Date of Award
MS in Biological Sciences
Riparian corridors in California are known to be an important but reduced and degraded resource for landbirds. In spite of previous research, the habitat characteristics that correlate with high landbird abundance remain poorly understood. In particular, the scale at which predictive models are useful (fine scale, watershed, sub-region or region) is ill defined. Herein, point count-based abundance indices for 8 riparian associated/obligate species with uniform and high detection probabilities are correlated with biotic and abiotic habitat variables: a sums of squares procedure is used to select the top 5 predictive variables for each species, best fit linear models are selected in an information theoretic framework, and the relative importance of individual variables assessed. These analyses identified site and vegetation characteristics that could serve as targets for restoration and conservation efforts within this coastal central California region. The specific characteristics vary somewhat across the 8 species I surveyed. In addition, the characteristics that I have found important as predictors are distinct from analyses that others have conducted. Therefore, just as we should probably accept regional variation in the composition of riparian avifaunas, we should also probably expect regional variation in the relationship between habitat variables and avian abundance. It appears that important habitat characteristics vary at the fine, watershed, sub-region and regional scales thus reducing the generality of all of the currently available models.