Modeling Soybean Development in the Brazilian Amazon.

Seasonal climate variability and future climate change scenarios can affect crop development and yield in several ways, which can be assessed using crop models. The use of simple models to simulate crop growth stages can be critical to plan and assist farmers in making management decisions during plant developmental phases. The objective of this research was to simulate soybean phenological phases using temperature and photoperiod data for the Amazon region of Brazil. The experiment was carried out in Paragominas, Northeastern Pará, Brazil (2^o59’08’’S, 47^o19’57’’W). The cultivar used was BRS Tracajá (maturity group 9.2). The experimental sowing dates were 2/5/2006, 2/23/2007, and 2/7/2008, using a planting density of 26-28 seeds per square meter. The row space was 0.5 m and fertilized using NPK formula of 7:70:63 kg ha^-1 applied to the soil surface. The harvesting dates were 6/15/2006, 6/21/2007, and 6/6/2008. The phenological development data was assessed daily using the scale described by Fehr and Caviness. The R program was used to implement the phenology model taking into account temperature and photoperiod effects allowing the easy simulation of planting dates and a number of varieties options to investigate climate indicators during different phenological phases. The R-phenological model will be used to develop a web-based tool for accessing crop development under different climate scenarios. The CROPGRO-Soybean model was also calibrated for the BRS Tracajá cultivar and then used to evaluate the performance of the R-phenology model. The model was able to simulate soybean phenology stages with good precision. However, more tests in different regions and weather conditions are necessary.