Several investigations of N, P, and K uptake by Upland cotton (Gossypium hirsutum L.) under dryland production have been conducted in the southeast, however few of these types of studies have been conducted under irrigation in more arid environments, with Upland cotton. None of these types of experiments have been conducted with American Pima cotton (G. barbadense L.). In addition, there has been a shift in production management of cotton from scheduled operations on a calendar basis (or days after planting, DAP) to the use of heat unit (HU) accumulations to project cotton growth and development. For these reasons, we conducted a study to describe the total nutrient (N, P, and K) uptake and the partitioning of each nutrient into various plant parts for Upland and Pima cotton. Heat unit accumulations were used to describe (model) the uptake and partitioning of nutrients over time throughout the development of the plant.

This study was conducted throughout the growing seasons of 1990, 1991, and 1992 at two locations in southern Arizona (six site years of data). All locations received adequate moisture via flood irrigation and N, P, and K fertility were maintained throughout, including in-season additions of N as needed (based on petiole feedback and crop monitoring). Both Upland (var. DPL 90) and Pima (var. S-6) cotton were grown each year at each location. Beginning 14 to 20 d after emergence, whole cotton plants were removed from a 1-m section of row (1-m rows) within two replications of each study. This was continued on approximately 14 d intervals until maturity. Cotton plants were then separated into stems, leaves (including petioles), and fruiting forms (mature and immature bolls). Mature bolls were separated into lint, seeds, and burs (carpel walls). The bur fraction, as denoted in the results, included squares, flowers, immature bolls, and burs from mature bolls. The dry weight of each fraction (leaf, stem, bur, seed, and lint) and the appropriate analyses for total N, P, and K were determined on each fraction (except lint).

Cotton yields at each location were determined at maturity by mechanically picking the center four rows of each eight row plot. All plant samples were removed from the four rows which were not harvested. Basic plant measurements, including height-to-node ratios and nodes above the top white bloom (first or second fruiting site on mainstem fruiting branches) were made, on 14 d intervals. Heat unit accumulations after planting (HUAP, 30/13 C thresholds) were obtained from weather stations at each location. Regression analyses was used to model nutrient uptake as a function of both DAP and HUAP. A stepwise procedure was used that would not allow a variable to remain in the model unless it produced an F statistic significant at an value of 0.05 or less.

Regression analyses indicated that HUAP was equally good, and in most cases superior to using DAP to model total nutrient uptake and partitioning within both Upland and Pima cotton. In every case there was close agreement between the predicted and actual total nutrient uptake. For Upland cotton the actual total N, P, and K uptake was 199, 29, and 250 kg ha^-1 and the predicted total N, P, and K uptake was 199, 29, and 250 kg ha^-1, respectively. For Pima cotton the actual total N, P, and K uptake was 196, 29, and 215 kg ha^-1 and the predicted was 184, 29, and 199 kg ha^-1, respectively. The pattern of nutrient partitioning in Upland cotton were similar to the findings of others and Pima showed the same general patterns of partitioning as Upland cotton. Seeds were a major sink of nutrients. Nutrient uptake in seeds resulted in decreasing uptake in leaves and stems. Presumably, due to mobilization of nutrients from those parts to the seeds during seed development. The nutrient requirements to produce 100 kg lint ha^-1 for Upland cotton was 14, 2.1, and 19 kg ha^-1 for N, P, and K, respectively and was 19, 3.0, and 21 kg ha^-1, respectively for Pima cotton.