In landscape genetics, model selection procedures based on Information Theoretic and Bayesian principles have been used with multiple regression on distance matrices (MRM) to test the relationship between multiple vectors of pairwise genetic, geographic, and environmental distance. Using Monte Carlo simulations, we examined the ability of model selection criteria based on Akaike’s information criterion (AIC), its small-sample correction (AICc), and the Bayesian information criterion (BIC) to reliably rank candidate models when applied with MRM while varying the sample size. The results showed a serious problem: all three criteria exhibit a systematic bias toward selecting unnecessarily complex models containing spurious random variables and erroneously suggest a high level of support for the incorrectly ranked best model. These problems effectively increased with increasing sample size. The failure of AIC, AICc, and BIC was likely driven by the inflated sample size and different sum-of-squares partitioned by MRM, and the resulting effect on delta values. Based on these findings, we strongly discourage the continued application of AIC, AICc, and BIC for model selection with MRM.

No abstract

The goal of this research was to use the long-term fishery data set and DNA from archived scales of walleye Sander vitreus in Escanaba Lake, WI, U.S.A., to improve the understanding of the underlying mechanism(s) influencing genetic diversity in naturally recruiting populations. The introduced population of S. vitreus in Escanaba Lake has a low mean effective population size (N(E)) between 124.6 and 185.5 despite a mean census size (N(C)) of 4659 (N(E)/N(C)c. 0.04), suggesting an accelerated rate of genetic drift between 1952 and 2002. These values are smaller than the median N(E) range of several studies suggesting typical N(E)/N(C) ratios of 0.11-0.16 in a wide range of taxa. N(E) increased steadily during the past two sampled decades (1992 and 2002) and was consistent with a lowering of the variance in S. vitreus reproductive success, possibly linked to a large, sustained exploitation (mean 28%) rate. Variance in reproductive success is one of the most important factors influencing N(E) in species, like S. vitreus, which have a potential for large fecundities and large juvenile mortalities (type III survivorship). The N(B) estimates across six sequential cohorts (age classes of S. vitreus, assayed from 1994 to 1999) was consistent with estimates of N(E) reported for 1992-2002. These results, coupled with in-depth census and exploitation data, show that the genetic characteristics of Escanaba Lake S. vitreus have changed substantially and that management activities, such as supplemental stocking and harvest practices, have profoundly influenced the genetic dynamics of S. vitreus in this lake.

Rehabilitation of migratory (''coaster'') brook trout Salvelinus fontinalis along Lake Superior's south shore is a topic of high interest among resource stakeholders and management agencies. Proposed strategies for rehabilitation of this brook trout life history variant in Wisconsin include supplemental stocking, watershed management, habitat rehabilitation, harvest regulations, or a combination thereof. In an effort to evaluate the success of coaster brook trout rehabilitation efforts, we collected genetic data from four populations of interest (Whittlesey Creek, Bois Brule River, Bark River, and Graveyard Creek) and the hatchery sources used in the Whittlesey Creek supplementation experiment. We characterized the genetic diversity of 30 individuals from each of four populations using 13 microsatellite DNA loci. Levels of genetic variation were consistent with those in similar studies conducted throughout the basin. Significant genetic variation among the populations was observed, enabling adequate population delineation through assignment tests. Overall, 208 of the 211 sampled fish (98.6%) were correctly assigned to their population of origin. Simulated F 1 hybrids between two hatchery strains and the Whittlesey Creek population were identifiable in the majority of attempts (90.5-100% accuracy with 0-2.5% error). The genetic markers and analytical techniques described provide the ability to monitor the concurrent coaster brook trout rehabilitation efforts along Wisconsin's Lake Superior south shore, including the detection of hybridization between hatchery and native populations.

No abstract