Three-dimensional post-glacial expansion and diversification of an exploited oceanic fish.

Vertical divergence in marine organisms is being increasingly documented, yet much remains to be carried out to understand the role of depth in the context of phylogeographic reconstruction and the identification of management units. An ideal study system to address this issue is the beaked redfish, Sebastes mentella - one of four species of 'redfish' occurring in the North Atlantic - which is known for a widely distributed 'shallow-pelagic' oceanic type inhabiting waters between 250 and 550 m, and a more localized 'deep-pelagic' population dwelling between 550 and 800 m, in the oceanic habitat of the Irminger Sea. Here, we investigate the extent of population structure in relation to both depth and geographic spread of oceanic beaked redfish throughout most of its distribution range. By sequencing the mitochondrial control region of 261 redfish collected over a decadal interval, and combining 160 rhodopsin coding nuclear sequences and previously genotyped microsatellite data, we map the existence of two strongly divergent evolutionary lineages with significantly different distribution patterns and historical demography, and whose genetic variance is mostly explained by depth. Combined genetic data, analysed via independent approaches, are consistent with a Late Pleistocene lineage split, where segregation by depth probably resulted from the interplay of climatic and oceanographic processes with life history and behavioural traits. The ongoing process of diversification in North Atlantic S. mentella may serve as an 'hourglass' to understand speciation and adaptive radiation in Sebastes and in other marine taxa distributed across a depth gradient.

Intense habitat-specific fisheries-induced selection at the molecular Pan I locus predicts imminent collapse of a major cod fishery.

Predation is a powerful agent in the ecology and evolution of predator and prey. Prey may select multiple habitats whereby different genotypes prefer different habitats. If the predator is also habitat-specific the prey may evolve different habitat occupancy. Drastic changes can occur in the relation of the predator to the evolved prey. Fisheries exert powerful predation and can be a potent evolutionary force. Fisheries-induced selection can lead to phenotypic changes that influence the collapse and recovery of the fishery. However, heritability of the phenotypic traits involved and selection intensities are low suggesting that fisheries-induced evolution occurs at moderate rates at decadal time scales. The Pantophysin I (Pan I) locus in Atlantic cod (Gadus morhua), representing an ancient balanced polymorphism predating the split of cod and its sister species, is under an unusual mix of balancing and directional selection including current selective sweeps. Here we show that Pan I alleles are highly correlated with depth with a gradient of 0.44% allele frequency change per meter. AA fish are shallow-water and BB deep-water adapted in accordance with behavioral studies using data storage tags showing habitat selection by Pan I genotype. AB fish are somewhat intermediate although closer to AA. Furthermore, using a sampling design covering space and time we detect intense habitat-specific fisheries-induced selection against the shallow-water adapted fish with an average 8% allele frequency change per year within year class. Genotypic fitness estimates (0.08, 0.27, 1.00 of AA, AB, and BB respectively) predict rapid disappearance of shallow-water adapted fish. Ecological and evolutionary time scales, therefore, are congruent. We hypothesize a potential collapse of the fishery. We find that probabilistic maturation reaction norms for Atlantic cod at Iceland show declining length and age at maturing comparable to changes that preceded the collapse of northern cod at Newfoundland, further supporting the hypothesis. We speculate that immediate establishment of large no-take reserves may help avert collapse.