Relating potato yield and quality to field scale variability in soil characteristics

Causes of within-field spatial variability in potato (Solanum tuberosum L.) yield are not well understood. To address this, a study was conducted from 1998 to 2000 on a commercial farm in southeastern Washington. Soil samples were collected from four center-pivot-irrigated, uniformly fertilized fields on a 0.4-ha grid interval prior to potato planting and analyzed for nitrate-N, ammonium, P, K, organic matter, pH, and texture. The elevation of each grid point was also recorded. Four to five days before commercial harvest, potatoes were collected from a 3-m row length at each original grid point using a one-row digger. The potatoes were weighed, sorted into five weight classes, and evaluated for specific gravity. Correlation and stepwise regression analyses were conducted to test relationships between soil-based and yield variables. Factors driving yield varied between fields. Soil texture components (sand, silt, clay) had stronger impact on yield than with the soil chemical properties we measured. However, all four fields showed an inverse relationship between specific gravity and soil test K, although the correlation coefficients and contributions to regression models were relatively low. Finding a general prescription formula for goals other than higher yield (e.g., nutrient-leaching potential) may be feasible. The consistent relationship of soil textural components in our models suggest that monitoring available soil water, a factor closely related to soil texture, should be included in any future work.