Substrate Cover and Nooksack Dace

The final curve is largely based on the relationship between % cobble substrate cover and Nooksack Dace density identified in Gray et al. (2024) using empirical data (Fig. 1a below), supported by the habitat suitability curve for Longnose Dace (Edwards et al. 1983; Fig. 2), which also generally supports a positive association between % cobble cover and system capacity. However, we modified the density vs. % cobble curve from Gray et al. (2024) to increase the initial slope to make it closer to the Edwards et al. (1983) Longnose Dace curve. This was because almost no dace critical habitat reaches have the 80% cobble in riffle habitat required to achieve 100% capacity as predicted by the Gray et al. (2024) function (Fig. 1a). Even the best Bertrand Ck. high-density Nooksack Dace habitat (which we are using as the reference benchmark for 100% capacity) is in the range of 50-60% cobble cover, which is not consistent with the Gray et al. 2024 function, since we assume that high-density Bertrand Creek reaches represent a reference condition of 100% capacity. The final curve parameterization closer to the Edwards et al. curve ensures that current maximum dace densities are consistent with upper model predictions, while still allowing for increased cobble abundance to incrementally improve habitat. Note the large scatter of data in Gray et al. (2024) Figure 1a relationship broadly overlaps with both the fitted Gray et al. curve and the final SR function relationship. In addition, at the expert elicitation workshop held in November 2024, John Gray indicated that Dace may even prefer larger substrate (i.e. boulder) and that a SR function for boulder would be similar to the one for cobble. Therefore, for the x-axis of this SR function %cobble and boulder combined is more appropriate for more general application to streams/reaches that actually may have boulder-sized substrate present.