Consumer-resource interactions: Seeds, mice and cats on Guadalupe Island, Mexico

Background/Question/Methods

Understanding the population dynamics of introduced mammals, and the interactions between these species and their food resources, is important for management purposes. Few studies have researched the functional and numerical responses of both house mice (Mus musculus) and feral cats (Felis catus) to food availability, as usually these species are elements of complex trophic webs. As part of this research, the factors that influence the population dynamics of these two invasive species on the semi-arid Guadalupe Island (24,171 ha), Mexico were investigated. Data was collected quarterly for 2.5 years in two of the main island habitats: grassland (introduced European grasses) and maritime desert scrub (low shrub native vegetation). This research estimated the numerical response of mice to seed availability using an index of mouse abundance, and mark-recapture live-trapping to estimate density. Seed production was estimated using seed traps. Cat exclosures were used to investigate the influence of cat predation on mouse numerical response to seed. The study also focused on estimating functional and numerical responses of cats to changes in mouse abundance. The functional response of cats was estimated from scats, and the numerical response from cat relative abundance (scent-station and spotlight indices).

Results/Conclusions

Results showed a strong numerical response of mice to changes in seed availability in both habitats. However, mouse density differed between grassland and maritime desert scrub, which could be related to dissimilarity in habitat structure. There was no indication of an influence of cat predation on mouse numerical response to seed. Results suggest that mice are limited and regulated by food availability while predation by feral cats seems to be limiting but not regulatory. Cats showed a hyperbolic (Type II) functional response to mice in both habitats, a strong numerical response in grassland and an ill-defined numerical response in maritime desert scrub. This suggests that feral cat numerical response is strongly associated with mice in grassland. The study contributes to our knowledge of the functional and numerical responses of house mice and feral cats on insular environments, particularly where these are the only invasive species. In addition, this research provides useful information for planning ecological management decisions on Guadalupe Island such as feral cat eradication.