Scottish Marine and Freshwater Science Vol 6 No 12: The demography of a phenotypically mixed Atlantic salmon (Salmo salar) population as discerned for an eastern Scottish river

Footnotes

1. If one has X calendar years of spawner-production data, this can only be modelled in full (to calculate the production of their ova) once (a) their oldest-aged smolts and (b) the oldest-sea-aged adults of those smolts, have all returned. Thus data is lost from the end of the time-series of X spawner-production years until those census results have been collated. To increase our sample size, X, we here ignore the rare extremes of smolt-ages (4,5) and sea-ages (3 or older). The data loss is reduced if the more extreme age-classes are amalgamated.

2. None of the sub-stocks had SR data particularly close to the origin, so none had data points near the origin exerting particularly strong leverage on the SR estimates of β. So all estimates of β should be similarly (un)affected by data near the origin, and hence their relative ratios robust to that (widespread) deficiency.

3. Usually spawner to spawner S-R data, with or without a year-trend to allow for changing marine survival

4. Or, if details of times spent at sea within sub-stocks were known to differ between the donor and recipient rivers, then one could instead transfer the overall monthly survival rate (as here defined and estimated) and then scale that to be appropriate to the alternative periods spent at sea by the local sub-stocks.

5. If densities were considered comparable between the catchments, but smolt ages were not, then the 'survival' aspect might plausibly be transferred by including a risk-rate adjustment for the mortality period, according to the appropriate smolt-ages (if known), thus allowing for differences in mean age at smolting between the parameter 'donor' and 'recipient' catchments.