Abstract
We undertook a common garden experiment in the Burrishoole catchment, western Ireland, to test for heritable life-history differences among neighboring brown trout (Salmo trutta L.) populations that exhibit neutral genetic divergence. Experimental crosses were made using either local females (obtained from a below-waterfalls section of the Rough River within the Burrishoole) or females from the Erriff River-a neighboring catchment that currently produces a stronger run of anadromous migrants than the Burrishoole. Each female was mated to three different types of males: Rough Below-Falls, Rough Above-Falls (resident males obtained from above the waterfalls), and Erriff. Offspring from the resulting six crosses were introduced as unfed fry into a stretch of the Rough River bounded upstream by the waterfalls and downstream by a Wolf-type fish trap (Rough River Downstream Trap, RRDT). Genetic parentage analysis (16 microsatellite markers) was then used to assign offspring sampled at various time points and locations back to cross type. No differences in parr survival rates (electrofishing in the Rough River) were found among the crosses, but parr moving downstream (intercepted at the RRDT) were skewed toward the Erriff female × Erriff male cross, with a deficit assigning to the Rough Below-Falls female × Rough Above-Falls male cross. Smolts leaving fresh water (sampled at two sea-entry traps) were assigned disproportionately to crosses involving one or two Erriff parents. Offspring from pure Burrishoole crosses were more likely to become putative spawners than those from crosses involving one or two Erriff parents, pointing toward possible local adaptation. These results are consistent with heritable variation in migratory tendencies-a key aspect of intraspecific biodiversity that warrants protection-and with previous suggestions that the Burrishoole system may have evolved recently toward reduced anadromy following a novel and catastrophic anthropogenic change.