Use of anal gland secretion to distinguish the two beaver species Castor canadensis and C. fiber

Introduced North American beavers Castor canadensis pose a potential threat to the continuing recovery of Eurasian beaver C. fiber populations in several European countries. For management purposes, it is necessary to be able to identify and distinguish the two species. This, however, is difficult because the two species are morphologically, ecologically and behaviourally similar. To find a method for species identification, we examined the possibility of using anal gland secretion (AGS) collected from the two beaver species. We asked 20 inexperienced volunteers to inspect the colour and viscosity of the AGS. When we provided the volunteers with the sex of each beaver and descriptions of the colour and viscosity of AGS from males and females of the two species, all volunteers could 100% correctly identify the two species. We therefore conclude that the colour and viscosity of the AGS can be used for a quick and easy identification of the two beaver species.

T he tw o extant species o f the genus Castor, the E ura sian beaver C. fib e r and the N orth A m erican beaver C. canadensis, are sim ilar both m orphologically and behaviourally (e.g. W ilsson 1971(e.g. W ilsson , Patenaude 1984(e.g. W ilsson , N ovak 1987 and w ere once classified as one species (H ill 1982). T he tw o-species status w as finally estab lished after L avrov & O rlov (1973) found that the diploid chrom osom e num ber is 48 in C. fib e r and 40 in C. canadensis, follow ing R obertsonian fusion of eight chrom osom e pairs in C. canadensis.  and also resulted in several unexpected eco logical consequences. T he N orth A m erican beaver w as introduced into F inland during 1935-1937 and into several other E uropean countries at various tim es in the 20th century (see N olet & R osell 1998). M ost o f the introductions w ere successful, and sever al populations have since thrived (e.g. Lahti 1995Lahti , N olet & R osell 1998. Interspecific com petition may have profound effects on the population dynam ics o f com peting species, and influence the species' distri bution and evolution (Begon, H arper & Tow nsend 1990). Field observations suggest that the N orth Am erican beaver m ight be a stronger com petitor than the E urasian beaver (Erm ala, H elm inen & Lahti 1989). C urrently, N orth A m erican beavers can be found along the Seine in France, in R ussian K arelia, and m ay also exist in A ustria, the C zech Republic, H ungary, P oland and Slovakia (see Sieber 1989, K ollar & Seiter 1990, N olet & Rosell 1998. Since beavers are capable o f undergoing long-distance dis persal (Sun, M iiller-Schw arze & Schulte in press) tim e m ay be the only problem before the N orth A m erican beavers in F inland invade other N ordic countries. It has becom e a m ajor task for w ildlife m anagers there to prevent this from happening (N um m i 1996). A lso, other sporadic populations o f the N orth A m erican beaver should be exterm inated before they grow and spread further. To handle these problem s, identification o f the tw o beaver species is essential. T he only reliable m ethod currently avail able is to check for the chrom osom e num ber (Lavrov & O rlov 1973). This m ethod is slow and expensive and requires support from chrom osom e experts and sophisticated technology. M anagers o f beaver popu lations need a quick and easy alternative to identify and distinguish the tw o beaver species.

Introductions o f N orth A m erican beavers, reintro ductions o f E urasian beavers and translocations o f the tw o species have played an im portant role in the recovery o f beavers in E urasia (see N olet
B eavers live in fam ily groups, w hich defend terri tories against other beavers (e.g. B radt 1938, D joshkin & Safonow 1972, N olet & Rosell 1994, R osell & N olet 1997. Both sexes o f the tw o species scent m ark the borders o f their territories w ith secretions from the castor sacs and/or the anal glands (e.g. H odgdon 1978, Svendsen 1980, B uech 1995, Rosell & B ergan 1999, R osell, Bergan & P arker 1998. The anal gland is a holocrine secretory gland, but the cas tor sac is only a pocket lined w ith a layer o f nonsecretory epithelium (Svendsen 1978, Valeur 1988). T he lipid com position, colour, viscosity and odour of anal gland secretion (A G S) can be used for sexing beavers (G rpnneberg 1979, G rpnneberg & L ie 1984, Schulte, M uller-Schw arze & Sun 1995. F or the E ura sian beaver, the AGS is a thick paste o f a greyish colour w ith an unpleasant odour in fem ales, and is an oily fluid w ith a w hitish or pale straw colour and a stronger and different odour in m ales (O w esen 1979, G rpnneberg & Lie 1984, Valeur 1988). T he variation in the com position o f A G S sam ples collected from individuals o f the sam e sex is sm all, and the appear ance o f the A G S is independent o f live-or deadtrapped beavers (G rpnneberg & Lie 1984). F or the N orth A m erican beaver, the AGS o f the tw o sexes show a consistent difference regardless o f age: it is brow n and viscous in m ales, b ut it is w hitish or light yellow and runny in fem ales (Schulte et al. 1995). In this study, w e exam ined the possibility o f using the colour and viscosity o f the A G S as a quick and easy w ay to identify and distinguish the tw o beaver spe cies.

Material and methods
We collected 46 fresh A G S sam ples from legally killed Eurasian beavers (23 m ales and 23 fem ales) during spring (M arch-A pril) 1997 in the m unicipali ty o f B 0, the county o f Telem ark, Norway. T he ani m als w ere sexed by presence or absence o f the os penis (O sborn 1955), We assigned beavers to three age classes based on body w eight (see H artm an 1992): 1-year-olds (< 12 m onths, 0-10 kg), 2-yearolds (12-24 m onths, 10-15 kg), and a3-year-olds (>24 m onths, >15 kg). We cut the anal glands open w ith a surgical blade and com pared the colour of m ale and fem ale AGS using the standardised colour chart system , PAN TO NE (Letraset) (Schulte et al. 1995). We placed secretion sam ples in glass vials and stored them in a freezer (-20°C). T he 46 secretion sam ples w ere brought frozen to a laboratory in USA. Tw enty sam ples w ere random ly selected, thaw ed at room tem perature and used in the experim ent d e scribed below.
We also collected 19 A G S sam ples from livetrapped N orth A m erican beavers (10 m ales and 9 fem ales) in different age classes near E llensburg in the state o f W ashington, and in the state o f N ew York, U SA , during autum n and spring in the years 1995-1997. L ive-trapped beavers w ere anaesthetised w ith a 1:2 m ixture by volum e o f xylazine and ketam ine (0.67 and 6.7 m g/kg body m ass, respectively). We collected A G S by palpating anal gland and papillae. We placed all sam ples in glass vials and stored them in a freezer at -20°C. We determ ined the age, sex and colour o f A G S by sim ilar m ethods as used for the E urasian beaver.

Experiment: identification of species by visual cues
We put ca 0.5 m l o f each A G S sam ple into a 10-ml 120 plastic test tube and pooled the 20 AGS sam ples ran dom ly selected from the 46 E urasian beavers w ith the 19 AGS sam ples from N orth A m erican beavers. A fter random ising the 39 sam ples, w e asked 20 volunteers (10 m en, 10 w om en; age range: 18-40; all inexperi enced w ith beaver secretions) to classify these sam ples into four groups, corresponding to the tw o sexes o f the tw o species. W e provided the volunteers with descriptions o f the colour and viscosity o f AGS from the m ales and fem ales o f the tw o species, and also inform ation about the sex o f the beaver, from w hich each secretion sam ple w as collected. T he experim ent w as carried out under a fluorescent light (several 40 w bulbs) in a room w ith a w hite sheet as the back ground.

Colour and viscosity of the AGS of Eurasian and North American beavers
T he A G S characteristics o f our sam ples w ere accord ing to earlier findings. T he sex difference in the colour o f A G S in the E urasian and N orth A m erican beaver w as consistent and w as not affected by age (Table 1).

Species identification using visual cues
In the experim ent, we pooled data because there was no significant difference in identification ability b e tw een m ale and fem ale subjects. V olunteers could 100% accurately classify the 39 sam ples into four groups corresponding to the tw o sexes o f the tw o species using the colour and viscosity o f AGS.

Discussion
Several researchers have noticed the sexual dim or phism in colour and viscosity o f A G S in the Eurasian beaver (O w esen 1979, G rpnnebcrg & Lie 1984, Valeur 1988 and the N orth A m erican beaver (A llred 1986(A llred , Schulte et al. 1995. F urtherm ore, in the N orth A m erican beaver, the chem ical com position o f the A G S changes little as an individual grow s older or by season, and it does not change if an individual m oves to a new site (Sun 1996). This could also be true for the E urasian beaver, although no data to support this are currently available. Sexing live beavers is easy by palpating for the presence or absence o f the os penis (O sborn 1955). H ow ever, the sex o f young beavers m ay be difficult to determ ine in the field (O sborn 1955(O sborn , L arson & K napp 1971(O sborn , B rady & Svendsen 1981, especially for inexperienced researchers. In this situation, the species o f the young beavers in question can be determ ined by identifying the species o f an older beav er living in the sam e colony because both species live in discrete fam ily units and extrem ely young beavers do not w ander far aw ay from their lodges.

Management implications
T he tw o m ain threats for the endem ic E urasian bea ver in E urope are habitat destruction and the spread ing o f introduced N orth A m erican beavers (N olet & R osell 1998). T he N orth A m erican beaver has a greater ecological flexibility, survives better under severe clim atic conditions, adapt m ore easily to sub stitute foods and regulate hydrological regim es o f w ater bodies better w ith a stronger building instinct than its E urasian counterpart (D anilov & K a n 'shiev 1983). T he N orth A m erican beaver is also m ore fecund (H ill 1982) w ith a significantly larger colony size (5.2 ± 1.4 SD) than E urasian beaver (3.8 ± 1.0 SD) (Rosell & P arker 1995). D ispersal o f the N orth A m erican beaver from F inland is thus a pending threat to the endem ic Scandinavian population o f the E urasian beaver. Sim ilar pending threats also call for control o f the N orth A m erican beaver in several other areas in Europe. B ecause beavers are relatively easy to trap, this threat can be relieved through extensive trapping. T he effectiveness o f this m anagem ent prac tice, how ever, depends on the availability o f a quick and easy m ethod to identify and distinguish the two species. O ur results show ed that the colour and vis cosity o f the A G S can be used for a quick and easy identification o f the tw o beaver species.