Spatio-temporal transpiration patterns reflect vegetation structure in complex upland terrain

Topography exerts control on eco-hydrologic processes via alteration of energy inputs due to slope angle and orientation. Further, water availability varies with drainage position in response to topographic water redistribution and the catena effect on soil depth and thus soil water storage capacity. Our understanding of the spatio-temporal dynamics and drivers of transpiration patterns in complex terrain is still limited by lacking knowledge of how systematic interactions of energy and moisture patterns shape ecosystem state and water fluxes and adaptation of the vegetation to these patterns. To untangle the effects of slope orientation and hillslope position on forest structure and transpiration patterns, we measured forest structure, sap flux, soil moisture, throughfall and incoming shortwave radiation along two downslope transects in a forested head water catchment in south-east Australia. Our plot locations controlled for three systematically varying drainage position levels (topographic wetness index: 5.0, 6.5 and 8.0) and two levels of energy input (aridity index: 1.2 and 1.8). Vegetation patterns were generally stronger related to drainage position than slope orientation, whereas sap velocity variations were less pronounced. However, in combination with stand sapwood area, consistent spatio-temporal transpiration patterns emerged in relation to landscape position, where slope orientation was the primary and drainage position the secondary controlling factor. On short temporal scales, radiation and vapor pressure deficit were most important in regulating transpiration rates, whereas soil water limitation only occurred on shallow soils during summer. The importance of stand structural parameters increased on longer time scales, indicating optimization of vegetation in response to the long-term hydro-climatic conditions at a given landscape position. Thus, vegetation patterns can be conceptualized as a 'time-integrated' predictor variable that captures large fractions of other factors contributing to transpiration patterns.