The Effect of Site Environment on Forest Productivity in the Illinois Shawnee Hills

The relationship of total basal area to soil and topographic factors was studied in 47 undisturbed, mature, compositionally stable (climax) stands. Forward stepwise multiple linear regression was used to develop two predictive models. Effective soil depth, percentage stone, aspect, and slope position were predictor variables in model I. Soil profile available water capacity, slope position, and aspect were predictor variables in model II. Model I is the more practical and easily applied model in stands where soil depth can be rapidly determined; model II requires additional soil measurements but can be used to assess the more direct biological relationship between soil water and basal area. Models I and II accounted for 91 and 93% of the variation in stand basal area, respectively, and were validated with data from seven additional mature stands. Subsequently the models were used to predict potential (maximum) basal area in eight stands where it had been reduced by disturbance; the models also were used to evaluate the effect of farming on the productivity of fields abandoned in the 1930s. Presettlement basal area was estimated to have been between 18 and 26 m²/ha but soil loss due to erosion has reduced the potential basal area to 8—21 m²/ha should hardwood stands redevelop. For comparison with a standard site evaluation technique, 27 undisturbed Quercus alba stands were used to develop another multiple linear regression model to predict site index from site factors. Soil profile available water holding capacity, slope position, and aspect were predictor variables in model III, which accounted for 73% of the variation in site index, a substantial reduction in predictive value from model II, which predicts potential stand basal area using the same three variables. Simple regression indicated a strong positive linear relationship between stand basal area and site index (r = 0.85, P < .01). The strong relationship of stand basal area to soil and topographic factors and to standard site index measurements indicates that models to predict potential basal area may be used to evaluate site productivity in the Illinois Shawnee Hills.