You are on page 1of 6

Molecular Ecology Notes (2007) doi: 10.1111/j.1471-8286.2007.01796.

x

TECHNICAL ARTICLE
Blackwell Publishing Ltd

Species identification of birds through genetic analysis of
naturally shed feathers
J A M I E A . R U D N I C K ,* T O D D E . K A T Z N E R ,† E V G E N Y A . B R A G I N ‡ and J . A N D R E W D e W O O D Y *
*Department of Forestry and Natural Resources, Purdue University, 195 Marsteller Street, West Lafayette, IN 47907-2033, USA,
†Department of Conservation and Field Research, National Aviary, Allegheny Commons West, Pittsburgh, PA 15212-5248, USA,
‡Science Department, Naurzum National Nature Reserve, Kostanay Oblast, Naurzumski Raijon, Karamendy 459730, Kazakhstan

Abstract
Genetic analysis of noninvasively collected bird feathers is of growing importance to avian
ecology; however, most genetic studies that utilize feathers make no mention of the need
to verify their species of origin. While plumage patterns and collection location often are
indicative of species identity, broad-scale feather collections may require definitive species
identification prior to analysis. Genetic species identification has been applied to noninva-
sively collected samples from a wide range of taxa but, to date, these techniques have not
been widely used on bird feathers. Here, we develop and test a polymerase chain reaction
(PCR)-based technique for identifying eastern imperial eagle (Aquila heliaca) samples
among a vast number of noninvasively collected feathers. Species identification is accom-
plished by amplifying a fragment of the mitochondrial cytochrome c oxidase I gene, then
digesting that fragment with a restriction enzyme. The resulting species-specific restriction
fragment length polymorphisms (RFLPs) are easily visualized by gel electrophoresis. We
tested this PCR-RFLP assay on over 300 individuals that had been genetically identified
from noninvasively collected feathers and demonstrated that the assay is both reliable and
robust for DNA of low quality and quantity. The genetic methods of species identification
used to develop this assay can readily be applied to other bird assemblages, making them
particularly relevant to a broad range of future avian research.
Keywords: cytochrome c oxidase I, eagle, feather, noninvasive, RFLP, DNA barcoding

Received 2 February 2007; revision accepted 12 March 2007

suitable sources of DNA for avian research (Pearce et al.
Introduction
1997; Segelbacher 2002; Horvath et al. 2005), most genetic
A growing number of studies are utilizing noninvasively studies that exploit these types of samples make no
collected samples (feathers, hair, scat, etc.) to characterize mention of the need to verify the feathers’ species of origin.
ecological or demographic attributes of populations that In some cases, plumage patterns or site-specific data are
are either difficult or impossible to investigate by more diagnostic for species, but in many cases, they are not. In
traditional means. Because the donor species of non- these situations, the species origin of noninvasively collected
invasively collected samples must often be verified, feathers must be verified if results are to be credible.
genetic methods of species identification have become The species origin of noninvasively collected feathers
invaluable resources (e.g. Lucchini et al. 2002; Palomares must be carefully evaluated when a number of species
et al. 2002; Kurose et al. 2005; Smith et al. 2006). Although with similar plumage patterns frequent areas where
naturally shed feathers have long been identified as feathers have been collected. For example, consider feathers
collected haphazardly from a productive wetland: if
Correspondence: Jamie A. Rudnick, Department of Conservation
multiple species with similar plumage patterns are present,
Science, Chicago Zoological Society, 3300 Golf Road, Brookfield, the feathers’ donor species should be ascertained before
IL 60513, USA. Fax: 708-485-6048; any species-specific analyses can occur. Genetic species
E-mail: jarudnic@brookfieldzoo.org identification also can be useful in situations where

© 2007 The Authors
Journal compilation © 2007 Blackwell Publishing Ltd

tailed eagle (Haliaeetus albicilla. greatest number of sampled nesting sites was identified. In total. while avian communities (e. Many of the samples we collected were that directly pre. between breeding seasons (Rudnick et al. 2005). our several aspects of eastern imperial eagle (Aquila heliaca. When those feathers lack distinctive plumage buffer (100 mm Tris-HCl pH 8. Scribner & Bowen 1998). each group of DNA extractions.g. In general. Because breeding EIEs are highly territorial. a small number of feathers often are the the Naurzum Zapovednik. Our previous work focused To minimize pseudoreplication among samples. (2005). Gilbert & Nancekivell 1982. Herein. Our ongoing work. 2003). Predator dietary patterns often are characterized by Reference samples examining faeces or regurgitated pellet castings (Korschgen 1980). we also we have collected thousands of feathers from communal supplemented these samples with chicks from the year roosting areas. the feathers of which sometimes nests to the analyses. eagles are socially monogamous techniques of species identification prove inconclusive. When a bird collides period (1998–2004) from EIE. the breeding season with the EIE nesting sites. it was necessary The Bald and Golden Eagle Protection Act prohibits for us to genetically verify the donor species of samples known import of GE samples into the United States. WE and SE chicks hatched at with an aircraft. and ultimately assign species identity when more conventional. © 2007 The Authors Journal compilation © 2007 Blackwell Publishing Ltd . 9 WE chicks and 6 SE chicks for use as SE). Manville 1963. To the assumption that eagles exhibit high nesting site fidelity study this nonterritorial component of the EIE population. Specimens were originally collected between our study site. For each species.g. 2005) and reference individuals for EIEs. species identification assay: developing blood feathers Genetic methods of species identification are capable of plucked directly from eagle chicks and toepad tissue taken determining the species identity of ambiguous remains. one chick from each nest was randomly chosen for feather’s donor species. discriminating among the four eagle species present at IL (USA). Farrell et al. Because our previous work empirically supported focusing on nonbreeding EIEs present at our field site. WEs and SEs. To year by the following means. we obtained tissue samples resemble those of EIEs: the steppe eagle (Aquila nipalensis. GE) and the white. genetic methods of species identification could NaCl. is analysis. 10 mm patterns. Genetic museum specimens were used as reference samples for GEs. there were few cases for which we had reason to doubt a Then.and proceeded the primary sampling body feathers that lacked distinctive plumage patterns. A how methods of genetic species identification can negative control containing no eagle tissue was included in contribute to future avian research. 100 mm EDTA. 2% SDS) immediately upon collection. reference samples. Methods but for which species identity is nonetheless ambiguous. then devel- oped a restriction fragment length polymorphism (RFLP) DNA methods assay to distinguish among each species’ mitochondrial DNA (mtDNA) haplotypes. this reason. Thus. but these studies rarely catalogue avian prey by Two types of reference samples were used to design the species (e. we describe samples were first soaked in 100% ethanol for 24– 48 h to the molecular assay we used to identify EIEs and discuss remove polymerase chain reaction (PCR) inhibitors. We first used reference samples of known 1908 and 1941 throughout both the United States and species origin to characterize COI sequences (haplotypes) Central Asia. Instead. collection of chick samples was replete with full-siblings. DNA was similarly and the mtDNA results were evaluated in light of micros. This RFLP assay was tested DNA was isolated from chick samples following the on several hundred noninvasively collected eagle feathers. isolated from toepad tissue. from 30 EIE chicks. with the exception that atellite data on the same set of samples. we chose on a breeding population of EIEs (Rudnick et al. For prior to intraspecific analyses of individuals. We identified all nesting sites complicate matters. the golden eagle (Aquila chrysaetos. GE chicks were not directly sampled. then added one chick from each of those species (Katzner et al. Consequently. methods in Rudnick et al. EIE) biology in Central Asia.0. however. 1994). approximately 25 raptor species breed that were sampled in the supplementary years but not in within our study site including three additional eagle the primary year. We used a fragment of the mitochondrial cytochrome c we obtained samples of toepad tissue from 20 GE specimens oxidase I gene (COI) to develop a robust assay for genetically located at the Field Museum of Natural History in Chicago. both within and among these eagle species. species identification also can be helpful for studies that Developing blood feathers were collected over a 7-year investigate bird–aircraft collisions.2 TECHNICAL ARTICLE feathers are collected by less traditional ‘noninvasive’ means. from museum specimens. visual stored at –80 °C.g. WE). a national nature reserve in only evidence remaining (e. Laybourne north-central Kazakhstan. SEs and WEs in the following utilized feathers that were noninvasively collected from manner. Developing blood feathers were which better characterizes a predator’s impact on local used as reference samples for EIEs. 2000). Samples were placed in a lysis 1974). and exhibit high nesting site fidelity both within and among We have used naturally shed feathers to investigate breeding seasons (del Hoyo et al.

and generate consensus sequences for all a 746-bp fragment of COI in our EIE.6 (Genecodes Corporation) was used to span at least 19 avian orders (Hebert et al.3 µm each primer. buffer system. and we validated the use of thermal profiles for all primer pairs were identical to this enzyme empirically. (New England BioLabs). four communal raptor roosts at the Naurzum Zapovednik. PCR conditions and among the four species in silico. following the Naturally shed feathers were noninvasively collected from manufacturer’s protocol modified to one-eighth reactions. Consequently. WE and SE chick individuals. Primers BirdF1 and EagleR2b (the redesigned 4. 0. modeltest 3. The thermal profile included an initial denaturation of species divergence. Table 1 Primers designed to amplify the Fragment target fragment of COI in four short. samples.008 U Vent DNA polymerase paup* 4.0b10 (Swofford 2004) was used to calculate pairwise BirdR1 primer. 1. followed by 32 cycles of 94 °C for COI sequence was downloaded from GenBank and used 1 min. ensure sequences contained open initially screened the BirdF1/BirdR1 primers published reading frames (to guard against amplification of nuclear by Hebert et al. 5′-ATTGATRGCGGTTGTGATAAA-3′) maximum-likelihood (ML) distances for all unique COI robustly amplify a 485-bp fragment of COI from both chick haplotypes. (2004) and found them to robustly amplify pseudogenes). COI sequences from all four eagle species were compared DNA isolated from GE toepad tissue was highly degraded with one another to identify a restriction enzyme that and COI failed to reliably amplify for these samples with would produce species-specific RFLPs. TECHNICAL ARTICLE 3 The mtDNA COI gene has previously been shown to Developing a method for species identification exhibit adequate interspecific sequence divergence to discriminate among a wide variety of bird species that sequencher 4. Restriction digests were performed those listed for BirdF1/EagleR2b. All PCR products were cleaned with a sodium acetate- ethanol precipitation and sequenced in both directions with Testing samples of unknown species origin the amplification primers. step of 72 °C for 5 min concluded the profile. and 72 °C for 1 min.6 (Posada & Crandall 1998) was samples. However. Primer Sequence (5′–3′) size (bp) overlapping amplicons. 1000 bootstrap replicates. Digests were incubated at 37 °C for 1 h. taking into account BirdF1/EagleR2b. 0. Brody & Kern 2004) and stained with ethid- Negative control PCRs (those without template DNA) ium bromide for visualization. The restriction enzyme NlaIII easily discriminated overlapping amplicons (Table 1). with the exception of in a final volume of 30 µL and contained 10 µL COI PCR the following annealing temperatures: 45 °C annealing product. A final extension as an outgroup. then these values were used to calculate a BirdF1/EagleR2b were performed in a final volume of grand mean for intraspecific ML distance observed across 12 µL and contained 1× Thermopol PCR buffer. known species origin were run on each gel for comparison. PCRs for for each species. paup* of COI. Sequencing reactions utilized BigDye version 3. Mean interspecific ML distance was calculated each dNTP. We align amplicons. 50 °C annealing temperature for primer pairs GE04/GE05 then electrophoresed in 2% agarose gels (sodium boric acid and GE06/GE07. 3 µg BSA temperature for primer pairs BirdF1/GE01 and GE02/GE03. 59 °C for 30 s. to provide a graphic representation 1991). Fragment sizes are given in number of base pairs (bp) Amplicon 1 BirdF1 sequence from Hebert et al. and 0.0b10 also was used to generate an ML tree with (New England BioLabs) to reduce Taq error (Mattila et al.0 U NlaIII. A peregrine falcon (Falco peregrinus) step of 94 °C for 2 min. we redesigned BirdR1 to amplify a shorter fragment was then used for all subsequent sequence analyses. 2004). because amplification proved inconsistent used to identify the most appropriate model of evolution for noninvasively collected feathers that yield suboptimal for analysing the amplified region of COI. and 1. Reference samples of were routinely conducted to rule out sample contamination. 1× NEBuffer 4 (New England BioLabs). (2004) 184 GE01 CTATGAAGAAGATTATTACRAAAGCA Amplicon 2 GE02 TCYTAGGCGAYGACCAAATC 157 GE03 TTGTTTATRCGTGGGAAGGCTA Amplicon 3 GE04 GGAAACTGACTTGTYCCMCTC 218 GE05 TGWAGGGARAAGATRGCTAGGTC Amplicon 4 GE06 CCCAYTAGCSGGYAACATA 158 GE07 GGGTGTYTGGTATTGGAGA © 2007 The Authors Journal compilation © 2007 Blackwell Publishing Ltd . Mean intraspecific ML distances were calculated and noninvasively collected feather samples. and that model DNA. we designed a series of the intraspecific variation observed in our reference primers to amplify the target fragment of COI in four short.0 U Taq DNA Polymerase from all possible interspecific pairs of COI haplotypes.1 (Applied Biosystems).2 mm all species.

Mean ML distance among these individuals.15%.4 TECHNICAL ARTICLE Sampling locations were chosen because they were in areas where significant numbers of EIEs have been regularly observed. A total of 262 individuals were in 1939 (specimen 228823). and found to distances within species was 0. tion for the relevant region of COI. GE 50/110–116/209. submitted). Robertson et al. DNA sequences from the six birds exhibiting anomalous pairwise maximum-likelihood (ML) distances were RFLP profiles revealed a unique COI haplotype shared calculated among unique haplotypes. four haplotypes were found in EIEs (n = 30). Results Initial characterization of COI When the 485-bp fragment of COI was considered. There was no intraspecific RFLP variation among COI haplotypes in any Discussion species (i. haplotype (n = 9). species identity of the remaining 32 individuals (10%) Using the Akaike Information criterion. these individuals were identified as EIEs that enzyme NlaIII produced diagnostic RFLPs for all four shared a previously unrecognized haplotype. quantities. and all WEs shared a single consideration. Sequences representing as a GE. only two Restriction fragment sizes in base pairs (bp) are as follows: EIE produced sequences for the complete fragment of COI 160/325. 1984. texture Our newly developed RFLP analysis was used to identify and size. but many samples were body feathers that lacked distinctive plumage patterns. 1 A gel image of COI RFLP profiles for all eagle species under were found in SEs (n = 6). Dove © 2007 The Authors Journal compilation © 2007 Blackwell Publishing Ltd . feathers obviously not from EIEs were ignored).84% and the grand mean of ML compared to all known eagle haplotypes. Of the 20 GE samples obtained. 1). eagle species under consideration (Fig.g. The restriction site. but microscopic attributes of the plumulaceous the species of 314 individuals identified from noninvasively barbs also can be utilized (e. Uncut PCR product is indicated by a plus-sign. All samples in this study were also genotyped at a suite of seven microsatellite loci that were used to assign individual-specific DNA fingerprints. Thus. NlaIII discriminates among species.6 could not be ascertained because of insufficient DNA picked GTR + I as the most appropriate model of evolu. Sampled feathers were placed in paper envelopes and stored dry at room temperature. under consideration. WE 94/160/231. no COI haplotypes identified as EIEs. and a total of 314 individuals were identified among the noninvasively collected feathers (Rudnick et al.e. Using this model. One or two feathers were screened per individual. 13 individuals as WEs. two haplotypes Fig. Feather characteristics Testing samples of unknown species origin include macroscopic attributes like plumage pattern. Traditional methods of identifying a feather’s species of origin involve matching species-specific feather characteristics to reference individuals (Chandler 1916). SE 30/160/295. The database under the accession nos DQ834322–DQ834329. These two samples share a single haplotype and represent individuals collected from the United States in 1935 (specimen 100792) and Afghanistan collected feathers. and the species origin of all individuals was elucidated by the newly developed RFLP assay. and 1 individual were shared among eagle species. ML trees exhibited differ from one of the EIE haplotypes by only a single strong bootstrap support (ranging from 86% to 99% across nucleotide substitution that created an NlaIII restriction nodes) for intraspecific monophyly.e. Six individuals exhibited a single RFLP profile that all unique haplotypes have been deposited into the GenBank failed to match one of the expected eagle profiles. Only feathers that were likely to be EIE in origin were collected (i. This unique sequence was among species was 10. Overall. DNA was isolated from single feathers. modeltest 3. and any sample producing an RFLP profile that failed to match one of the known eagle profiles was subsequently sequenced. not among haplotypes within a species).

they found that spinner sharks from Species identification by genetic means is usually the North Atlantic produce different RFLPs than spinner accomplished through one of three techniques: phylogenetic sharks from the Pacific coast of Australia. 2003. phylogenetic reconstruction would have been burdensome. 2003. The mtDNA fragment we DNA isolated from noninvasively collected samples is employed was previously identified as an appropriate tool often of low quantity and quality. only two Alacs et al. there are only a few experts in feather haplotypes. when our RFLP assay was applied microstructure. nucleotide differences were found between the two groups specific primer amplification (Shivji et al. a PCR-RFLP approach would provide a more robust assay. a significant disadvantage of these characterization of intraspecific variation. Malik et al. these results highlight the importance RFLP approach for several reasons. As the number of studies that utilize naturally shed With species-specific primer amplification. techniques will supplement if not supplant other means tification. 282 were definitively manuscript. IL) for providing us with toepad tissue identified as one of the four eagle species under consideration. The primer pair produces a PCR product of unique size. PCR primers feathers continues to rise. Identifying the presence of quokkas (Setonix brachyurus) and © 2007 The Authors Journal compilation © 2007 Blackwell Publishing Ltd .e. 2002. However. TECHNICAL ARTICLE 5 1997). fragment of mtDNA evaluated in their study. and our results enhance that claim. Because assemblage of bird species. submitted). genetic species identification are designed for all likely species of origin such that each will become increasingly important for avian research. of spinner sharks. Even so. we felt more sequences are deposited in public databases (e. Alpers D. When the PCR-RFLP assay was Latch for their helpful comments on earlier versions of this tested on a data set of 314 individuals. as with our work. In the ~400-bp reconstruction (Baker & Palumbi 1994. Traditional methods of species identification are One important aspect of our approach to designing a advantageous because they require little specialized PCR-RFLP species identification assay was the initial equipment. Domingo-Roura et al. 2006). disparity. Much The species identity of the remaining 32 individuals could of the fieldwork and a portion of the laboratory work for this not be determined because of lack of DNA as opposed to a project was supported by the Wildlife Conservation Society and flaw in the species identification assay. It is worth noting the National Geographic Society. This research is ARP publication that we successfully identified the donor species of ~90% 2006-17940 from Purdue University. species. and supported in part by an NSF International Postdoctoral Research those samples were easily resolved with DNA sequencing. an uncharacterized EIE haplotype where these types of assays are routine). it was possible that the for genetic species identification in birds (Hebert et al. demonstrate the use of genetic analyses to successfully Heist & Gold (1999) reported similar results when identify 268 eastern imperial eagles from among thousands developing a genetic method of species identification for a of noninvasively collected feathers. Gomez-Moliner et al. de Tores PJ. GenBank) and molecular assays for species identification We used reference samples to develop and phylo. we produced an RFLP profile unexpected for the species. of the noninvasive samples we assayed and furthermore. suite of shark species. or PCR-RFLP analysis (Hold et al. Armstrong differences coincidentally resulted in a restriction site et al. We initially screened a previously published of feather identification in situations where positive primer set that amplified a 746-bp fragment of COI. 2004). method we have described for developing a PCR-RFLP Samples that fail to amplify are assumed to come from assay for species identification can be applied to any a donor species not included in the assay. We redesigned one of the Acknowledgements primers to amplify a shorter fragment of COI (485 bp) and found this new primer pair to be both reliable and robust We thank members of the DeWoody laboratory group and Emily for even suboptimal DNA. that our results were entirely consistent with species-specific References differences observed in microsatellite data collected for Alacs E. However. we envision that molecular genetically groundtruth a PCR-RFLP assay for species iden. 1997. one of these 2005). but identification by visual means is cryptic. We chose a PCR. Dillon M. Katzner was Only six individuals exhibited anomalous RFLPs. We also thank David Willard at the Field Museum of Natural History (Chicago. become easier to develop. As more and high proportion of false negatives. Consequently. Because of the large of characterizing intraspecific variation prior to devel- number of samples requiring analysis (~300 individuals). species-specific primer approach would have produced a 2004). found this fragment did not reliably amplify in DNA of low quantities and qualities. 2001. from every golden eagle specimen we requested. Spencer PBS (2003) these samples (Rudnick et al.g. but there are many genetic laboratories to a large data set. Kurose et al. and we were careful to are a valuable alternative to traditional techniques because choose a restriction enzyme that would produce the same they are less dependent on user interpretation and are species-specific RFLP across the range of intraspecific more portable (i. oping a PCR-RFLP assay for species identification. Fellowship (INT-0301950) and by Arizona State University. Here. Genetic methods of species identification eagle species under consideration. Collectively. Multiple COI methods is that accurate species identification can require haplotypes were discovered in two out of the four extensive experience.

Posada D. Hurt A et al. Biosecurity. 13. Kalmar L. Atlantic large coastal shark fishery. 105. Condor. Molecular Ecology. Pank M et al. identifying species and sex of sympatric carnivores: a noninvasive Massachusetts. University of California Publications in 75. DeWoody JA (submitted) A Hold GL. Walsh PJ Laybourne RC (1974) Collision between a vulture and an aircraft (2003) Tussock moth species arriving on imported used vehicles at an altitude of 37 000 ft. 385– Central Asia. vison). Sunderland. 84–88. 9. American mink (M. 53–61. 1365–1374. Wilson Bulletin. 2. Japan. Martinez-Cruz B. Bowen TD (1998) Microsatellites identify depredated Horvath MB. 688–697. Lucchini V. 471–481. Fields RL. Korpela J. Roman J. New World Vultures to Guineafowl. of genetic data for avian ecological studies. 86. Smith DA. 2. Dove CJ (1997) Quantification of microscopic feather characters Palomares F. and polecat (M. 16. Spain. Tenkanen T. (2006) Badger hair in DNA synthesis by the Thermococcus litoralis DNA polymerase — shaving brushes comes from protected Eurasian badgers. Rudnick JA. taxonomic significance. 24. Palumbi SR (1994) Which whales are hunted? A molecular tracking of colonizing wolf (Canis lupus) packs in the western genetic approach to monitoring whaling. and population monitoring in an 1657–1663. Brody JR. Crandall KA (1998) modeltest: testing the model of Canadian Journal of Zoology. Kern SE (2004) Sodium boric acid: a Tris-free. Journal of Zoology. with reference to their Manville RH (1963) Altitude record for mallard. McHugh P.S. Gomez-Moliner BJ. Barcelona. Katzner TE.. Journal of Field Gilbert FF. Science. Swofford DL (2004) PAUP*: Phlyogenetic Analysis Using Parsimony Kurose N. Negro JJ. 15. 7. Schemnitz SD). determined by DNA analysis. 60. 3. 2. Zoology. 16 –20. Ralls K. Pitkanen K (1991) Fidelity of Domingo-Roura X. vison) and otter (Lutra canadensis) in northeastern Alberta. cooler Malik S. 97. Frampton ER.C. DNA substitution. pp. 387–406. (2001) Validation of a PCR. Godoy JA waterfowl remains from glaucous gull stomachs. 214–216. Bragin EA. 538–551. 11. Chandler AC (1916) A study of feathers. sympatric carnivores identified by molecular analysis of scats. 1401–1405. Condor. method for conservation on the Tsushima Islands. the World. Pryde SE et al. Piriz A. Lynx Edicions. 36. Katzner TE. (2006) Assessing reliability of Korschgen LJ (1980) Procedures for food-habits analyses. 243–446. Chinn W. Molecular (2005) An overlooked DNA source for non-invasive genetic Ecology. 1538–1539. Wilson PJ. Sinauer & Associates. Nancekivell EG (1982) Food habits of mink (Mustela Ornithology. (2002) Noninvasive molecular Baker CS. Molecular Ecology. Faecal genetic analysis to determine the presence and distribution 47–57. Farrell LE. analysis in birds. 262. 1036–1047. endangered Eastern imperial eagle (Aquila heliaca) population Heist EJ. from Kazakhstan. Vol. 14. Molecular Ecology. Zemlak TS. genetic and GIS analyses. Sunquist ME (2000) Dietary separation of Molecular Ecology. Bragin EA. Bragin EA. Molecular Ecology Notes. Marucco F et al. analysis of excremental DNA. Johnson WE (2002) used in the identification of North American plovers. 367–369. 212. 311–316. Segelbacher G (2002) Noninvasive genetic analysis in birds: del Hoyo J. noninvasive Genetic Evaluation of Population Size and Natal RFLP based method for the identification of salmon species in Philopatry of Non-breeding Imperial Eagles. Cabria MT. (2004) PCR-RFLP Robertson J. 6. 36. 389. of Heredity. (2002) Genetic identification of Katzner TE. Harkin C. European Food Research and Technology. 4967–4973. 1282–1288. Journal Wildlife Research. 2171–2182. putorius) by 85–98. BioTechniques. 817–818. of elusive carnivores: design and feasibility for the Iberian lynx. 11. Clarke S. Nucleic Acids Research. D. Fishery Bulletin. (Aquila Heliaca) in food products. Biological Conservation. 99. Masuda R. 128. Elliott A. 2959–2967. Stoeckle MY. Washington. 56. Journal of Avian Biology. O’Brien SJ. Mattila P. Sargatal J (eds) (1994) Handbook of the Birds of testing reliability of feather samples. 11. White BN (1997) conductive medium for DNA electrophoresis. Rudnick JA. 425 – 430.6 TECHNICAL ARTICLE other macropods using cytochrome b analyses from faeces. lutreola). Shivji M. Rhodes OE Jr. an extremely heat stable enzyme with proofreading activity. Tatara M (2005) Fecal DNA analysis for (*and Other Methods). Wilson Bulletin. PLoS Biology. Ferrando A et al. Govan J (1984) The identification of bird identification of mustelid species: European mink (Mustela feathers: scheme for feather examination. Knick ST. Francis CM (2004) (2005) Using naturally shed feathers for individual identification. Scribner KT (1997) Nest materials as a source Molecular Ecology. Lavigne DM. Pinniped penises in trade: a molecular-genetic investigation. 92. 857–868. DeWoody JA Hebert PDN. 265. Bioinformatics. 41– 47. Russell VJ. Scribner KT. 68. Smith RJ. Conservation Biology. 461–462. © 2007 The Authors Journal compilation © 2007 Blackwell Publishing Ltd . The Wildlife Society. 14. a multi-species assemblage of eagles in central Asia. Identification of birds through DNA barcodes. Rubines J et al. 113–127. Armstrong KF. Italian Alps. Pearce JM. Gold JR (1999) Genetic identification of sharks in the U. Science and Justice. Conservation Biology. Godoy JA. 19. Smith AT (2003) Coexistence in pelagic shark body parts for conservation and trade monitoring. Fabbri E. In: microsatellite genotypes from kit fox faecal samples using Wildlife Management Techniques Manual (ed. genetic parentage analyses. Marmi J. 1583 –1590.