I'm part of a team of researchers conducting a comprehensive assessment of carbon stocks and fluxes in terrestrial ecosystems of the Hawaiian Islands as part of the USGS LandCarbon program. This multi-agency effort seeks to estimate how ecosystem carbon balance will respond to projected future changes in climate, wildfire, biological invasions, and land use. We will also provide estimates of uncertainty and identify knowledge gaps to improve the accuracy of future assessments in Hawaii and in other regions worldwide.
Rising temperature and tropical forest carbon balance
Selmants et al., in prep.
Tropical wet forests have a disproportionate influence on the global carbon cycle, yet it remains unclear how these highly productive ecosystems will respond to climate warming. I am collaborating with Creighton Litton and Christian Giardina to examine the temperature sensitivity of ecosystem carbon pools and fluxes across a 5.2 ºC mean annual temperature (MAT) gradient in tropical montane wet forests on the Island of Hawaii. This MAT gradient is unusually well-constrained, with constant species composition, disturbance history, soil type, geology, and soil water balance. Isolating the effect of temperature using this model system approach provides a unique opportunity to examine the long-term, integrated response of tropical wet forest ecosystems to climate change.
Realistic species losses and ecosystem functioning
Selmants et al. 2014, Ecology
Human-caused declines in biological diversity have inspired a large body of research on biodiversity and ecosystem functioning (BEF). However, most of these experiments have randomly manipulated species richness, a design element that contrasts with the non-random patterns of species losses observed in real ecosystems. In collaboration with Erika Zavaleta and others, we designed a field-based experiment directly comparing how randomized and realistic (non-random) plant species losses affect invasion resistance, productivity and plant nitrogen use. In general, the functional consequences of realistic species losses differ distinctly from randomized losses. This suggests that incorporating more realistic species loss scenarios into BEF experiments can increase their relevance to conservation.