Using historical captive stocks in conservation. The case of the lesser white-fronted goose

Transcription

1 Conserv Genet (2007) 8: DOI /s ORIGINAL PAPER Using historical captive stocks in conservation. The case of the lesser white-fronted goose Minna Ruokonen Æ Anna-Carin Andersson Æ Håkan Tegelström* Received: 2 November 2005 / Accepted: 28 April 2006 / Published online: 5 July 2006 Ó Springer Science+Business Media B.V Abstract Many captive stocks of economically or otherwise valuable species were established before the decline of the wild population. These stocks are potentially valuable sources of genetic variability, but their taxonomic identity and actual value is often uncertain. We studied the genetics of captive stocks of the threatened lesser white-fronted goose Anser erythropus maintained in Sweden and elsewhere in Europe. Analyses of mtdna and nuclear microsatellite markers revealed that 36% of the individuals had a hybrid ancestry. Because the parental species are closely related it is unlikely that our analyses detected all hybrid individuals in the material. Because no ancestral polymorphism or introgression was observed in samples of wild populations, it is likely that the observed hybridisation has occurred in captivity. As a consequence of founder effect, drift and hybridisation, captive stocks were genetically differentiated from the wild populations of the lesser white-fronted goose. The high level of genetic diversity in the captive stocks is explained at least partially by hybridisation. The *Deceased 20 March M. Ruokonen (&) Department of Biology, University of Oulu, P.O. Box 3000, FIN-90014, Oulu, Finland minna.ruokonen@oulu.fi A.-C. Andersson Evolutionary Functional Genomics, Department of Evolution, Genomics and Systematics, EBC, Uppsala University, Norbyvägen 18 D, SE Uppsala, Sweden H. Tegelström Conservation Biology and Genetics, EBC, Uppsala University, Uppsala, Sweden present captive stocks of the lesser white-fronted goose are considered unsuitable for further reintroduction, or supplementation: hybridisation has involved three species, the number of hybrids is high, and all the investigated captive stocks are similarly affected. The results highlight the potential shortcomings of using captive-bred individuals in supplementation and reintroduction projects, when the captive stocks have not been pedigreed and bred according to conservation principles. Keywords Hybrid Æ Captive Æ Supplementation Æ Reintroduction Æ Lesser white-fronted goose Anser erythropus Introduction Captive breeding programmes may be used to conserve species or populations that are incapable of surviving in their natural habitats, and such programmes act as an insurance against extinction in the wild (Frankham et al. 2002). In addition to the role they play in research and education, captive stocks sometimes provide material for the purposes of reintroduction when a species has become locally or globally extinct, or supplementation to enhance the survival when a population is considered e.g. too small and/or genetically impoverished (Kleiman et al. 1994). Captive-bred individuals that are to be released should be similar to the original population, in terms of both genetic relatedness and ecological characteristics in order to maximise the possibilities for a successful reintroduction or supplementation (IUCN 1995). Often a captive stock is founded when decline of the wild population is

2 198 Conserv Genet (2007) 8: severe and extinction in the near future is a realistic threat, in which case the origin of the individuals, and thus suitability for reintroduction or supplementation, is known to some extent. However, in some species, captive stocks may have been kept for economical or recreational reasons already prior to the decline of the population (e.g. O Brien et al. 1987; Negro et al. 2001; FitzSimmons et al. 2002; Brightsmith et al. 2005). These stocks are potentially valuable sources of genetic variability, but often the origin or the taxonomic identity of the individuals is unknown. The lesser white-fronted goose (Anser erythropus) is a globally threatened species, and the most vulnerable of the populations breeds in Fennoscandian Lapland (Madsen 1996; Lorentsen et al. 1999). The world estimate is currently individuals and the Fennoscandian population is estimated to be breeding pairs (Tolvanen et al. 2004). To enhance the survival of the Fennoscandian wild population, supplementation was carried out in Finland during the years using captive individuals with unknown origins (Markkola et al. 1999). The poor fitness of the released birds and the ambiguous genetic background of the captive stock contributed to the decision to stop the restocking project. In particular, it was discovered that some birds in the Finnish captive stock had the mitochondrial DNA (mtdna) of a congeneric species, the greater white-fronted goose Anser albifrons (Ruokonen et al. 2000b). In Sweden, captive-bred lesser white-fronted geese were released into the wild in (Andersson 2004). Because the captive stock in Finland originated from Swedish farms, it was considered likely that at least some of the Swedish birds would also prove to have a hybrid origin (Ruokonen et al. 2000b; Tegelström et al. 2001). In 2000, the release of captive birds was also stopped in Sweden until the genetic background of the stock could be investigated with the help of both mitochondrial and nuclear markers. At the start of the 1970s, there were few captive stocks of the lesser white-fronted goose in Sweden. The original breeding stock for reintroductions consisted of seven wild individuals collected from Swedish Lapland (Tegelström et al. 2001). During and after the late 1970s additional geese were obtained from collections in the Netherlands, Germany, England, Denmark and Sweden, and the number of breeding pairs was increased to (Tegelström et al. 2001). Some of these birds probably originated from the wild population, but most of them have a long history in captivity (Tegelström et al. 2001). As a consequence of repeated mixing of captive breeding groups to avoid inbreeding all European stocks are likely interrelated to some extent. However, in absence of detailed stud-books, the actual degree of relatedness is not known. The aim of this study was to assess the genetic status of the captive lesser white-fronted goose in Sweden, using biparentally and uniparentally inherited genetic markers. The occurrence of greater white-fronted goose genes in the captive stock of the lesser whitefronted goose may be explained by ancestral polymorphism, or by interspecific hybridisation in the wild or in captivity (Ruokonen et al. 2000b). Both species have been raised in captivity, often in mixed flocks, for decades (Delacour 1954) and are known to produce fertile hybrids under captive conditions (Gray 1958). Despite the ability to hybridise, they are considered to be true species (Cramp and Simmons 1977). In order to study the possible hybridisation or ancestral polymorphism of the species in the wild, we included greater and lesser white-fronted geese from the natural populations. Material and methods Blood or other tissue was collected from a total of 128 captive lesser white-fronted geese held in Swedish farms (Fig. 1). Two of the farms (Sweden 1 and Sweden 2, Fig. 1) were involved in the production of captive stock for reintroduction in Sweden. Our sample contained 52 adult individuals and 19 juveniles from these two farms. The other farms (Sweden 3 11, Fig. 1) are maintained by private organisations or breeders, and 45 birds that were not included in the breeding programme were sampled from these farms. An experimental reintroduction, in which an ultra-light aeroplane was used to guide a group of captive birds (originating from a Belgian farm) to a wintering area in Germany, was carried out in Sweden in 1999 (Tegelström et al. 2001). The 12 birds that returned to the reintroduction area were recaptured and sampled for the study. Ninety one wild greater white-fronted goose individuals were sampled in the Netherlands, Sweden, Ireland, Bulgaria, Kazakhstan and Russia. Three captive greater white-fronted geese from a Finnish zoo were also included in the study. An additional 50 wild lesser white-fronted geese from Fennoscandia, Bulgaria, Russia and Kazakhstan were used for the microsatellite analysis. The mtdna results of the wild lesser white-fronted geese have been published earlier (110 individuals; Ruokonen et al. 2004).

3 Conserv Genet (2007) 8: Unknown Sweden Hailuoto Sweden wild Sweden4 5 Netherlands 2 Sweden8 1 Sweden2 Sweden1 Denmark Sweden9 Sweden10 Sweden7 Sweden6 Sweden5 England Germany Belgium Fig. 1 Sampling scheme, presence of alien DNA and connections between the sampled farm stocks. Translocations of individuals between farms are shown by arrows. Sampling took place in farms with boxed names and the numbers indicate the number of sampled individuals. Some geese were known to originate from another farm than that where they were kept at the time of sampling. In the figure, those individuals are placed under the farm from which they originated (e.g. two geese originated from a farm in the Netherlands, but were sampled in the farm Sweden 1). The presence of alien DNA has been detected in all circled farms. Horizontal lines show the presence of alien mtdna and vertical lines indicate the presence of alien microsatellite alleles. The genetic composition of lesser whitefronted geese in the Hailuoto stock in Finland has been studied earlier using mtdna (Ruokonen et al. 2000b) Mitochondrial DNA Total DNA was extracted using the phenol chloroform method (Sambrook and Russell 2001). A 221 bp fragment of the mitochondrial DNA control region was amplified using primers L16642 and H411-AL (Ruokonen et al. 2000a). Double-stranded sequencing of PCR products was carried out using BigDye 3.0 and ABI PRISM 377 according to manufacturer s instructions and PCR primers were used for sequencing. Haplotype sequences have been submitted to the GenBank with accession numbers AF , AF and DQ The mitochondrial origin of the sequences was ascertained from four captive lesser white-fronted geese from the Belgian farm and three wild greater white-fronted goose individuals from Sweden by amplifying mtdna-enriched isolates with primers flanking the mtdna control region (see Ruokonen et al. 2000a for detailed methodology). The same verification method has been used previously for the lesser white-fronted goose (Ruokonen et al. 2000a). mtdna sequences were aligned manually. The program Modeltest 3.06 (Posada and Crandall 1998) was used to choose an appropriate DNA substitution model and, based on this, the HKY85 model (unequal base frequencies, unequal transition-transversion rates; Hasegawa et al. 1985) was used to estimate genetic distances. A neighbour-joining tree of the haplotypes was constructed with PAUP*4.0b10 (Swofford 2003) using 1000 bootstrap replicates. Bayesian analysis was performed with MrBayes3.0b4 (Huelsenbeck and Ronquist 2001). The MCMC search was run with four chains for 10 6 generations using default priors. Sampling frequency was set to 100 and the first 1000 trees were discarded as burn-in, yielding a total of 9001 trees for constructing the consensus tree. A nuclear copy of the mtdna control region (accession number AF159970) was used as an outgroup, and two greylag goose (Anser anser) sequences were used for comparison (accession numbers AF ). Microsatellites For the microsatellite study, total DNA was extracted using a salt procedure (Miller et al. 1988). A total of 25 microsatellite loci developed for several other goose and duck species (Fields and Scribner 1997; Buchholz et al. 1998; Cathey et al. 1998; Maak et al. 2000) were tested on both the lesser and greater white-fronted goose. Ten loci were found to be suitable for the present study (Table 1). The remaining 15 loci were

4 200 Conserv Genet (2007) 8: Table 1 Annealing temperature, number of alleles, allelic richness, observed and expected heterozygosity for the microsatellite loci used in the present study Locus Annealing temp. ( C) No. of alleles Allelic richness gwfg lwfg Captive gwfg lwfg Captive gwfg lwfg Captive H O H E H O H E H O H E Bcal2 a Bcal6 a Bcal7 a Bcal8 a Bcal9 a Hhil1 a Hhil3 a Aall1 b TTUCG5 c APH11 d Total/average a Buchholz et al. (1998), b Fields and Scribner (1997), c Cathey et al. (1998), d Maak et al. (2000) The observed heterozygosities marked in bold deviate significantly from Hardy Weinberg equilibrium. Gwfg = greater white-fronted goose; lwfg = lesser white-fronted goose; captive = captive stock of the lesser white-fronted goose excluded because they were monomorphic, had too low levels of variation or because of problems with amplification. Amplification was carried out in 10 ll reaction volumes containing 1 ll template DNA (approximate concentration 50 ng/ll), 1 PCR Buffer (10 mm Tris HCl, 50 mm KCl, 0.08% Nonidet P40), 1.5 mm MgCl 2, 200 lm dntp, 400 nm of each primer and 2.0 U of Taq Polymerase (MBI Fermentas). PCR cycling conditions were 94 C for 3 min followed by 35 cycles of 94 C for 35, 35 s of annealing at different temperatures for each locus (Table 1), 72 C for 35 s and a final cycle of 72 C for 5 min. PCR products were run on 6% polyacrylamide gels and visualised using standard silver staining (Bassam et al. 1991). The observed (H O ) and expected (H E ) heterozygosity, statistical tests of Hardy Weinberg expectations and linkage disequilibrium were calculated using Genepop 3.3 (Raymond and Rousset 1995) and allelic richness was calculated with FSTAT (Goudet 1995). To reduce type I error the sequential Bonferroni procedure of Holm (1979) was applied when testing for the non-random association between loci. Pairwise F ST values between sampling localities and captive populations of the lesser white-fronted goose were calculated with the software FSTAT (Goudet 1995). The captive stock was divided into two for the F ST analysis: (i) stock from the breeding programme and (ii) private farms. The software Structure 2.1 (Pritchard et al. 2000), originally designed to infer population structure for a single species, was used to test if the microsatellite variation supported the two goose species. The programme identifies genetic structure in a dataset without using prior information on where the individuals are sampled (or, in this case to which goose species each individual belongs). An assignment test was performed with the software GeneClass (Cornuet et al. 1999). Assignment probabilities to either the lesser or greater white-fronted goose reference group were calculated for each captive lesser white-fronted goose. The option of Bayesian likelihood method with simulation was used. For each reference group, GeneClass simulated a large population of multilocus genotypes according to the allele frequencies in the sample. Then GeneClass estimated the probability that a specific individual belonged to each of the reference groups. Results Eight different mtdna haplotypes were found among the 106 captive lesser white-fronted geese (Fig. 2). Three of the haplotypes (W1, W2 and E1) were also found in wild lesser white-fronted geese. A fourth haplotype (W10) clustered with the haplotypes of the lesser white-fronted goose, although it has not, so far, been found in the wild. Another three haplotypes (ALB17, ALB18, ALB19) in the captive stock clustered with the greater white-fronted goose haplotypes (Fig. 2). A total of 17 individuals (16%) of the captive lesser white-fronted geese had mtdna typical for the greater white-fronted goose. All three greater whitefronted goose haplotypes found in the captive population of the lesser white-fronted goose were found in wild greater white-fronted goose individuals sampled in Sweden (ALB17, ALB19), and the Netherlands (ALB17, ALB18, ALB19). Three captive greater

5 Conserv Genet (2007) 8: Fig. 2 Phylogenetic tree showing the relationships of mtdna haplotypes of the lesser white-fronted (W, E), greater white-fronted (ALB) and greylag (B, ANSRUB, ANSANS) goose. MtDNA haplotypes found in captive lesser white-fronted geese individuals are indicated with asterisks. A nuclear copy of the mtdna control region derived from a greylag goose (numt) was used as an outgroup. Support exceeding 50% in Bayesian and neighbour-joining replicates is shown in each branch ALB6 ALB10 ALB9 78/- ALB5 71/- ALB4 ALB11 52/- ALB8 ALB12 ALB20 ALB13 75/72 ALB19 ALB1 * 83/85 ALB3 ALB15 76/61 ALB16 greater whitefronted goose 100/92 ALB2 98/66 ALB17* ALB18 * ALB14 E2 -/52 E5 E1* E3 100/59 98/72 E4 E6 W10 * W9 W2 * W8 W3 W6 W1* W7 lesser whitefronted goose 100/99 81/63 B1 60/80 W5 * W4 ANSRUB ANSANS numt greylag goose 0.02 white-fronted geese from a Finnish zoo were also found to carry the mtdna haplotype ALB18. Stud book records show that the ancestors of these three geese originated from a Central European captive stock. Four of the captive birds originating from a Belgian farm were found to carry a mtdna haplotype that was not typical for either of the species (B1 in Fig. 2). A comparison to GenBank sequences indicated that this haplotype is most closely related to the greylag goose A. anser haplotypes (Fig. 2). None of the wild lesser white-fronted geese (N = 110, Ruokonen et al. 2004) or greater whitefronted geese (N = 91) was found to carry a mtdna haplotype typical for the other species or for the greylag goose. The two species were monophyletic with respect to mtdna. The average genetic distance between the greater and the lesser white-fronted goose haplotypes was Within the greater white-fronted goose the genetic distances between haplotypes averaged 0.022, and within the lesser white-fronted goose

6 202 Conserv Genet (2007) 8: In total, 88 microsatellite alleles were observed among the wild geese of the two species. The lesser white-fronted (N = 50) and the greater white-fronted (N = 87) goose had 56 and 87 alleles, respectively (Table 1). Private alleles were found for both the wild greater (32 alleles) and the wild lesser (1 allele) whitefronted geese. Thus, the species shared 55 (63%) alleles. Most private alleles were observed in low frequencies and only 11 alleles (all private for the greater white-fronted goose) had a frequency of 5%. The captive lesser white-fronted geese (N = 128) had altogether 58 alleles (Table 1). Ten of the greater whitefronted goose private alleles were observed among the captive lesser white-fronted geese, seven of these alleles had a frequency of 5% in the greater whitefronted goose. Three of the alleles in the captive stock were not found in either species in the wild. One of the three alleles was observed in seven captives originating from the Belgian farm, three of which also carried greylag goose mtdna, and in one captive goose from a Swedish private farm. After correcting for sample sizes, the average allelic richness (Table 1) was greatest for the greater white-fronted goose (7.97). Allelic richness for the wild material of lesser white-fronted goose (5.60) slightly exceeded that of the captive lesser white-fronted geese (5.05). The captive, wild lesser and greater white-fronted geese all showed significant deviations from Hardy Weinberg equilibrium at several loci (Table 1). Sampling for both species in the present study was performed over large geographic areas (in order to detect as much allelic variation as possible) or from captive stocks, which made it likely that the samples would not meet Hardy Weinberg expectations. All ten loci were in linkage equilibrium for both the wild lesser and greater white-fronted geese. Pairwise F ST values were found to be significant between all groupings where populations of wild lesser white-fronted geese were compared with the two groups of captive geese (Table 2). For example, genetic differentiation between the wild Fennoscandian population and the captive breeding stock in Sweden (F ST =0.0562) is four times larger than that between the wild populations of Fennoscandia and Russia (F ST =0.0120). The lowest pairwise F ST value was between the two groups of captive geese (Table 2). When all wild geese of both species were pooled, the programme Structure identified the two species. All individuals were assigned to the correct species, and the probability of the data was highest for two groups (data not shown). When the captive and wild geese were pooled for the analysis, all captive lesser whitefronted geese were assigned to the wild lesser whitefronted group except for three individuals. Two captive geese were assigned to the greater white-fronted goose group (probability > 97%), and the third individual was found to have a 40% probability of belonging to the greater white-fronted geese (data not shown). However, caution should be used when results from the Structure runs are interpreted, because the assumption of Hardy Weinberg equilibrium is violated in our dataset. Geneclass analysis revealed that captive lesser white-fronted geese in general had a higher probability of being assigned to the group of wild lesser whitefronted geese than to the greater white-fronted goose group (Fig. 3). However, for a majority of the individual captive lesser white-fronted goose (55%), the probabilities of being assigned to the wild lesser whitefronted goose group were nonetheless below 50%. Most of the captive lesser white-fronted goose individuals carrying alien mtdna or microsatellite alleles showed low assignment probabilities to the lesser white fronted goose group (Fig. 3). Combined results of mtdna and private greater white-fronted goose nuclear alleles are shown in Table 3. Out of all the studied captive lesser white-fronted geese, 13% had both mtdna and private alleles of the greater white-fronted goose. The proportion decreases to 4% if only private alleles that have frequencies of 5% in the wild greater white-fronted goose are considered. Four percent of the individuals had only mtdna and 19% of the individuals had only nuclear private alleles, from the greater white-fronted goose. In the breeding programme stock, 14% of the individuals had greater white-fronted goose mtdna Table 2 Pairwise F ST values between the wild populations and the captive lesser white-fronted geese Population Fennoscandia (N = 15) Russia (N = 20) Captives, breeding programme (N = 52) Russia (N = 20) Captive, breeding a a programme (N = 52) Captive, private farms (N = 45) a a Wild lesser white-fronted geese were grouped according to sampling locality. Captive geese were divided in two groups, (i) the breeding programme, geese that were involved in the production of captive stock for reintroduction in Sweden (Sweden 1 and 2) and (ii) the private farms, the remaining Swedish captive geese (Sweden 3 11). Bold F ST values are significant after Bonferroni correction ( a P < 0.05)

7 Conserv Genet (2007) 8: P (lesser white-fronted goose) G G 0.2 G G P (greater white-fronted goose) Fig. 3 Results from the GeneClass (Cornuet et al. 1999) assignment test of the captive lesser white-fronted geese. Each bar represents the probability for a captive individual of being assigned to either of the two species. The filled arrows indicate captive geese carrying mtdna from the greater white-fronted goose. The arrows denoted with G show individuals carrying mtdna from the greylag goose. Arrows with broken lines show captives that carried the W10 haplotype, not found among wild lesser white-fronted geese. Circles indicate captive individuals with private greater white-fronted goose microsatellite alleles, filled circles are alleles with a frequency of 5% in the wild population. Triangles show individuals that carried microsatellite alleles private for the captive geese from the Belgian farm, where the greylag mtdna was observed and all but one of them had also rare private alleles from the same species. Additionally, 20% of the individuals had private alleles (2% if considering only private alleles with a frequency of 5%) from the greater white-fronted goose, but carried the mtdna of the lesser white-fronted goose. A majority of the individuals had lesser white-fronted goose mtdna and no private alleles from the greater white-fronted goose. A positive association between the type of mtdna and presence or absence of greater whitefronted goose private nuclear alleles was found in most cases (Table 3). More often than by chance an individual had either greater white-fronted goose mtdna and private alleles or lesser white-fronted goose mtdna and no greater white-fronted goose private alleles. Discussion Existence of greater white-fronted goose mtdna in the captive lesser white-fronted goose stock might be explained by ancestral polymorphism of the species, or by hybridisation in the wild or captivity. The species are closely related (Ruokonen et al. 2000a), but our results from wild-sampled individuals indicate that the species are monophyletic with respect to their mtdna. Suspected morphological hybrids have been reported in the wild (Panov 1989). However, because mtdna of the greater white-fronted or greylag goose was not observed in the wild-sampled lesser white-fronted geese it can be concluded that hybridisation between greylag or greater white-fronted goose females and lesser white-fronted goose males is rare or has not led

8 204 Conserv Genet (2007) 8: to interspecific introgression in the wild. However, if only male mediated gene flow occurs between the species hybridisation will not be detected with mtdna markers. Incongruence between mitochondrial and nuclear markers has been observed in several taxa, for example savanna and forest elephant (Loxodonta africana and L. cyclotis; Roca et al. 2005), as well as white-tailed and mule deer (Odocoileus virginianus and O. hemianus; Ballinger et al. 1992). In the present study, the variation of the nuclear microsatellites can be used to separate the two species of white-fronted geese at the species level. Although differences in allele frequencies between the species do not rule out occasional hybridisation events, our data indicate that hybridisation does not appear to have been an important process in the evolutionary history of the lesser white-fronted goose. All three greater white-fronted goose mtdna haplotypes that were present in the captive lesser white-fronted geese were also found in wild greater white-fronted geese, which indicates that the hybridisation events are recent relative to the split between the two species (cf. Roca et al. 2005). One of Table 3 Captive lesser white-fronted geese in Sweden grouped according to mitochondrial DNA haplotypes and the presence (+) or absence ( ) of greater white-fronted goose private microsatellite alleles Private alleles gwfg mtdna lwfg mtdna (A) All private alleles All captive birds Breeding programme Private farms (B) Private alleles with a frequency 5% All captive birds Breeding programme NS Private farms Individuals were classified both on the basis of (a) all private alleles and (b) private alleles with a frequency 5% in the wild greater white-fronted geese. The total number of captive lesser white-fronted geese analysed with both markers was 102 (individuals with greylag goose mtdna were excluded here). Captive geese were divided in two groups, (i) the breeding programme, geese that were involved in the production of captive stock for reintroduction in Sweden (Sweden 1 and 2) and (ii) the private farms, the remaining Swedish captive geese (Sweden 3 11). The values are shown in percentages. Association between the mtdna haplotype and presence or absence of greater whitefronted goose private alleles was tested with Fisher s exact test ( 1 P < 0.001, 2 P < 0.01). Gwfg = greater white-fronted goose; lwfg = lesser white-fronted goose the haplotypes was also present in a captive stock of the greater white-fronted goose. The most likely explanation for the presence of alien mtdna in the captive lesser white-fronted geese is that interspecific hybridisation and introgression have occurred in captivity. The fact that four different alien mtdna haplotypes were found among the captives indicates that at least four independent hybridisation events between greater white-fronted or greylag goose females and lesser white-fronted goose males have taken place. Because most of hybridisation events take place between closely related species it is often difficult to find species-specific differences among the parental species. The low levels of microsatellite differentiation between the two species of white-fronted geese prevent the detailed investigation of levels of nuclear introgression in the captive stocks. Differences in microsatellite allele frequencies between the species allow the use of frequency based methods to discriminate between the two white-fronted geese at the species level, but fail to identify all individuals with hybrid ancestry. Both the frequency of alien mtdna and private microsatellite alleles among the captive geese indicate that there are many individuals with hybrid ancestry present in both the breeding programme and private captive stocks. A total of 36% of the captive lesser white-fronted goose stock was shown to be contaminated: 4% of the captive individuals had greater white-fronted goose mtdna, 19% carried private microsatellite alleles from the greater whitefronted goose and 13% of the individuals showed both greater white-fronted goose mtdna and microsatellite alleles. In reality, levels of genetic contamination may be much higher, but the similarities in nuclear alleles between the species make more detailed estimates impossible. In this study the sampled individuals were traced back to 11 farm stocks of which 6 are in Sweden and five elsewhere in Europe (Fig. 1). In eight of the stocks, either nuclear or mitochondrial genes (or both) from an alien species were detected (Fig. 1). Because captive lesser white-fronted geese have been exchanged among breeding units, in an attempt to avoid possible inbreeding, it is probable that genes from the greater white-fronted and greylag goose are widespread in captive stocks of the lesser white-fronted goose throughout Europe. Nevertheless, microsatellite variation provides us with information on the genetic differences between captive stocks in Sweden and the wild populations of the lesser white-fronted goose. Despite the fact that the original breeding stock, at least in part, consisted of wild Swedish birds, genetic differentiation between the

9 Conserv Genet (2007) 8: captive breeding programme stock and the wild Fennoscandian population was four times greater than the differentiation observed between the wild populations in Fennoscandia and western Russia. Significant differentiation between the captive stock and the wild populations can be explained in several ways. Founder effect combined with random genetic drift could create a situation where a captive stock differentiates rapidly from the wild populations. According to the Swedish breeding records, the size of the captive stock at the main farm (Sweden 1) mostly varied from only 10 to 15 breeding pairs (Tegelström et al. 2001). Differentiation between captive and wild populations is probably further reinforced by genetic drift in the declining wild populations, in particular the Fennoscandian population of the lesser white-fronted goose. Known hybridisation events between the captive lesser whitefronted goose and the greater white-fronted and the greylag goose might also explain the genetic differences between the captive and wild populations. We believe that a combination of the above scenarios explains the observed genetic differentiation between the captive and wild populations of the lesser white-fronted goose. Genetic diversity in captive stocks is often reduced as a result of founder effects and genetic drift acting on small populations (Fuerst and Maruyama 1986). In the present study, genetic variability (measured as allelic richness and expected heterozygosity) of the captive lesser white-fronted goose population was at the same level as in the wild population (Table 1). The high level of genetic variability of the total captive stock, despite the low effective population sizes of the individual stocks, can be at least partly explained by subdivision into breeding groups in separate farms and the recent introduction of new individuals into the farms. Further, hybridisation with the greater white-fronted and greylag goose, of which at least the former was shown to be genetically more diverse than the lesser white-fronted goose, may have increased the level of genetic variability in the captive stocks compared to the wild population. The proportion of non-specific alleles is potentially high in the captive lesser white-fronted goose stock. Of the eight female founders of the captive stock (shown by the eight different mtdna haplotypes), three founders were greater white-fronted goose and one greylag goose. How useful are the captive stocks of currently threatened species that have not been originally managed for conservation purposes? The present study demonstrates the need to use molecular methods to evaluate the genetic suitability of captive stocks. Because prezygotic constraints to interbreeding may be relaxed in ex-situ conditions, hybridisation and introgression are apparently common in captive stocks that have not been managed for conservation purposes. Previous, similar cases are known, for example those of the Asiatic lion (O Brien et al. 1987), yak and bison (Ward et al. 1999), Vietnamese sika deer (Thevenon et al. 2003), red-legged partridge (Negro et al. 2001) and Siamese crocodile (FitzSimmons et al. 2002). These examples demonstrate that unintentional or even intentional interbreeding of populations or species during captive propagation is rather common. Often, the recommended management implication has been to exclude or isolate either affected captive stocks or hybrid individuals from the breeding programme (O Brien et al. 1987; Ward et al. 1999; Negro et al. 2001; Olech and Perzanowski 2002; Thévenon et al. 2003). The results from the present study indicate that the existing captive stocks of the lesser white-fronted goose are unsuitable for further reintroduction or supplementation for three reasons. First, hybridisations with two other closely related species have taken place in captivity. Second, the proportion of individuals with hybrid ancestry is high, at least 36%. Third, the fact that most of the investigated captive lesser whitefronted goose stocks showed signs of hybrid ancestry, suggests that in fact most of the captive stocks, even elsewhere in Europe, may be similarly affected (Fig. 1). Following the recommendations of IUCN (1995) and bearing in mind that there still are many lesser white-fronted goose populations in the wild (cf. Allendorf et al. 2001), supplementation of the small Fennoscandian population, if considered necessary, with birds originating from the other wild populations seems preferable to further attempts to supplement these populations with the existing captive bred material. However, it is not known if the wild lesser white-fronted goose populations are ecologically exchangeable (cf. Crandall et al. 2000) and whether translocations may compromise adaptive differences between the populations (cf. Merilä and Crnokrak 2001; McKay and Latta 2002). Acknowledgements We thank Tomas Aarvak, Åke Andersson, Bart Ebbinge, Toni Eskelin, Jens Kristiansen, Kimmo Lahti, Gerard Müskens, P.-A. Ohlsson, Ingar J. Øien, Jari Peltomäki, Sami Timonen & Petteri Tolvanen for assistance with sampling, and Hannele Parkkinen & Kerstin Santesson for assistance in the lab. Honor C. Prentice is gratefully acknowledged for many discussions and support. We thank Tomas Aarvak, Åke Andersson, Juha Merilä and Honor C. Prentice for critical reading of the manuscript. This work was supported by grants from WWF Sweden, Swedish Environmental Protection Agency and Swedish Association for Hunting and Wildlife Management (HT, MR), Maj & Tor Nessling Foundation and

10 206 Conserv Genet (2007) 8: Academy of Finland (MR), and the Swedish Research Council for Environment, Agricultural Sciences and Spatial Planning (HT, A-CA). References Allendorf FW, Leary RF, Sruell P, Wenburg JK (2001) The problems with hybrids: setting conservation guidelines. TREE 16: Andersson A (2004) The reintroduction of the lesser whitefronted goose in Swedish Lapland a summary for In: Fennoscandian Lesser White-fronted Goose conservation project. Annual Report 2003 (eds. Aarvak T, Timonen S), pp WWF Finland Report 20 & NOF Rapportserie Report no Ballinger SW, Blankenship LH, Bickham JW, Carr SM (1992) Allozyme and mitochondrial DNA analysis of a hybrid zone between white-tailed deer and mule deer (Odocoileus) in West Texas. Biochem Genet 30:1 11 Bassam BJ, Caetano-Anolles G, Gresshoff PM (1991) Fast and sensitive silver staining of DNA in polyacrylamide gels. Anal Biochem 196:80 Brightsmith D, Hilburn J, del Campo A, Boyd J, Frisius M, Frisius R, Janik D, Guillen F (2005) The use of hand-raised psittacines for reintroduction: a case study of scarlet macaws (Ara macao) in Peru and Costa Rica. Biol Conserv 121: Buchholz WG, Pearce JM, Pierson BJ, Scribner KT (1998) Dinucleotide repeat polymorphisms in waterfowl (family Anatidae): characterization of a sex-linked (Z-specific) and 14 autosomal loci. Anim Genet 29: Cathey JC, DeWoody JA, Smith LM (1998) Microsatellite markers in Canada geese (Branta canadensis). J Hered 89: Cornuet JM, Piry S, Luikart G, Estoup A, Solignac M (1999) New methods employing multilocus genotypes to select or exclude populations as origins of individuals. Genetics 153: Cramp S, Simmons KEL (1977) Handbook of the birds of Europe, the Middle East and North Africa. The birds of the Western Palaearctic, vol 1. Oxford University Press, Oxford Crandall KA, Bininda-Emonds ORP, Mace GM, Wayne RK (2000) Considering evolutionary processes in conservation biology. TREE 15: Delacour J (1954) The waterfowl of the world. Country Life Ltd, London Fields RL, Scribner KT (1997) Isolation and characterization of novel waterfowl microsatellite loci: cross species comparisons and research applications. Mol Ecol 6: FitzSimmons NN, Buchan JC, Lam PV, Polet G, Hung TT, Thang NQ, Gratten J (2002) Identification of purebred Crocodylus siamensis for reintroduction in Vietnam. J Exp Zool 294: Frankham R, Ballou JD, Briscoe DA (2002) Introduction to conservation genetics. Cambridge University Press, Cambridge Fuerst PA, Maruyama T (1986) Considerations on the conservation of alleles and of genic heterozygosity in small managed populations. Zoo Biol 5: Goudet J (1995) FSTAT (version 1.2): a computer program to calculate F-statistics. J Hered 86: Gray AP (1958) Bird hybrids. A checklist with bibliography. Commonwealth Agricultural Bureaux, Bucks Hasegawa M, Kishino K, Yano T (1985) Dating the human-ape splitting by a molecular clock of mitochondrial DNA. J Mol Evol 22: Holm S (1979) A simple sequentially rejective multiple test procedure. Scand J Stat 6:56 70 Huelsenbeck JB, Ronquist F (2001) MrBayes: Bayesian inference of phylogenetic trees. Bioinformatics 17: IUCN (1995) IUCN/SSC guidelines for re-introductions. IUCN Council, Gland, Switzerland. ssc/pubs/policy/reinte.htm Kleiman DG, Stanley Price MR, Beck BB (1994) Criteria for reintroductions. In: Olney PJS, Mace GM, Feistner ATC (eds) Creative conservation: interactive management of wild and captive animals. Chapman & Hall, London, pp Lorentsen S-H, Øien IJ, Aarvak T, Markkola J, von Essen L, Farago S, Morozov V, Syroechkovsky E Jr, Tolvanen P (1999) Lesser white-fronted goose Anser erythropus. In: Goose populations of the Western Palearctic. A review of status and distribution (eds. Madsen J, Cracknell G, Fox AD), pp Wetlands International Publ. No. 48, Wetlands International, Wageningen, The Netherlands. National Environmental Research Institute, Rønde, Denmark Maak S, Neumann K, Von Lengerken G, Gatterman R (2000) First seven microsatellites developed for the Peking duck (Anas platyrhynchos). Anim Genet 31:233 Madsen J (1996) Lesser white-fronted goose. In: Heredia B, Rose L, Painter R (eds) Globally threatened birds in Europe Action Plans. Council of Europe Publishing, Strasbourg, pp Markkola J, Timonen S, Nieminen P (1999) The Finnish breeding and restocking programme of the lesser whitefronted goose: results and current situation in In: Fennoscandian Lesser White-fronted Goose onservation project. Annual Report 1999 (eds. Tolvanen P, Øien IJ, Ruokolainen K), pp WWF Finland Report 12 & NOF Rapportserie Report no McKay JK and Latta RG (2002) Adaptive population divergence: markers, QTL and traits. TREE 17: Merilä J, Crnokrak P (2001) Comparison of genetic differentiation at marker loci and quantitative traits. J Evol Genet 14: Miller SA, Dykes DD, Polesky HF (1988) A simple salting procedure for extracting DNA from human nucleated cells. Nucl Acids Res 16:215 Negro JJ, Torres MJ, Godoy JA (2001) RAPD analysis for detection and eradication of hybrid partridges (Alectoris rufa A. graeca) in Spain. Biol Conserv 98:19 24 O Brien SJ, Roslin P, Smith GL III, Wolfe R, Schaffer N, Heath E, Ott-Joslin J, Rawal PP, Bhattacharjee KK, Martenson JS (1987) Evidence for African Origins of founders of the Asiatic lion species survival plan. Zoo Biol 6: Olech W, Perzanowski K (2002) A genetic background for reintroduction program of the European bison (Bison bonasus) in the Carpathians. Biol Conserv 108: Panov EN (1989) Hybridisation and ethological isolation in birds. Nauka, Moscow (In Russian) Posada D, Crandall KA (1998) Modeltest: testing the model of DNA substitution. Bioinformatics 14: Pritchard JK, Stephens M, Donnelly P (2000) Inference of population structure using multilocus genotype data. Genetics 155: Raymond M, Rousset F (1995) GENEPOP (version 1.2): population genetics software for exact tests and ecumenicism. J Hered 86:

11 Conserv Genet (2007) 8: Roca AL, Georgiadis N, O Brien SJ (2005) Cytonuclear genomic disassociation in African elephant species. Nature Genet 37: Ruokonen M, Kvist L, Aarvak T, Markkola J, Morozov VV, Øien IJ, Syroechkovsky EE Jr, Tolvanen P, Lumme J (2004) Population genetic structure and conservation of the lesser white-fronted goose Anser erythropus. Conserv Genet 5: Ruokonen M, Kvist L, Lumme J (2000a) Close relatedness between mitochondrial DNA from seven Anser goose species. J Evol Biol 13: Ruokonen M, Kvist L, Tegelström H, Lumme J (2000b) Hybrids, captive breeding and restocking of the Fennoscandian Lesser White-fronted Goose (Anser erythropus). Conserv Genet 1: Sambrook J, Russell DW (2001) Molecular cloning: a laboratory manual, 3rd edn. Cold Spring Harbor Laboratory Press, New York Swofford DL (2003) PAUP*. Phylogenetic Analysis Using Parsimony (*and Other Methods). Version 4.b10. Sinauer Associates, Sunderland, Mqassachusetts Tegelström H, Ruokonen M, Löfgren S (2001) The genetic status of the captive lesser white-fronted geese used for breeding and reintroduction in Sweden and Finland. In: Fennoscandian Lesser White-fronted Goose conservation project. Annual Report 2000 (eds. Tolvanen P, Øien IJ, Ruokolainen K), pp WWF Finland Report 13 & NOF Rapportserie Report no uploads/pdf/ar%2000%20genetic.pdf Thevenon S, Bonnet A, Claro F, Maillard JC (2003) Genetic diversity of captive populations: the Vietnamese sika deer (Cervus nippon pseudaxis) in zoological parks. Zoo Biol 22: Tolvanen P, Aarvak T, Øien IJ, Timonen S (2004) Introduction. In: Fennoscandian Lesser White-fronted Goose conservation project. Annual Report (eds. Aarvak T, Timonen S), pp 5 8. WWF Finland Report 20 & NOF Rapportserie Report no www/uploads/pdf/lesser_white_fronted_goose_report01-03_part1.pdf Ward TJ, Bielawski JP, Davis SK, Templeton JW, Derr JN (1999) Identification of domestic cattle hybrids in wild cattle and bison species: a general approach using mtdna markers and the parametric bootstrap. Anim Conserv 2:51 57