Are characteristics of introduced salmonid fishes biased by propagule pressure?

Many recent studies of nonindigenous species (NIS) have used life history and morphological characteristics of invaders to either (i) build statistical models that predict new invaders or (ii) test ecological and evolutionary hypotheses. However, species characteristics may be confounded if NIS are transported or introduced nonrandomly with respect to the chosen contrast group, which typically consists of native or globally available species. For example, deliberately introduced NIS are often chosen according to economic rather than ecological factors. Here, I use stocking records of salmonid species introduced into Nevada, USA, to test for propagule biases within this system. I find that established salmonids are introduced significantly more times, and in greater numbers, than are those that fail and that species chosen for introduction are a nonrandom sample of the global salmonid species pool. Statistical differences among characteristics of salmonid species that established, those that were introduced, and those from the global source pool suggest that maximum reported size and weight as well as latitudinal range and midlatitude all represent propagule-biased characteristics of salmonids introduced into Nevada. These results highlight the need for caution when using characteristics of invaders to develop statistical models or to test hypotheses relevant to ecology and evolution.