Rising input costs, particularly for fertilizer, have been an ongoing concern for agriculture in general but more recently have become increasingly serious as global supplies have been disrupted

For specialty crop farmers in particular, this adds pressure to an industry that faces multiple challenges that, if left unaddressed, could jeopardize the viable production of many vegetable, fruit and nut crops in the country. We propose to demonstrate a plant-by-plant precision fertilization capability for cool season vegetable crops that will increase harvest yields while reducing the overall amount of fertilizer applied. The solution will help farmers increase margins in the face of escalating production costs while meeting increasingly stringent environmental regulations. The overarching objectives of our variable rate application (VRA) approach are to determine and apply an optimal amount of fertilizer for each plant, to adjust for the inherent field and seed variations so that most plants meet customer specifications and are sold at harvest.an additional benefit of the anticipated capability will be less food waste, employing the EPA food recovery hierarchy's most preferred method of reducing waste where the food is produced. Our innovation includes obtaining a priori information on a field's plant size distribution from high resolution aerial imagery of the entire field. The size distribution will be used to determine an application prescription, which will be sent to a mechanical controller on a tractor-mounted sprayer. Over time, associated plant growth models will be augmented to account for the effects of additional fertilizer, possibly even accounting for different, future varieties that are more drought tolerant or less fertilizer hungry. We will leverage recent experimental results indicating that different fertilizer application amounts lead to different growth rates of iceberg lettuce, so that applying different amounts to different sized plants willincrease field uniformity at harvest and also reduce total required fertilizer amounts per crop cycle. The proposed effort will validate and extend these preliminary findings to develop a unique capability for vegetable production. During phase i, we will determine geolocation accuracy requirements and evaluate if system components can meet them. We will assess timing and coordination requirements of aerial imagery collection, processing and conversion to a prescription map, and we will evaluate spray controller requirements and performance for the VRA approach. We will assess the overall technical feasibility of the proposed concept and perform a preliminary cost-benefit analysis for VRA commercialization in the vegetable industry. We plan to sell the precision fertilization capability as an add-on to subscription services we have already developed and commercialized for the fresh produce industry.