Ecohydrological separation in low-seasonality tropics: Evidence of absence or absence of evidence?

Abstract:

Recent meta-analysis and remote-sensing studies have shown an apparent ecohydrological separation where tree-transpired water is of a character different to mobile water found in soils, groundwater and streams. This suggests that soil water pools are poorly mixed and likely the rule than the exception. As such, water contributing to groundwater recharge is often (but not always) isolated from water used in plant transpiration. However, the mechanisms that underpin these phenomenological observations remain elusive. Wetness-dependent interconnectivity (i.e. hydrologic input being temporally out of phase with primary productivity) was originally hypothesized as a possible explanation. Here we seek to test such “wetness-dependent interconnectivity hypothesis” in two tropical environments in Puerto Rico where precipitation seasonality is relatively low and where precipitation is positively correlated with primary productivity. We then determine the sources of water uptake by trees (Swietenia spp.) at different landscape positions using a simple linear mixing model, implemented in a Bayesian inference framework. Our results suggest that ecohydrological separation might be related less to the temporal phasing of hydrologic inputs and primary productivity, and more to the fundamental processes that drive evaporative isotopic enrichment of residual soil water within the soil matrix. The lack of an evaporative signature in both groundwater and streams in the study area suggests that these water balance components have a water source that is transported quickly to deeper subsurface storage compared to waters that trees use. Notwithstanding, a mechanistic model that unifies water extraction processes across the soil-plant-atmosphere continuum, vis-à-vis ecohydrological separation, remains elusive. We provide a roadmap for future work in this regard, aimed especially at controlled studies at the Biopshere-2 Tropical Research Biome.