- NINA Report 226

This report suggests spawning targets for Atlantic salmon (Salmo salar) populations for 80 major rivers in Norway. The rivers were chosen among rivers prioritised by Norwegian management authorities for protection of wild salmon populations. As a starting point, we used information from nine rivers where data are available for modelling of stock-recruitment (SR) relationships. We evaluate different types of reference points for spawning populations, based on either parameter estimates from SR models or non-parametric values from the SR data points. Well functioning reference points seem to be the spawning stock giving maximum recruitment in the Shepherd SR model (Shep), and the average spawning stock giving the five highest recruitment data points (Max5R). In these analyses, spawning stock and recruitment were measured as the number of eggs (S) and the density of juveniles (R), respectively. Based on SR relationships in the nine rivers, we suggest that spawning targets for salmon populations in Norway can be grouped into four categories of egg densities from <> 5 eggs/m2 (group averages being, respectively, 1, 2, 4 and 6 eggs/m2). Shep and Max5R, which were the most robust spawning targets in an analysis of process error (“model uncertainty”) for SR data points from River Alta, give a relatively similar ranking of the spawning targets for the nine populations. Spawning targets for the large majority of salmon populations must be determined from limited information on the particular river. We transferred spawning targets from data rich rivers to data deficient rivers through considerations of scale (wetted area of the salmon-producing stretch) and of local population dynamics and productivity. Wetted area was determined by GIS methods from digital geographic data in the N50 series (scale 1 : 50 000) of Statens Kartverk, and mapping of obstacles to upstream migration registered by the Directorate for Nature Management. ssessment of productivity has for most rivers been carried out based on catch statistics (onverted to catch per area), age distribution (from scale readings), and general knowledge aout the river. In addition, we have evaluated more detailed information about the salmon population (e.g. density of smolts and/or pre-smolts) in the few rivers where this has been reported. From the spawning target (eggs/m2) we have estimated the number of eggs necessary to seed the whole river, and the corresponding number of females needed to meet that number. For some large watercourses, we provide calculations of spawning target by considering parts of the watercourse (tributaries) separately. For all rivers, we estimate the number of smolts produced, assuming that the spawning target is met for the whole watercourse and that the mortality rates from eggs to smolts are near average values for Norwegian populations. The spawning targets for most of the 80 watercourses treated in this report, lie between 2 eggs/m2 and 4 eggs/m2. The highest total spawning target is found in the River Tana, where the female spawning biomass should be around 55 000 kg to meet the target for the whole river system. Other watercourses with a high total spawning target are the rivers Gaula, Orkla and Namsen where the female biomass should exceed 18 000 kg per year, and the rivers Numedalslågen and Alta the female biomass should exceed 12 000 kg per year. We emphasize that the report provides a first-generation spawning target for the populations in question. Most of the variables used in the models are subject to measurement and process error, and for most of the rivers, limited knowledge exists about the salmon population. Currently, the two major limitations to setting precise spawning targets are believed to be estimation of productive area (as part of the wetted area) and estimation of the number of spawners from information on catch. We suggest that these limitations are targeted in future work. When due considerations are taken with respect to the limitations of the SR modelling and other methodology used in this report, we believe that the methods developed here can be used for producing spawning targets for all of Norway’s salmon-producing rivers.