THE IMPACT OF POPULATION KIN-STRUCTURE ON NESTLING SURVIVAL IN TOWNSEND VOLES, MICROTUS-TOWNSENDII

1. In many small rodent species, females are philopatric and may repro duce in proximity to their female relatives. The prevalence of cluster s of relatives in populations reflects previous patterns of immigratio n, local recruitment and mortality. Kin-structuring among females will be most prevalent in populations which have experienced successful re production and philopatric recruitment. If population kin-structure in fluenced key demographic parameters this would introduce a potential t ime-delay in the regulation of population size by social factors. 2. W e tested the influence of population kin structure among female Townse nd's voles on nestling survival, using data from a replicated experime nt with two non-enclosed populations. Most nestlings were marked befor e weaning, so that we knew which recruits were related to one another and to which adult female. The average size of matrilineal groups was experimentally manipulated by removing immigrants, and preventing pred ation by birds, on the high kinship treatments (Kin), and by removing members of family groups on the low kinship treatments (Non-kin). We c ompared preweaning survival of pups marked in the maternal nest betwee n high and low kinship treatments using the recent generalisations of the Cormack-Jolly-Seber statistical models as implemented in the softw are SURGE. 3. The manipulation was effective in that the number of bre eding females with relatives among the six nearest breeding female nei ghbour was higher on the high-kinship treatment grids than on the low- kinship treatment grids. Most first degree related females were neares t neighbours and the prevalence of any relatedness relationship among the six nearest neighbours declined with neighbourhood class. The degr ee of matrilineal relatedness influenced the distance between the nest s (and pups) of breeding females; first degree relatives and second de gree relatives were, respectively, 1.32 and 1.19 times closer to each other than unrelated females. Unrelated females and more distant relat ives had similar internest distances. The relationship between matrili neal relatedness and distance was similar for the six nearest breeding female neighbours and was not influenced by year, season or treatment . 4. The parsimonious capture-recapture models selected included a gra dual increase of trappability during the first 5-7 weeks of life and t hree age-specific survival rates. Survival over the first 2 weeks of l ife was estimated as a single parameter and was relatively low. Weekly survival probabilities were very high (0.96, SE = 0.012) from week 3- 5 and decreased later in life (week 6-15, 0.85, SE = 0.012) possibly a s a result of emigration. 5. The experimental manipulation of female k in-structure had a significant influence on the weekly survival probab ilities of nestlings during their first 2 weeks of life. Pup weekly su rvival was higher on the high-kinship treatment (0.76, SE = 0.018) tha n on the low-kinship treatment (0.67, SE = 0.020). 6. Pairs of first d egree relatives among the six nearest neighbours were present, althoug h in different proportion, within both treatments and we asked whether this within-treatment variability in relatedness was reflected in wit hin-treatment variability in early age survival. Estimates of pup surv ival for the three replicates with the largest sample size suggest tha t this was the case in 1990 (on both treatments) but not in 1991 (high kinship treatment). This suggests that between-treatment differences probably reflect differences in the prevalence of related and unrelate d females rather than the presence of immigrant females on either trea tments. 7. A simple age-structured demographic model, using values of juvenile and adult survival measured for populations with high and low kinship, indicate that this factor may have a substantial impact on t he rate of growth of microtine populations (predicted monthly multipli cation rate lambda = 1.48 vs. 1.72 for low and high kinship, respectiv ely). Thus, the impact of kinship on demographic parameters may introd uce powerful time lags in the social regulation on vole populations.