Modeling small-scale spatial interaction of shortgrass prairie species

Native grasses interact spatially with themselves and their environment and can therefore be thought of as a system of dependent random variables. One method of modeling the spatial dependence of a multi-species population is a Gibbsian pairwise potential model. Since natural selection operates at the level of individual plants, the information obtained from such a model should provide a greater understanding of the intraspecific interactions in plant populations, while providing a theoretical basis for determining a plants' 'competitive zone' of influence. In this paper we fit a pairwise potential model to describe the spatial dependency of dominant grasses and forbs measured on a 1.5 x 1.5 m study plot located on a shortgrass prairie site near Fort Collins, Colorado. Dominant grasses included blue grama (Bouteloua gracilis), western wheatgrass (Agropyron smithii), Indian ricegrass (Oryzopsis hymenoides), and needle-and-thread grass (Stipa comata). Procedures for introducing spatial heterogeneity in the model is also discussed. © 1997 Elsevier Science B.V.