Document vB8qvpdmpZLL7VMyZdqY7LB4Z

ORGANOCHLORINE RESIDUES IN BROWN PELICAN EGGS: RELATION TO REPRODUCTIVE SUCCESS Lawrence J. Blus US Bureau ofSport Fisheries and Wildlife, Patuxent Wildlife Research Center, Laurel, Maryland 208)0, USA Buhkctt S. Neely, Jr US Bureau of Sport Fisheries and Wildlife, Capa Ronxain Wational Wildlife Refuge, Awendaw, South Carolina 29429, USA Andre A. Belisle 4t Richard M. Prouty US Bureau of Sport Fisheries and Wildlife, Patuxent Wildlife Research Center, Laurel, Maryland 20810, USA ABSTRACT This study M'Of conducted to determine the influence of residues of organochlorine poflulents on reproductive success in the brown pelican. A sample egg was takenfrom each of 93 marked nests In the nesting colony in South Carolina. Periodic visits were made to determine the fate of marked nests. Each sample egg was analysedfor residues of organochlorine pollutants. Nest success in the brown pelican was related to residues of DDE and dieldrin in sample eggs. DDE seemed primarily responsible for nestfailure; however, deleterious effects of this pollutant on nest success were not satisfactorily separatedfrom those induced by dieldrin. Significant intercorrelation of ailfive organochlorine residues identified in the eggs complicated the relationship ofresidues to nest success. Maximum DDE residues in an eggfrom a successful nest were 2-37 pgjg and 8 48 pgjg in an egg from an unsuccessful nest. Comparable maximum residues for dieldrin in sample eggs were 0-54 pgjg (successful) and 0-99 Ugfg (unsuccessful). Nineteen of the 26 eggsfrom unsuccessful nests contained DDE residues that exceeded 2*37 pglg. and 14 of then 26 eggs contained dieldrin residues that exceeded 0-54 pgjg, Residues of DDD, DDT or PCBs in sample eggs were not significantly related to nest success. Reproduction success in the brown pelican colony war subnormal in the years ofstudy--1971 and 1972--but reproductive success was normal in those nests In which the sample egg containtd either 2-5 pgfg of DDE or 0-54 pg/g ofdieldrin. 81 Environ. Pollut, (7) 0974)-- Applied Science Publishers Ud, England, 1974 Primed in Great Britain 82 LAWRENCE ). ALUS el at. INTRODUCTION , The eggshell thinning phenomenon has been well documented in wild birds (Ratcliffe, 1967a; Hickey & Anderson, 1968; Ratcliffe, 1970; Anderson & Hickey, 1972). DDE has been shown to be the principal agent inducing eggshell thinning in the brown pelican {Pelecamts occidentalis) (Blus ef at., 1971, 1972a, 1972b); and captive birds of several species fed DDE have exhibited eggshell thinning (Heath et at., 1969; Wiemeyer & Porter, 1970; Longcore et at., 1971; McLane 8t Hall, 1972), Eggshell thinning of 15 to 20% over a period of years is generally considered to result in declining populations of birds (Anderson & Hickey, 1972). This is some what corroborated in short-term studies with captive birds fed DDE in that deleterious effects on reproductive success occur when eggshell thinning is below 18% (Heather at., 1969; Longcore et at., 1971). It is assumed that eggshell thinning in wild birds is related to lowered reproductive success and this assumption is supported by results obtained with experimental birds. However, the degree of response of wild birds (o a pollutant may be different from that of captive birds. The brown pelican seems much more susceptible to DDE-induced eggshell thinning than any species used experimentally. It is possible to calculate the level of DDE that induces 15% eggshell thinning (between 4 and 5jjg/g in the brown pelican (Blus el at., 1972a)), and extrapolation from experiments with captive birds suggests that this level would result in a declining population if maintained over a period of years. However, a more direct measure of the association of residues with reproductive success is desirable because eggshell thinning is only one of the factors that may influence productivity. MATERIALS AND METHODS We obtained a more direct measurement by comparing residues in sample eggs to nest success on Marsh Island, Cape Romain National Wildlife Refuge (South Carolina), in 197) and 1972. Nests were selected for sampling over as wide an area as possible at different times during the nesting season. Numbered aluminium stakes with attached flagging or plastic (ape were used to mark 93 nests; the stakes were driven into the ground near the nests. We visited the nesting colony on seven occasions in 1971 and on 16 occasions in 1972. Our visits were short--usually not much more than an hour in duration. Prolonged observations were never localised in a particular segment of the colony except on two visits in 1972 when a blind (hide) was employed. One egg was collected from each of the marked nests. The sample egg was collected from a complete clutch in most cases, but an egg was sometimes taken from an incomplete clutch. A nest was considered successful if one or more downy young (3 weeks of age) were produced and unsuccessful if HONS 082180 ORGA NOCHLORINE RESIDUES IN PELICAN EGGS 83 no downy young were produced. Young in two nests disappeared soon after hatching, and these were classified as unsuccessful. There were 50 nests whose fates were unknown; four which were visited loo infrequently to determine nest success, 35 where the nest markers could not be located and 11 where no additional eggs were produced after a single egg was collected. Six nests were flooded by saltwater; these were excluded from the analysis. A freshly laid egg was taken from each of 19 nests, and an incubated egg, the stage of incubation varying from several days to about 30 days (near hatching), was collected from each of 18 nests. Contents of individual eggs were analysed by electron capture gas chromato graphy; certain residues were confirmed by either thin layer chromatography or mass spectrophotometry. The residues found were p,p'-DDE, p,p'-DDT, p,p'-DDD, dieldrin and PCBs. The PCBs in these pelican eggs resembled Aroclor 1260 or those Aroclors that lie near 1260. Some of the eggs contained small quantities of mirex, heptachlor epoxide, and metabolites of chlordane. A full description of the residue analyses is presented by Blus ( at. (in press) and Mulhern er at. (1970, 1971). Residue values were transformed to logarithms (log10) for statistical computations. A two-way analysis of variance (fixed model) was used to compare residues in eggs in the four treatment groups that consisted of freshly laid and incubated eggs further divided into eggs from successful and unsuccessful nests. RESULTS AND DISCUSSION Eleven nests produced downy young, 26 produced no downy young, and the fate of 50 nests was unknown. Residues of the five organochlorines tended to be lower in those eggs collected from successful nests when compared with those collected from unsuccessful nests (Table 1); but analysis of variance revealed that these differences were significant only for DDE, dieldrin, and DDD (Table 2). Analysis of variance revealed that DDT and DDD were the only residues altered by the stage of incubation of the sample egg--DDT in eggs decreased significantly (P < 0 025) during incubation and DDD increased significantly (P < 0-05). These changes were apparently due to metabolism of DDT by the developing embryo. This was previously reported by Abou-Donia & Menzel (1968). The DDT residues were so small that metabolism of DDT to DDE, also reported by Abou-Donia & Menzel (1968), would result in a very small increase in the DDE level. While we considered combining DDT with DDD for the statistical analysis, this may have tended to oversimplify a complicated problem. There was a positive, significant intercorTelation among each of the residues (Table 3). These intercorrelations among residues in brown pelican eggs were noted previously (Blus et at., 1971). Determination of the relation of egg residues to nest success was made more difficult because of the intercorrelations. LAWRENCE i. BLUS Ct at. TABLE) OROANOCHLORINE RUIDUES IN }7 NNOLE COOS FROM INDIVIDUAL PILICAN NUT* Mglg {fifth wtt weight) Successful nests Unsuccessful netti DDE Dtetdrin PCBt DDT DDD DDE Ditidrin PCBs DOT DDD 120 190 2-0S 2-26 G* 1-B0 028 052 0-17, 029 0-29 Eggt fifthly laid when collected 3-9 005 016 1*96 0-19 6-0 0-12 015 2-30 037 3-9 0-24 018 2-33 005 3-9 0-42 0-61 2-46 043 2-47 015 2-66 0-64 3-49 0-75 3-30 074 3-75 063 3-92 044 3-97 0-61 4 14 071 5-24 077 60S 092 8-48 0-40 4-34 016 023 3 43 0-43 37 3-5 6-0 365 4-4 6-5 78 61 7-6 94 9-2 98 167 9-9 32-3 9-06 032 013 005 032 OIO 018 027 021 023 0)5 018 025 045 079 0-39 024 029 029 026 047 Oil 030 027 030 063 033 043 0 36 061 1 04 095 039 130 0-33 1-39 0-19 158 019 1-76 0-29 190 0-54 2 25 042 2*37 0-35 c ns 031 33 30 3-9 3# 17-3 18-6 3-1 6-28 Eggs Incubated when collected 0-03 048 1-77 036 0-19 0-27 1-93 0-58 003 019 196 052 005 017 225 032 0)8 071 2-88 0 30 0 13 0-37 3-32 058 013 0-61 3 97 084 399 052 399 0-83 399 0-99 4-37 064 010 035 2-97 058 Geometric Mean. 34 45 19 90 6-6 90 34 9-2 92 83 10-4 6-67 005 037 005 005 005 018 0-47 030 005 033 005 012 025 057 054 038 060 070 077 055 1-04 039 083 056 TABLE 2 GEOMETRIC MEANS (WCIOHTED) Of OROANOCHLORINE RESIDUU IN 37 loot FROM SUCCESSFUL AND UNSUCCESSFUL NUT* OF THE BROWN PELICAN Nett status DDE Geometric Mean (jig/g, fresh wet weight) Dieidrin PCBt DDT DDD Succeiiful Unsuccessful Error Mean Squire (Log Residue) 1-77 3*23f 0-0250 030 049* 0 0671 3 50 7-94 0-0700 0-11 017 01214 030 0 46* 00532 MOMS O0<il02 ^f>*-<` rx - ' !.>. ' * * /****.; * v '' ` .,7y- -w '* X V *U. -- ~ ORGANOCHLORINE RESIDUES IN PELICAN EGOS 85 TABLE 3 ASSOCIATION OP PIVt OROANOCHLOAIN! RESIDUES FOUND tN 37 SROWN PELICAN IOOS Chemical DftUtin Simple correlation coefficient (r) PCBs DDT ODD DDE PCBi DDT 0-564** 0-585** 0-351* 0-480$ 0-400T 0*4071 0-599** 0-363** 0-478* 0*337* P < 0*03. i p < ooi; P <0001. The maximum DDE residue in an egg from a successful nest was 2-37 pgfg. The maximum DDE residue in an egg from an unsuccessful nest was 8-48 /jg/g and 19 of 26 eggs from unsuccessful nests (73-1 %) contained more than 2-37 pg/g. Use of binomial tables (Mainland *t ai, 1956) to test these data further revealed that production of downy young seemed related primarily to DDE in the egg (Fig. I). Of 20 nests in which the sample egg contained 2-5 ^g/g of DDE. II produced at least one downy young (probability of a successful nest 0-550; NgmWf ef ftatfi. r*Miniftq and downy yaang Fig. \. Nest*, remaining eggs, and downy young related to DDE content of sample eggs. MQNS 082183 I'XiStt 0RGAN0CHL0R1NE RESIDUES IN PELICAN EGGS 87 downy young it the time rate is the remaining eggs (17 downy young from 33 eggs SI S/'), production can be calculated to be 27 young or 1*36 young per nest, which is considered a normal rate of productivity for the brown pelican (Henny, 1972). Using the same analysis for dieldrin, 17 young were produced in 23 nests in which the sample egg contained 50-54 pg/g of dieldrin. Assuming that sample eggs would have produced downy young at the same rate as the remaining eggs (17 downy young from 37 eggs - 45-9%), production can be calculated to be 28 young or 1*22 young per nest, which is also considered a normal rate of productivity. The relationship of DDE or dieldrin residues to nest success followed the same trend in each of the 2 years. Student's /-test was used to compare residue means; there were insufficient eggs to consider residues in relation to stage of incubation for each year. DDE or dieldrin residues in eggs from successful nests in each year were not significantly different (/* > 0 05) from each other nor were DDE or dieldrin residues in eggs collected each year from unsuccessful nests significantly different (P > 0 05) from each other. DDE or dieldrin residues in eggs collected from unsuccessful nests in each year were significantly higher (P < 0-05) than those residues in successful nests in either year. Residues of each organochlorine in eggs from nests where fates were unknown were lower than those found in eggs from unsuccessful nests, but they were higher than in (hose sample eggs from unsuccessful nests. Because of the similiarity of the relationship of egg residues to nest success in each year, data from both yean were combined for statistical analysis. Upon examination of residues of DDE and dieldrin in individual eggs (Table 1), the close intercorrelation of these two chemicals is evident. There is only one egg from an unsuccessful nest in which dieldrin (0- 58 vgjg) exceeds the maximum level found in an egg from a successful nest, and the DDE level (1-93 pg/g) falls below the maximum level found in an egg from a successful nest. Otherwise, eggs from unsuccessful nests contain levels of DDE and dieldrin that are both below the maximum levels found in an egg from a successful nest (6 eggs), both above the maximum levels (13 eggs), or with DDE above the maximum level and dieldrin below the maximum level (6 eggs). This relationship tends to indicate that nest success is primarily dependent upon the DDE level in the female that is reflected by the residue in the egg. However, the effects of dieldrin and DDE were not completely identified. AH of the pollutants found in the pelican eggs are presumably toxic to the embryo or hatched young, but their effects on toxicity may not be strictly additive with regard to the quantitative effect induced by a given unit of chemical. Thus, little would be gained and much might be lost in assuming straightforward additivity without first demonstrating this relationship. There is also the problem of separating embryotoxic effects from eggshell thinning effects, although, In the final analysis, eggshell thinning may result in cracking or crushing of the eggshell that is toxic to the embryo. DDE appears to induce all or most of the eggshell thinning found in the brown pelican (Blus et al., 1971, 1972a, 19726); and this chemical would be expected to contribute to all of the deaths of embryos caused by crushing or cracking of the eggshell attributable to environ mental pollution. There seemed to be some indication that DDE is embryotoxic in the intact egg at the low levels found in this study; but this is difficult to determine, particularly in the presence of dieldrin. The principal effect of dieldrin is probably embryotoxicity in those eggs with shells with sufficient strength to withstand the weight stresses associated with incubation. There is also the possibility that dieldrin may be associated with post-hatching mortality (Koeman er al., 1967) or adverse effects on behaviour of young birds (Baxter et al., 1969). Although influences of pollutants on the breeding behaviour of birds have been mentioned as one of the factors adversely affecting reproductive success (Snyder et al., 1973), there is little experimental evidence to confirm this. Snyder et al. (1973) mention possible abnormal parental behaviour in American kestrels (Falco sparverius) given dietary DDE plus PCBs. Although there was no significant (P > 0-10) interaction between stage ofincuba tion and nest success in relation to levels of five chemicals in pelican eggs, there seemed to be a trend for DDE levels to vary with stage of incubation (Table I). Of the five highest DDE residues, four occurred in freshly laid eggs collected from unsuccessful nests. This trend was also reflected in two freshly laid crushed eggs that contained 3-61 and 7-04yg/g of DDE; these nests were not included it) the analysis because only one egg was known to be (aid in each nest. Two other freshly laid eggs which had been crushed were collected from unsuccessful nests, each of which contained three eggs. These eggs contained 3-52 and 3-08 ;ig g of DDE, but were excluded from the analysis due to subsequent flooding by salt water. The dieldrin levels in these four eggs ranged from 0 32 to 0-50 jrg/g. Crushed eggs were observed in two of the 26 unsuccessful nests included in the analysis. Thus, there is a tendency (not statistically significant) for freshly laid eggs from unsuccessful nests to contain more DDE than incubated eggs from unsuccessful nests. This may be related to eggs with higher DDE levels having a shorter nest life than those with lower levels because of cracking and crushing of shells thinned by DDE. We cannot completely rule out (he possibility that some unidentified pollutant may have had significant adverse effects on nest success of the pelicans, but this possibility seems highly unlikely. Six of the eggs were analysed for residues of chromium, nickel, selenium, arsenic, mercury, lead, cadmium, copper, zinc, and magnesium. Concentrations of these metals did not seem unduly high--residues of Pb and Cd seemed unusually low--but little is known concerning the influence of low levels of these metals on reproductive success in birds. It is also necessary to consider that some nests were unsuccessful for reasons other than residues in the eggs. Nest desertion is the most likely cause of lack of success because it may occur for various reasons such as inclement weather, food shortage, or human disturbance. Young may also die from diseases, inclement ORGANOCHLORINE RESIDUES IN PELICAN EGGS 89 weather, and other factors. This might alter the relationship of residues to nest success by inclusion of eggs with low residues in the sample from unsuccessful nests and by possible desertion of some nests in which the sample egg contained >2-5 ^g/g of DDE or >0-54 pg/g of dieldrin. This bias, if it occurred, was not sufficient to obscure the relationship between residue levels and nest success. Our calculations assume that residues in one egg from a clutch of brown pelican eggs (mean clutch size is near three eggs) is representative of residues of other eggs in the clutch. We have residue data from two eggs from each of four clutches collected in Florida in 1969 that bear on this point. Analysis by a paired Mest (Simpson era/., 1960) revealed no significant difference (P > 0 05) in organochlorine residues relative to the order in which eggs were laid in the same clutch. Potts (1968) also found no significant differences in organochlorine residues in eggs of the same clutch of the shag (Phalacrocorax aristotelis). Vermeer A. Reynolds (1970) reported highly significant correlations of residue levels between eggs of the same clutch of the California gull (Larus cali/omicus). It appears valid to use residues in one egg of the clutch to represent residues in other eggs within the same clutch. The evidence for influence of DDE and dieldrin residues on nest success is similar to evidence for the effects of DDE on eggshell thinning in wild birds; it is correlative evidence and should not be construed as exhibiting a cause-and-effect relationship. Cause-and-effect evidence may be drawn from experimental demon strations that birds fed DDE experienced lowered reproductive success (Heath et ai. 1969; Longcore era/., 1971). In the study of black ducks (Anas rubripts) fed DDE (Longcore et /., 1971), lowered reproductive success occurred; however, DDE residues in the egg were much higher than those in the brown pelican eggs. There have been few studies of effects of residues in eggs on success of marked nests. In several studies, dieldrin was reported to have a deleterious effect on nest success (Potts, 1968; Lockie era/., 1969). In our study, dieldrin levels approximated those found in the Scottish golden eagle (Aquita chrysaetos); a mean level of 0 86 M8/S in the eagle eggs was associated with lowered reproductive success and reproductive success improved markedly when the average dieldrin level dropped to 0*34 pg/g (Lockie et a/., 1969). Snyder et ai. (1973) found DDE related to nest success in Cooper's hawks {Acciplter cooperi), but there was little mention of residue levels in eggs from successful or unsuccessful nests. The average dieldrin level in Cooper's hawk eggs was low but the range was not given. RatclifTe (19676) found no significant difference in organochlorine residues in eggs from successful and unsuccessful nests ofthe peregrine (Falcoperegrfnus); however, residue analyses were made at a time when PCBs were not taken into acoount. PCBs may seriously interfere with analysis for certain other organochlorines unless suitable analytical procedures are followed. Brown pelicans are reproducing subnormally in the colony involved in this study, and the number ofyoung produced in each of the last four years is insufficient to maintain a stable population (Blus et ai, 1972a; Henny, 1972). The population MQNS QaZU7 90 LAWRENCE J. BLUS tl at. in South Carolina declined from over 5000 breeding pairs in the 1950s to the current population of slightly over 1000 pairs (Beckett, 1966; Blue, 1970; Blue ei at., 1972c). The data presented in this study show a strong correlative relationship between DDE and dieldrin in eggs with reproductive success and decline of the brown pelican population in South Carolina. ACKNOWLEDGEMENTS We thank George A. Shegogue, Steve Joyner, Daniel Doshier, and others who helped with the field work. We thank Robert Heath and Lucille Stickel for their critical evaluation of the manuscript and Mr Heath for advice on statistical aspects. Aeou-DoNtA, M. B, 4 Mtsm, D. B. 0968). The metabolism tit eti-o of l,),l-trich(oro*2,2-bii` (<hlorophenyJ)ethanc (DDT). l.l-dichloro-2,2'bis(p-chl0rophcnyl)ethinc(DDD)and 1,1* aichloro-2,2-bt(p'ChIorophenyl)ethylene (DDE) in the chick by embryonic injection and dietary Mention. Biochem. Fharmac., 17,2)43-6). Anderson, D. W. 4 Hictciv, i. i. (1972). Eggshell changes in certain North American birds. Aw, fnt. orn. Congr., 15th. 514-40. Baxtcr, W. L., Linder, R. L. 4 Dahloren, R. B. (1969). Dieldrin effects in two gmcraiwns of penned hen pheasants. J. WUdl. Mgmt, 33,96-102. Bickctt. T. A.. Ftl (1966). DevetuK Bank--196* end 1965. Chat, 30,93-100. Btus, L. J. (1970). Measurements of brown pelican eggshells from Florida and South Caroline. BioSeienee, 20,967-9. Bius, L. J.t Bclisu, A. A. 4 PRolty, R. M. (in press). Relations of the brown pelican to certain environmental pollutants, Ftstic. Montt.J, Btus, J.. J., Gish. C. D.f Bslislc. A. A. 4 P*outy, R. M. (1972a). Logarithmic relationship of DDE residues to eggshell thinning. Nature, Laud., 235,376-7. Blus. L. 3., Gish, C. D.. Beusle, A. A. 4 Paourv, R. M. (19726). Further analysis of the logarithmic relationship of DDE residues to nest success. Nature, Loud., 240,164-6. Bun, L. J., Heath, R. G., Gish, C. D., Beusle, A. A. 4 Paourr, R. M (I97I). Eggshelt thinning in the brown pelican: Implication of DDE. BtoSclente, 21,1213-15, Heath, R. G., Spann, i. W. 4 Krk(T2er, ). F. (1969). Marked DDE impairment of maUafd reproduction in controlled studies. Nature. Laud., 224,47-1. Hcnnv. C. 3. (1972). An analysis of the population dynamics of selected avian species -with special reference to changes during the modem pesticide era. Wttdt. Bet. Rep. US Fisk Wild/. Sera., No. 1,99 pp. Hickey, i. J. 4 Anocmon, D. W. (1968). Chlorinated hydrocarbons and eggshell changes ia raptorial and Ash-eating birds. Science, N. Y., 162,271-3. Kocman, J. H., Oodemnb, R. C. H. M. 4 Huisman, E. A. (1967). Danger of chlorinated hydrocarbon insecticides in birds' eggs. Nature, Land.. 2)9,1094-6. Lockji, J. D,, RATCum, D. A. 4 Balxarrv. R. 0969). Breeding success and organo<IUorin residues in golden eagles in 'Vest Scotland. J. appl. Ecol., 6, 3e)-9. Lonocore. J. R., Samson, F. B. 4 Wkittesdau,T. W.. Jr (1971). DDE thins eggshells and lowers reproductive success of captive black ducks. Bull. Environ. Contam. 4 Toxicol., 6, 413-90. Mainland, D., Herrera, L. 4 ScTCum. M, I. (1936). Statistical tables for use with binomial samples--contingency tests, confidence limits, and sample she estimates. New York, New York University Colleac of Medicine. McLane, M. A. R. 4 Hall, L. 0972). DDE thins screech owl eggshells. Bull. Environ. Contam. A Toxicol., I. 63-8. AO/V.S V 3sUr* OROANOCHLORINE RESIDUES IN PELICAN EOGS 91 MUIHIRN, B. M., CROmartii, E., Reichiu W. L. A Beuilb, A. A. (1971). Semiquantltetive determination of polj^hlormattrt biphenyls in tissue samples by thin layer chromatography. MwiHiR^ B^rRiiCHSU W'. L.'. Lockir L. N., Lamont, T. O.. Beluu, A. A., Cromartii, E . Baouv^O. E. A Prouty, R. MJ1970). Organochlorine residue* and autopsy data from bald PornijXA. U96S). Success of tha shagon the Fame Islands, Northumberland, in relation to thairconient ofdieldrin and p.p'-DDE. Nature. Lond., 217, 1282-4. Ratclipr^D. A. (1967a). Decrease in eggshell weight in certain birds of prey. Saiurr, loud., 21S, Ratclippb, D. A. (19676). The peregrine situation in Great Britain 1965-1966. Bird Study, 14, 231-46. Raiclipr. D. A. (1970). Change* attributable to pesticides in ctg breakage frequency and eggshell thickness in tome British birds. J.appl, Ecol., 7,67-107. Simpson, 0.0., Roe, A. A Uwontin, R. G. (1960). Quontltativt xoohgy. New York, Harcourt, gMVMaTN. F. R.. Snvdw, H. A.. Lincsr. J. L. A Riynou* R. T. (1973). Organochlorine^ heavy metals, and tha biology of North American acciptten. B/eSefww,300-5. Virmccr K. A Riynoum, L. M. (1970). Organochlonna residues m aquatic birds m the Canadian prairie provinces. Con. Fid Nut., 64,117-30. . WiiMvaa, s! N. A Rontib, R. D. (1970). DDE thins esssbells of captive American kestrels.