Skip booklet index and go to page content

COSEWIC assessment and status report on the coho salmon (Interior Fraser population) in Canada

Limiting Factors and Threats

Anthropogenic and natural factors can limit populations of coho salmon. Human population growth has resulted in increased demands for water, waste disposal, and because of altered land-use patterns, habitat degradation. Since many coho live in near shore marine environments, they are susceptible to natural and man-made changes to the marine ecosystem. However, impacts of human activities on salmon populations are often difficult to quantify. This is especially true in the coastal ecosystem where interrelationships among physical and biological processes are not well understood. In contrast, various studies have documented the role of climate change in altering the marine ecosystem and related this to shifts in ocean survival for salmon (e.g. Beamish et al. 1999a, b; Coronado and Hilborn 1998).

Since coho salmon spend a full year in freshwater, they are also susceptible to freshwater habitat degradation. Bradford and Irvine (2000b) related the rate at which the abundance of coho returning to individual spawning streams declined to the extent of human activity in the corresponding watershed. The hypothesis being tested was that average rates of decline (for years 1988-1998) for individual spawning populations would be negatively related to land use in the catchment. It was assumed that all spawning populations were experiencing the same rates of fishing and ocean mortality so that variability among spawning populations might be related to freshwater productivity. Bradford and Irvine showed that rates of decline were correlated with agricultural land use, road density, and a qualitative index of stream habitat status (Fig. 10).8

Figure 9: Projected Escapements of Female North Thompson Coho Under Improved (3.21 Recruits/Spawner), Average (1.47), and Poor (0.96) Survival Conditions Assuming Current Low Exploitation Rates (7%)

Figure 9: Projected escapements of female North Thompson coho under improved (3.21 recruits/spawner), average (1.47), and poor (0.96) survival conditions assuming current low exploitation rates (7%).

Horizontal lines indicate boundaries between abundance categories.

Land use patterns may be one reason why the abundance of spawners in the last 25 years in the South Thompson declined at a greater rate than those of the North Thompson did (Irvine et al. 2000). Watersheds in the South Thompson are more impacted by human activities; average scores for the three measures of land use that are correlated with coho declines (Fig. 10) were higher for the South than the North Thompson basin.

Productive freshwater habitats can help sustain salmon populations during periods of adverse marine conditions (or overexploitation) because they maximize the number of smolts produced per spawner. The analysis of Bradford and Irvine (2000b) shows that spawning populations are at greater risk when the watershed is subject to extensive human modification. Populations from healthy watersheds experienced the smallest declines, and are likely to recover at a faster rate if ocean conditions improve. Thus, the recovery and sustainability of coho will be improved through a balanced program of habitat protection and watershed restoration.

To evaluate the role of fishing in the decline of interior Fraser coho salmon, mean annual estimates of r for the North and South Thompson were used to calculate the harvest rate that would have maintained wild spawner abundances at levels similar to those of the parental escapement:

h* = 1- e-ran

where h* = 0 if r ≤ 0 (Bradford and Irvine 2000b). For years when r > 0, h* would have maintained populations at stable levels (i.e. St = St-3 ) assuming all other mortality factors remained constant. Fishing contributed to declines in abundance when exploitations exceeded h*.

Figure 10: Relationships between three land use measures and productivity of coho salmon (i.e. r) for 40 Thompson tributaries

Figure 10: Relationships between three land use measures and productivity of coho salmon (i.e. r) for 40 Thompson tributaries.

Adapted from Bradford and Irvine 2000b. (a) proportion of land in each catchment dedicated to agricultural or urban use, (b) density of forest, agricultural and hard surface roads in each catchment, and (c) index of habitat concerns. Open circles are streams that have had hatchery programs.

When actual exploitation rates were compared to estimates of h*, it was found that harvest rates were excessive from 1989 until 1998 (Fig. 11). In 1999 and 2000, exploitations were low enough that populations were above replacement levels.

Figure 11: Differences Between Exploitation Rates that would have Maintained Coho Production at the Brood Year Escapement Level (i.e. h*; St = St-3 ) and Calculated Exploitations for North and South Thompson Indicator Stream Aggregates

Figure 11: Differences between exploitation rates that would have maintained coho production at the brood year escapement level and calculated exploitations for North and South Thompson indicator stream aggregates.

Negative values indicate that populations were overexploited.


8 Note that in Fig. 7, ran is the annual mean of two indicator stream time series (1978-1998) while in Fig. 10, individual stream values of r are plotted for 40 streams, averaged over 1988-1998.