OOS 40-9
Long-term monitoring at the San Joaquin Experimental Range: Avian population trends and predicting response to climate change based on 27 years of data

Thursday, August 14, 2014: 4:20 PM
202, Sacramento Convention Center
Kathryn L. Purcell, Pacific Southwest Research Station, USDA Forest Service, Fresno, CA
Sylvia R. Mori , Pacific Southwest Research Station, USDA Forest Service, Albany, CA
Background/Question/Methods

Experimental forests and ranges are living laboratories that are invaluable for their capacity to foster long-term research and monitoring. The San Joaquin Experimental Range (SJER) was established in 1934 to develop appropriate land management practices on foothill rangelands in California and has a rich history of avian research. Using 27 years of point count data collected at 210 count stations at SJER, we examined population trends for 34 oak woodland bird species, modeled the importance of weather and climate variables on annual variability in bird abundance, and assessed the response of birds to changing climate conditions, using the variability inherent in weather and bird abundance.  We used both linear regression and non-parametric regression (spline-smoothing the year effect) to examined population trends. To model responses to weather and climate variables, we used generalized additive models for Poisson-distributed response (counts) with non-parametric smoothing functions as a first exploratory approach. We considered three parametric mixed generalized linear models, Poisson, Quasi-Poisson, and negative binomial, to estimate the overdispersion variance. The parametric functions (polynomials) were inferred by visual inspection of the smoothed partial residuals. We used Akaike information criterion and the percent deviance explained to select the best models, considering correlations among the explanatory variables.

Results/Conclusions

Abundance of most species varied greatly over the study period. For many species, the spline curves revealed oscillations in abundance over time suggestive of healthy populations, or evidence of recovery following a decrease (for declining species). Overall, twelve species appeared to have increased at SJER over the study: Red-tailed Hawk, American Kestrel, California Quail, Anna’s Hummingbird, Acorn Woodpecker, Nuttall’s Woodpecker, Common Raven, Violet-green Swallow, White-breasted Nuthatch, Bewick’s Wren, European Starling, and House Finch. Five species decreased: Western Scrub-jay, Blue-gray Gnatcatcher, Western Bluebird, Western Meadowlark, and Bullock’s Oriole. Four species were sensitive to hot temperatures but more species were sensitive to cold temperatures. Species sensitive to warming included Western Scrub-jays, Oak Titmice, and Anna’s Hummingbirds. More species decreased following El Niño years than increased. Our results provide information on current trends in avian abundance in California oak woodlands and help predict how species might respond to locally changing climatic conditions. Our results underscore the importance of long-term research and monitoring. Long-term datasets such as there are particularly valuable in providing early warnings of undesirable environmental changes. Experimental ranges and forests are ideally suited for long-term multidisciplinary research exploring biotic responses to changing environmental conditions.