ABSTRACT
Several types of models have been developed to describe the competitive interactions between plants, and they can be classified into five types based on the assumptions they make about the utilization of resources by individual plants. There is no type of model in which the underlying set of assumptions is entirely realistic. In this paper, we fit models representative of the different types to a standard data set from an extensive row spacing experiment using carrots. The comparative agreement between observed and fitted values when different models were fitted was dependent on the data. An overlapping domain model gave the best description of root weights derived from plants growing in the inner rows of plots, but a nonoverlapping domain model gave better fits to root weights derived from plants growing in edge rows. Possible explanations of this observation are discussed.
The persistence of several contrasting spacing models in the literature might thus be explained because the fit of models is sensitive to the data to which they are fitted.
