- Informed conservation of small mammals, ecosystems, and predators requires a detailed understanding of how small mammals species and communities vary in both space and time, as well as the relative cyclicity and synchrony of this variation. This variation can be especially informative to land managers interested in manipulating the abundance or density of populations, as it often provides inference into the habitat and weather factors to which species and communities are sensitive. While this inference is often informed by spatial or temporal variation in species-specific abundance or density, in some cases, species-specific metapopulation and sink-source dynamics can confound the interpretation of this variation. Thus, more informed inference is based not only on the spatiotemporal variation in abundance or density, but also on the spatiotemporal variability on species-specific vital rates. Here, in Chapter 2, I estimated the abundance, temporal synchrony, and spatiotemporal associations of small mammal populations in a late-successional forest in western Oregon. In Chapter 3, I estimated the vital rates (apparent annual survival, population growth rate, and recruitment) of Humboldt’s flying squirrels (Glaucomys oregonensis) and Townsend's chipmunks (Neotamias townsendii). For both analyses, I used live-trapping data collected on 9 sites from 2011–2016, in a late successional forest.
In Chapter 2, I used Huggins closed-capture models to estimate site-specific abundance of Humboldt’s flying squirrels, Townsend's chipmunks, western red-backed voles (Myodes californicus), and deer mice (Peromyscus maniculatus). Additionally, I used generalized linear mixed effects models to investigate effects of 4 spatial and 3 temporal covariates on species-specific mean abundance estimates. I then estimated the linear correlation between individual counts of the 8 most commonly-caught species of small mammals using Pearson’s correlation coefficients. Lastly, I assessed spatial variation in sex ratios and body mass of focal species among grids. Focal species abundances varied by as much as 4-fold among years and 6-fold among sites, with only slight evidence of linear correlation between species. Humboldt’s flying squirrel abundance was positively autocorrelated at 1 and 5-year intervals, whereas western red-backed vole abundance was negatively autocorrelated at 4 and 5-year intervals. Sex ratios and body masses did not vary widely across grids. Humboldt’s flying squirrels were more abundant on low elevation sites with high berry producing plant cover than on high elevation sites with low berry producing plant cover. Townsend’s chipmunks and western red-backed voles were more abundant on high elevation, open canopy sites than on low elevation, closed canopy sites. Deer mice were slightly more abundant on sites with high berry producing plant cover than on sites with low berry producing plant cover. Minimum winter temperature was positively related to the mean abundance of Humboldt’s flying squirrels and Townsends chipmunks and negatively related to the mean abundance of western red-backed voles and deer mice, while western red-backed voles and deer mice were less abundant after periods of drought. Counts of the 8 most commonly captured species were only weakly correlated.
In Chapter 3, I used robust design Pradel models to estimate site-specific apparent annual survival, population growth rate, and recruitment for Humboldt’s flying squirrels and Townsend's chipmunks. I then used Pearson’s correlation coefficients to estimate the species-specific linear correlations among population growth rate, abundance, apparent annual survival, and recruitment. My estimates were generally intermediate to previous estimates of vital rates. I was able to link abundance-associated covariates with the vital processes most associated with population growth rate. Changes in Humboldt’s flying squirrel population growth rate were strongly correlated with apparent annual survival, while the population growth of Townsend’s chipmunks was strongly correlated with both apparent annual survival and recruitment. But, for both species, abundance was only moderately correlated with vital rates. Apparent annual survival was nearly constant among years and grids for Humboldt’s flying squirrels, but was consistently lower and highly variable among years for Townsend’s chipmunks. Recruitment and population growth rates for both species were similar and variable among years.
Taken together, the results of this study further our understanding of the spatial and temporal variation of small mammal population attributes and vital processes in late-successional forests in Oregon’s western Cascades. Co-occurring species abundance varied independently even when exposed to similar spatial and temporal drivers, and 2 of the 4-focal species abundances exhibited cyclical population dynamics. There has been considerable debate about the usefulness of abundance variability in determining habitat quality, but these results suggest that in this study system, inference based on abundance variation and inference based on vital rate variation are consistent.