Publication

Abstract : Vegetation growth drives many of the interactions between the land surface and atmosphere including the uptake of carbon through photosynthesis and loss of water through transpiration. In arid and semi-arid regions water is the dominant driver of vegetation growth. However, few studies consider the fact that water can move laterally across the landscape as runoff via streams and floodplains, termed hydrologic connectivity. Using multiple observations alongside models and a hydromorphology dataset for Australia, we examine how ecosystems with high hydrologic connectivity differ in their vegetation response to water availability, soil properties, and interannual variability and extremes in vegetation productivity. We find that the average interannual variability of vegetation productivity is 21–34% higher in ecosystems with high hydrologic connectivity, with skewed annual anomalies showing larger extremes in carbon uptake. This is driven by a higher average and more variable surface soil moisture and significantly higher soil available water capacity and soil depth. These spatially small ecosystems, covering 14% of the study region, contribute 15–22% (median = 17%) to regional-scale carbon uptake through higher rates of gross photosynthesis, especially evident during wet years, and 3–37% (median = 19%) to annual anomalies. Current global land surface models do not reproduce the observed spatial patterns of interannual variability in carbon uptake over regions where hydrologic connectivity is high as they lack the mechanism of connectivity of water between discrete land surface elements. This study highlights the significant role of riparian and floodplain vegetation on the interannual variability and extremes of the regional carbon cycle.