Applying the CROPGRO Perennial Forage Model for long-term estimates of Marandu palisadegrass production in livestock management scenarios in Brazil

Abstract

The use of long-term forage production simulation is important to studies of strategic scenarios, yield-gap predictions and production risk analysis. However, forage models have only been tested using short period data. Therefore, this study aimed to apply the CROPGRO Perennial Forage Model (CROPGRO-PFM), a mechanistic model from Decision Support System for Agrotechnology Transfer (DSSAT) platform, to simulate long-term forage production of Marandu palisadegrass {Urochloa brizantha (Hochst. ex A. Rich.) R.D. Webster [syn. Brachiaria brizantha (Hochst. ex A. Rich.) Stapf]} in different locations of Central Brazil, under different fertilization levels. We chose nine locations distributed in Amazon, Cerrado, and Atlantic Forest Biomes to simulate Marandu palisadegrass production over 36 years (1980–2016). The simulations were carried out using prior calibrations for DSSAT’s CROPGRO-PFM and, as a reference point of growth and herbage simulation, model results were initially contrasted against an agrometeorological model, the Integrated Agrometeorological Model (IAM), to check the consistency between these two different models. Then the CROPGRO-PFM was applied to simulated scenarios represented three beef cattle livestock production systems: (1) Potential production water-not-limiting and highly fertilized, (2) rainfed with high N fertilization, and (3) rainfed with low N maintenance fertilization. In general, the long-term simulations of annual forage production and seasonality were similar for the CROPGRO-PFM and IAM models. As expected for the Central Brazil region, historical averages of monthly herbage accumulation rate of potential yield scenarios were relatively constant throughout the year, declining in the winter period for most locations due to mild temperature reduction and decline of solar radiation. In general, rainfed scenarios simulations showed a seasonal cycle following the rainfall pattern, with reduction in growth rates in the dry season. During the winter season climate is the major limiting growth factor, simulated growth rates were low and similar regardless of N fertilization. Although the CROPGRO-PFM presented consistent results for long term grazing systems, further model changes, calibration, and validations are recommended to improve the process-based model as well as the knowledge about these systems.