Evaluating existing movement hypotheses in linear systems using larval stream salamanders

Because of their linear nature, streams provide a restrictive framework to understand the movement ecology of many animals. Stream movements have been characterized under two competing hypotheses. The colonization hypothesis dictates that small individuals experience passive drift, but concurrent, upstream movement by larger individuals replaces the loss of small individuals. Alternatively, the production hypothesis suggests that downstream movements are a consequence of limited resource availability. Previous research suggests that large larvae should move upstream and vice versa for small larvae, which should therefore be found downstream more often. We conducted a mark–recapture study of larval red salamanders ( Pseudotriton ruber (Sonnini de Manoncourt and Latreille, 1801)) to assess the validity of these hypotheses. We found that no larvae exhibited downstream movement (skew = 0.361, p = 0.019; biased upstream), and large larvae were the only size cohort to exhibit directional movement upstream (skew = 0.901, p = 0.035). Contrary to predictions under the colonization hypothesis, small larvae were found upstream more frequently than large larvae (N = 871, H = 16.29, df = 2, p < 0.001). Our results suggest that larval movements are related to abiotic stream conditions, and we conclude that neither hypothesis fully explains stream movement. In the absence of drift, new movement hypotheses are necessary to describe persistent upstream movement in streams. These hypotheses should consider individual causes of movement and the direction of movements that will improve the fitness of the organism.