Case Study: Kerber Creek Restoration Project: Employing Statistical Techniques to Analyze Effects of Restoration Activities, Saguache, Co

Abstract

Abstract:The Kerber Creek watershed, located in Saguache County, CO has been undergoing restoration since 1991 to address the impacts of historic mining activities in the upper watershed. These efforts have produced reasonable, albeit limited, quantities of data on stream morphology, water quality, fish, macroinvertebrate populations, and vegetation cover within the watershed. However, to date there has not been a concentrated attempt to evaluate the effects of restoration on these variables. The objectives of this case study are to employ robust statistical techniques to analyze the effects of restoration on sinuosity in the Kerber Creek watershed and to assess the validity and feasibility of using these statistical methods as project evaluation tools. Sinuosity was measured at five restored sites using National Agricultural Imagery Program one-meter resolution aerial imagery from 2005, 2009, and 2011. The phytostabilization index was used to represent the extent to which the floodplain was restored at each site through in-situ treatment of mine waste deposits, termed phytostabilization. Repeated measures analyses of variance were subsequently performed to evaluate the effects of time (i.e., natural channel evolution) within sites and the extent of restoration among sites on changes in sinuosity. Simple linear regression analysis was then employed to elucidate the nature of the relationship between extent of phytostabilization and within-sites sinuosity means. No treatment was found to have a significant effect on sinuosity at the 0.05 or 0.10 levels of significance. Similarly, the regression coefficient for the phytostabilization index was not significant (p >0.20), and the correlation coefficient was relatively low (r 2 = 0.357). Although these results indicate that restoration activities in the Kerber Creek watershed have not significantly improved sinuosity, a number of methodological issues, including the suitability of statistical models and the phytostabilization index, lack of sufficient data, and the presence of outliers, require cautious interpretation. Most importantly, this case study reveals the necessity for intensive monitoring regimes to accurately analyze project results and identifies numerous variables that must be considered when designing a statistically valid restoration project evaluation technique. _______________________ 1 Paper presented at the