Base temperature determination of tropical Panicum spp. grasses and its effects on degree-day-based models

•Classic methods for calculating base temperature from field data are limited to annual row crops.•They are based on phenological stages, not as well defined in forages as in row crops.•New approaches for determining base temperature in forages and methods of calculation are studied.•Standard deviation in growing degree-days and in number of days resulted in high base temperature and did not detect cultivar differences.•Best methods were iteration, coefficient of variation in growing degree-days and b-coefficient. Development of management tools is essential to explore the potential of grassland systems and such tools include simulation models used for management, planning and research purposes. The simulation models account for temperature effects on forage growth in various ways, and most of them use degree-day-based sub-models to simulate plant growth. Little or no growth is expected for tropical grasses when temperatures are between 10 and 15°C; thus, the assumption of 15°C as the base temperature for growth of these plants is not uncommon. The objective of this paper is to test an approach, commonly used for row crops, for determination of pasture grasses base temperature using a Panicum spp. dataset, and to compare different methods of calculation. Data was collected from well-established plots (4m×10m) of five Panicum spp. cultivars (Atlas, Massai, Mombaça, Tanzânia and Tobiatã), arranged in four randomized complete blocks, sampled from December 2002 to April 2004 in Piracicaba, SP, Brazil. Light interception measurements from three summer and one winter growth cycles were used to determine thermal time to reach 95% canopy light interception. Base-temperature was calculated using iteration method, the b-coefficient method, minimum coefficient of variation of accumulated degree-days, and minimum standard deviation in degree-days and in days. The order of best methods was iteration, coefficient of variation of accumulated degree-days and b-coefficient method, respectively. The standard deviation method in degree-days and in days resulted in high base temperatures and was not able to detect differences among cultivars. Overall base temperatures were different among cultivars: Massai: 16°C, Atlas: 15°C, Mombaça: 11°C, Tobiatã: 10°C and Tanzânia: 7°C.