This dissertation explores sensor-based and modeling approaches for site-specific prediction of grassland biomass yields under small-scale agricultural conditions. Its goal is to develop practical models for yield prediction to support optimized and sustainable grassland management.

Drawing on a systematic literature review and multi-year field studies in southwestern Germany, the study compares terrestrial and remote sensing systems. Results demonstrate that ground-based sensors, particularly height-based measurements, outperform remote sensing for yield prediction. Incorporating sward composition and local weather data further improves model accuracy. The economic feasibility of these technologies is also evaluated, showing that simple terrestrial measurement methods are practical for small-scale farms, while more complex approaches remain of limited economic use.

Overall, the dissertation provides an empirically grounded foundation for advancing site-specific grassland management, offering tangible potential to enhance both efficiency and sustainability.