Submitted to: Agronomy Abstracts
Publication Type: Abstract only
Publication Acceptance Date: 10/15/2005
Publication Date: 11/8/2005
Citation: White, J.W., Batchelor, W.D., Boote, K.J., Hoogenboom, G., Jones, J., Singh, U. 2005. Modeled responses to temperature: a multi-species comparison. Agronomy Abstracts. CD-Rom (P6598) Interpretive Summary:
Technical Abstract: Processed-based models are widely used to analyze how temperature affects crop growth and development. Ideally, each process being simulated should be rigorously tested using field or controlled environment conditions that represent the full range of temperatures of interest and consider potentially important interactions with factors such as water and nutrient supply. In practice, models often are parameterized using data from an incomplete range of conditions, and responses may be extrapolated from related species or other arguably less-reliable sources. Models are usually tested for a range of environments to ensure that overall performance is acceptable. However, given the difficulties in conducting field trials under extreme temperatures, such testing seldom reveals the full range of modeled temperature responses. Sensitivity analyses provide an alternative method for assessing model performance. A procedure was recently developed for standardized assessments of temperature responses, using conditions of non-limiting water and nitrogen (White et al., 2004. Agron. J. 97:426-439). Models are tested using constant daily temperature regimes from 3°C to 40°C with a 10°C daily range. Analyses focus on major categories of processes such as overall growth, phenology and resource use efficiency. We apply the procedure to various crop species, including rice, peanut, soybean, and tomato, using the CSM model as released in DSSAT4. Most differences in responses reflect expected differences among species. Harvest index and unit grain weight were more responsive to temperature than expected, reflecting the difficulty of balancing declining overall growth against partitioning to reproductive tissue, including allocation among yield components.