Seasonal movement and distribution of fluvial adult bull trout in selected watersheds in the mid-Columbia River and Snake River basins

Abstract

From 1997 to 2004, we used radio telemetry to investigate movement and distribution patterns of 206 adult fluvial bull trout (mean, 449 mm FL) from watersheds representing a wide range of habitat conditions in northeastern Oregon and southwestern Washington, a region for which there was little previous information about this species. Migrations between spawning and wintering locations were longest for fish from the Imnaha River (median, 89 km) and three Grande Ronde River tributaries, the Wenaha (56 km) and Lostine (41 km) rivers and Lookingglass Creek (47 km). Shorter migrations were observed in the John Day (8 km), Walla Walla (20 km) and Umatilla river (22 km) systems, where relatively extensive human alterations of the riverscape have been reported. From November through May, fish displayed station-keeping behavior within a narrow range (basin medians, 0.5-6.2 km). Prespawning migrations began after snowmelt-driven peak discharge and coincided with declining flows. Most postspawning migrations began by late September. Migration rates of individuals ranged from 0.1 to 10.7 km/day. Adults migrated to spawning grounds in consecutive years and displayed strong fidelity to previous spawning areas and winter locations. In the Grande Ronde River basin, most fish displayed an unusual fluvial pattern: After exiting the spawning tributary and entering a main stem river, individuals moved upstream to wintering habitat, often a substantial distance (maximum, 49 km). Our work provides additional evidence of a strong migratory capacity in fluvial bull trout, but the short migrations we observed suggest adult fluvial migration may be restricted in basins with substantial anthropogenic habitat alteration. More research into bull trout ecology in large river habitats is needed to improve our understanding of how adults establish migration patterns, what factors influence adult spatial distribution in winter, and how managers can protect and enhance fluvial populations.

Figure 1. Map of the study area.

Map of the study area in the Columbia and Snake river basins.

Figure 2. Comparison of migration distance among the study watersheds.

Box plots display the median (solid line), two middle quartiles (box), 5^th and 95^th percentiles (whiskers), and outliers (black dots) for fish from each study watershed. Plots with the same letter are not significantly different (Q = 5.5–11.7; P<0.001).

Figure 3. Seasonal distribution of bull trout in Grande Ronde River tributaries and the Imnaha River.

Seasonal distribution of bull trout in Grande Ronde River tributaries and the Imnaha River, which include the estimated spawning locations (solid triangle) and modal winter locations (hollow).

Figure 4. Seasonal distribution of bull trout in the Walla Walla, Umatilla, and John Day river basins.

Seasonal distribution of adult bull trout in the Walla Walla (Mill Creek), Umatilla, and John Day River basins, which include the estimated spawning locations (solid triangle) and modal winter locations (hollow).

Figure 5. Temporal pattern of annual life history phases of fluvial bull trout in relation to discharge.

Temporal pattern of annual life history phases of fluvial bull trout in relation to discharge. Boxplots, consisting of the observed start dates of the prespawning migration, spawning, and wintering for individual fish, were overlaid on mean daily discharge (gray lines; log scale) of the Imnaha River (RK 31; 2001–02), Grande Ronde River (RK 70; 1998–99), Wallowa River (RK 7; 2001–2), Mill Creek (RK; 1998–99), Umatilla River (RK 94; 2002–03), and John Day River (RK 404; 1998–99). Sample sizes are in parentheses.

Figure 6. Pre- and postspawning migration rates of fluvial bull trout from the study watersheds.

Pre- and postspawning migration rates of fluvial bull trout from the study watersheds.

Figure 7. Comparison of winter range distance among the study watersheds.

Comparison of winter range distance among the study watersheds. Sample sizes are in parentheses. Plots with the same letter are not significantly different (P>0.05).