Document: CAR-3-91-7

Statistical thinking and eoclogical release: More on the weight of evidence.

TUCKFIELD, R.C.* ¹ and M.C.BELK ²

Savannah River Technology Center, Savannah River Site, Aiken, SC 29808 USA ¹
Brigham Young University, ²

Abstract:
Statistical thinking is essential to getting an understanding of ecological phenomena. It is more than methods for data analysis. It is a paradigm for prospective data designs and for weighing the empirical evidence. An important and global ecological problem is that of ecological release, i.e. what to do about the introduction of non-native species which threaten by predation and competition, the ultimate demise of native flora and fauna. Statistical thinking in the form of an a priori power analysis was provided to ecologists studying the non-native brown trout (Salmo trutta) and its potential to prey upon native Utah fish species to extinction. A sampling plan was recommended and implemented in a study sponsored by the Provo River Mitigation Project, Utah. In addition to a randomized blocking factor, a habitat complexity covariate was used to model the effects of predator size and abundance on prey mortality. Weighing the evidence with effective data displays shows that juvenile survival in native fish may depend on the availability of refugia. Recommendations are provided to investigators that will reduce the variation in subsequent experiments, improve statistical power, and add more weight of evidence for taking potent action to restore the Provo River.

Keywords: statistical thinking, power analysis, ecological release

This abstract is being presented at: 4:00 PM in session:
Oral Session #46: Modeling Populations and Statistical Ecology.