Seasonal and temperature-related movement of Colorado River cutthroat trout in a low-elevation, Rocky Mountain stream.

Abstract

Mobile species will migrate considerable distances to find habitats suitable for meeting life history requirements, and stream‐dwelling salmonids are no exception. In April–October 2014, we used radio‐telemetry to examine habitat use and movement of 36 Colorado River cutthroat trout Oncorhynchus clarkii pleuriticus (CRCT) in a 14.9‐km fragment of Milk Creek, a relatively low‐elevation stream in the Rocky Mountains (Colorado). We also used a network of data loggers to track stream temperature across time and space. Our objectives were to (1) characterize distribution and movement of CRCT, (2) evaluate seasonal differences in distribution and movement of CRCT, and (3) explore the relationship between stream temperature and distribution and movement of CRCT. During the course of our study, median range of CRCT was 4.81 km (range = 0.14–10.94) and median total movement was 5.94 km (range = 0.14–26.02). Median location of CRCT was significantly further upstream in summer than in spring, whereas range and movement of CRCT were greater in spring than in summer. Twenty‐six of the 27 CRCT tracked through mid‐June displayed a potamodromous (freshwater migratory) life history, migrating 1.8–8.0 km upstream during the spring spawning season. Four of the seven CRCT tracked through July migrated >1.4 km in summer. CRCT selected relatively cool reaches during summer months, and early‐summer movement was positively correlated with mean stream temperature. Study fish occupied stream segments in spring and fall that were thermally unsuitable, if not lethal, to the species in summer. Although transmitter loss limited the scope of inference, our findings suggest that preferred habitat is a moving target in Milk Creek, and that CRCT move to occupy that target. Because mobile organisms move among complementary habitats and exploit seasonally‐unsuitable reaches, we recommend that spatial and temporal variability be accounted for in delineations of distributional boundaries.