Soil Chemical Attributes, Soil Type, and Rainfall Effects on Normalized Difference Vegetation Index and Cotton Fiber Yield Variability

Abstract

Understanding of how crop yield is influenced by interactions between the soil and environment is needed to minimize agricultural inputs and reduce the impacts of agriculture on the environment. The objective of this study was to determine the effect of soil chemical attributes, soil type, and rainfall on cotton (Gossypium hirsutum L.) yields and yield variability. In this study, the average cotton fiber yield was collected from 101 fields over nine crop seasons. In these fields, located in central Brazil, the average and standard deviation of normalized difference vegetation index (NDVI) were determined. Growing season rainfall and soil nutrients were measured. The variables analyzed were: average cotton fiber yield, NDVI from Landsat images calculated at the boll development phenological stage (enlargement and fiber filling phases), CV of the NDVI, precipitation during the cycle, and soil chemical attributes. The average field NDVI values collected at the boll development phenological stage were correlated to yield (r = 0.50; p < 0.05). Fields with a higher CV for the NDVI values generally had lower yields. The effect of precipitation depends on the cotton phenological stage and soil type. A high rainfall volume at the beginning of the cropping cycle increased plant height variability, which harmed the cotton fiber yield. The highest average values of aluminum content, aluminum saturation, and potential acidity promoted variability in cotton plant growth. Soil chemical attributes related to aluminum were negatively correlated with yield (r = −0.34; p > 0.05). Core Ideas Annual averages of NDVI and yield presented a significant correlation. The effect of precipitation depends on the cotton phenological stage and soil type. Soil chemical variables related to aluminium are negatively correlated with the yield.