Crop productivity and sustainability of agro-photovoltaic systems in response to climate change

Agrivoltaic systems (AVS) represent an innovative land-management strategy that enables the simultaneous production of food and renewable energy on the same area of land. By integrating crops and photovoltaic modules, they increase land-use efficiency and contribute to achieving climate neutrality as well as food and energy security targets. In the context of climate change, water scarcity, and growing energy demand, AVS therefore offer a concrete solution capable of mitigating conflicts between agricultural production and electricity generation. However, the presence of photovoltaic structures alters the crop microclimate by reducing solar irradiance, lowering air temperature, and increasing relative humidity, with variable effects on photosynthesis, plant growth, and yield depending on the cultivated species. It is thus essential to test the effectiveness of AVS on crops, understanding how they respond to heterogeneous shading and how system design can be adapted accordingly.Reliable assessment of these crop responses requires continuous, high-resolution monitoring. Traditional field measurements, while accurate, are labor-intensive and difficult to scale up; hence, there is a need for integrated methodologies that combine optical sensing, process-based modeling, and energy-flow analysis. Against this background, the present thesis aims to provide tools and knowledge for the design of Agro-photovoltaic installations that maximize agricultural production without compromising energy efficiency.In the framework of the PhD project, the research focused on the development of monitoring and modelling tools for crop production in AVS. Building on these premises, Chapter 3 presents an image-driven modelling approach for fAPAR-derived biomass estimation in alfalfa cultivated under fixed photovoltaic panels in central Italy. A hybrid method combining proximal sensing (2D time-lapse RGB and 3D LiDAR scans) with the process-based SSM-iCrop2 model was implemented to quantify daily biomass accumulation. Although both 2D and 3D imaging reproduced fAPAR dynamics well, 2D imagery ensured the best trade-off between temporal resolution and accuracy. By forcing the crop model with 2D-derived fAPAR and radiation data simulated in GroIMP, the system successfully captured growth cycles, mowing events and shading impacts across two seasons, demonstrating the feasibility of low-cost, high-frequency biomass monitoring in AVS. Building on these findings, Chapter 4 integrates 3D radiation modelling with crop growth simulations to analyse shading effects on alfalfa across different AVS designs (fixed and dynamic panels). A detailed GroIMP reconstruction of each AVS layout enabled precise calculation of direct and diffuse radiation and its spatial variability. Coupling these data with SSM-iCrop2 allowed the prediction of light interception, water balance and yield under variable shading regimes. The study provided useful design elements to minimize productive losses while simultaneously optimizing renewable energy generation, demonstrating how the integration of data and modelling can support site-specific agronomic management and the optimal configuration of photovoltaic panels. Chapter 5 moves beyond crop modeling and investigates how microclimatic gradients induced by photovoltaic panels (shading and redistribution of rainwater runoff) influence pasture composition at the field scale. Along transects laid perpendicular to the panel rows, the seasonal composition of the vegetation and the aboveground biomass were monitored. Grasses predominated overall, whereas legumes were favored in the inter-row areas with lower shading. Yield patterns reflected the combined control of light availability and redistributed water, showing maxima at the panel edges and reductions in the most shaded zones, thereby highlighting management opportunities for targeted mowing and grazing. In line with this approach, Chapter 6 explores the ecological implications of microclimatic gradients generated by AVS, with a specific focus on their effects on vegetation composition and soil biological quality. Through seasonal surveys conducted in a Mediterranean AVS pasture, the chapter integrates vegetation and mesofaunal indicators to analyse the functional responses of plant communities and soil microarthropods along the shading gradient induced by photovoltaic panels. The analysis is based on the combined use of the Pasture Value (PV) index and the QBS-ar index, which assess forage quality and soil biological fertility, respectively. The results reveal clear seasonal and spatial variations. In spring, when contrasts in radiation and temperature are most pronounced, inter-row zones exhibit significantly higher PV and QBS-ar values. These areas support a co-dominance of light-demanding legumes and a richer mesofaunal community. Conversely, the under-panel zones, subjected to combined shading and altered water regimes (either drought or waterlogging), are dominated by stress-tolerant forb species and display lower soil biological quality. In autumn, although overall biological activity tends to decrease, a partial recovery of soil fauna is observed, favoured by the accumulation of litter and greater thermo-hygrometric stability.