Black-shouldered kite -- small mammal -- vegetation relationships in northwestern California

This thesis is written in 2 sections to divide the major themes of the study. The second section is partially based on the work presented in the first. I studied the relationships between populations of both California voles (Microtus californicus) and Western harvest mice (Reithrodontomys megalotis) and vegetation structure and composition in northwestern California. Both species exhibited annual fluctuations in abundance during the 19 months of the study. In general, M. californicus populations were largest in areas with lower cover height and larger percent brown grass cover. This finding is contrary to the majority of Microtus spp. studies that have examined population - vegetation relationships. The discrepancy is likely due to the fact that all areas that I trapped had fairly abundant cover (no areas were grazed). The increase of M. californicus populations (from the annual low) coincided with grass seeds being more available (i.e., dropping off of grasses). To my knowledge I recorded the largest annually fluctuating M. californicus population ever found (range = 0 - 1414 ind/ha, Jolly-Seber estimate). I propose that resource limitation is a primary factor in M. californicus population fluctuations. R. megalotis populations fluctuated synchronously with those of M. californicus, but at much reduced numbers. Increasing populations of R. megalotis were associated with an increased availability of grass seeds. The largest populations of R. megalotis were associated with relatively low cover height of vegetation. Each month I developed discriminant function analysis models to predict both species of small mammal abundances using vegetation (structure and composition) variables. Mean percent correct classification was 70.03% and 88.2% for M. californicus and R. megalotis, respectively (n = 19). I also studied the relationship between black-shouldered kite (Elanus caeruleus) territory size and both prey abundance and competitor abundance in northwestern California. Kite territory size ranged from 1.6 - 21.5 ha (n = 26). Estimated mean number of M. californicus per territory was 1483 (SE = 163, n = 25). Competitor abundance (i.e. total raptor abundance) ranged from 4.8 - 31.0 individuals/km2 and was strongly correlated with abundance of M. californicus. Both estimated prey abundance and competitor abundance were negatively correlated with kite territory size. After developing a multiple regression model using both variables, partial correlation analysis revealed that once the effects of prey abundance were statistically controlled, competitor abundance continued to be significantly correlated with kite territory size. When the effects of competitor abundance were statistically controlled, prey abundance was no longer significantly correlated with kite territory size. I conclude that kite territory size is proximately regulated by competitor abundance and ultimately regulated by prey abundance because M. californicus abundance regulates numbers of raptors.