Flowering phenology is often related to climate, but trait based phenological responses to climate are understudied. To address this we used a historical dataset of flowering observations from the 1990s along with resurveys from 2011 through 2015 of 10 sites within Santa Cruz County, CA. Study sites represent two habitat types within the coastal climate zone; four inland sandhills sites, and three maritime and three inland grassland sites. Surveys recorded first and last flowering date, and duration of flowering of all species at each site. First, we grouped species by traits that characterize drought tolerance, heat tolerance, blooming season, and nativity to California. For each group of species we used a random forest analysis to determine which climate variables best predicted first flowering date. Then we used generalized additive models to determine how first flowering date responds to the six most important climate variables.
Results/Conclusions
We found that in grasslands early season annual species advance, but their late season counterparts delay in responses to temperature; however at sandhills early season annuals delay in response to temperature, but late season annuals only respond to precipitation. We found delays in response to temperature in both native and non-native early season species, but precipitation only affected non-native species. This approach allows us to predict phenological responses based on species’ traits.