The Impact of Large, Manipulated Rainfall Events on Soil Greenhouse Gas Fluxes Is Mediated by Warming, Spatial Heterogeneity and Hurricane Disturbance Recovery in a Puerto Rican Wet Tropical Forest

Substantial uncertainty surrounds how terrestrial ecosystems may respond to the interaction of multiple global change drivers, especially the interaction of atmospheric warming and more frequent or severe extreme weather events. Climate models suggest that the Caribbean may experience more frequent and/or heavier rainfalls under future climate, which is likely to also interact with increased atmospheric temperatures. In this thesis, I investigate how long-term (“press”) soil moisture and temperature conditions interact with acute (“pulse”) extreme precipitation events to influence CO₂ fluxes, labile carbon, and microbial carbon in a wet tropical forest in Puerto Rico. I conducted a fully-factorial soil incubation experiment manipulating antecedent moisture and temperature before applying a simulated extreme rainfall event. I found that drier, warmer antecedent conditions led to the largest post-wet-up CO₂ pulses, while saturated soils consistently showed suppressed CO₂ flux regardless of temperature, reflecting oxygen limitation. I also conducted a field rainfall manipulation across a topographic gradient to test how ecosystem heterogeneity influences flux responses. CO₂ fluxes varied by both rainfall treatment and plot, suggesting that underlying moisture and disturbance history shaped ecosystem responses. Together, these results show that antecedent conditions and spatial heterogeneity mediate soil carbon dynamics under interacting global change drivers. This highlights the need to incorporate ecosystem variability into predictions of tropical forest carbon balance under climate change, with implications for climate mitigation and ecosystem management.