PREDICTING PROBABLE EFFECTS OF URBANIZATION ON FUTURE ECOLOGICAL INTEGRITY IN THE UPPER ILLINOIS RIVER BASIN, USA

摘要

A study was undertaken to predict the probable effects that future urbanization may have on ecological integrity in the Upper Illinois River Basin (Chicago area), USA. Biotic indices and sediment trace-element concentrations for 43 streams, determined by Illinois State agencies and as part of the U.S. Geological Survey's National Water Quality Assessment program, were examined along an agricultural-to-urban land-use gradient. The relations found among biotic integrity, sediment chemistry, and urbanization were associated with annual samples collected from 1982 through 1993. Because these annual samples were from different tributary basins with different urban percentages and geologic settings, the trends along the gradient suggest the absence of bias. Analytical equations were fit to bivariate relations, and probability density functions fit to residuals for use with the Monte Carlo technique so that stochastic modeling could be performed. Stability of stochastic modeling required 1,500 Monte Carlo trials; reliability of stochastic modeling was evaluated by comparing statistical summaries of measured to simulated biotic indices, and future predictions approximately validated against an independent AIBI score for Long Run Creek. Stochastic modeling of future urbanization-induced changes in ecological integrity for basins (Big Rock Creek, Des Plaines River, Mill Creek, and Flag Creek) along an urban gradient (1990 percent urban land use of 1, 5, 10, and 87 percent) resulted in a broad range of probable biotic resource quality (excellent to very poor). Predictors used to simulate changes in basin ecological integrity from 1990 to 2000 and 2000 to 2010 included fish and invertebrate biotic indices, and streambed sediment nickel concentration. Using these predictors, the degradation of ecological integrity in tributary basins occurred at differential rates and with a probable distribution of likely outcomes. For example, the AIBI median predictions of ecological integrity from 1990 and 2010 was 2 quality classes (good to poor) in the Big Rock Creek and Des Plaines tributary basins, and 1 quality class (poor to very poor) in the Mill Creek and Flag Creek tributary basins. A scale was devised for converting MBI scores to biotic resource quality classes for interchanging results with AIBI scores. This calibrated scale should be useful in more urbanized streams where it is not always possible to compute AIBI scores, and for comparison between biotic indices in other studies. Bed sediment nickel concentration was a useful predictor of ecological integrity and basin percent urban land use (and population density). Because the time and costs for determining nickel concentrations are much less than for determining biotic integrity scores, future studies could use this scale or other correlated variables as predictors.