Why Fishing Does Not Magnify Temporal Fluctuations in the Population Abundance of Stream-Living Salmonids

The hypothesis that size-selective fishing induces magnified temporal variations in recruitment of fished, relative to un-fished, populations is explored by comparing the recruitment of two stream-resident populations of brown trout Salmo trutta inhabiting two tributaries of Rio Esva drainage (northwestern Spain). One population is exploited by angling; the other has never been fished. Fishing truncated the length structure of the fished population. In some years, fishing extirpated the two older reproductive year-classes (age 2 and 3), and the reproductive potential was limited to the age 1 spawners. Nevertheless, the temporal variation in recruitment over a 20-year period was lower in the fished population. The inter-annual variation in recruitment of the two populations closely tracked inter-annual environmental variation with a parabolic relationship between recruitment and stream discharge. Year-to-year variation in the carrying capacity to sustain recruits implies that annual recruitment only requires the survival of a few spawners to buffer the combined effect of environmental variability and fishing-induced mortality. Conventional fishing theory is not compatible with such processes, suggesting that new strategies are required to make fisheries and conservation goals compatible, with the importance of environmental stochasticity replacing the deterministic character of density-dependent population growth rates inherent to conventional fishery models.