Conservation planning is often hampered by the lack of causal quantitative links between landscape characteristics, restoration actions, and habitat conditions that impact the status of imperiled species. Here we present a first step toward linking actions on the landscape to the population status of endangered stream-type Chinook salmon (Oncorhynchus tshawytscha). We developed relationships between land use, landscape characteristics and freshwater habitat of spring Chinook salmon in a Pacific Northwest river basin. Available data allowed us to find relationships that described water temperatures at several life stages (prespawning, egg incubation, and summer rearing) and substrate characteristics, including fine sediments, cobble, and embeddedness. Predictors included elevation, gradient, mean annual precipitation, total and riparian forest cover, road density, impervious surface, and alluvium. We used a model averaging approach to account for parameter and model selection uncertainty.
Results/Conclusions
Key predictors for prespawning and summer rearing temperatures were total forest cover and impervious surface area. Precipitation and stream gradients were important predictors of the percent of fine sediments in stream substrates. We estimated habitat conditions using these relationships in three alternative landscape scenarios: an estimate of historical conditions, a no restoration scenario, and a scenario that included a set of restoration actions from local conservation planning. We found that prespawning and summer temperatures were estimated to be slightly higher historically relative to current conditions in dry sparsely forested areas, but lower in some important Chinook salmon spawning and rearing areas and lower in those locations under the restoration scenario. Fine sediments were lower in the historical scenario and were reduced as a consequence of restoration actions in two areas currently unoccupied by Chinook salmon that contain reaches with some potential for high quality spawning and rearing. Cobble and embeddedness in general were predicted to be higher historically and changed little as a result of restoration actions relative to current conditions. This modeling framework converts suites of restoration actions into changes in habitat condition, thereby enabling restoration planners to evaluate alternative combinations of proposed actions. It also provides inputs to models linking habitat conditions to population status. This approach represents a first step in estimating impacts of restoration strategies, and can provide key information to conservation managers and planners.
