Forests are critical to the terrestrial carbon and water cycle, but droughts are predicted to increase in frequency and intensity, threatening forest function through lower productivity and increased mortality. Identifying tree species differences in drought response is critical to predicting the extent of this threat. Species strategies for handling water stress have been explored almost entirely in terms of above-ground responses, but not how water stress might be avoided or tolerated through differential access to the highly dynamic below-ground water environment.
We present a recently developed approach to quantifying species-specific below-ground strategies for water uptake, which we believe could be applied globally to advance predictive models of drought response in forests. This framework, which we call the hydrological and demographic assessment of drought (HADAD), formally bridges the gap between functional ecology, life history theory, and population ecology.
HADAD uses a hydrological model to estimate belowground water availability by depth and links it to long-term records of species-specific growth patterns to back infer—via a model of water-stress—the depths at which co-existing species in the forest accessed water. Finally, we analyse whether coexisting species partition the vertical gradient in water availability and drought frequency via trade-offs in performance.
Results/Conclusions
The results of the application of HADAD to 12 common species from the tropical forest inventory data at Mudumalai, Western Ghats, India and 57 species from Barro Colorado Island in Panama showed diverging water-uptake depths at both sites. Our analyses showed species partitioning of vertical hydrological niche, rejecting a null model of no niche differentiation. Although deeper uptake species showed greater attainment of growth-potential in years when they had access to water, when they did not, their growth decreased more than that for shallow water-uptake species. These results identify modes of coexistence distributed through water uptake depth in soils.
This has important implications for the assessment and prediction of drought impacts. The expansion of such a community wide approach through the inclusion of remote sensing technologies, more intensive and targeted monitoring systems, and the integration of this approach into dynamic global vegetation models could vastly improve our knowledge and prediction of drought impacts on forests.