In arid landscapes, landscape-scale vegetation heterogeneity is often tightly coupled to underlying variation in soils. These correlated spatial patterns portend to differences in relationships between climate, moisture availability, and vegetation productivity among soil units. Understanding the processes that are driving these patterns may be useful in understanding ecosystem vulnerability to multi-year climatic events, such as extreme drought. It is not fully clear how fine-scale (< 1 km2) heterogeneity in soils and lifeform dominance influences these patterns, especially during periods of multiple dry and wet years. We investigated relationships between precipitation and potential evaporation, aboveground net primary productivity (ANPP) and soil moisture for 15 grassland and shrubland sites on heterogeneous soils at the Jornada USDA-LTER site in southern New Mexico, USA, from 1989-2015. Our primary goal was to disentangle the effects of soil and vegetation differences at these sites on relationships between potential evaporation and soil moisture at shallow (30 cm), moderate (60-90 cm), and deep (120-180 cm) depths in the soil profile, and to relate these differences to field measurements of ANPP, both within and between grasslands and shrublands.
Overall, we found greater soil moisture coupling to potential evaporation in grassland soils compared to shrubland soils, and in shallow compared to deeper soil depths in both grasslands and shrublands. Similarly, coupling of ANPP to both soil moisture and potential evaporation was stronger in grasslands than in shrublands, especially during dry years. We also observed lower, but important, within-ecosystem variability in coupling across soil types and differences in vegetation structure, especially in grasslands. Although landscape-scale differences in soil and lifeform dominance are the dominant drivers of patterns in ANPP, within-ecosystem heterogeneity may have an important influence on ecosystem responses to climate – especially for ANPP in grasslands during periods of dry years. Our results suggest that gaining insight on the ecosystem types most likely to exhibit sensitivity to climate-driven disturbances, we can better identify locations and time periods in the future when these ecosystems will be most vulnerable to multi-year climatic events.