1 CONCEPTS & SYNTHESIS EMPHASIZING NEW IDEAS TO STIMULATE RESEARCH IN ECOLOGY Ecology, 83(1), 2002, pp. 1–13  2002 by the Ecological Society of America RESTRICTED MOVEMENT IN STREAM FISH: THE PARADIGM IS INCOMPLETE, NOT LOST MARCO A. RODRI ´ GUEZ 1 De ´partement de chimie-biologie, Universite ´ du Que ´bec a ` Trois-Rivie `res, C.P. 500, Trois-Rivie `res, Que ´bec G9A 5H7, Canada Abstract. The ‘‘restricted-movement paradigm’’ (RMP) states that adult fish in streams are sedentary and spend most of their lives in short (20–50 m) reaches of stream. In mark– recapture studies, however, many fish initially marked are often never recaptured. As well, turnover rates of individuals in the home section (where fish were originally marked) can be high when marked fish moving out are rapidly replaced by unmarked ones. Recent challenges to the RMP have been based on the inference that high turnover indicates high mobility. However, when the home section is small many individuals may leave (high turnover) but not move far away (low displacement). I present two models for the frequency distribution of displacement distances: one represents populations as homogeneous ensembles with a single mobility parameter; the second represents populations as a mixture of stationary and mobile individuals. Both readily distinguish the turnover and displacement components of movement and show that high turnover rate is compatible with low displacement. The models were then fit to dispersal curves for six species of stream salmonids in 27 populations. Empirical estimates of turnover rate were high (median: 0.53), variable among populations (range: 0.15–0.78), but unrelated to displacement distance. Median displacement was 100 m for 24 populations and was typically 50 m. The proportion of mobile individuals was low in most populations (median: 19%) and exceeded 50% in only five of the populations. Brook trout, a species central to studies critical of the RMP, appeared to be exceptionally mobile relative to other salmonids. The compatibility of high turnover rates with short dis- placement distances and the finding that median displacement was usually limited support the notion that restricted movement is the norm in populations of stream salmonids during nonmigratory periods. However, the finding of considerable intra- and interspecific hetero- geneity in the extent of movement underscores the potential importance of the mobile com- ponent to population processes. By providing an analytical framework that yields quantitative measures of different components of movement and allows for standardized comparisons, these models can bring needed rigor to analysis and design in movement studies. Key words: animal movement; dispersal behavior; displacement distance; intrapopulation var- iation; mark–recapture; models, exponential; restricted-movement paradigm (RMP); salmonid fish; stream habitats; turnover rate. INTRODUCTION There is growing interest among ecologists in quan- tifying animal movements and evaluating their biolog- ical consequences (Gaines and Bertness 1993, Porter and Dooley 1993, Turchin 1998, Okubo and Levin 2001). For example, knowledge of fish movement in streams contributes to our understanding of energy transfers (Hall 1972), longitudinal size patterns (Hughes and Reynolds 1994), links between ‘‘source’’ Manuscript received 19 June 2000; revised 15 February 2001; accepted 16 February 2001; final version received 15 March 2001. 1 E-mail: marco rodriguez@uqtr.ca and ‘‘sink’’ populations (Schlosser 1995), colonization of isolated or newly available habitats (Taylor 1997, Lonzarich et al. 1998), and whole-stream patterns of age segregation (Hughes 1998). Stream fishes provide an excellent model system for analyses, because lon- gitudinal movement can be conveniently modeled along a single dimension, reducing model complexity and simplifying parameter estimation. Furthermore, fish can usually be sampled efficiently in streams, and rapid progress is being made due to improvements in tagging and tracking technology, which allow for non- destructive identification of small individuals over many recapture events.