Behavioral Ecology Advance Access originally published online on March 9, 2007
Behavioral Ecology 2007 18(3):579-589; doi:10.1093/beheco/arm010
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Inbreeding is reduced by female-biased dispersal and mating behavior in Ethiopian wolves
a Wildlife Conservation Research Unit, University of Oxford, Tubney House, Abingdon Road, Tubney OX13 5QL, UK b Department of Ecology and Evolutionary Biology, University of California, 621 Charles E. Young Drive South, Los Angeles, CA 90095, USA
Address correspondence to D.A. Randall. E-mail: deborahrandall{at}fzs.org.
Received 5 August 2006; revised 11 January 2007; accepted 3 February 2007.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONSUPPLEMENTARY MATERIALREFERENCES |
|---|
Molecular tools have enabled wildlife researchers to obtain accurate information on the kinship, mating behavior, and dispersal of individuals. We genotyped 192 Ethiopian wolves (n = 29 packs) in the Bale Mountains for 17 microsatellite loci to 1) elucidate kinship within and between packs, 2) assess parentage of pups, and 3) evaluate whether inbreeding is avoided by dispersal and/or mating behavior. Mean pairwise relatedness within packs (R = 0.39) was significantly greater than that estimated from random assignment of individuals to packs. However, breeding pairs were most often unrelated, suggesting that female-biased dispersal reduces inbreeding. We assigned maternity to 49 pups and paternity to 47 pups (n = 12 litters) using a combination of exclusion, likelihood analyses (using CERVUS software), and sibship reconstruction. Multiple paternity occurred in 33% of litters; extrapack paternity accounted for 28% of all resolved paternities, occurring in 50% of litters. We found no evidence that extrapack copulations reduce inbreeding; however, more detailed analyses may elucidate the effect of recent population declines and demographic disturbances due to recurring disease outbreaks. The adaptive advantages of female-biased dispersal and the observed mating system are discussed in relation to Ethiopian wolf sociobiology and ecology.
Key words: canids, CERVUS, dispersal, EPC, kinship, microsatellites, paternity.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONSUPPLEMENTARY MATERIALREFERENCES |
|---|
The consequences for fitness of group formation are central to the evolution of sociality (Macdonald et al. 2004
), and natal philopatry is an important mechanism in the formation and maintenance of social groups (Waser 1996). Philopatric young may benefit directly from group living ("benefits of philopatry hypothesis," Waser 1988; Stacey and Ligon 1991), or they may be ‘forced’ to remain in groups as a result of the high costs or risks of dispersal and independent breeding ("ecological constraints hypothesis," Selander 1964; Brown 1969; Woolfenden and Fitzpatrick 1978; Koenig 1981; Emlen 1982). The adaptive advantages of dispersal versus philopatry are thought to be related to inbreeding avoidance, access to mates, and competition for resources (Clutton-Brock 1989; Johnson and Gaines 1990; Creel and Macdonald 1995; Pusey and Wolf 1996). Sex-biased dispersal is often hypothesized as a means of avoiding inbreeding with opposite sex parents (Wolff 1992, 1993). Among social canids, this prompts the question of which sex should disperse given that subordinates are typically the offspring of the dominant pair (Moehlman 1989). Theoretical models predict that juveniles will be philopatric only if philopatry maximizes their lifetime reproductive success (Brown 1987). As such, mating systems play a key role in determining the degree and direction of bias in dispersal (Greenwood 1980; Dobson 1982). Greenwood's (1980) "resource competition hypothesis," predicts that male philopatry predominates among species that are monogamous, require paternal investment in offspring care, and exhibit resource (i.e., territory) defense by males. Under such conditions, male philopatry increases the chances of a male acquiring a high-quality territory with which to attract a female.
The Ethiopian wolf (Canis simensis) lives in packs of up to 13 individuals, which cooperatively defend an exclusive territory (Sillero-Zubiri and Gottelli 1995b). As in many social canids, well-defined hierarchies exist and the dominant female usually monopolizes breeding, producing a litter of up to 6 pups annually (Sillero-Zubiri et al. 2004). All group members contribute to parental care, but there is little evidence that helpers accrue inclusive fitness benefits through increased survival of related offspring (Sillero-Zubiri et al. 2004). Dispersal appears to be extremely costly for Ethiopian wolves due to habitat saturation in their Afroalpine "island" habitats and the lack of unoccupied territories on which to breed independently (Sillero-Zubiri and Gottelli 1995b; Sillero-Zubiri, Gottelli, and Macdonald 1996). Observational data suggest that the structure of packs and formation of breeding units occur by 3 main mechanisms: 1) males are predominantly philopatric, and the few females that remain in the natal pack may eventually acquire breeding status 2) most females disperse around 2 years of age, and some become "floaters" that may successfully immigrate into existing packs, 3) subordinate females form alliances with same-pack males to split a natal pack (Sillero-Zubiri 1994; Sillero-Zubiri, Gottelli, and Macdonald 1996; Marino 2003; Sillero-Zubiri et al. 2004).
Previous studies have explained Ethiopian wolf sociality in terms of access to resources (Sillero-Zubiri 1994; Marino 2003). Ethiopian wolves forage solitarily on the abundant rodents of the Afroalpine ecosystem (Sillero-Zubiri and Gottelli 1995a). Where resources are particularly rich, Ethiopian wolves appear to tolerate relatives that remain in the pack, thereby enabling group size increase and territory expansion (sensu Kruuk and Macdonald 1985). Under these conditions, wolves in larger groups gain the benefit of more prey-rich habitat per individual (Marino 2003). Where resources are scarce, groups are smaller, typically consisting of pairs of individuals (Marino 2003). Thus, the benefits of sociality appear to be primarily acquisition, inheritance, and/or defense of a high-quality territory (Marino 2003). Resource-based explanations for sociality may also underlie female-biased dispersal in Ethiopian wolves, which contrasts with the predominantly male-biased dispersal pattern in other mammals (Greenwood 1980). Based on the assumptions of the resource competition hypothesis (Greenwood 1980), males may have a greater stake in resource defense and the prospect of future territory inheritance. These predictions for Ethiopian wolves are supported by the pattern of male philopatry and greater male investment in territorial defense (Sillero-Zubiri and Macdonald 1998).
It is well recognized that multiple factors often determine the relative costs and benefits of dispersal versus philopatry (Macdonald and Carr 1989). Among Ethiopian wolves, inbreeding avoidance may be an additional adaptive advantage of female-biased dispersal and may underlie the observed mating behavior. Male philopatry and the long tenure of breeding females increase the potential for incest within groups, which may be countered in part by female dispersal (Sillero-Zubiri et al. 2004). However, given the extent of habitat saturation, constraints on dispersal, and occasional female philopatry, alternative behavioral mechanisms may have evolved to avoid mating with close kin. A previous study found that 70% of observed copulations were between dominant females and males in neighboring packs (Sillero-Zubiri, Gottelli, and Macdonald 1996), and genetic analyses provided preliminary evidence for multiple paternity (Gottelli et al. 1994). It has been suggested that such extrapack copulations (EPCs) may be an inbreeding avoidance strategy for Ethiopian wolves (Sillero-Zubiri, Gottelli, and Macdonald 1996; Sillero-Zubiri et al. 1998), enabling females to gain access to genetically dissimilar males (see also Jennions and Petrie 2000). However, this remains unsubstantiated, and alternative explanations for EPCs are possible (e.g., deterring infanticide by neighbors, Wolff and Macdonald 2004).
We combined demographic and spatial information collected during field observations of focal Ethiopian wolf packs in the Bale Mountains with multilocus microsatellite data to 1) determine genetic relatedness within and between social groups and 2) assess whether inbreeding avoidance underlies dispersal and/or mating behavior. Based on previous observational data, we expected packs to be predominantly close-kin groups (Sillero-Zubiri 1994; Marino 2003; Sillero-Zubiri et al. 2004). However, if sex-biased dispersal is a mechanism to avoid inbreeding, then we predicted that breeding individuals would be unrelated and incestuous matings would be uncommon. We identified mated pairs based on genetic evidence of parentage, which was also used to quantify the extent of multiple- and extrapack paternity (EPP) in litters. Paternity and relatedness data were used to test the hypothesis that EPCs enable females to obtain genetically dissimilar mates outside their social group.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONSUPPLEMENTARY MATERIALREFERENCES |
|---|
Study population
We focus predominantly on 2 subpopulations of Ethiopian wolves in the Bale Mountains: the Web Valley (WV) during the 2002–2003 breeding season and the Sanetti Plateau (SN) during the 2003–2004 breeding season (Figure 1). Packs were identified easily by different group sizes and territory locations. Pack compositions were determined by complete enumeration of individuals observed on multiple occasions around the den, during social greetings, and on boundary patrols (as in Marino et al. 2006
). To account for the possibility of missing pack members from a given group sighting, pack size and age/sex structure were deduced from multiple observations of the pack over the course of the breeding season until no new individuals were recorded for that pack. Individuals were classified as adults (>2 years), yearlings (1–2 years), or pups (<1 year). Litter size was defined as the maximum number of pups at emergence from dens (Sillero-Zubiri et al. 1998). Territories were represented by minimum convex polygons after removing 5% of outliers in ARCVIEW version 3.0 (Environmental Systems Research Institute, Inc., Redlands, CA) using yearly (April–March) GPS locations (Garmin 12 GPS software with map datum WGS 84) pooled for each pack (as in Sillero-Zubiri and Gottelli 1995b).
|
Sample collection and analysis
Fecal samples were used as our primary source of DNA because we favored a noninvasive approach to DNA sampling. Because individuals without ear tags or other artificial marks were not always individually identifiable, a surplus of fecal samples was collected from each pack and each litter of pups so as to maximize the chance that each animal was sampled at least once. Fecal samples were collected from wolves seen defecating so that age, sex, and pack could be determined. Pup feces were collected around the entrance to dens. We also genotyped tissue samples from 27 wolves that died during a rabies outbreak in WV in 2003 (Randall et al. 2004
) and 85 wolves livetrapped during the subsequent emergency vaccination campaign (Knobel D, Fooks AR, Brookes S, Randall DA, Williams SD, Argaw K, Shiferaw F, Tallents LA, Laurenson MK, unpublished data). All samples were preserved in 96% ethanol.
We extracted DNA from feces using QIAamp DNA Stool Mini Kits (QIAGEN, Valencia, CA) and from tissue using QIAamp DNA Mini Kits (QIAGEN) according to the manufacturer's protocol. Negative controls were included in each set of extractions to monitor for contamination. We selected 17 microsatellites from the dog genome (Breen et al. 2001) that were polymorphic in Ethiopian wolves, including 16 tetranucleotide repeat loci (FH2001, FH2054, FH2119, FH2137, FH2138, FH2140, FH2159, FH2174, FH2226, FH2293, FH2320, FH2422, FH2472, FH2537, PEZ17, and PEZ19) and 1 dinucleotide repeat locus (C05.377). Sex allocation was confirmed using a Y-chromosome microsatellite locus and the primer pairs MS34 or MS41 (Olivier et al. 1999). Microsatellites were amplified using the QIAGEN Multiplex PCR Kit in 10 µl volumes containing 1.5 µl DNA, 1.0 µl primer mix, 0.4 µl 10 mg/ml bovine serum albumin, 5.0 µl QIAGEN master-mix, and ddH2O. Amplifications were performed on a programmable Peltier Thermal Cycler (MJ Research [Waltham, MA] PTC-200). Genotyping reactions were run on an automated sequencer with fragment sizes checked manually with reference to a size standard. Fecal samples were genotyped using a multiple-tubes approach (Navidi et al. 1992; Taberlet et al. 1996) in which both alleles at a heterozygous locus were amplified at least twice and the single allele of a homozygous locus was amplified at least 3 times. Consensus genotypes were completed for a minimum of 14 loci. Complete details of the genotyping methods and validation techniques are available elsewhere (Randall 2006; Randall DA, Pollinger JP, Tallents LA, Macdonald DW, Wayne RK, unpublished data).
We used CERVUS version 2.0 (Marshall et al. 1998) to calculate allele frequencies. The Web version of GENEPOP (Raymond and Rousset 1995) was used to test for deviations from Hardy–Weinberg equilibrium per locus within subpopulations and overall. We also used GENEPOP (Raymond and Rousset 1995) to test for linkage disequilibrium (LD) within subpopulations and overall using Fisher's exact tests. We excluded yearling and pup genotypes to reduce the effect of different generations and applied adjusted P values at the 0.05 nominal level using Bonferroni corrections to account for multiple tests (Rice 1989).
We calculated the probability that 2 siblings would have identical genotypes by chance in our population (PIDsib, as in Evett and Weir 1998; Waits et al. 2001) using the program GIMLET version 1.3.1. (Valière 2002). PIDsib is a conservative estimate of the power to resolve between individuals where populations exhibit substructure or where comparisons are made between related individuals (Waits et al. 2001). The program IDENTITY (Allen et al. 1995) was used to check for duplicate genotypes in the data set.
Parentage
Maternity and paternity were assigned according to specific rules (Figure 2). Parentage was assigned initially using the likelihood-based approach in CERVUS version 2.0 (Marshall et al. 1998). This method accounts for the proportion of candidate parents sampled, allows for genotyping errors, and calculates statistical confidence based on the difference in LOD (i.e., the logarithm of the likelihood ratio) scores of candidate parents. However, this method is limited by its inability to assign parentage to unsampled individuals within or outside the study population or to reliably resolve correctly between close relatives (Marshall et al. 1998; Morrissey and Wilson 2005). To account for this and further enhance our parentage assignments, we used COLONY version 1.2 (Wang 2004) to predict the most likely number of parents within each subpopulation (and social group) by inferring sibling relationships among offspring.
|
Because it is unlikely for yearlings to breed (Sillero-Zubiri 1994
), we excluded them from the candidate parent pool. Furthermore, long female breeding tenure (Sillero-Zubiri, Gottelli, and Macdonald 1996) meant that yearlings in subpopulations were probably full or half sibs of the pups in their pack and might produce higher likelihood scores than the true parents (Marshall et al. 1998; Morrissey and Wilson 2005). Litters were defined as one or more pups with the same mother in the same birth year. Maternity was assigned by exclusion if only one adult female in a pack had zero mismatches with one or more pups in her pack (as in Baker et al. 2004). Otherwise, we used CERVUS to analyze the likelihood of paternity for candidate males with respect to all nonexcluded mothers.
First, we performed an open paternity analysis in which all males in a given subpopulation and, to account for EPCs, males in packs adjacent to each subpopulation were considered candidate fathers (Figure 1). We also performed a restricted analysis in which only within-pack adult males or adult males in neighboring packs were considered candidate fathers. Packs with adjoining territories were considered neighbors in the restricted analysis (Figure 1). Males (or mother–father pairs where multiple mothers were not excluded) were allocated parentage if they were assigned at 95% or higher confidence. Alternatively, males (or mother–father pairs) were allocated parentage if they were assigned at 80% or higher confidence if they produced only one or zero parent–offspring mismatches, which were accepted to allow for the possibility of genotyping errors and/or mutations. However, assignments with more than one mismatch at the lower confidence level were considered unreliable and rejected. If the results of the open and the restricted paternity analyses differed, the father (or mother–father pair) with the higher confidence was accepted. If the confidences were equal, the result of the restricted analysis was accepted because within-pack males or males from neighboring packs were more likely to mate with the dominant female than males from nonneighboring packs (Sillero-Zubiri, Gottelli, and Macdonald 1996). If more than one mother–father pair was still not excluded, then we accepted the results with the higher confidence or with fewer mismatches if the confidences were equal. If the results for different mother–father pairs were equal, then parentage was unresolved using CERVUS and sibship reconstruction using the program COLONY was used to infer genealogical relationships further.
We then used COLONY to assess parentage of pups with yet unresolved maternity or paternity by inferring full and half siblings among pups. Each subpopulation was analyzed separately with no parents assigned a priori in the first instance and with mothers assigned a priori where possible in the second instance (Figure 2). Pups with unresolved mothers (or fathers) were assigned maternity (or paternity) if they were identified as full siblings of one or more pups for which maternity (or paternity) was resolved based on exclusion and/or CERVUS analysis.
Simulation parameters
The error rate for the parentage simulations (termed eg in Morrissey and Wilson 2005) was set at the default of 0.01, which is slightly more conservative than the error rate (0.008) estimated for our data set (Randall 2006; Randall DA, Pollinger JP, Tallents LA, Macdonald DW, Wayne RK, unpublished data). The error rate used in our parentage likelihood assignments (termed el in Morrissey and Wilson 2005) was set to 0.001. Although the advanced options in CERVUS allow eg and el to vary, most studies set eg = el. Morrissey and Wilson (2005) demonstrated that setting 0 < el < eg (i.e., underestimating the true error rate in the likelihood tests) produced higher assignment success while remaining robust to genotyping errors (i.e., not excluding true parents on the basis of mismatches). Open and restricted simulations of paternity were run separately for each pack using subpopulation- and pack-specific parameters, including the number of candidate fathers, the proportion of candidate fathers sampled, and the effect of relatives in the population (see Supplementary Table S1). We simulated the effect of relatives on the paternity results by including average relatedness among within-pack adult males and the average number of adult males in each pack as relatives of the true parent. Simulations were run for 10 000 cycles, and statistical significance was determined at 80% and 95% confidence. In COLONY, we used subpopulation-specific allele frequencies and genotyping error rates (Randall 2006; Randall DA, Pollinger JP, Tallents LA, Macdonald DW, Wayne RK, unpublished data).
Relatedness, kinship, and inbreeding
We applied rarefaction analysis (by Geffen E, as in Girman et al. 1997) to determine the number of microsatellites needed for accurate estimates of relatedness (Queller and Goodnight 1989). Pairwise relatedness (R) was determined using RELATEDNESS 5.0 (Queller and Goodnight 1989) after including a bias correction of population allele frequencies by pack. Standard errors (SEs) were determined by jackknifing over loci. Unrelated pairwise R values were estimated from 15 dyads of males in nonneighboring packs in a subpopulation. We also tested whether individuals in the same pack were more related to each other than to individuals in other packs and whether individuals in neighboring packs were more related to each other than to individuals in nonneighboring packs. To do this, we compared observed pairwise relatedness values for dyads within packs (and for dyads between neighboring packs) to permutations that randomly assigned all individuals to packs with the same size and age/sex structure as those observed (after Girman et al. 1997; de Ruiter and Geffen 1998). Permutation tests were carried out separately for the 2 subpopulations using SAS macros (SAS 1990). Results were considered statistically significant if the observed values were in the range of either the lower or upper 2.5% of the simulated values in 1000 permutations (Manly 1994). Likelihood analyses implemented in KINSHIP (Goodnight and Queller 1999) were used to assess kinship between adults within packs.
To test for inbreeding avoidance, we compared the relatedness of breeding females to within-pack fathers versus extrapack fathers using a pairwise t-test (SPSS Inc., Chicago, IL). In order to determine whether EPCs might enable females to obtain genetically unrelated mates, we compared the average relatedness between breeding females and adult males in their packs to simulations in which males were randomly assigned to packs, keeping the same number of males in each pack as that observed. Because relatedness may be higher between neighboring packs, we also compared relatedness between breeding females and adult males in their packs to simulations in which the males randomly assigned to packs were restricted to those in neighboring groups.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONSUPPLEMENTARY MATERIALREFERENCES |
|---|
Genetic data set
There were no significant heterozygote deficits for any locus or subpopulation following Bonferroni correction (P > 0.05). Mean FIS values were not significantly different from zero but slightly negative (–0.04 to –0.09, see Supplementary Table S2), suggesting random mating within subpopulations (Dobson et al. 1997
). Seventeen pairs of loci showed significant LD following Bonferroni correction (P < 0.05), but patterns were not consistent across subpopulations suggesting that LD is probably due to population structure rather than physical linkage. Because weak linkage is unlikely to affect the analyses (Meagher 1986), we included all 17 loci in our analyses. The 17 microsatellites in our data set had a low overall probability of identity among siblings (PIDsib = 3.79 x 10–6). This is equivalent to approximately 0.0004% of full siblings sharing the same genotype by chance, implying that identical genotypes were unlikely in our population. Even if the 3 most informative loci (i.e., those with the highest per locus PIDsib values) were excluded due to missing data, the overall PIDsib was 1.55 x 10–5 (approximately 0.002% chance of identical genotypes). Our relatedness values also changed little after 12 loci were used, implying that the 17 loci used in our analyses provided high resolution (see Supplementary Figure S1). Locus C05.377 had small, positive null allele frequencies in WV and SN according to CERVUS, thus we ignored homozygous–homozygous parent–offspring mismatches at this locus.
Parentage
Our study population consisted of 179 individuals in 17 focal packs from which we obtained 148 individual genotypes (Table 1). Pack sizes (adults and yearlings) ranged from 2 to 11 with a mean ±SE of 7.2 ± 0.7 (n = 17). Seven packs (41%) had multiple adult females (2 adult females in 5 packs; 3 adult females in 2 packs). Only 13 of 17 (76%) packs reproduced successfully (i.e., pups emerged from the den), producing a total of 57 pups. Litter sizes at emergence ranged from 1 to 6 pups (mean ± SE = 4.4 ± 0.5, n = 13).
|
We analyzed parentage for 52 of 57 (91%) offspring born in our study population (5 pups were either not sampled or had incomplete genotypes, Table 1). Maternity was determined for 49 pups in 12 litters, of which 48 were assigned to the dominant female and 1 to a subordinate female (GAR, Table 2). In all cases, dominant females were confirmed behaviorally by pregnancy and/or lactation, although the subordinate breeder showed no such evidence of having bred. Only one pup (MEG) had more than one adult female that could not be excluded as the true mother, and both adult females showed signs of pregnancy and lactation suggesting that they both reared pups that year. Mother–offspring mismatches occurred for only 3 pups (2 in DAR, 1 in DOD), but maternity assignments were determined by sibship reconstruction and confirmed by mother–father–offspring trios assigned at 95% confidence. Given a mutation rate of 10–2 to 10–3 per generation for canid dinucleotide and tetranucleotide markers (Francisco et al. 1996
), we could expect approximately 8 mother–offspring mismatches in our data set due to mutation alone. Maternity was unresolved for 2 pups in MUL pack because the only adult female in that pack was not genotyped (Tables 1 and 2).
|
Forty-seven of 49 pups with resolved maternity were sired by males we genotyped in this study, and a minimum of 14 fathers were identified for the 12 litters with resolved maternity (Table 2). Thirty-seven paternities were assigned at 95% confidence, 6 paternities were assigned at 80% confidence, and 4 paternities were assigned by sibship reconstruction. Paternity was unresolved for 3 pups with unresolved maternity (1 in MEG and 2 in MUL) and 2 pups for which no genotyped male could be assigned (BBC). A single father sired all the pups in 7 litters for which all pups were genotyped, and a single father was assumed in an eighth litter (BAT) with one unsampled pup. Two fathers were confirmed in 3 litters in which all pups had resolved paternity and suspected in a fourth litter (BBC) in which 2 pups had unresolved paternity. Thus, the number of sires per female was 1–2, and the rate of multiple paternity was 33% (4 of 12 litters). Overall, 13 pups with resolved paternity were sired by extrapack males (Table 2); in all cases the father was from a neighboring pack (Figures 1 and 5). Three breeding females (DOD, KOT, BAD) had both within pack sires and extra-pack sires. A fourth breeding female (BBC) is assumed to also have both a within pack and extra-pack sire (if the 2 pups with unresolved paternity were sired by the unsampled within pack male). Six females had only within pack sires, and 2 females (both in GAR) had only extra-pack sires. Our paternity results suggest an EPP rate of 28% (13 of 47) of pups and 50% (6 of 12) of litters.
|
Overall, adult males sired 0 to 6 pups in a breeding season, but reproductive success among males was highly skewed with 67% and 75% of males analyzed siring zero offspring in WV and SN, respectively. Dominance appeared to be an important predictor of reproductive success. In 4 of 6 packs for which social status within packs was deduced, the dominant male was assigned paternity of all pups (DAR, n = 6; TAR, n = 5, WOL, n = 1), and in the other 2 the dominant male was assigned paternity for the majority of pups (KOT, n = 5 of 6; DOD, n = 3 of 4).
Relatedness
Mean (±SE) pairwise R values calculated for dyads with known relationships were similar to the theoretical values of 0.5, 0.5, and 0 for parent–offspring, full sibling, and unrelated dyads, respectively (mother–offspring, n = 49, R = 0.494 ± 0.029; father–offspring, n = 47, R = 0.511 ± 0.032; full siblings, n = 78, R = 0.522 ± 0.025; unrelated, n = 15, R = –0.031 ± 0.042), but half siblings had a higher mean R than the expected value of 0.25 (n =24, R = 0.313 ± 0.050, Figure 3).
|
Mean (±SE) pairwise R values within packs were 0.298 ± 0.016 for adult and yearling dyads (n = 16), 0.392 ± 0.065 for adult females (n = 4), and 0.333 ± 0.023 for adult males (n = 15) (Figure 3). All values were significantly higher than relatedness determined by random distribution of individuals (P < 0.001, Table 3), except for adult female comparisons in WV for which we had only one dyad. Mean relatedness between individuals in neighboring packs was also lower than that between individuals in the same pack for all age/sex categories but did not consistently differ significantly from relatedness determined by randomly allocating wolves to neighboring packs (Table 3). In WV, relatedness was significantly higher between neighbors than a random distribution for adult and yearling comparisons (R = 0.069, P < 0.0001) and male-only comparisons (R = 0.104, P < 0.0001) but not female-only comparisons (R = –0.084, P = 0.90). In SN there was no significant difference between neighbor relatedness compared with a random distribution for adult and yearling comparisons (R = 0.014, P = 0.90) and male-only comparisons (R = –0.026, P = 0.090) but significantly higher relatedness for female comparisons (R = 0.054, P < 0.0001).
|
We used likelihood tests (Goodnight and Queller 1999
) to determine if we could classify kinship among adults within packs. For the full sibling/parent–offspring (primary hypothesis) versus half sibling (null hypothesis), the estimated proportion of tests resulting in type II errors (falsely accepting the null hypothesis) at P = 0.05 was 0.348 and 0.323 in WV and SN, respectively, even assuming R = 0.25 for half siblings. Thus, substantial overlap in the distribution of simulated R values makes the results of such a test equivocal (Figure 4). Predicted type II error rates were similarly high (0.426 and 0.450) for the significance test of half siblings versus unrelated individuals. Therefore, we tested whether adult dyads within packs were significantly more likely to be full siblings/parent–offspring or unrelated as this produced lower type II error rates (0.038 and 0.033). Overall, 100% of female dyads (n = 8) and 77% (36 of 47) male dyads were significantly more likely to be full siblings/parent–offspring than unrelated. On the other hand, only 43% (24 of 56) female–male dyads were significantly more likely to be full siblings/parent–offspring than unrelated. Within packs, mean relatedness between opposite sex adults (n = 14, R = 0.103) was also lower than relatedness for same-sex comparisons (Figure 3), and the pattern was fairly consistent across packs (data not shown).
|
Adult kinship patterns within packs varied (Figure 5). Three packs (ALA, DOD, BAT) consisted of 2 or more related males (R = 0.23–0.72; KINSHIP full sibling/parent–offspring test, all P < 0.05) paired to an unrelated female (R = –0.49 to –0.08, all P > 0.05). Four packs (DAR, KOT, TAR, BAD) consisted of only an unrelated dominant pair (R = –0.20 to 0.06, all P > 0.05) and philopatric male offspring related to both the dominant individuals (R = 0.27–0.53, all P < 0.05) or to the dominant female only (one pup in DAR, R = 0.27, P < 0.05). In BBC and GAR, adults had mixed kinship: BBC had 2 related males (R = 0.54, P < 0.05) unrelated to the breeding female (R = –0.27, both P > 0.05) and 1 adult male that may have been the offspring of the dominant female (R = 0.34, P < 0.05); GAR had 1 adult male unrelated to the dominant female (R = –0.16, P > 0.05), 1 adult male that may have been the offspring of the dominant female (R = 0.16, P < 0.05), and one adult male that was unrelated to the dominant female (R = 0.04, P > 0.05) and unrelated to the other 2 adult males (R = –0.11 to 0.22, both P > 0.05) but related as a full sibling/parent–offspring to a subordinate female (R = 0.31, P < 0.05). Two subordinate adult females in GAR were related as full sibling/parent–offspring to the dominant female (R = 0.50 and 0.54, both P < 0.05) and one or more males (R = 0.31–0.72, all P < 0.05), implying female philopatry. We also observed strong kinship between 2 adult females in both MEG (R = 0.13, P < 0.05) and QUA (R = 0.50, P < 0.05), although we had insufficient evidence to infer sister or mother–daughter relationships. WOL and QUA appeared inbred with R values consistent with full sibling/parent–offspring kinship in almost all pairwise comparisons (R = 0.27–0.71, all P < 0.05). Only 2 adult males were unrelated to all other adults suggesting possible male immigration (BAD, R = –0.10 to 0.05, all P > 0.05; NYA, R = –0.28 to 0.08, all P > 0.05). All yearlings were related to the dominant female in their pack as parent–offspring (R = 0.31–0.67, all P < 0.05).
Inbreeding
Kinship analysis showed that 11 of 15 (73%) of successful matings (7 of 9 within pack and 4 of 6 extrapack) were between unrelated individuals (R = –0.26 to 0.22, P > 0.05). Four apparently incestuous matings resulted in 9 of 47 (19%) pups being potentially inbred. In 2 packs (WOL and QUA), females may have been related to the within-pack sires of their pups (R = 0.18 and 0.44, respectively; KINSHIP full sibling/parent–offspring test, both P < 0.05). In KOT and GAR (subordinate female), females mated with extrapack males that were significantly more likely to be related to them as full siblings/parent–offspring than unrelated (R = 0.41 and 0.44, P < 0.01). On average, breeding females were more related to adult males in their pack than to adult males in other packs in the same subpopulation. For WV, the mean R value was 0.093 (mean of randomizations = –0.017, permutation test, P = 0.02); for SN, the mean R value was 0.034 (mean of randomizations = –0.123, P < 0.01). However, breeding females were similarly related to males in their pack and to males in neighboring packs in WV (mean of randomizations = 0.077, P = 0.65) and SN (mean of randomizations = 0.020, P = 0.60). Actual mean R values between breeding females and neighboring males were 0.048 and –0.039 in WV and SN, respectively. In 2 of 3 packs where breeding females mated both within and outside their pack, females were more (not less) related to extrapack sires of their offspring than within-pack sires, but the difference was not significant (pairwise t-test: t = –0.944, df = 2, P = 0.445), but our sample size was small.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONSUPPLEMENTARY MATERIALREFERENCES |
|---|
Our results demonstrate higher relatedness within Ethiopian wolf packs than between packs, implying that social cohesion and cooperation are associated with kinship ties. Values were similar for adults (females R = 0.39, males R = 0.33) and yearlings (R = 0.30), all of which approximate relatedness among half siblings (R = 0.31). These results have also been found in other social canids, such as African wild dogs (Lycaon pictus, R = 0.27 overall within packs, R = 0.28 for adult females, and R = 0.35 for adult males, Girman et al. 1997
). Although adults of the same sex are typically related, breeding pairs were typically unrelated, including both within-pack mates (7 of 9 breeding pairs) and extrapack mates (4 of 6 breeding pairs). However, like most canids, Ethiopian wolves also show large variation in mean kinship within packs, which is probably due to variation in the number of EPPs among litters and degree of natal philopatry among offspring. Observational data suggest that packs and breeding units are usually formed by 1) alliances between one or more males and an immigrant floater female, 2) pack fission by a subordinate female and one or more same-pack males that together split from their natal pack, and 3) subordinate female ascendance to breeding status after the death of the incumbent (Sillero-Zubiri 1994; Marino 2003; Sillero-Zubiri et al. 2004). Our genetic data on kinship within packs lend support to the occurrence of all 3 mechanisms, although the latter 2 are difficult to discern from genetic evidence alone. In packs for which we had adequate data on kinship (n = 11), breeding females were either unrelated to their within-pack mates (n = 6) or unrelated to at least one male in their pack (n = 3). In the remaining packs (n = 2), we can infer the following. In QUA pack, the breeding female was related to all other individuals in her pack, except one adult male but including a subordinate female. Thus, she ascended to breeding status in her natal pack (in which case the subordinate female is her sister or daughter), split from her natal pack with other individuals, including the subordinate female (alternatively the subordinate female may be her daughter from a subsequent litter), or mated with her son (in which case the unrelated male may be her mate from a previous breeding season and the other individuals their offspring). In WOL pack, the 2 related individuals are probably siblings that split from their natal pack together. These results provide genetic evidence that breeding females are most likely to be immigrants, resulting in low relatedness between mates and reducing the potential for inbreeding. On the other hand, female philopatry appears to increase the likelihood of inbreeding within packs (e.g., QUA, WOL) or between packs (as a result of EPCs, as in the case of the GAR subordinate female).
Our genetic results also support behavioral observations that only one female usually breeds per pack (Sillero-Zubiri et al. 2004). Only 1 of 7 multifemale packs in our study had a subordinate female breeder (GAR pack); she reared a single pup in the same den as 3 pups produced by the dominant female. Behavioral observations suggest that the subordinate female in MEG pack at least attempted to raise pups in the same den as the dominant female, although maternity of the lone pup analyzed in this study was unresolved. Polyandry and EPCs were not uncommon, with multiple paternity and EPP occurring in 33% and 50% of litters, respectively, and EPP accounted for 28% of pups. Similar results were found in a study of red foxes (Vulpes vulpes), which demonstrated high levels of polygynous and polyandrous matings within and among social groups (Baker et al. 2004). However, the levels of multiple paternity and EPP in Ethiopian wolves and red foxes are higher than those reported in other canids (Girman et al. 1997; Roemer et al. 2001). As expected from behavioral observations of EPCs in Ethiopian wolves (Sillero-Zubiri, Gottelli, and Macdonald 1996), all EPPs were obtained by males in neighboring packs in this study, including at least one EPP obtained by a subordinate male.
We found no evidence that EPCs reduce inbreeding in this population. There was a trend toward higher (not lower) relatedness with extrapack sires, although sample sizes were too small for reliable statistical tests. However, these results at least provide evidence that females cannot or do not consistently choose unrelated males during EPCs. Furthermore, although breeding females were significantly more related to males within their pack than to a random distribution of males, there was no difference in breeding female relatedness to males in her pack and males in neighboring packs. Because EPPs are restricted to males in neighboring packs, EPCs do not appear to increase the probability of producing outbred offspring. Thus, females may seek EPCs for other reasons such as infanticide avoidance (Wolff and Macdonald 2004, but see below).
Near-zero FIS values imply that inbreeding within subpopulations is neither promoted nor minimized by dispersal and mating behavior (i.e., mating is random). Similar findings have been reported in other social species such as alpine marmots (Marmota marmota, Goossens et al. 2001) and black-tailed prairie dogs (Cynomys ludovicianus, Foltz and Hoogland 1983). Under such conditions, the number of incestuous matings observed (4 of 15 mated pairs in this study) should reflect that predicted by chance in a population of this size. However, inbreeding may be higher than it was historically because population sizes have decreased from historic levels (Gottelli et al. 2004). Furthermore, outbreaks of infectious disease (Sillero-Zubiri, King, and Macdonald 1996; Randall et al. 2004), which disrupt social cohesion and appear to mediate male dispersal, may increase the chances of inbreeding in a population, for instance, if females are more philopatric or disperse shorter distances, males disperse more often, or both. For example, Sugg et al. (1996) demonstrated that inbreeding increased in black-tailed prairie dogs when natal dispersal of the normally philopatric sex was increased as this increased the probability of mating between close kin. Given the frequency of disturbances in this population as a result of recurring disease outbreaks (reviewed in Randall et al. 2006), it is also conceivable that EPCs may have evolved as an inbreeding avoidance strategy but are no longer functioning as such. In particular, demographic events and changes in dispersal behavior may have resulted in neighboring packs and breeding pairs being more closely related than they would have been historically. This may explain the pattern of high relatedness between neighboring packs in WV but not SN as WV has had higher number of pack disruptions and/or extinctions in 16 years of monitoring (1988–2004), primarily due to disease outbreaks (Sillero-Zubiri 1994; Randall et al. 2004; Marino et al. 2006). In summary, due to recent demographic changes, the level of inbreeding recorded in this study may be higher than is typical for this population historically.
A fair amount of research has been devoted to understanding variation in mating systems and social structure among canids (Bekoff et al. 1981; Moehlman 1989; Geffen et al. 1996). Often empirical data demonstrate that social structure results from a number of selective pressures operating simultaneously and teasing apart the various processes can be difficult. Previous studies suggested that group formation and group augmentation in Ethiopian wolves were mediated by 1) the low costs of tolerating relatives when resources are abundant and 2) retention or acquisition of high-quality territories (Sillero-Zubiri and Gottelli 1995b; Sillero-Zubiri and Macdonald 1998; Marino 2003). We show that a reduction in inbreeding may be an adaptive advantage of female-biased dispersal and mating behavior. However, inbreeding occurs by chance when dispersal and mating are random as we observed among Ethiopian wolves in this study. Genetic data from additional breeding seasons and populations are being collected as part of the long-term monitoring plan for Ethiopian wolves in the Bale Mountains and elsewhere in Ethiopia to enable a more accurate assessment of the level of inbreeding in this species. Given its potential deleterious effects on fitness and long-term population viability (Hedrick and Kalinowski 2000), inbreeding among Ethiopian wolves should be monitored regularly, especially if population sizes fluctuate or are further reduced overall.
|
SUPPLEMENTARY MATERIAL |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONSUPPLEMENTARY MATERIALREFERENCES |
|---|
Supplementary Figure S1 and Tables S1 and S2 can be found at http://www.beheco.oxfordjournals.org/.
|
ACKNOWLEDGEMENTS |
|---|
We thank the Wildlife Conservation Department, the Oromiya Rural Land and Natural Resources Administration Authority, and the Bale Mountains National Park for permission to undertake research in Ethiopia. We are grateful for the support of the Ethiopian Wolf Conservation Programme for their assistance in the field and to Jorgelina Marino, Claudio Sillero-Zubiri, and Hannah Dugdale for comments on an earlier draft of the manuscript. Thanks also to Toby Dunn for assistance with the SAS code for the randomizations. Fieldwork was funded by the Iris Darnton Trust, Jesus College, and the Sophie Danforth Conservation Biology Fund (Rhode Island Park Zoo). D.A.R. was supported by an Overseas Student Award (UK) and a Clarendon Fund Award from the University of Oxford.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONSUPPLEMENTARY MATERIALREFERENCES |
|---|
Allen PJ, Amos W, Pomeroy PP, Twiss SD. Microsatellite variation in grey seals (Halichoerus grypus) shows evidence of genetic differentiation between two British breeding colonies. Mol Ecol (1995) 4:653–662.[Medline]
Baker PJ, Funk SM, Bruford MW, Harris S. Polygynandry in a red fox population: implications for the evolution of group living in canids? Behav Ecol (2004) 15:766–778.
Bekoff M, Diamond J, Mitton JB. Life-history patterns and sociality in canids: body size, reproduction, and behavior. Oecologia (1981) 50:386–390.[CrossRef][Web of Science]
Breen M, Jouquand S, Renier C, Mellersh CS, Hitte C, Holmes NG, Cheron A, Suter N, Vignaux F, Bristow AE, et al. Chromosome-specific single-locus FISH probes allow anchorage of an 1800-marker integrated radiation-hybrid/linkage map of the domestic dog genome to all chromosomes. Genome Res (2001) 11:1784–1795.
Brown JL. Territorial behaviour and population regulation in birds. Wilson Bull (1969) 81:293–329.
Brown JL. Helping and communal breeding in birds: ecology and evolution (1987) Princeton, NJ: Princeton University Press.
Clutton-Brock TH. Mammalian mating systems. Proc R Soc Lond B Biol Sci (1989) 236:339–372.[Medline]
Creel S, Macdonald D. Sociality, group size, and reproductive suppression among carnivores. Adv Study Behav (1995) 24:203–257.
de Ruiter JR, Geffen E. Relatedness of matrilines, dispersing males and social groups in long-tailed macaques (Macaca fascicularis). Proc R Soc Lond B Biol Sci (1998) 265:79–87.[Medline]
Dobson FS. Competition for mates and predominant juvenile male dispersal in mammals. Anim Behav (1982) 30:1183–1192.[CrossRef][Web of Science]
Dobson FS, Chesser RK, Hoogland JL, Sugg DW, Foltz DW. Do black-tailed prairie dogs minimize inbreeding? Evolution (1997) 51:970–978.[CrossRef][Web of Science]
Emlen ST. The evolution of helping. I. An ecological contraints model. Am Nat (1982) 119:29–39.[CrossRef][Web of Science]
Evett IW, Weir BS. Interpreting DNA evidence: statistical genetics for forensic scientists. (1998) Sunderland (MA): Sinauer.
Foltz DW, Hoogland JL. Genetic evidence of outbreeding in the black-tailed prairie dog (Cynomys ludovicianus). Evolution (1983) 37:784–797.
Francisco LV, Langston AA, Mellersh CS, Neal CL, Ostrander EA. A class of highly polymorphic tetranucleotide repeats for canine genetic mapping. Mamm Genome (1996) 7:359–362.[CrossRef][Web of Science][Medline]
Geffen E, Gompper ME, Gittleman JL, Luh H-K, Macdonald DW, Wayne RK. Size, life-history traits, and social organization in the Canidae: a reevaluation. Am Nat (1996) 147:140–160.[CrossRef][Web of Science]
Girman DJ, Mills MGL, Geffen E, Wayne RK. A molecular genetic analysis of the social structure, dispersal, and interpack relationships of the African wild dog (Lycaon pictus). Behav Ecol Sociobiol (1997) 40:187–198.[CrossRef][Web of Science]
Goodnight KF, Queller D. Computer software for performing likelihood tests of pedigree relationship using genetic markers. Mol Ecol (1999) 8:1231–1234.[CrossRef][Medline]
Goossens B, Chikhi L, Taberlet P, Waits LP, Allaine D. Microsatellite analysis of genetic variation among and within Alpine marmot populations in the French Alps. Mol Ecol (2001) 10:41–52.[CrossRef][Medline]
Gottelli D, Marino J, Sillero-Zubiri C, Funk SM. The effect of the last glacial age on speciation and population genetic structure of the endangered Ethiopian wolf (Canis simensis). Mol Ecol (2004) 13:2275–2286.[CrossRef][Medline]
Gottelli D, Sillero-Zubiri C, Applebaum GD, Roy MS, Girman DJ, Garcia-Moreno J, Ostranders EA, Wayne RK. Molecular genetics of the most endangered canid: the Ethiopian wolf Canis simensis. Mol Ecol (1994) 3:301–312.[Medline]
Greenwood PJ. Mating systems, philopatry and dispersal in birds and mammals. Anim Behav (1980) 28:1140–1162.[CrossRef][Web of Science]
Hedrick PW, Kalinowski ST. Inbreeding depression in conservation biology. Annu Rev Ecol Syst (2000) 31:139–162.[CrossRef][Web of Science]
Jennions MD, Petrie M. Why do females mate multiply? A review of the genetic benefits. Biol Rev (2000) 75:21–64.[Medline]
Johnson ML, Gaines MS. Evolution of dispersal: theoretical models and empirical tests using birds and mammals. Annu Rev Ecol Syst (1990) 21:449–480.[CrossRef][Web of Science]
Koenig WD. Reproductive success, group size, and the evolution of cooperative breeding in the Acorn Woodpecker. Am Nat (1981) 117:421–443.[CrossRef][Web of Science]
Kruuk H, Macdonald DW. Group territories of carnivores: empires and enclaves. In: Behavioral ecology. Ecological consequences of adaptive behaviour—Sibly RM, Smith RH, eds. (1985) Oxford: Blackwell Publishers. 521–536.
Macdonald DW, Carr GM. Food security and the rewards of tolerance. Comparative socioecology: the behavioural ecology of humans and other mammals. In: Spec Publ Bri Ecol Soc—Standen V, Foley RA, eds. (1989) 8:75–99.
Macdonald DW, Creel S, Mills MGL. Society: canid society. In: Biology and conservation of wild canids—Macdonald DW, Sillero-Zubiri C, eds. (2004) Oxford: Oxford University Press. 85–106.
Manly BF. Randomisation and Monte Carlo methods in biology. (1994) London: Chapman and Hall.
Marino J. The spatial ecology of the Ethiopian wolf, Canis simensis [D. Phil. thesis]. (2003) Oxford (UK): University of Oxford.
Marino J, Sillero-Zubiri C, Macdonald DW. Trends, dynamics and resilience of an Ethiopian wolf population. Anim Conserv (2006) 9:49–58.[Medline]
Marshall T, Slate J, Kruuk L, Pemberton J. Statistical confidence for likelihood-based paternity inference in natural populations. Mol Ecol (1998) 7:639–655.[CrossRef][Medline]
Meagher TR. Analysis of paternity within a natural population of Chamaelirium luteum I. Identification of most-likely male parents. Am Nat (1986) 128:199–215.[CrossRef][Web of Science]
Moehlman PD, Gittleman JL. Intraspecific variation in canid social systems. In: Carnivore behavior, ecology, and evolution (1989) Ithaca (NY): Cornell University Press. 143–163.
Morrissey MB, Wilson AJ. The potential costs of accounting for genotyping errors in molecular parentage analyses. Mol Ecol (2005) 14:4111–4121.[CrossRef][Medline]
Navidi W, Arnheim N, Waterman MS. A multiple-tube approach for accurate genotyping of very small DNA samples by using PCR: statistical considerations. Am J Hum Genet (1992) 50:347–359.[Web of Science][Medline]
Olivier M, Breen M, Binns MM, Lust G. Localization and characterization of nucleotide sequences from the canine Y chromosome. Chromosome Res (1999) 7:223–233.[CrossRef][Web of Science][Medline]
Pusey A, Wolf M. Inbreeding avoidance in animals. Trends Ecol Evol (1996) 11:201–206.[CrossRef]
Queller D, Goodnight K. Estimating relatedness using genetic markers. Evolution (1989) 43:258–275.[CrossRef][Web of Science]
Randall DA. Determinants of genetic variation in the Ethiopian wolf, Canis simensis [D. Phil. thesis]. (2006) Oxford (UK): University of Oxford.
Randall DA, Marino J, Haydon DT, Sillero-Zubiri C, Knobel DL, Tallents LA, Macdonald DW, Laurenson MK. An integrated disease management strategy for the control of rabies in Ethiopian wolves. Biol Conserv (2006) 131:151–162.[CrossRef]
Randall DA, Williams SD, Kuzmin IV, Rupprecht CE, Tallents LA, Tefera Z, Argaw K, Shiferaw F, Knobel DL, Sillero-Zubiri C, et al. Rabies in endangered Ethiopian wolves. Emerging Infect Dis. 10:2214–2217.
Raymond M, Rousset F. GENEPOP (version 1.2): population genetics software for exact tests and ecumenicism. J Hered (1995) 86:248–249.
Rice WR. Analyzing tables of statistical tests. Evolution (1989) 43:223–225.[CrossRef][Web of Science]
Roemer GW, Smith DA, Garcelon DK, Wayne RK. The behavioural ecology of the island fox (Urocyon littoralis). J Zool (2001) 255:1–14.[CrossRef][Web of Science]
SAS. SAS.STAT user's guide, version 6. (1990) 4th ed. Cary (NC): SAS Institute.
Selander RK. Speciation in wrens of the genus Campylorhynchus. Univ Calif Publ Zool (1964) 74:1–224.
Sillero-Zubiri C. Behavioural ecology of the Ethiopian wolf, Canis simensis [D. Phil. thesis]. (1994) Oxford (UK): University of Oxford.
Sillero-Zubiri C, Gottelli D. Diet and feeding behavior of Ethiopian wolves (Canis simensis). J Mamm (1995a) 76:531–541.[CrossRef][Web of Science]
Sillero-Zubiri C, Gottelli D. Spatial organization in the Ethiopian wolf Canis simensis: large packs and small stable home ranges. J Zool (Lond) (1995b) 237:65–81.
Sillero-Zubiri C, Gottelli D, Macdonald DW. Male philopatry, extra-pack copulations and inbreeding avoidance in Ethiopian wolves (Canis simensis). Behav Ecol Sociobiol (1996) 38:331–340.[CrossRef][Web of Science]
Sillero-Zubiri C, Johnson PJ, MacDonald DW. A hypothesis for breeding synchrony in Ethiopian wolves (Canis simensis). J Mamm (1998) 79:853–858.[CrossRef][Web of Science]
Sillero-Zubiri C, King AA, Macdonald DW. Rabies and mortality in Ethiopian wolves (Canis simensis). J Wildl Dis (1996) 32:80–86.[Abstract]
Sillero-Zubiri C, Macdonald DW. Scent-marking and territorial behaviour of Ethiopian wolves Canis simensis. J Zool (1998) 245:351–361.[CrossRef][Web of Science]
Sillero-Zubiri C, Marino J, Gottelli D, Macdonald DW. Ethiopian wolves. In: Biology and conservation of wild canids—Macdonald DW, Sillero-Zubiri C, eds. (2004) Oxford: Oxford University Press. 311–322.
Stacey PB, Ligon JD. The benefits-of-philopatry hypothesis for the evolution of cooperative breeding: variation in territory quality and group size effects. Am Nat (1991) 137:831–846.[CrossRef][Web of Science]
Sugg DW, Chesser RK, Dobson FS, Hoogland JL. Population genetics meets behavioral ecology. Trends Ecol Evol (1996) 11:338–342.[CrossRef]
Taberlet P, Griffin AS, Goossens B, Questiau S, Manceau V, Escaravage N, Waits LP, Bouvet J. Reliable genotyping of samples with very low DNA quantities using PCR. Nucleic Acids Res (1996) 26:3189–3194.
Valière N. GIMLET: a computer program for analysing genetic individual identification data. Mol Ecol Notes (2002) 2:377–379.[CrossRef][Web of Science]
Waits LP, Luikart G, Taberlet P. Estimating the probability of identity among genotypes in natural populations: cautions and guidelines. Mol Ecol (2001) 10:249–256.[CrossRef][Medline]
Wang J. Sibship reconstruction from genetic data with typing errors. Genetics (2004) 166:1963–1979.
Waser PM. Resources, philopatry, and social interactions among mammals. In: The ecology of social behavior—Slobodchikoff CN, ed. (1988) New York: Academic Press. 109–130.
Waser PM. Patterns and consequences of dispersal in gregarious carnivores. In: Carnivore behavior, ecology, and evolution—Gittleman JL, ed. (1996) Ithaca (NY): Comstock Publishing Associates. 267–295.
Wolff JO. Parents suppress reproduction and stimulate dispersal in opposite-sex juvenile white-footed mice. Nature (1992) 359:409–410.[CrossRef][Medline]
Wolff JO. What is the role of adults in mammalian juvenile dispersal. Oikos (1993) 68:173–176.[CrossRef][Web of Science]
Wolff JO, Macdonald DW. Promiscuous females protect their offspring. Trends Ecol Evol (2004) 19:127–134.[CrossRef][Medline]
Woolfenden GE, Fitzpatrick JW. The inheritance of territory in group-breeding birds. Bioscience (1978) 28:104–108.[CrossRef][Web of Science]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  G. Iossa, C. D. Soulsbury, P. J. Baker, K. J. Edwards, and S. Harris Behavioral changes associated with a population density decline in the facultatively social red fox Behav. Ecol., March 1, 2009; 20(2): 385 - 395. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||