Simplified Simulation Model of the Plant Producer Function in Shortgrass Steppe

Abstract

Models of grassland ecosystems generally have been either simple regression equations for certain processes or large complex simulation models consisting of numerous integrated sets of equations. The former do not account explicitly for interacting mechanisms and the latter are expensive and, because of their complexity, may be difficult to interpret. This paper presents a simple, 11-equation simulation model of the primary producer function in the shortgrass steppe. Model results closely approximate observed data from the Central Plains Experimental Range in northeastern Colorado. INTRODUCTION The U. S. Forest Service has the responsibility for managing vast areas of forests and grasslands. Developing comprehensive plans for these management responsibilities requires predictions about how these ecosystem types will respond to alternative management strategies (Joyce et. al., 1983). These plans must account for heterogeneous landscape units and be able to evaluate how management affects individual vegetation types within these units. Ecologists and managers are interested if generic models can be constructed for vegetation types such that these models will produce the required level of descriptive accuracy to be used in planning. In this paper we describe a grassland model designed to project realistic grassland responses to management practices, such as irrigation, fertilization and changes in the grazing regime. The resulting model is not intended to address detailed ecological questions, but rather, to provide general estimates of primary production over relatively wide geographical areas. Furthermore, the model is simple enough so that the equations capture the important ecological and physiological processes, but the model output is designed to be conveniently interpreted in relation to practical decisions about natural resource management options. During the past two decades, numerous mathematical models have been designed to simulate the essential characteristics and behavior of grassland ecosystems. These models have addressed various objectives, e. g, providing a format for improving understanding about ecosystems, providing a tool for communicating information, suggesting further experimental research, and improving management of grasslands. Many of the earlier mathematical models of grassland productivity were regression models or algebraic equations that fit patterns of observed data (Clayton et. al., 1983; Sneva and Hyder, 1962; Sims et al., 1978). Subsequently, some grassland models became very complex, consisting of hundreds of interrelated equations (e.g, Innis, 1978). While these complex models have proved useful for understanding grasslands (Parton and Risser, 1980; Risser et al., 1981; Risser and Parton, 1982), their size and complexity have made them expensive to run and, in some cases, difficult to evaluate. Therefore, the model described in this paper was designed to be relatively simple and to use independent driving variables that are easily measured.