels (Mytilus edulis) from Oregon, USA. Mar. Environ. Res. 9:193-209.

Murphy, D.J., R.M. Buchan, and D.G. Fox. 1982. Ambient particulate and benzo(a)pyrene concentrations from residential wood combustion, in a mountain community. Pages 567-574 in M. Cooke, A.J. Dennis, and G.L.

Fisher (eds.). Polynuclear aromatic hydrocarbons: physical and biological chemistry. Battelle Press, Columbus, Ohio.

Neff, J.M.
environment.
1979. Polycyclic aromatic hydrocarbons
Applied Science Publ. Ltd., London. 262
in
pp.
the aquatic
Neff, J.M. 1982a. Accumulation and release of polycyclic aromatic hydrocarbons from water, food, and sediment by marine animals. Pages 282-320 in N.L. Richards and B.L. Jackson (eds.). Symposium: carcinogenic polynuclear aromatic hydrocarbons in the marine environment. U.S. Environ.

Protection Agency Rep. 600/9-82-013.

Neff, J.M. 1982b. Polycyclic aromatic hydrocarbons in the aquatic environment and cancer risk to aquatic organisms and man. Pages 385-409 JH N.L.

Richards and B.L. Jackson (eds.). Symposium: carcinogenic polynuclear aromatic hydrocarbons in the marine environment. U.S. Environ. Protection Agency Rep. 600/9-82-013.

Neff, J.M. 1985. Polycyclic aromatic hydrocarbons. Pages 416-454 in G.M.

Rand and S.R. Petrocelli (eds.). Fundamentals of aquatic toxicology.

Hemisphere Publ. Corp., New York.

Neff, J.M., B.A. Cox, D. Dixit, and J.W. Anderson. 1976. Accumulation and release of petroleum derived aromatic hydrocarbons by, four species of marine animals. Mar. Biol. 38:279-289.

Pal, K. 1984. The relationship between the levels of DNA-hydrocarbon adducts and the formation of sister-chromatid exchanges in Chinese hamster ovary cells treated with derivitives of polycyclic aromatic hydrocarbons. Mutat.

Res. 129:365-372.

Pasquini, R., and S. Monarca. 1983. Detection of mutagenic/carcinogenic compounds in unused and used motor oils. Sci. Total Environ. 32:55-54.

Patton, J.F., and M.P. Dieter, 1980. Effects of petroleum hydrocarbons on hepatic function in the duck. Comp. Biochem. Physiol. 650:33-36. 75 Potvin, R.R., E.G. Adamek, and D. Balsillie. 1981. Ambient PAH levels near a steel mill in northern Ontario. Pages 741-753 in M. Cooke and A.J. Dennis (eds.). Chemical analysis and biological fate: polynuclear aromatic hydrocarbons. Fifth international symposium. Battelle Press, Columbus, Ohio.

Prahl, F.G., E. Crecellus, and R. Carpenter. 1984. Polycyclic aromatic hydrocarbons in Washington coastal sediments: an evaluation of atmospheric and riverine routes of introduction. Environ. Sci. Technol. 18:687-693.

Prein, A.E., G, M. Thie, G.M. Alink, and J.H. Koeman. 1978, Cytogenetic changes in fish exposed to water of the River Rhine. Sci. Total Environ. 9:287-291.

Pucknat, A.W. (ed.). 1981. Health impacts of polynuclear aromatic hydrocarbons. Environmental Health Review No. 5. Noyes Data Corp., Park Ridge, New Jersey. 271 pp.

Quaghebeur, D, E. DeWulf, M.C. Ravelingien, and G. Janssens. 1983. Polycyclic aromatic hydrocarbons in rainwater. Sci. Total Environ. 32:35-54.

Richards, N.L., and B.L. Jackson (eds.). 1982. Symposium: carcinogenic polynuclear aromatic . hydrocarbons in the marine environment. U.S.

Environ. Protection Agency Rep. 600/9-82-013. 409 pp.

Rogan, E., E. Cavalieri, A. Munhall, and S. Salmasi. 1983. Synergistic effect of the environmental contaminants cyclopenta(cd)pyrene and benzo(,a)pyrene in mouse skin carcinogenicity. Pages 1035-1046 in M. Cooke and A.J. Dennis (eds.). Polynuclear aromatic hydrocarbons: formation, metabolism and measurement. Battelle Press, Columbus, Ohio.

Rose, F.L. 1977. Tissue lesions of tiger salamander (Ambystoma tigrinum):
relationship to sewage effluents. Ann. N.Y. Acad. Sci. 298:270-279.

Sabourin, T.D., and R.E. Tullis. 1981. Effect of three aromatic hydrocarbons on respiration and heart rates of the mussel, Mytilus californianus. Bull.

Environ. Contam. Toxicol. 25:729-736.

Schafer, E.W., Jr., and W.A. Bowles, Jr. 1985. Acute oral toxicity and repellency of 933 chemicals to house and deer mice. Arch. Environ.

Contam. Toxicol. 14:111-129,
Schoeny, R., T. Cody, M. Radike, J. Caruso, J. Rogers, B. Smiddy, M.

Trentman, and D. Warshawsky. 1984. Metabolism, growth effects and mutagenicity of plant metabolites of benzo(a)pyrene. Pages 1197-1211 in M.

Cooke and A.J. Dennis (eds.). Polynuclear aromatic hydrocarbons:
mechanisms, methods and metabolism. Battelle Press, Columbus, Ohio. 76 Schoor, W.P. 1984. Benzo(a)pyrene metabolism in marine fish and some analytical aspects of its metabolites. Pages 391-396 in K.L. Hoover (ed.). Use of small fish in carcinogenicity testing. Natl. Cancer Inst.

Mono. 65. Bethesda, Maryland.

Schoor, W.P., and M. Srivastava. 1984. Position-specific induction of benzo(a)pyrene metabolism by 3-methylcholanthrene and phenobarbital in mullet (Mugil cephalus), a marine fish. Comp. Biochem. Physiol. 730:391-396.

Schultz, M.E., and J.R. Schultz. 1982. Induction of hepatic tumors with 7,12-dimethylbenz(a)anthracene in two species of viviparous fishes (genus Poeciliopsis). Environ. Res. 27:337-351.

Schwen, R.J., and G.J. Mannering. 1982a. Hepatic cytochrome P-450-dependent monooxygenase systems of the trout, frog and snake - II, Monooxygenase activities. Comp. Biochem. Physiol. 718:437-443.

Schwen, R.J., and G.J. Mannering. 1982b. Hepatic cytochrome P-450-dependent monooxygenase systems of the trout, frog and snake - III. Induction. Comp.

Biochem. Physiol. 718:445-453.

Shiraishi, H., N.H. Pilkington, A. Otuski, and K,. Fuwa. 1985. Occurrence of chlorinated polynuclear aromatic hydrocarbons in tap water. Environ. Sci.

Technol. 19:585-590.

Sims, P., and P.L. Grover. 1981. Involvement of dihydrodiols and diol epoxides in the metabolic activation of polycyclic hydrocarbons other than benzo(a)pyrene. Pages 117-181 in H.V. Gelboin and P.O. Ts'o (eds.).

Polycyclic hydrocarbons and cancer. Vol. 3. Academic Press, New York.

Sims, R.C, and R. Overcash. 1983. Fate of polynuclear aromatic compounds (PNAs) in soil-plant systems. Residue Rev. 88:1-68
Sirota, G.R., and J.F. Uthe. 1981. Polynuclear aromatic hydrocarbon contamination in marine shellfish. Pages 329-341 in M. Cooke and A.J.

Dennis (eds.). Chemical analysis and biological fate: polynuclear aromatic hydrocarbons. Fifth international symposium. Battelle Press, Columbus, Ohio.

Sirota, G.R., J.F. Uthe, A. Sreedharan, R. Matheson, C.J. Musial, and K.

Hamilton. 1983. Polynuclear aromatic hydrocarbons in lobster (Homarus americanus) and sediments in the vicinity of a coking facility. Pages 1123-1135 in M. Cooke and A.J. Dennis (eds.). Polynuclear aromatic hydrocarbons: formation, metabolism and measurement. Battelle Press, Columbus, Ohio.

Slaga, T.J., W.M. Bracken, A. Viaje, D.L. Berry, S.M. Fischer, D.R. Miller, W. 77 Levin, A.H. Conney, H. Yagi, and D.M. Jerina. 1978. Tumor initiating and promoting activities of various benzo(a)pyrene metabolites in mouse skin.

Pages 371-382 in P.W. Jones and R.I. Freudenthal (eds.). Carcinogenesis a comprehensive survey. Vol. 3. Polynuclear aromatic hydrocarbons:
second international symposium on analysis, chemistry, and biology. Raven Press, New York.

Smith, J.D., J. Bagg, and B.M. Bycroft. 1984. Polycyclic aromatic hydrocarbons in the clam Tridacna maxima from the Great Barrier Reef, Australia. Environ. Sci. Technol 18:353-358.

Smith, J.D., J.Y; Hauser, and J. Bagg. 1985. Polycyclic aromatic hydrocarbons in sediments of the Great Barrier Reef region, Australia.

Mar. Pollut. Bull. 16:110-114.

Solbakken, J.E., K. Ingebrigtsen, and K.H. Palmork. 1984. Comparative study on the fate of the polychlorinated biphenyl 2, 4, 5, 2', 4', 5' - hexa chlorobiphenyl and the polycyclic aromatic hydrocarbon phenanthrene in flounder (Platichthys flesus), determined by liquid scintillation counting and autoradiography. Mar. Biol. 83:239-246.

Southworth, G.R. 1979. Transport and transformation of anthracene in natural waters. Pages 359-380 in L.L Marking and R.A. Kimerle (eds.). Aquatic toxicology. ASTM STP 667. American Society for Testing and Materials, Philadelphia, Pennsylvania.

Southworth, G.R., J.J. Beauchamp, and P.K. Schmeider. 1978. Bioaaccumulation potential of polycyclic aromatic hydrocarbons in Daphnia pulex. Water Res. 12:973-977.

Srivastava, V.K., P.K. Srivastava, and U.K. Misra. 1985. Polycyclic aromatic hydrocarbons of coal fly ash: analysis by gas-liquid chromatography using nematic liquid crystals. J. Toxicol. Environ. Health 15:333-337.

Statham, CN., M.J. Melancon, Jr., and J.J. Lech. 1976. Bloconcentratlon of xenobiotics in trout bile: a proposed monitoring aid for some waterborne chemicals. Science 193:680-681.

Stegeman, J. 1981. Polynuclear aromatic hydrocarbons and their metabolism in the marine environment. Pages 1-60 in H.V. Gelboin and P.O. Ts'o (eds.).

Polycyclic hydrocarbons and cancer. Vol. 3. Academic Press, New York.

Stegeman, J.J., T.R. Skopek, and W.G. Thilly. 1982. Bioactivation of polynuclear aromatic hydrocarbons to cytotoxic and mutagenic products by marine fish. Pages 201-211 in N.L. Richards and B.L. Jackson (eds.).

Symposium: carcinogenic polynuclear aromatic hydrocarbons in the marine environment. U.S. Environ. Protection Agency Rep. 600/9-82-013. 78 stoker, P.W., M.R. Plasterer, R.L. McDowell, R. Campbell, S. Fields, R. Price, C Muehle, W.R. West, G.M. Booth, J.R. Larsen, and M.L. Lee. 1984.

Effects of coal-derived tars on selected aquatic and mammalian organisms.

Pages 1239-1257 in M. Cooke and A.J. Dennis (eds.). Polynuclear aromatic hydrocarbons: mechanisms, methods and metabolism. Battelle Press, Columbus, Ohio.

Strand, J.W., and A.W. Andren. 1980. Polyaromatic hydrocarbons in aerosols over Lake Michigan, fluxes to the Lake. Pages 127-137 in A. Bjorseth and A.J. Dennis (eds.). Polynuclear aromatic hydrocarbons: chemistry and biological effects. Battelle Press, Columbus, Ohio.

Suess, M.J. 1976. The environmental load and cycle of polycyclic aromatic hydrocarbons. Sci. Total Environ. 6:239-250.

Sugimura, T- 1985. Studies on environmental chemical carcinogenesis in Japan. Science 233:312-318.

Sullivan, T.J., and M.C. Mix. 1985. Persistence and fate of polynuclear aromatic hydrocarbons deposited on slash burn sites in the Cascade Mountains and coast range of Oregon. Arch. Environ. Contam. Toxicol. 14:187-192.

Szentpaly, L.V. 1984. Carcinogenesis by polycyclic aromatic hydrocarbons: a multilinear regression on new type PMO indices. J. Am. Chem. Soc. 105:5021-6028.

Thomas, W., A. Ruhling, and H. Simon. 1984. Accumulation of airborne pollutants (PAH, chlorinated hydrocarbons, heavy metals) in various plant species and humus. Environ. Pollut. 35A:295-310.

Tsang, W-S., and G.W. Griffin. 1979. Metabolic activation of polynuclear aromatic hydrocarbons. Pergamon Press, Oxford, England. 125 pp.

Valerio, F., P. Bottino, D. Ugolini, M.R. Cimberle, G.A. Tozzi, and A.

Frigerio. 1984. Chemical and photochemical degradation of polycyclic aromatic hydrocarbons in the atmosphere. Sci. Total Environ. 40:169-188. van Noort, P.CM., and E. Wondergem. 1985. Scavenging of airborne polycyclic aromatic hydrocarbons by rain. Environ. Sci. Technol. 19:1044-1048.

Varanasi, U., and D.J. Gmur. 1980. Metabolic activation and covalent binding of benzo(a)pyrene to deoxyribonucleic acid catalyzed by liver enzymes of marine fish. Biochem. Pharmacol. 29:753-762.
79 Varanasi, U., and D.J. Gmur. 1981. In vivo metabolism of naphthalene and benzo(a)pyrene by flatfish. Pages 367-376 in M. Cooke and A.J. Dennis (eds.). Chemical analysis and biological fate: polynuclear aromatic hydrocarbons. Fifth international symposium. Battelle Press, Columbus, Ohio.

Varanasi, U., D.J. Gmur, and M.M. Krahn. 1980. Metabolism and subsequent binding of benzo(a)pyrene to DNA in pleuronectid and salmonid fish. Pages 455-470 in A. Bjorseth and A.J. Dennis (eds.). Polynuclear aromatic hydrocarbons: chemistry and biological effects. Battelle Press, Columbus, Ohio.

Varanasi, U., M. Nishimoto, and J. Stover. 1984. Analyses of biliary conjugates and hepatic DNA binding in benzo(a)pyrene-exposed English sole.

Pages 1315-1328 in M. Cooke and A.J. Dennis (eds.). Polynuclear aromatic hydrocarbons: mechanisms, methods and metabolism. Battelle Press, Columbus, Ohio.

Varanasi, U., W.L. Reichert, J.E. Stein, D.W. Brown, and H.R. Sanborn. 1985.

Bioavailability and biotransformation of aromatic hydrocarbons in benthic organisms exposed to sediment from an urban estuary. Environ. Sci.

Technol. 19:836-841.
von Hofe, E., and H.W. Puffer. 1986. In vitro metabolism and in vivo binding of benzo(a)pyrene in the California killifish (Fundulus parvipinnis) and speckled sanddab (Citharicthys stigmaeous). Arch. Environ. Contam.

Toxicol. 15:251-255.

Wang, D.T., and 0. Meresz. 1982. Occurrence and potential uptake of polynuclear aromatic hydrocarbons of highway traffic origin by proximally grown food crops. Pages 885-896 in M. Cooke, A.J. Dennis, and G.L. Fisher (eds.). Polynuclear aromatic hydrocarbons: physical and biological chemistry. Battelle Press, Columbus, Ohio.

Ward, E.C, M.J. Murray, and J.H. Dean. 1985. Immunotoxicity of nonhalogenated polycyclic aromatic hydrocarbons. Pages 291-313 in J.H.

Dean, M.I. Luster, A.E. Munson, and H. Amos (eds.). Immunotoxicology and immunopharamacology. Raven Press, New York.

Warshawsky, D., T. Cody, M. Radike, B.A. Smiddy, and B. Nagel. 1983. Toxicity and metabolism of benzo(a)pyrene in the green alga Selenastrum capricornutum. Pages 1235-1245 in M. Cooke and A.J. Dennis (eds.).

Polynuclear aromatic hydrocarbons: formation, metabolism and measurement.

Battelle Press, Columbus, Ohio.

West, W.R., P.A. Smith, G.M. Booth, S.A. Wise, and M.L. Lee. 1986.

Determination of genotoxic polycyclic aromatic hydrocarbons in a sediment from the Black River (Ohio). Arch. Environ. Contam. Toxicol. 15:241-249. 80 ..if,:!„KV.-i^«-^^<=-.^-.7.,„^,^:^^-.^
West, W.R., P.A. Smith, P.W. Stoker, G.M. Booth, T. Smith-Oliver, B.E.

Butterworth, and M.L. Lee. 1984. Analysis and genotoxicity of a PAC-polluted river sediment. Pages 1395-1411 in M. Cooke and A.J. Dennis (eds.). Polynuclear aromatic hydrocarbons: mechanisms, methods and metabolism. Battelle Press, Columbus, Ohio.

Whitehouse, B. 1985. The effects of dissolved organic matter on the aqueous partitioning of polynuclear aromatic hydrocarbons. Estuar. Coastal Shelf Sci. 20:393 •402.

Williams, U.P., J.W. Kiceniuk, and J.R. Botta. 1985. Polycyclic aromatic hydrocarbon accumulation and sensory evaluation of lobsters (Homarus americanus) exposed to diesel oil at Arnold's Cove, Newfoundland. Canad.

Tech. Rep. Fish. Aquatic Sci. 1402. 13 pp.

Yan, L.S. 1985. Study of the carcinogenic mechanism for polycyclic aromatic hydrocarbons - extended bay region theory and its quantitative model.

Carcinogenesis 6:1-6.

Yoshikawa, T., L.P. Ruhr, W. Flory, D. Giamalva, D.F. Church, and W.A. Pryor. 1985. Toxicity of polycyclic aromatic hydrocarbons. I. Effect of phenanthrene, pyrene, and their ozonized products on blood chemistry in rats. Toxicol. Appl. Pharmacol. 79:218-226.
81 50?72-10'
I. REPORT NO. 1. R*clpl«nt'« Accamslon No. REPORT tXX:UMENTATION PAGE Biological Report 85 (1.11)

4. TItl* and Subtltl* 5. Raport Ditt
Polycyclic aromatic hydrocarbon hazards to fish, wildlife, and May 1987 invertebrates: a synoptic review

7. Author<t) e. P»r(onT>lng Ori»nti»lton Reot. No.

Ronald Eisler

9. Parformlnc Orssniiitlon Nam* and Addraia 10. Proiact/Taik/Work Unit No.

Patuxent Wildlife Research Center

11. Contract(C) or Grant(G) No. U.S. Fish and Wildlife Service Laurel, MD 20708 (C)
<G)

12. Sponsoring Organization Nama and Addraii 13. Typa of Report & Period Covered Division of Wildlife and Contaminant Research
Fish and Wildlife Service
U.S. Department of the Interior
Washington, n,r. ?n?4n
IS. Supplementary Notet

16. Abitract (Limit: 200 word*)
This report synthesizes technical literature on ecological and toxicological aspects of polycyclic aromatic hydrocarbons (PAH) in the environment, with special reference to fisheries and wildlife resources. Subtopics include:
chemical properties, sources, and fate; background concentrations in biological and nonbiological samples; toxic and sublethal effects of PAH to flora and fauna; and proposed criteria and research needs for the •protection of sensitive, nonhuman organisms.
1 17. Document Analy«l( a. D*«chptor«
Contaminants
Toxicity
Fishes
Wildlife
Natural resources
Invertebrates
b. ld*frtlft«r«/Op*n-Cnd«d Tarmt
Polycyclic aromatic
PAH
Carcinogenesis
hydrocarbons Mutagenesis
Teratogenesis
c. COSATI Field/Group
IS. Availability Statement
Release unlimited

19. Securtty Clati (This
Unclassified
2a S«:urity Cl»i« ahi i
Unclassified
Report)
Page)

21. 22.

No. of
81 Price
Paget 1
!
(S M ANS(-Z39.18 ) See /natrucfione on Aevene OPTIONAL FORM 272 (4-77) (Formerly NTIS-35)
Oepartment of Commerce
a U.S. GOVERNMENT PRINTING OFFICE: 1987—7 7 5 -2 17'6512 7

Long Beach City CouncilMarch 21, 2023

Long Beach City Council
March 21, 2023