Content area
Abstract
Nitrogen (N) is an essential nutrient to potato growth, yield, and quality. However, over-application of nitrogen can lead to nitrate leaching and groundwater contamination, particularly in Wisconsin’s primary potato production region, which is characterized by sandy soil and frequent field season precipitation. This agronomic study investigated effects of different N application treatments on the in-season crop N status (indicated by N levels in different plant tissues), growth, yield and quality as well as at-harvest yield, size distribution and quality of two potato varieties (Russet Burbank and Soraya) grown over two field seasons. We found that 1) petiole nitrate-N, whole leaf total N, and whole vine total N typically responded to N fertilization at early and mid-tuber bulking by showing an increase one week after each application; 2) as the season progresses, N is translocated from the aboveground biomass to the belowground tubers; 3) no significant effects of N on total yield were found in this study, but N treatments significantly influenced marketable yield of Russet Burbank in both years; 4) the varying weather conditions across years resulted in different size distributions of Russet Burbank, by producing more small tubers (<113 g) under all N treatments when early-season drought and heat stress occurred; 5) compared to Russet Burbank, Soraya is a more efficient N user, as this variety maintained high yield under the low N treatment. Our agronomic study concluded that potato responses to N fertilization were variety- and year-dependent. Responses to N in plant tissues during the growing season did not always result in yield and quality responses to N at harvest.
Recent rapid advancement in remote sensing of vegetation offers great opportunities to utilize cutting-edge technology to improve sustainability in vegetable production systems. Timely and accurate monitoring of crop N status is essential to achieving maximum yield potential while minimizing over- or under-application of N fertilizer, reducing nitrate leaching and groundwater contamination potential. The most commonly used method of potato N status monitoring is to test nitrate-N levels in petiole tissue, however, this method is destructive, time-consuming, labor-intensive, and cannot cover spatio-temporal field variation. Our remote sensing study used an alternative method, full-spectrum hyperspectral imaging (400-2500nm) and partial least-squares regression (PLSR), to model the relationship between different in-season and at-harvest potato traits from two potato varieties (Russet Burbank and Soraya) grown over two field seasons. We found that this innovative method provides moderately to highly accurate predictions of some traits, such as the whole leaf total N level, tuber N uptake, immediate yield during the growing season (R2 between 0.52 and 0.71, NRMSE between 0.15 and 0.24), and the marketable yield and specific gravity at harvest (R2 between 0.61 and 0.78, NRMSE between 0.19 and 0.24). We have concluded that the best PLSR models for predicting potato in-season and at-harvest traits associated with crop nitrogen use will depend on year-to-year growing condition variability, varietal difference, and the reflectance data from different regions of the full spectrum. Overall, this hyperspectral imaging and computer-assisted modeling approach has the potential to create powerful tools to assist potato growers in monitoring crop N status accurately and efficiently and in guiding their N fertilization to be more timely and informed. These tools will contribute to more sustainable potato N management and cleaner groundwater.
