Conservation and Society

: 2008  |  Volume : 6  |  Issue : 2  |  Page : 190--194

Forage Preferences of the European Beaver Castor fiber: Implications for Re-introduction

MJ O'Connell1, SR Atkinson2, K Gamez3, SP Pickering4, JS Dutton2,  
1 University of Bath, Department of Biology and Biochemistry, Claverton Down, Bath, BA2 7AY, United Kingdom
2 University of the West of England, Hartpury College, Hartpury, Gloucestershire, GL9 3BE, United Kingdom
3 Lower Mill Estate, Somerford Keynes, Gloucestershire, GL7 6BG, United Kingdom
4 Cotswold Water Park Society, Spratsgate Lane, Shorncote, Gloucestershire, GL7 6DF, United Kingdom

Correspondence Address:
M J O«SQ»Connell
University of Bath, Department of Biology and Biochemistry, Claverton Down, Bath, BA2 7AY
United Kingdom


By the beginning of the twentieth century, hunting and land use change had reduced the European beaver Castor fiber to a relict population of no more than 1200 individuals. In some European states, re­introduction has successfully established viable populations, whilst other schemes have failed. Environ­mental, social and economic issues associated with beaver re-introduction have given rise to a range of information needs in relation to the species«SQ» ecology. In 2005, six European beavers were translocated from Bavaria to a re-introduction site in southern England. The following year, a 6 month study was un­dertaken to investigate which tree species and size classes were most frequently utilised by the beavers, and to determine if the utilised species and size classes reflected resource availability within the home range. These questions were answered by comparing the use and the availability of foraged tree species in two ways: (1) using a survey of tree stumps; and (2) as a food choice experiment. The results showed that the re-introduced beavers were highly selective in relation to both the species and the size of the trees they used i.e. the re-introduced beavers selected species in significantly different proportions to their availability. However, they selected similar size classes between the preferred tree species, and did not utilise human «SQ»timber«SQ». The relevance of the study is discussed with respect to information needs as­sociated with the re-introduction of a «SQ»keystone«SQ» species, overcoming negative perceptions, population viability, welfare and «SQ»soft release«SQ» strategies.

How to cite this article:
O'Connell M J, Atkinson S R, Gamez K, Pickering S P, Dutton J S. Forage Preferences of the European Beaver Castor fiber: Implications for Re-introduction.Conservat Soc 2008;6:190-194

How to cite this URL:
O'Connell M J, Atkinson S R, Gamez K, Pickering S P, Dutton J S. Forage Preferences of the European Beaver Castor fiber: Implications for Re-introduction. Conservat Soc [serial online] 2008 [cited 2020 Mar 29 ];6:190-194
Available from:

Full Text


HISTORICALLY, THE EUROPEAN BEAVER Castor fiber oc curred throughout the Western Palaearctic (Hartman 1996; Halley & Rosell 2002). By the beginning of the twentieth century, a relict population of no more than 1200 individuals remained in France, Norway, Poland, Russia and Sweden (Zurowski & Kasperczyk 1988). This substantial reduction resulted from the combined effects of changes in land use and hunting. The latter was either for fur (pelts) or for castoreum, a secretion of the scent gland believed to have medicinal properties (Nolet & Rosell 1998). During the second half of the twentieth century, beavers were re-introduced into many areas across Europe. In some countries, this resulted in popula­tions known to be self-sustaining (e.g. Estonia, Latvia, Lithuania and Poland), while the long term viability is less clear in others (e.g. Denmark, the Netherlands, Bel­gium, Spain, Czech Republic, Hungary, Slovenia and Slovakia) (Halley & Rosell 2002). The total European population is currently estimated at around 300,000 indi­viduals, but is not considered to be occupying all of its former range (Halley & Rosell 2002). There are also 12,000 American beavers C. canadensis established in parts of Scandinavia.

In the United Kingdom (UK), hunting exterminated the last recorded population from Scotland in the sixteenth century (South et al. 2000). The geographical isolation of the UK from mainland Europe means that beaver popula­tions cannot be re-established through dispersal. Since 1995, the UK government's statutory conservation agen­cies have been investigating the potential for beaver re­introduction under the auspices of Article 22(a) of the European Council Directive 92/43/EEC. The UK gov-ernment has taken a strategic approach to re-introduction, using pilot programmes where animals are released on the understanding that should 'problems' arise they would be removed (Morrison 2004). They have also as­similated the issues raised by the re-introduction experi­ences of other European states, and undertaken extensive public consultation on potential environmental damage and economic loss (Wigley & Garner 1987; Jonker et al. 2006). Negative perceptions about beavers often persist in human communities, despite evidence of only rela­tively benign impacts, often balanced by benefits to biodiversity and local economies (Johnston & Naiman 1990; Collen & Gibson 2001; Rosell et al. 2005). Part of this perception is a result of the functional 'keystone' role of beavers within local ecosystems. This can be viewed as causing environmental 'impact' by land owners and the general public (South et al. 2000; Rosell et al. 2005).

The issues associated with re-introduction of the beaver in the UK (welfare, viability, land owner percep­tions, dispersal, ecosystem impacts etc.), have given rise to a number of information needs in relation to the spe­cies' ecology, particularly feeding habits. This is impor­tant in understanding the forage requirements and local habitat interactions of re-introduced groups in the short and long terms. It is also important in addressing some of the negative perceptions of local people and land owners. Although there is an extensive body of research on bea­vers, much of this has been conducted in North America for C. canadensis. Relatively little work has focused on food preferences and size selection for C. fiber in differ­ent parts of its newly established range (Belovsky 1984; Rosell et al. 2005; Haarberg & Rosell 2006). In addition to the re-introduction activities of statutory agencies, re­cent efforts in Europe have also involved 'private' indi­viduals undertaking release programmes under licence. Beavers are known to be able to feed on a very broad range of non-woody vascular plants (Parker et al. 2007), and so will have different diets at different sites. This study therefore focused on the use of tree species. Trees are felled by beavers for both food and construction of lodges, and the types and sizes of trees felled can have an impact on local ecosystem diversity and processes (Nolet et al. 1994; Hartman 1996; Haarberg & Rosell 2006). The present study was designed to answer two key questions in relation to the re-introduced beavers: (1) which tree species and size classes are most frequently utilised?; and (2) do the utilised species and size classes reflect resource availability within the home range?


Study Site and Re-introduced Animals

Six European beavers were translocated from Bavaria (southern Germany) and released under licence (DEFRA 2006) into the Flagham Fen lake on the Lower Mill Es­tate near Somerford Keynes, Gloucestershire, England. They were initially quarantined for a 6 month period from April 2005. The group comprised two families: a breed­ing pair and their male offspring from the preceding year, and a second breeding pair with a female offspring from the preceding year. The lake forms a part of a larger complex of 146 lakes called the Cotswold Water Park, re­sulting from the commercial extraction of sand and gravel. Nearly half of the lakes have some human activi­ties associated with them e.g. fishing, water recreation or houses (O'Connell et al. 2007). The translocated beavers were contained within a fenced wildlife reserve, compris­ing 15 ha of lake, and 5 ha of embankments and mature planted woodland.

Tree Species and Size Class Preferences

Tree Surveys

In August 2005, a survey of all trees (n=3105) was con­ducted in the Flagham Fen wood by staff of the Lower Mill Estate. All trees were identified to species level, and their diameter (cm) at 50 cm above the ground placed into one of three categories: 30 cm.

Canopy Experiments

Based on the protocol of Aldous (1938) four tree 'cano­pies' (branches) were provided as forage for the beavers. Two of the trees used were abundant in the study wood­land (willow Salix alba and alder Alnus incana), and two less common (ash Fraxinus excelsior and elder Sambucus nigra). Details of the data to quantify the availability of different tree species in the study woodland are provided below. Commercially available pine fence posts were used as a fifth 'canopy' to investigate the potential for foraging on this material. Canopies were approximately 1.5 m in length and 10 cm in diameter. A row of the five canopy types was placed in the ground (random order) amongst low vegetation, 1 m from the edge of the lake, and at four separate sites. A new set of canopies was pro­vided at each site on fifteen occasions between July 2006 and March 2007. Gloves were used when acquiring and placing the canopies to minimise contact with the forage material. For each experiment, the canopies were left in place for 24 hours.

Stump Surveys

Between July and December 2006, monthly tree stump surveys were conducted in the Flagham Fen wood. Trees felled by the beavers were recognised by the conical stump shape, height from the ground, arrangement of teeth marks and the presence of wood chippings (Johns­ton & Naiman 1990). The diameter (cm) of each tree stump was measured immediately below the gnawed area (Belovsky 1984). A total of 108 trees (sixty-one willows, twenty-nine alders, seventeen aspens, one hawthorn) were felled during the study.


The G statistic was used to identify significant differ­ences in frequency distributions using a goodness-of-fit approach to calculate expected values i.e. for species preferences and stump size class preferences, expected values were calculated as a function of resource availabil­ity. Williams' Correction was applied where appropriate (Sokal & Rohlf 1994). Analysis of variance was used to identify between-species differences in the size of felled trees. An Anderson-Darling test was used to test data for normality prior to this procedure. Linear regression on arcsine transformed data, was used to investigate tempo­ral changes in the proportions of the canopy types and stump sizes used (Cardinal & Aitken 2005).


Tree Preferences

Canopy Experiments

The proportions of canopies removed did not change in relation to month (6) or site (4). Twenty-five per cent of the 300 canopies provided were removed. Willow (57 per cent of canopies removed) and alder (43 per cent of ca­nopies removed) were preferentially selected over other canopies (G=136.1, P Felled Trees

Four species of trees (alder, aspen, hawthorn and willow) were felled by the re-introduced beavers. [Figure 1] shows the percentage of each species compared to their percent­age availability within the woodland reserve area.

The number of trees felled for the four species used was significantly disproportionate to their availability (G=72.7, P Stump Size

The mean, minimum and maximum diameters of the 108 felled trees measured are shown in [Table 1].

There was no significant difference in the mean diame­ter of the four species utilised, and the diameter of trees used did not change between July and December. The size class distribution for felled trees compared to the size classes available in the wood is shown in [Figure 2]. Hawthorn is not compared as only one tree of this species was felled. For the three species, the size classes utilised were significantly disproportionate to those available (al­der: G=15.8, P et al. 2000; Wanless et al. 2002; Hardman & Moro 2006). A major benefit of pilot studies is that they produce information to underpin and guide the implementation of larger scale releases. For ex­ample, baseline information on range size, feeding ecol­ogy and dispersal can be used for modelling the suitability of sites and landscapes for re-introduction. The information can be used either to quantify the number of individuals that could be maintained in a pre-defined area (South et al. 2000), or to identify sites where a pre­determined viable population size could be maintained (Macdonald et al. 2000). Information from pilot schemes can also be used to overcome a number of beaver re­introduction problems that conservation agencies face. One of the major issues is that the species has been ab­sent from the landscape for hundreds of years, and lacrustine and riparian habitats have reached new func­tional, economic and cultural equilibria in this time. Bea­vers are known to exert considerable ecosystem influences by altering environmental structure, diversity and function at local scales (see review by Rosell et al. 2005). This is potentially a difficulty for the conservation agencies involved in beaver re-introductions, because they also have statutory responsibility to ensure the main­tenance of local species and habitats that may be vulner­able to the environmental changes resulting from beaver activity. Of course many elements of beaver influence have been shown to be environmentally 'positive', for example, they can increase plant diversity (Nolet et al. 1994; Rosell et al. 2005). Nevertheless, whilst overall di­versity may increase, individual species of concern may be negatively impacted. Experiences from re-introduc­tions in other countries may be of some use in this regard, but information from local pilot studies can provide a more species relevant indication of beaver impacts.

Another issue for re-introduction programmes is to en­sure the welfare of the re-introduced animals over a range of timescales by ensuring the ongoing provision of ade­quate food and construction resources. Pilot projects can provide such information by identifying the range of tree species utilised, their relative preference in comparison to availability and the distribution of size classes preferred. At the Cotswold Water Park, the re-introduced beavers showed significant selectivity in relation to both species and size class. In particular, the importance of willow Salix spp. was identified. Nolet et al. (1994) also high­lighted the importance of willow, and demonstrated that beavers re-introduced in a willow dominated wood in the Netherlands, reduced tree numbers by about 5 per cent (equating to about 30 per cent of total annual willow re­growth). Fryxell et al. (1994) suggested that food reten­tion time in the digestive track and the resultant energy pay off is an important factor in food selection by the Eu­ropean beaver, and that willows are an important group because they allow beavers to optimise energy intake. The distribution and relative abundance of willow is therefore a key issue for re-introduction attempts. Fur­thermore, re-introduction programmes on mainland Eu­rope suggest that post-release beaver population growth growth is secondary to an initial period of spatial expan­sion within a watershed (Halley & Rosell 2002). The im­plication of this is that suitable resources (including willow) need to be available over larger spatial scales be­yond the release site.

The last issue for beaver re-introduction is in relation to negative perceptions of the species by landowners and the general public. Responsible statutory agencies often attempt to measure and assuage local negative opinions prior to releases, using questionnaires and public aware­ness techniques. The use of information from pilot stud­ies can also play a key role in this area. For example at the Cotswold Water Park, the stump surveys and canopy experiments of the present study, showed that the beavers did not gnaw human constructions (e.g. fence posts), of­ten a major concern of local people prior to re­introduction (Wigley & Garner 1987; Jonker et al. 2006). It has been known for some time that European beavers have strong olfactory senses, which they utilise in select­ing preferred forage (Muller-Schwarze et al. 1994). How­ever, the provision of new, locally gathered evidence is perhaps more persuasive for local communities. Impor­tantly, this finding also has the management implication that timber fences can be used either to construct con­tainment barriers at the initial release site, or to fence off areas from where the beavers need to be excluded.


We are very grateful to Jeremy Paxton for allowing ac­cess to the Flagham Fen Nature Reserve. Thanks also to the Lower Mill Estate and Cotswold Water Park Society staff for their help, advice and guidance; particularly to Will Vicary.[27]


1Aldous, S.E. 1938. Beaver Food Utilization Studies. Journal of Wildlife Management 2: 215-222.
2Belovsky, G.E. 1984. Summer Diet Optimization by Beaver. American Midland Naturalist 11: 209-222.
3Bright, P.W. and P.A. Morris. 1994. Animal Translocation for Conser­vation: Performance of Dormice in Relation to Release Methods, Origin and Season. Journal of Applied Ecology 31(4): 699-708.
4Cardinal, R.N. and M.R.F. Aitken. 2005. ANOVA for the Behavioural Sciences Researcher. Lawrence Erlbaum Associates, New Jersey, USA.
5Collen, P. and R.J. Gibson. 2001. The General Ecology of Beavers (Castor spp.), as Related to Their Influence on Stream Ecosystems and Riparian Habitats, and the Subsequent Effects on Fish: A Re­ view. Reviews in Fish Biology and Fisheries 10: 439-461.
6DEFRA (Department for Environment, Food and Rural Affairs) 2006. DEFRA Landscape Protection, Recreation and Public Access: Coun­tryside and Rights of Way Act 2000. URL: (accessed 13 May 2008).
7Fryxell, J.M., S.M. Vamosi, R.A. Walton et al. 1994. Retention Time and the Functional Response of Beavers. Oikos 71: 207-214.
8Haarberg, O. and F. Rosell. 2006. Selective Foraging on Woody Plant Species by the Eurasian Beaver (Castor fiber) in Telemark, Nor­way. Journal of Zoology 270: 201-208.
9Halley, D.J. and F. Rosell. 2002. The Beaver's Re-conquest of Eurasia: Status, Population Development and Management of a Conserva­tion Success. Mammal Review 32: 153-178.
10Hardman, B. and D. Moro. 2006. Optimising Reintroduction Success by Delayed Dispersal: Is the Release Protocol Important for Hare­wallabies? Biological Conservation 128(3): 403-411.
11Hartman, G. 1996. Habitat Selection by the European Beaver. Journal of Zoology 240: 317-325.
12Johnston, C.A. and R.J. Naiman. 1990. Browse Selection by Beaver: Effects on Riparian Forest Composition. Canadian Journal of For­est Research 20: 1036-1043.
13Jonker, S.A., R.M. Muth, J.F. Organ et al. 2006. Experiences with Beaver Damage and Attitudes of Massachusetts Residents Toward Beaver. Wildlife Society Bulletin 34(4): 1009-1021.
14Letty, J., S. Marchandeau, J. Clobert et al. 2000. Improving Translocation Success: An Experimental Study of Anti-stress Treatment and Release Method for Wild Rabbits. Animal Conservation 3(3): 211-219.
15Macdonald, D.W., F.H. Tattersall, S. Rushton et al. 2000. Reintroduc­ing the Beaver (Castor fiber) to Scotland: A Protocol for Identify­ing and Assessing Suitable Release Sites. Animal Conservation 3(2): 125-133.
16Morrison, A. 2004. Trial Re-introduction of the European Beaver to Knapdale: Public Health Monitoring 2001-3. Scottish Natural Heri­tage Commissioned Report No. 077 (ROAME No. F02AC327).
17Muller-Schwarze, D., B.A. Schulte, L. Sun et al. 1994. Red Maple (Acer rubrum) Inhibits Feeding by Beaver (Castor canadensis). Journal of Chemical Ecology 20: 2021-2034.
18Nolet, B.A., A. Hoekstra and M.M. Ottenheim. 1994. Selective Foraging on Woody Species by the Beaver Castor fiber, and its Impact on a Riparian Willow Forest. Biological Conservation 70: 117-128.
19Nolet, B.A. and F. Rosell. 1998. Comeback of the Beaver Castor fiber: An Overview of Old and New Conservation Problems. Biological Conservation 83: 165-173.
20O'Connell, M.J., M.L. Yallop, R.M. Ward et al. 2007. Collecting and Integrating Multi-scale Waterbird and Lentic Data to Model Trade­offs Between Forage Quality and the Risk of Disturbance. Ibis 149: 65-72.
21Parker, J.D., C.C. Caudill and M.E. Hay. 2007. Beaver Herbivory on Aquatic Plants. Oecologia 151(4): 616-625.
22Rosell, F., O. Bozser, P. Collen et al. 2005. Ecological Impact of Bea­vers Castor fiber and Castor canadensis and Their Ability to Mod­ify Ecosystems. Mammal Review 35: 248-276.
23Sokal, R.R. and F.J. Rohlf. 1994. Biometry: The Principles and Practice of Statistics in Biological Research. 3rd edition. Freeman, New York, USA.
24South, A., S. Rushton and D. Macdonald. 2000. Simulating the Pro­posed Reintroduction of the European Beaver (castor fiber) to Scotland. Biological Conservation 93: 103-116.
25Wanless, R.M., J. Cunningham, P.A.R. Hockey et al. 2002. The Suc­cess of a Soft-release Reintroduction of the Flightless Aldabra Rail (Dryolimnas [cuvieri] aldabranus) on Aldabra Atoll, Seychelles. Biological Conservation 107(2): 203-210.
26Wigley, T.B. and M.E. Garner. 1987. Landowner Perceptions of Beaver Damage and Control in Arkansas. Proceedings of the Third Eastern Wildlife Damage Control Conference. 8: 34-41.
27Zurowski, W. and B. Kasperczyk. 1988. Effects of Reintroduction of the European Beaver in the Lowlands of the Vistula Basin. Acta Theriologica 33: 325-338.