Dispersal pattern and effective population size of the beaver Lixing Sun, Dietland Müller-Schwarze, and Bruce A. Schulte Abstract: The dispersal pattern of the beaver (Castor canadensis) was studied by intensive livetrapping, tagging, and observation in Allegany State Park and its vicinity in New York from 1984 to 1996. The majority (74%) of dispersing beavers (n = 46) initiated dispersal in a downstream direction after spring ice-out. Females dispersed significantly far- ther away from their natal colonies than males (10.15 ± 2.42 (SE) km vs. 3.49 ± 0.86 km). Movements to neighboring sites were common (16 of 46 dispersers), indicating that beavers, especially males, may prefer to disperse to the near- est available sites. Most (64%) natal dispersers were 2-year-olds. Three-year-olds also constituted a considerable pro- portion (21%) of the dispersers, but 1-year-old dispersers were relatively rare (14%). Many adults underwent secondary dispersal after successful natal dispersal in our study area. Male secondary dispersers were more inclined to take over neighboring sites than were male natal dispersers (10 of 13 vs. 3 of 13). The effective population size in a 250-km 2 area was estimated to be 161–228 individuals by the areal method and 267–378 individuals by the 85th percentile method. Résumé : La dispersion a été étudiée chez le Castor du Canada (Castor canadensis) par capture d’animaux vivants, par marquage et par observation, dans le parc Allegany State Park et ses environs dans l’état de New-York de 1984 à 1996. La majorité (74 %) des castors qui ont quitté leur lieu d’origine (n = 46) se sont d’abord orientés vers l’aval du ruisseau après la disparition des glaces. Les femelles se sont éloignées significativement plus de leur colonie d’origine que les mâles (10,15 ± 2,42 (erreur type) km vs. 3,49 ± 0,86 km). Les déplacements vers des endroits avoisinants ont été observés fréquemment (16 animaux sur 46), ce qui indique que les castors, plus particulièrement les mâles, préfè- rent gagner les sites disponibles les moins éloignés. La plupart des castors non sédentaires (64 %) étaient âgés de 2 ans. Les animaux de 3 ans constituaient aussi une proportion importante (21 %) des animaux nons sédentaires, mais la dispersion des castors de 1 an s’est avérée relativement rare (14 %). Dans notre zone d’étude, plusieurs adultes ont quitté de nouveau leur lieu de séjour après une première tentative réussie. Ces mâles non sédentaires secondaires avaient plus tendance à s’approprier des sites avoisinants que les mâles non sédentaires primaires (10 sur 13 vs. 3 sur 13). L’effectif de la population dans un rayon de 250 km 2 a été évalué à 161–228 individus par la méthode d’évaluation en fonction de la surface et à 267–378 individus par la méthode du 85 e percentile. [Traduit par la Rédaction] 398 Sun et al. Introduction Dispersal pattern has profound effects on populations demographically, socially, and genetically (Lidicker 1985; Chepko-Sade et al. 1987; Stenseth and Lidicker 1992). For this reason, dispersal pattern is of critical importance in understanding the dynamics of populations. A significant portion of our current knowledge about mammalian disper- sal patterns has been gleaned from in-depth studies of small (typically with body mass <0.5 kg) rodents (e.g., Stenseth and Lidicker 1992). Information about mammalian dispersal patterns in other groups is more or less sporadic. In rodents, prominent differences in aspects of life history exist between small and large (typically with body mass >0.5 kg) species, and this may result in sharply contrasting dispersal patterns between large and small rodents. Key aspects of dispersal in large rodents pose a strong test for the generality of theories and hypotheses about mammalian dispersal patterns that have been based on studies of small rodents. Information about dispersal patterns of large rodents, however, has so far been limited. To our knowledge, information is available for several species of marmots (e.g., Armitage 1991; Meier 1992; van Vuren and Armitage 1994; Blumstein and Arnold 1998), the capybara (Hydrochaeris hydrochaeris; Herrera 1992), and the porcupine (Erethizon epixanthum; Roze 1989); there have also been a few studies of the beaver (Castor canaden- sis), most of which have very limited sample sizes and are anecdotal in nature (e.g., Beer 1955; Leege 1968; Hodgdon 1978, see below). Beavers live in family units and defend a common terri- tory, which includes a series of connected ponds. Each fam- ily typically consists of an adult pair and their offspring of both the current and previous years. The two breeding adults normally give birth to 3–4 kits each summer. Both male and female offspring disperse from their native family at the age Can. J. Zool. 78: 393–398 (2000) © 2000 NRC Canada 393 Received December 15, 1998. Accepted October 27, 1999. L. Sun. 1 Department of Biological Sciences, Central Washington University, Ellensburg, WA 98926-7537, U.S.A. D. Müller-Schwarze. State University of New York, College of Environmental Science and Forestry, Syracuse, NY 13210, U.S.A. B.A. Schulte. Department of Biology, Providence College, Providence, RI 02918-0001, U.S.A. 1 Author to whom all correspondence should be addressed (e-mail: lixing@cwu.edu).