Ecohydrology of change: Dynamics of climatological, hydrological, and ecological change across Hawaiʻi's landscapes

Abstract

Across the earth, human activities are driving transformative vegetation changes to vegetation through both direct landscape alteration to the indirect effects of climate change and species invasion. These shifting systems provide an opportunity to examine the climate-vegetation interface and ecohydrological processes that shape partitioning of incoming solar energy and precipitation. This dissertation consists of three studies focused on different aspects of transformative change at this interface: (1) What are the key variables controlling near surface air temperatures across the Hawaiian Islands in space and time? (2) How does conversion of biodiverse mesic forest to forest monotype alter ecohydrologic processes? (3) How does the removal of invasive grass affect rainfall use and growth of tropical dry forest trees? In the first study, by comparing existing daily near surface air temperature records and environmental covariates, I identified that elevation and atmospheric variables explained over 90% of the spatial variance in diurnal temperatures across the Hawaiian Islands and used these to develop improved air temperature models and maps. I then used model-derived temperature indices to examine temperature changes over time. This revealed strong nighttime warming at the coast and suppressed daytime warming at upper elevations into the cloud zone pointing to atmospheric drying within the marine boundary layer and changes to low cloud properties. In the second study, I monitored sapflow and soil moisture in two adjacent forest plots with contrasting biodiversity and structure and identified strong influence woody invader Psidium cattleyanum var. littorale on ecohydrologic processes. Contrasting soil moisture heterogeneity between a diverse mesic native stand and the Psidium monotype suggests stemflow results in lateral redistribution soil moisture during small rain events and preferential infiltration during high intensity rainfall. I observed higher leaf gas exchange rates and sap velocities for Psidium and, unexpectedly, nighttime sapflow indicating nocturnal stomatal conductance. These together suggest Psidium’s competitive advantage in water-limited environments is linked to its ability to capture soil water resources through stemflow which offsets costs of high water use. Besides explaining the success of this monotype-forming woody invader, the findings point to Psidium’s potential to alter larger scale hydrological processes. In the third study, by comparing the microclimate and physiological responses of tropical dry forest trees in grass-invaded and grass-removal treatment plots, we identified that differences in tree species’ responses to grass removal and pulse rain events were ultimately due to differences in physiological traits. Shallow-rooted alaheʻe (Psydrax odorata) competes directly with invasive C4 fountain grass (Cenchrus setaceus, CS) for moisture, and grass removal allowed trees to respond more quickly and more strongly to soil moisture pulses via increased leaf level carbon assimilation. Deeper-rooted lama (Diospyros sandwicensis, DISA) benefits from CS removal at longer timescales, producing more and thinner leaves with grass removal. Elevated DISA canopy leaf area in removal plots lowered understory air temperature, decreasing the understory VPD driving subcanopy PSOD transpiration. These three studies collectively highlight the power of ecohydrological perspectives for understanding change driven by both abiotic factors such as climate and biotic factors linked to invasion and ecological restoration. Ecohydrological frameworks have the potential to inform management decisions by connecting ecologically-grounded, conservation-oriented watershed protection with efforts in hydrology and water resource management. Hawaiʻi’s environmental gradients and sustained efforts in conservation and ecological restoration across these point to opportunities for innovative ecohydrology research in the next decade.