Sourcing Eurasian beaver Castor fiber stock for reintroductions in Great Britain and Western Europe



  • 1Eurasian beavers Castor fiber, formerly threatened with extinction, have been widely reintroduced since the 1920s. Reintroductions and studies of possible reintroductions are continuing.
  • 2The International Union for Conservation of Nature (IUCN) guidelines for reintroductions state that ‘the source population should ideally be closely related genetically to the original native stock’.
  • 3Palaeoecological studies suggest that the species survived the last Ice Age in two refugia: in the west in Iberia and Southern France and in the east in the Black Sea region. The post-Ice Age population of Western Europe, including Great Britain, recolonized from the western refugium. Recent mitochondrial deoxyribonucleic acid studies strongly support this view, and extant beaver populations are clearly divided into eastern and western evolutionarily significant units (ESUs).
  • 4The western ESU is composed of three stocks which survived the 19th and early 20th century as very small, isolated populations. They are very closely related to each other. Each is genetically depauperate, apparently as a result of genetic drift at low population levels.
  • 5There is evidence of inbreeding depression and of phenotypic abnormalities in beaver populations descended from unmixed stocks.
  • 6The evidence suggests three coherent management options for sourcing reintroduction stock for Great Britain and for unoccupied areas of western continental Europe. These are (i) use animals from a single western ESU stock; (ii) intentionally mix animals from two or all three of the surviving western ESU stocks; (iii) make an informed exception to the IUCN guidelines and reintroduce animals of mixed eastern and western ESU provenance.
  • 7These options are discussed with regard to IUCN guidelines, conservation biology and animal welfare considerations. It would be advantageous if a common policy on the origin of reintroduction stock were agreed by the national agencies responsible.


The Eurasian beaver Castor fiber, threatened with extinction in the 19th and early 20th centuries due to overhunting for fur and castoreum, has been widely reintroduced across Europe and the species is now out of conservation danger (Halley & Rosell 2002). The first known reintroduction, in 1922, was to Sweden using animals from Telemark, Norway (Hartman 1995). Since then, at least 203 translocations to distinct sites have been recorded outside the former Soviet Union, within which translocations also took place on a vast scale. Reintroductions are continuing throughout Europe; England, Wales, Portugal, Italy and the countries of the southern Balkans (Bulgaria, Greece, Macedonia, Albania, Montenegro, Kosovo and Turkey) are the only areas of the beaver's natural range in Europe in which reintroductions have not yet taken place (Fig. 1; Halley & Rosell 2002, 2003 for reviews).

Figure 1.

Distribution in 2010 of Eurasian beavers Castor fiber and North American beavers Castor canadensis in Europe. Dark grey = Eurasian beaver distribution; mid grey = North American beaver distribution, light grey = no beavers present. ‘F’ = feasibility study published or in preparation. Black = refugia where Eurasian beavers never became extinct. 1: Telemark, Norway; 2: Elbe, Germany; 3: Rhône, France; 4: Pripet, Belarus/Ukraine/Russia; 5. Voronezh (Russia). Three other refugia in Siberia (Russia) and Xinjiang China/Mongolia are not shown. Updated from Halley and Rosell (2003).

A trial reintroduction to Scotland commenced in May 2009 (S. Jones, Scottish Wildlife Trust, personal communication). Feasibility studies have recently been published for England and for Wales (Gurnell et al. 2008, Halley et al. 2009), both of which conclude that reintroduction is biologically feasible. The provenance of potential reintroduction stock has been a topic of some discussion, and in this review, I discuss the issue with regard to International Union for Conservation of Nature (IUCN) criteria, morphometric studies, recent deoxyribonucleic acid (DNA) evidence, genetic diversity and animal welfare perspectives. I suggest that these considerations may provide a case for modifying the source stock strategy for reintroductions in Great Britain. Although the focus in this paper is on Great Britain, the rationale and conclusions apply in principle to much of Western Europe (Great Britain, Scandinavia, Germany outwith the Danube watershed, Switzerland outwith the Danube and Po watersheds, the Benelux countries, France and Iberia.). The main options and their rationales are described, and it is suggested that a common policy on source stock among the various agencies responsible for regulating any future reintroductions would be desirable.


The IUCN guidelines for reintroductions (Anonymous 1998) state that ‘the source population should ideally be closely related genetically to the original native stock and show similar ecological characteristics (morphology, physiology, behaviour, habitat preference) to the original subpopulation.’

In the case of Eurasian beavers, there seems to be little or no variation in physiology, behaviour or habitat preference, and only very small variations in morphology, between the extant populations (Heideke 1986, Rosell & Pedersen 1999, Halley & Rosell 2002, Rosell et al. 2005). The problem therefore mainly consists of assessing how to obtain a source population as close as possible genetically to the original subpopulation.


The status of the various Eurasian beaver subspecies, and even the number of acceptable subspecies, was formerly far from clear (Heideke 1986, Saveljev 1997). The most usual classifications divided the species into eight subspecies, largely on the basis of very small differences in cranial morphology (Heideke 1986), one for each of the eight 19th–20th century refugia where the species never became extinct. These were C.f. fiber, from the Telemark refugium (several small watersheds), Norway; C.f. albicus, from the Elbe, Germany; C.f. galliae, from the Rhône, France; C.f. belorussicus (belarusicus), from the Pripet marshes in Belarus/Ukraine/Russia; C.f. osteuropaeus (orientoeuropaeus), from the Voronezh region, Russia; C.f. pohlei, surviving on south-west tributaries of the Ob; C.f. birulai in South-west Mongolia and adjacent parts of Xinjian; and C.f. tuvinicus, on the upper Yenesei (see Table 1 for summary).

Table 1.  Summary of the extant stocks of Eurasian beaver Castor fiber, named for the refugia in which they survived, with former subspecies (Heidecke 1986, Saveljev 1997) and modern evolutionary significant unit (Durka et al. 2005) classifications
StockFormer subspecies classificationESUPopulation at minimum (estimated)Population estimate source
TelemarkC.f. fiberWestern60–120Collet 1897
ElbeC.f. albicusWestern200Heidecke & Hörig 1986
RhôneC.f. galliaeWestern30Richard 1985
PripetC.f. belorussicusEastern<300Lavrov & Lavrov 1986
VoronezhC.f. osteuropaeusEastern70Lavrov & Lavrov 1986
ObC.f. pohleiEastern300Lavrov & Lavrov 1986
Mongolia/XinjiangC.f. birulaiEastern<100–150Lavrov & Hao-Tsuan 1961
YeneseiC.f. tuvinicusEastern30–40Lavrov & Lavrov 1986

All of these subspecies, except belorussicus and pohlei, derive from tiny populations of fewer than 200 animals (Table 1), and it has been argued that the small differences in average morphology (mainly of the skull) that were the basis of the current classification might plausibly be ascribed to recent founder effects and genetic drift, and to local adaptations to prevailing conditions (Halley & Rosell 2002).

Reintroduced populations in Western Europe are mainly of single stock origin, with the very important exception of the Bavarian population and the many reintroduced populations derived from it. These are a mix of Telemark, Pripet, Rhône and probably Voronezh stock, with Pripet stock predominating and so of both mixed stock and mixed evolutionarily significant unit (ESU; see below) origin. Central and eastern European (former Soviet bloc) reintroduced populations are mainly mixed Pripet and Voronezh stock, and so pure eastern ESU, origin. See Halley & Rosell 2002 for a detailed discussion of the origins of the various reintroduced populations.


Kitchener and Lynch (2000) analysed subfossil remains of the species from Great Britain and concluded that overall morphology of skull and mandibles was most similar to beavers from the Telemark refugium in Norway. However, as one might expect, the detailed picture is complex. To simplify somewhat, British (Scottish and English; no specimens were available from Wales) beavers were most similar to Telemark beavers in mandible variation while English skull specimens lay between Elbe and Telemark beavers in average morphology, slightly closer on average to Telemark beavers. Only three Scottish skull specimens were available for study, and of these two lay, in terms of morphology, halfway between Elbe and Rhône beavers, and furthest from Telemark beavers, while the third was clearly most similar to Rhône animals.

The task Kitchener and Lynch (2000) were asked by Scottish Natural Heritage to perform was ‘to identify which extant population of Eurasian beavers (morphologically) resembles the extinct British population the most’, and their answer from the available data is firmly grounded. The underlying premise is that unmixed stock from a single refugium should be used for reintroduction. If, however, the data were assessed in terms of ‘% similarity’ rather than simply to identify the single stock most similar to British subfossils, one might conclude that, overall, the original British stock was most similar to Telemark, Elbe and Rhône beavers in that order, but not identical to any of them; and with individual indicators not all pointing in a uniform direction.

It must also be recalled, as Kitchener and Lynch (2000) point out, that the extant populations of beavers all descend from very small groups: c. 60–120 in Telemark (Collet 1897), 30 in the Rhône (Richard 1985) and <200 in Elbe (Heideke & Hörig 1986), of which perhaps 25% were breeding females. Genetic drift (the random loss of genetic diversity which occurs in small populations) may therefore have been a significant feature in determining average morphometric characteristics. Cranial and mandibular measurements may plausibly reflect this rather than that Telemark beavers are the most closely related surviving population to the extinct British stock.

Further, morphological similarity may not be congruent with genetic similarity. Beavers reintroduced to a taiga environment from the Voronezh refugium, in the temperate forest zone, showed distinct morphological and physiological changes within 40 years of the first release (Solovjov 1991, cited in Ulevicius & Paulauskas 2003); in Lithuania, skulls of beavers of Pripet refugium provenance (determined both from historical records and phenetic and biochemical measures) were significantly larger than those of the parent stock in 11 of 15 parameters (Ulevicius & Paulauskas 2003) after only a few decades, though the reasons for this are unclear. Similar rapid morphological changes have been recorded in a genetically homogenous muskrat Ondatra zibethicus population introduced in Russia (Vasil’ev et al. 1999).


Temperate forests suitable for beaver survival remained on the Rhône and Gironde catchments in Southern France (and in much of Iberia) throughout the most severe phase of the last Ice Age, and it is from there that recolonization northwards presumably occurred as the climate warmed (Coles 2006; Fig. 2); the other main refuge on the lower Danube and Black Sea area is much further away and is separated from Western Europe by the Alps and Carpathians, in which Ice Age conditions lingered longer than on the surrounding lower ground. Modern evidence on patterns of spread indicate that beavers would have easily been able to keep pace with the expansion of suitable vegetation zones northwards (Hartman 1995, Halley & Rosell 2002), and so would have entered Great Britain as soon as suitable vegetation became available. Beavers are known to have recolonized Great Britain in the Windermere interstadial, 14700–12700 Before Present (BP), at which time the Thames was a tributary of the Rhine. In the subsequent Loch Lomond stadial, some woodland remained in the South of England and presumably also along the lower Rhine watershed in what is now the English Channel; beavers probably survived the stadial there (Coles 2006). In modern Norway, beavers live in the montane willow scrub zone at 900m above sea level on the extreme tree line (personal observation, Rosell & Pedersen 1999), where iceover often lasts for 7 months of the year.

Figure 2.

Sites of probable Ice Age maximum beaver refugia on the Gironde (1), Rhône (2) and lower Danube (3). Courses of major rivers (grey lines), locations of mountain massifs (dark grey), and modern coastlines (grey lines) are shown. Areas of current seabed which were land in the late Devensian (mid grey), and early Holocene (light grey) are indicated. Figure from Coles (2006).

The final post-Ice Age change to warmer conditions in around 11500 BP was followed within a few centuries at most by species typical of temperate woodland, such as roe deer Capreolus capreolus and aurochs Bos primigenius (Coles 2006). It can be assumed that beavers expanded through what is now Great Britain as soon as conditions were suitable; the earliest dated beaver remains in Great Britain from the post-Ice Age period are from Starr Carr in Yorkshire, 10900–10700 BP (Coles 2006).

At this time, although the Rhine and Thames had become separated, the rivers of modern East Anglia flowed east across the ‘Doggerland’ plain into either the Rhine, or a north-flowing river sharing a long watershed divide with the Rhine in low-lying terrain. This contact was maintained until the final isolation of Britain about 3000 years later. During this period, the British beaver population was therefore part of a wider continental stock. As the dominant vegetation from the early Holocene was woodland, increasingly deciduous, the beaver population of the Doggerland river systems must have been very large. ‘Founder effects’ of colonization by only a few subsequently genetically isolated individuals seem very unlikely.

More widely in Western Europe, there are no barriers to expansion from the western refugia northwards in the continuously low-lying terrain, and recolonization would have kept pace with suitable climatic and habitat conditions to the Elbe, and into Scandinavia across the land bridge between modern Denmark and Sweden and around the then freshwater Baltic Lake, which persisted until c .7500 BP. Italy may have supported a third refugium through and beyond Ice Age times, or have been recolonized from either or both of the known Ice Age refugia.


Significant new evidence has recently become available from DNA analysis, which strongly supports the inferences on patterns of post-Ice Age spread derived from climatology, archaeology and ecology presented above. In 2005, an international group headed by Walter Durka (Durka et al. 2005), published a seminal paper on Eurasian beaver mitochondrial DNA (mDNA) collected from animals in all of the 19th/20th century refugia. The results indicate that the traditional subspecies classification is not tenable, and that the species can be divided into two ESUs: ‘western’ and ‘eastern’. Even in the more genetically differentiated eastern refugia, ‘the degree of divergence detected is below values usually reported for subspecies differentiation’ (DuCroz et al. 2005). Beavers from the Telemark, Elbe and Rhône refugia belong to the much less differentiated western ESU, which is particularly tightly clustered, forming a ‘clear monophyletic group’ (Figs 3 and 4; Table 1; Durka et al. 2005). In short, by modern mDNA analysis criteria, Elbe, Telemark and Rhône beavers belong to the same subspecies. Recent work on ancient DNA confirms this pattern, with the addition that there are indications that the former Danube population may have been a third ESU, now extinct (Horn, personal communication). Work on beaver DNA aimed at identifying individuals, parentage, population structure, diversity and gene flow is currently underway at the Royal Zoological Society of Scotland (Ogden, personal communication); the data obtained may also have relevance to studies of differences between populations.

Figure 3.

(a) Strict consensus of 16 most parsimonious trees constructed from 16 Eurasian beaver Castor fiber mtDNA d-loop haplotypes; (b) neighbour joining tree constructed from the matrix of pairwise p distances. In both cases, Telemark-refuge beavers (fi), Rhône beavers (ga) and Elbe beavers (al) form a clear monophyletic group (bi = Mongolia/Xinjiang refuge stock; po = Ob; tu = Yenesei; in = hybrid Pripet/Voronezh stock; c. = North American beaver Castor canadensis). Figure from Durka et al. (2005).

Figure 4.

Median-joining network for Eurasian beaver Castor fiber mtDNA control region haplotypes. Circle areas are proportional to haplotype frequencies. The median-joining network shows clearly that the haplotypes from the three western refugia stocks (fi, ga, al) are close to each other, whereas the haplotypes from the eastern populations form four clusters connected by relatively long branches. See Fig. 3 for nomenclature. Figure from Durka et al. (2005).

Intrapopulation mDNA haplotype variation in each western population is either zero (Telemark, Rhône) or very low (Elbe, two haplotypes, one very rare), apparently as a result of the severe, and compared to that of eastern ESU populations long-lasting, population bottleneck in each refugium in the 19th and early 20th centuries (Durka et al. 2005). None of the haplotypes was shared between populations. This supports the interpretations that the earlier classifications were influenced by genetic drift, and that the three extant stocks represent fragments of an originally much more variable genome. Ancient DNA investigations currently underway confirm that many haplotypes formerly present in western beavers are now extinct (Horn, personal communication).

Telemark-descended beavers are, in addition, known to have little variation in nuclear DNA fingerprinting profiles, in contrast to the considerable variability found in beavers from a Russian population (of undefined parent stock, but from an area drained by tributaries of the Voronezh river, so most likely to be of Voronezh stock) and in mammal populations generally (Ellegren et al. 1993). A standard measure of nuclear DNA diversity is average percent difference. In Telemark stock beavers, this value is between 10.8 and 23.6 depending on the probe and enzyme combination used; for Russian beavers, it is 47.2–55.3, a higher score equalling greater diversity. Ellegren et al. (1993) conclude that ‘the Scandinavian beaver population is depauperate of genetic variation’ and that ‘it is an open question whether the loss of genetic diversity in the Scandinavian population may influence its capability to cope with future environmental changes’, although its ability to expand in numbers and range under present (Scandinavian) conditions is evident.

Details of the differentiation between eastern and western ESUs strongly suggest differentiation during the last Ice Age in two distinct geographical ranges. This east–west split between lineages is paralleled in many other species, including mammals, birds, amphibians, freshwater fishes and insects, and appears to reflect the existence of two large Ice Age refugia, probably one in Iberia and Southern France and another in the Black Sea region from which populations spread post-Ice Age (Durka et al. 2005). Data from palaeoclimatology support this interpretation (Coles 2006).

DNA evidence therefore indicates that for reintroductions to Western Europe including Great Britain, only beavers from the western ESU should be used if IUCN criteria are to be adhered to. This would imply, for example, avoiding the use of beavers from the recently established beaver population in Bavaria as source stock. Beavers from Bavaria are sourced from a number of refugia and mainly from the eastern ESU. On the other hand, differences between the three western stocks are tiny, so that each of them would be ‘closely related genetically’ to the original stock of any part of Western Europe.


The severe bottlenecks and evidence of low genetic diversity in all three of the stocks descended from western refugia raise two related questions of relevance to reintroductions: is there any evidence of inbreeding depression? And do unmixed stocks have increased incidence of genetic abnormalities? The latter is relevant not only from a conservation but also an animal welfare perspective.

The Scandinavian beaver population, descended entirely from Telemark beavers, has increased from c. 100 (Bevanger 1995) to over 170000 (Halley & Rosell 2003) in less than 100 years. Satisfactory population growth of reintroduced unmixed stocks has also been recorded for Rhône beavers (Richard 1986) and Elbe beavers (Elmeros et al. 2003). Failure of reintroductions caused by inbreeding depression seems therefore very unlikely.

On the other hand, data from Russia do strongly suggest that a significant degree of inbreeding depression nevertheless occurs in unmixed stocks. Beavers from the Pripet marsh and Voronezh refugia are each more diverse genetically than western stocks (DuCroz et al. 2005, Durka et al. 2005). Nevertheless, reintroduced populations of beavers in Russia, of mixed Pripet and Voronezh stock origin, show a considerably higher average litter size, 3.4, than stocks from only one of these refugia, whether living in the refugium (2.8 and 2.9) or reintroduced elsewhere (2.9 and 3.0; Saveljev & Milishnikov 2002). Unmixed stocks from western refugia show a similar pattern: beavers from the Elbe have an average litter of 1.9 kits (Nolet et al. 1994), while Swedish (Telemark-descended) beavers have on average 2.5 kits (Rosell & Pedersen 1999). It is likely therefore that mixing of beavers from two or more of the refugia of the western ESU would similarly increase average fecundity.

Genetic bottlenecks and loss of genetic diversity also appear to be related to phenotypic abnormalities. 8.7% of Elbe beavers have jaw abnormalities, apparently of genetic origin (Piechocki 1977). Similarly, the introduced North American beaver C. canadensis population of the Obor and Ussuri rivers in the Russian Far East derives from a repeatedly bottlenecked population. Dental abnormalities (Saveljev 2001, Saveljev & Milishnikov 2002) are extremely common: 15% of beavers have fewer, and 7% have more than the normal number of teeth, and 30% have tooth caries and gum abnormalities. Defects of the jaws and teeth are relatively easy to detect; it is likely that other, less evident, abnormalities also occur with heightened frequency in unmixed stocks.

This is of significance from an animal welfare, as well as from a conservation perspective. Reintroducing an unmixed stock of Elbe beavers certainly, and by inference Rhône and Telemark beavers probably, will result in a descendant stock more prone to phenotypic abnormalities among kits which are born, than if the stock were intentionally mixed from two or all three of the western refugia or were of mixed east and west ESU origin. This may impact mortality rates and longevity of adults, assuming the phenotypic abnormalities which occur are deleterious. In addition, fecundity will probably be lower. Greater genetic diversity in the founder populations would also provide more variation, on which natural selection can act to produce a stock adapted to local conditions.


The data indicate clearly that reintroductions in Western Europe, including in Great Britain, must be carried out using stock from the western ESU, if IUCN guidelines are to be adhered to; the case for a management decision to make an informed exception to the guidelines is presented below. Beyond this, the picture is less clear. Morphologically, it has been established that the former British population in particular differed from all three of the extant stocks, though overall was most similar to Telemark beavers. Genetically, nothing is currently known, though it seems certain that the former British and western European populations would have been much more diverse genetically than any of the current western stocks, as the diversity of Pripet and Voronezh stocks, which had less severe bottlenecks in numbers and/or time, would suggest.

All three of the extant western stocks are very closely related and genetically depauperate. The genetic history is of a single source stock spreading north from the western Ice Age refuge. Genetic interchange occurred until c. 7500 BP for Scandinavia (and so the population from which the Telemark refugium stock derives), when the Scandinavian peninsula separated from NW Europe; and until c. 8000 BP for Great Britain, when the English Channel formed.

The severe population bottlenecks experienced by all three western stocks have led to considerable reductions in their genetic diversity. It seems very likely therefore that each of these stocks contains genetic material originally much more widespread within the ESU, and lost in other stocks due to genetic drift. It is likely that most of the differences observable between stocks are due to this. Composing a founder population of a mix of western ESU stocks would be more likely therefore to result in a population resembling in genetic diversity the original genetic material found in, e.g. Great Britain, than reintroducing animals from any one stock would. It is certain that reintroducing a single stock would be to reintroduce only a fragment of the genetic variability formerly present.

In addition to genetic criteria, population viability and animal welfare should be taken into account. While it is clear that inbreeding depression and the incidence of genetic defects are by no means sufficient in this species to prevent the development of viable populations from unmixed stocks, it seems also clear that a degree of inbreeding depression and a heightened incidence of phenotypic abnormalities does occur in unmixed stocks. Mixing of the fragments of the original western ESU diversity represented by the three extant stocks or reintroducing beavers from hybrid west and east ESU stock would make reintroduced populations more robust, perhaps particularly with respect to epidemic disease; lower the incidence of genetic disorders in the reintroduced population; and provide greater genetic diversity from which natural selection could produce locally adapted subpopulations. Against this, Durka et al. (2005) consider that the ‘geographically nearest form should be used for reintroductions unless the genetic identity of the original population is known’, in order to preserve the identity of stocks, at least until the detailed genetic history of beavers is better known. While a valid consideration for continental Eurasia, where genetic introgression between stocks is likely in the longer term, this would not apply to reintroductions to Great Britain, and it does not take into account wider issues of population robustness and animal welfare.

In summary, there are three options available for source stocks for reintroductions:

  • 1Unmixed stock from one of the western refugia. Alternative options (below) would be irrevocable short of removing all reintroduced beavers and starting again; the conservative case for an unmixed stock is that it would be possible to augment its diversity later if this appeared advisable, thus leaving options open pending further evidence, while nevertheless not hindering the practical implementation of any planned reintroduction or affecting the ecological role of the species. Beavers of unmixed Telemark stock are the source of the current Scottish trial population (S. Jones, personal communication).
  • 2A mixed western ESU founder stock. There is a case, from a detailed consideration of morphometric data; new genetic evidence on the relatedness and genetically depauperate nature of the populations descended from the various refugia; considerations of genetic diversity; population robustness to shocks such as disease outbreaks; and animal welfare, all while retaining adherence to IUCN guidelines for mixing surviving genetic material from the western ESU. This would be accomplished in practice by reintroducing animals from two or all three of the western refugia as founder populations. It is probable that such a mixed stock would resemble more closely the genetic diversity of the original stock than reintroduced populations derived from a single, genetically depauperate, western refugium. Greater genetic diversity in the founder population would also allow greater scope for natural selection to adapt the reintroduced population to local conditions.
  • 3A mixed eastern and western ESU founder stock. The third option is to make an informed exception to the IUCN guidelines in this case, and intentionally reintroduce beavers of mixed east and west ESU origin, as suggested by Saveliev and Schwab (2006). This would provide maximum genetic diversity, arguably including some material lost from all three of the western stocks by genetic drift (but also undoubtedly some material not present in the original western ESU), and so perhaps greater flexibility to current conditions. Mixed ESU populations would undoubtedly maximize levels of heterozygosity in individual animals (most phenotypic abnormalities result from the expression of recessive genes in individuals homozygous for that gene). Mixing ESU genetic material would have no impact on the wider ecological role of the species, as behaviour appears identical regardless of origin. For Great Britain, genetic introgression with unmixed stocks in continental Europe is not a problem (at least until the next Ice Age). And as a practical point, beavers of Bavarian mixed ESU origin are the easiest and cheapest to obtain and are the source of most of the various captive populations of beaver currently present in Britain.

The fundamental choice, or perhaps trade-off, reduces therefore mainly to recognizing that our knowledge remains incomplete and keeping options open by restricting the stock(s) used, vs. the animal welfare implications of reintroducing inbred animals likely to suffer from a heightened level of genetic abnormalities than would a stock of mixed western or mixed western and eastern ESU origin. On the balance of probabilities, adherence to IUCN guidelines and restoration (as far as possible) of the genetic variation present in the original British/west European stock would be maximized by options 2, 1 and 3 in that order; animal welfare considerations would be best addressed by options 3, 2 and 1; the ability to revise stock origin decisions retrospectively based on new evidence by options 1, 2 and 3. Mixing stocks would raise the issue of the appropriate balance of founder stock. For option 2, the evidence indicates that each stock has very little diversity, so a roughly equal split would approximate an even diversity in the founder population. For option 3, considerations as to the best eastern/western ESU balance come into play. Pripet stock is the largest founder of the Bavarian population, but this is likely to vary geographically since the species was released at a number of sites in the area, and the founder stock makeup varied between them (G. Schwab, personal communication).

It has recently become clear that a population of wild-living beavers is established on the Tay watershed in Scotland and is spread quite widely. Young of the year have been recorded on film (P. Ramsay, personal communication). The number of known sites at which activity has persisted over two or more years, strongly suggesting an established pair, is c. 7 (unpublished data); the true figure may be greater as no systematic survey has been performed. Data from European reintroductions (Halley & Rosell 2002) indicates that in the absence of strong intervention, a population of this size normally survives and increases to the biological limits of the habitat, which would be several hundred individuals on the Tay. The population may originate from escapes from any of several sites where captive animals have been held near the river (sightings date back to 2001, H. Chalmers, personal communication). These are known to be mainly of Bavarian origin, plus a few of Telemark stock. However, the possibility of an illegal release at some time in the last decade cannot be excluded. If these animals are mixed ESU (which should be checked), this would be equivalent in its effects to an intentional release under option 3.

Genetic analysis of subfossil remains from Europe is currently underway, though so far permission to analyse material from Great Britain has been difficult to obtain (Horn, personal communication). Given the very significant addition to our understanding of the relationship of the former British population to extant stocks such an analysis would represent, analysis of these remains should be a priority. Pending the results of such an investigation, the argument for maintaining a policy of reintroducing unmixed stock in the interim is strengthened.

Apart from the current Scottish trial reintroduction, reintroductions to Wales and England are under active consideration. While it is unlikely that beavers from any of these possible future reintroductions would be in genetic contact except perhaps in the longer term, it would nevertheless save much complication and potential delay in the future if a thought-through common policy on sources of reintroduction stock were agreed by the various national agencies responsible.


Andrew Kitchener, National Museums of Scotland and University of Edinburgh, Frank Rosell of Telemark University College, and three anonymous reviewers contributed valuable constructive criticism to earlier drafts of this paper.

Editor: KH