Comparison of linear and nonlinear overstory-understory models for ponderosa pine

Abstract

Linear and nonlinear regression models of total understory biomass in relation to overstory canopy cover and several abiotic site factors were tested using data from 31 ponderosa pine ( Pinus ponderosa Dougl.) sites along the Front Range of Colorado and Wyoming. At the forest level, site factors were insignificant in explaining understory biomass. Canopy cover, however, was a significant predictor. The relationship between canopy cover ( X) and grass production ( Y) was best expressed by the nonlinear model, Y = A B X . Total understory production could be predicted with equal precision using simple linear and nonlinear regressions with canopy cover. In this case, the simple linear model captured the interaction of the nonlinear grass response and a suspected quadratic relationship between shrub production and overstory cover as well as any other model.