Effectiveness of using vegetation index to delineate zones of different soil and crop grain production characteristics

Abstract

Cost-effective methods to map differences in productivity across fields have potential application for site-specific management of fertilizer and pesticides. In this study, zones were delineated for a field with hummocky topography in southwestern Saskatchewan by clustering the normalized difference vegetation index (NDVI) derived from Landsat TM information. Zones uniqueness were confirmed if zones differed in grain yield. Two different zones were delineated in the field. These zones had significant (P < 0.05) differences in soil factors related to productivity: average solum depth, spring soil nitrogen (N), phosphorus (P), and spring soil moisture. Over 4 yr, the two zones also had different spring wheat (Triticum aestivum L.) grain yield and protein content. The yield response to N-P fertilizer blend within each zone was different with no significant response to N-P fertilizer in the zone with higher NDVI values versus a significant response to N-P in the zone with lower NDVI. The results indicate a potential economic advantage to reducing fertilizer application to the zone without fertilizer response. Further, the residual soil NO3-N in the zone without fertilizer response was positively correlated with N application in the previous year. Therefore, there is a potential environmental benefit to reducing fertilizer application to that zone to decrease residual NO3-N, which can leach and contaminate ground water or can be denitrified to the greenhouse gas, N2O. Hence, this relatively simple and low-cost method of zone delineation has potential practical application to realize economic and environmental benefits from site-specific management of fertilizer.