That scale is the key point. The response scale then becomes the question. Productivity per unit area would be constrained by algae productivity per unit area. But productivity per unit area is not a continental scale measure. I suspect at a continental scale, the key constraints on salmon would be temperature regime and disturbance regime. You could look at productivity at a larger scale, but I wouldn't consider productivity to be a continental scale response variable. Productivity instead responds to local scale constraints (food, prim. prodn.) in rivers where continental scale constraints are lacking (i.e. the temperature and disturbance regime are suitable).

Thank you Thomas for your answer.

To clarify, I meant more "between river" scale (up to large region).

For instance, we observe major differences in productivity between rivers with a similar temperature regime but we suspect food availability might differ. Food availability (aquatic and drifting terrestrial invertebrates) is generally correlated to temperature, but also to other regional scale variables such as water chemistry and geology . 

In a management perspective, to build a 'between river' scale productivity model, we would like to add salmon food availability, but because it is variable and labor intensive, we would be interested in finding proxy variables (such as primary production?).

Cheers   

You might try some measure of river morphology. As you are interested on large scale differences, you could try sinousity. A more heterogenous habitat might reduce the effects of intraspecific competition.

Thanks Irene,

Physical habitat is indeed quite relevant, but I am looking for biological proxy variables for food availability.

Cheers

Hi, Mathieu,

sorry, obviously I did not read your question carefully enough!

Primary production seems an appropriate candidate as Proxy variable.

However, it may be better to use terrestrial productivity instead of instream productivity. As you know, small rivers tend to be heavily dependent on subsidies from the shoreline Vegetation. However, there might be a trade-off between the cooling effect of shoreline vegetation and increased supply of leaf litter. You might get a curvilinear relationship between percentage of shoreline covered by trees and trout production.

Good luck!

Irene

Just to put it out there, I wonder if stream gradient, along with the ept (ephemeroptera, plecoptera and trichoptera)  ratio would help.