Density dependence in voles and mice: A comparative study

The presence of direct and delayed density dependence in populations of three sympatric rodent species (Clethrionomys rufocanus, Apodemus speciosus, and A. argenteus) in Hokkaido, Japan, was evaluated using triannual census data (spring, summer, and fall) spanning 30 years (1963-1992) on 79 populations for each species. The average abundance and population variability (the s-index) generally increased from spring to fall in C. rufocanus but were typically highest in summer for the Apodemus populations. Based on a comprehensive and comparative review of the population and community biology of the species we made four explicit predictions about the pattern of density dependence: (1) the three species were expected to exhibit socially induced direct density dependence, but (2) this was expected to be weaker in the Apodemus species than C. rufocanus; (3) delayed density dependence caused by predation was only expected in C. rufocanus; thus (4) time series of C. rufocanus were expected to reflect a second-order dynamic process, and those of the Apodemus species were expected to reflect a first-order process. Dennis and Taper's method based on the Gompertz model was used to test for direct and delayed density dependence and thereby to test the predictions. Direct density dependence was detected in most series (81.0-97.5%) for all three seasons and for all three species. A significant proportion of the time series of C. rufocanus (11.8-18.5%) exhibited negative delayed density dependence, whereas detection rates in the two Apodemus species did not differ from that expected by chance alone. Autoregressive analyses corroborated this: a secondorder process was commonly found to be the appropriate model for the time series of C. rufocanus, whereas a first-order process was preferred for most time series of the Apodemus species. The high incidence of direct density dependence in all three species and the contrasting results on delayed density dependence between C. rufocanus and the Apodemus species are discussed with reference to social and trophic interactions. Territoriality, delayed maturation, and reduced pregnancy rates are probable causes for the high incidence of direct density dependence in all species. The more unpredictable variability in Apodemus food resource is argued to have a potential to disrupt social regulation and thus to lower the incidence of direct density dependence. A candidate mechanism for the incidence of delayed density dependence is differential vulnerability to predation: the demography of C. rufocanus is much more affected by predators than Apodemus.