Abstract: Ground-mounted solar arrays are increasing globally, and the extensive spatial footprint necessary for photovoltaic (PV) energy production can cause land use tension for certain types of ecosystems. This is true for semi-arid managed grassland and rangeland ecosystems because these water-limited ecosystems are characterized by abundant solar radiation, low topographic relief, and short-statured vegetation. An increasingly popular solution for resolving this tension is the adoption of an agrivoltaics (AV) approach, where the land beneath PV panels is co-utilized for agricultural production. The energy and ecosystem benefits of AV arrays are proposed to outweigh alternative approaches to utility scale PV installations, where soils are graded, and vegetation is either removed or mowed. However, how PV-induced alterations in patterns of sunlight availability and soil moisture might affect plant productivity, plant ecophysiological attributes, and ultimately ecosystem structure and health requires greater resolution to more accurately assess the cost and benefits of AV systems.
Here, in a 1.5 ha, 1.3 MW AV research site located near Longmont, CO, we quantified soil moisture and light availability throughout the growing season to determine how patterns of resource redistribution within a single axis (East-West) tracking PV array affected physiological processes and patterns of plant growth. We measured stomatal conductance (gs) and water potential (ψL) diurnally for leaves of a C3 perennial grass (Bromus inermis) located between rows of PVs (highest light), beneath rows of PVs (least light), and adjacent to panel edges that received primarily morning or afternoon light. End of season biomass was collected to determine if productivity differences could be related to resource availability and ecophysiological responses.
Surprisingly, we found that photosynthetic responses to light did not vary among locations and that light saturated values of photosynthesis were consistent across all sites in the PV array, indicating that sunlight did not strongly alter the ecophysiology of this C3 grass, even beneath PV panels receiving ~15% of full sunlight. As expected, leaves receiving primarily morning sunlight had drastically different diurnal patterns of gs and ψL than those receiving only afternoon sunlight, but these patterns were not simply mirror images of each other. Further, areas receiving primarily morning sunlight, with afternoon shade, were the most productive, despite soil moisture levels being greater in areas that received primarily afternoon sunlight. We discuss how this heterogeneity in environmental drivers develops over the growing season and how corresponding physiological responses may impact ecosystem resilience to climate extremes.
