Submitted to: Weed Research
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 12/31/2002
Publication Date: 2/1/2003
Citation: DEEN, W., COUSENS, R., WARRINGA, J., BASTIAANS, L., CARBERRY, P., REBEL, K., RIHA, S., MURPHY, C., BENJAMIN, L., CLOUGHLEY, C., CUSSANS, J., FORCELLA, F., HUNT, T., JAMIESON, P., LINDQUIST, J., WANG, E. AN EVALUATION OF FOUR CROP: WEED COMPETITION MODELS USING A COMMON DATA SET. WEED RESEARCH. 2003. v. 43. p. 116-129. Interpretive Summary: One aspect of global climate change that is of interest to weed researchers is whether the impact of weeds on crops will change in the future. To predict this future response, a number of crop:weed simulation models can be used. The Global Change Terrestrial Ecosystems component of the International Geosphere-Biosphere Programme convened a workshop to compare the accuracies of four of these models using a common data set involving a weed, annual ryegrass, competing with wheat in replicated field experiments. All models had positive and negative features, but what was apparent is that better data on both weed and crop biology will be necessary to increase accuracies of predictions by all models. Relatively simple methods to calculate competition were sufficient to predict crop yield losses, suggesting that complex models are not always needed. Lastly, specific processes that lead to negative weed:crop interactions, such as interception of sunlight by one species to the detriment of another, must be programmed in models so that they are in "modules" that can be evaluated and compared by researchers more easily. The results from this workshop and research should be useful for other scientists whose goals are to predict crop:weed interactions.
Technical Abstract: To date, several crop:weed competition models have been developed. Developers of various models were invited to compare model performance using a common data set. The data set consisted of wheat and Lolium rigidum grown in monoculture and mixtures under dryland and irrigated conditions. Models were ALMANAC, APSIM, CROPSIM, and INTERCOM. Deviations between observed and predicted values for monoculture wheat were slightly lower than for wheat with L. rigidum for all models. Most errors resulted from inaccurately simulating growth of individual species. Relatively simple competition algorithms could account for most competition responses. Increased model complexity did not improve accuracy of predictions. Comparison of specific competition processes, such as radiation interception, was difficult because effects could not be isolated in each model. Algorithms for competition processes should be modularized so that exchange, evaluation, and comparison across models is facilitated.