Predicting Success of Walleye Stocking Programs in the United States and Canada

Abstract

Abstract We developed a model to predict the probability of establishing a reproducing population of walleye (Stizostedion vitreum) in lakes and reservoirs in North America. A list of 542 lakes and reservoirs where walleye stocking occurred was obtained by questionnaire survey. Water quality and morphometric data for these systems were obtained from STORET, the national computer data base of the Environmental Protection Agency, and various reports. Two stepwise discriminate analyses were performed on normalized data; the variables area, maximum depth, and pH were found to be significantly related to walleye success in lakes. In addition, the variable date of dam closure was included in the reservoir data set. Using a jackknife discriminate analysis, 71% of the lakes and 72% of the reservoirs were classified correctly. Use of our model increases the probability of predicting the establishment of a self-sustaining walleye population by about 20%.