COS 2-4 - Effects of altered precipitation on green-up phenology in coastal sage scrub

Monday, August 8, 2016: 2:30 PM
305, Ft Lauderdale Convention Center
Ellen Esch, Ecology, Behavior & Evolution Section, University of California - San Diego, CA, David Lipson, Department of Biology, San Diego State University, San Diego, CA, John B. Kim, Corvallis Forestry Sciences Laboratory, USDA Forest Service Pacific Northwest Research Station, Corvallis, OR and Elsa Cleland, Ecology, Behavior & Evolution Section, University of California San Diego, La Jolla, CA
Background/Question/Methods

Global precipitation patterns are expected to change in the coming decades showing increased intensity and frequency of events, and southern California is predicted to be a regional global change hotspot. Rainfall scenarios impact invasion dynamics in this system, with native shrublands facing increasing invasion by exotic annual grasses in wet years. Here we focus on green-up date and growing season length as phenological responses to altered precipitation with implications on ecosystem level functioning. To do this, we measured normalized difference vegetation index (NDVI) on a plot-level throughout three growing seasons within a rainfall manipulation experiment. Paired plots were dominated either by native shrubs or exotic herbaceous species and subjected to treatments of 50% (low), 100% (ambient), or 150% (high) of ambient rainfall. The study site was located in a coastal sage scrub ecosystem, with patches dominated by native shrubs and exotic grasses located in San Diego County. 

Results/Conclusions

Green-up date was significantly affected by rainfall treatment, though the magnitude of this response depended on plot community composition and varied between growing seasons (3-way interaction, p=0.037). Green-up date was delayed in the 50% rainfall treatment when compared to the 150% treatment in both 2014 and 2015 (p<0.01 for both, by 55 and 14 days, respectively). In 2014, the green-up date of the 50% treatment was also significantly later than the 100% treatment (p=0.048, by 49 days). In both years, green-up date was negatively correlated with growing season length measured from onset of green-up to date to when peak NDVI was recorded, for both native and exotic dominated plots (p<0.01, r<-0.70 for all) suggesting that delayed green-up date does not shift the growing season but rather truncates the period of plant activity. Thus, later green-up may result in reduced ecosystem net primary production; this is supported by the finding that the low rainfall treatment significantly reduced maximum green-up for exotic dominated plots (p=0.001). Interestingly, native dominated communities did not vary in maximum green-up, indicating that native dominated communities may be more resilient to altered precipitation regimes with the capacity to increase the rate of green-up despite impacted growing season length.