The water relations of two evergreen tree species in a karst savanna
- First Online:
- 346 Downloads
The ecohydrology of karst has not received much attention, despite the disproportionally large contribution of karst aquifers to freshwater supplies. Karst savannas, like many savannas elsewhere, are encroached by woody plants, with possibly negative consequences on aquifer recharge. However, the role of savanna tree species in hydrological processes remains unclear, not least because the location and water absorption zones of tree roots in the spatially complex subsurface strata are unknown. This study examined the water sources and water relations of two savanna trees, Quercus fusiformis (Small) and Juniperus ashei (Buchholz) in the karst region of the eastern Edwards Plateau, Texas (USA). Stable isotope analysis of stem water revealed that both species took up evaporatively enriched water during the warm season, suggesting a relatively shallow water source in the epikarst, the transition zone between soil and bedrock. Q. fusiformis had consistently higher predawn water potentials than J. ashei during drought, and thus was probably deeper-rooted and less capable of maintaining gas exchange at low water potentials. Although the water potential of both species recovered after drought-breaking spring and summer rain events, associated shifts in stem water isotope ratios did not indicate significant uptake of rainwater from the shallow soil. A hypothesis is developed to explain this phenomenon invoking a piston-flow mechanism that pushes water stored in macropores into the active root zones of the trees. Epikarst structure varied greatly with parent material and topography, and had strong effects on seasonal fluctuations in plant water status. The study suggests that tree species of the Edwards Plateau do not commonly reduce aquifer recharge by tapping directly into perched water tables, but more likely by reducing water storage in the epikarst. A more general conclusion is that models of savanna water relations based on Walter’s two-layer model may not apply unequivocally to karst savannas.