Print PageMany studies of climate change impacts on ecosystems rely on seasonally or annually averaged data. These data potentially give an incomplete picture of the relationship between climate and species abundance. Here we investigate variation in timing and amount of precipitation at a low-elevation site in California, and how these factors might affect butterfly populations. Butterflies are an ideal taxon to study climate change and abiotic impacts as studies have indicated that butterflies are affected directly and indirectly (through host plants) by climatic variation. We used a long-term dataset of butterfly richness and daily weather records to ask the following questions 1) which climate factors affect butterfly richness? and 2) do important variables show any temporal trends? Richness and weather data were collected from a California Central Valley site from 1972 through 2007. Candidate variables derived included temporal patterns such as 1st precipitation event date and number of precipitation events, large scale climate patterns (e.g. Pacific Decadal Oscillation (PDO) index), and traditional weather variables such as total precipitation. All candidate variables were assessed using Bayesian model averaging (BMA) to identify the most supported variables. Finally, structural equation models (SEM) were used to investigate relationships among variables, especially in regards to year.
Results/Conclusions
The climate variables with the most support were: PDO index, mean precipitation per event, 1st precipitation event date, and lagged PDO index. Maximum and minimum temperature received little support. From the SEM, year had a significant effect on 1st event date, maximum, and minimum temperature and a marginal effect on number of events, suggesting temporal trends in these facets of climate. These results indicate temporal patterns of precipitation may be useful to explain butterfly richness, especially in Mediterranean climates. Additionally, butterfly richness was explained by patterns at multiple temporal scales, indicated by the support for PDO index and PDO index of the previous year. Not only was richness explained by temporal patterns, there is indication that these temporal patterns are shifting over time indicating potential climate change impacts. These results demonstrate the need to explore different facets of climatic variation in an effort to better understand phenology and patterns of biological richness.
