Evaluating Methods for Research in Physical Weed Control and Farm Asset Tracking

Effective weed control has long been recognized as critical for agricultural production, yet weeds remain a major constraint to production and economic return in many agroecosystems. Moreover, improvements in physical weed control are necessary to address increasing problems of herbicide resistance in weeds of grain and fiber crops and the high cost of hand weeding in vegetables. From tractor-mounted cultivation tools to autonomous weeders, weeding implements are affected by weeds, crops, soil conditions, and actuator effectiveness. In order to address these complex and often interacting factors concerning weed control, new and innovative tools must be designed and evaluated. Chapter one addresses a series of experiments designed to determine the functionality and efficacy of Franklin Robotics’ TertillTM and to explore its place in the growing field of robotic weeding. The TertillTM demonstrated high weed control efficacy, supporting its utility as a tool for home gardeners. However, in its current form, the TertillTM would require modification to be viable for farmscale use. Yet, its simple and effective design may offer insights to inform future development of farmscale weeding robots. Chapter two addresses an analysis of the early growth characteristics of wild radish (Raphanus raphanistrum L.) and four related Brassica species commonly used as surrogate weeds in physical weed control research. Plants of each species were grown in a greenhouse, destructively harvested at three distinct growth stages, and analyzed for anchorage force and root architecture. Wild radish and the selected Brassica surrogate weeds were comparable in biomass and root architecture. However, differences in anchorage force necessitates caution and field validation. Chapter three builds upon the previous chapter by making the explicit comparisons between surrogate weeds and their weedy counterparts that have hitherto been absent from the literature. Additionally, the viability of golf tees as artificial weeds was assessed. Field experiments were conducted in 2019 and 2020 using six flex-tine harrows to compare the reactions to cultivation of wild radish, two Brassica surrogate weeds, and golf tee artificial weeds. Rates of efficacy for both surrogate weed species were comparable to those of wild radish, indicating that these species are useful surrogates for this weed species. However, golf tees failed to accurately simulate weed seedling response to cultivation, and their response was highly variable. Chapter four addresses the challenges and inefficiencies apparent in diversified organic farming by evaluating the potential of inexpensive, wearable GPS watches to monitor farm labor. Labor data acquired with GPS watches was correlated with a reference system. However, elevated rates of error associated with commercially available GPS devices potentially limits their viability in tracking labor on small farms where error may result in significant inaccuracies