4.1 | Sympatry and population structure
One of the surprising findings of this study was the strong genetic clustering where the individuals of respective clusters were dispersed among the sampling locations, as well as conserved between the two years (Figure 3). This is consistent with the predictions from the DisMELS model (Stockhausen 2009), but in contrast with the findings of Kamin et al. (2014), where the collections were mapped to the closest adult groups and no genetic clustering was detected. However, their study only used twelve microsatellite markers and therefore may have lacked statistical power to detect the finer-scale genetic clustering as the RADseq approach we employed here. This inference is supported by the lowFST values detected here (FST ranging from 0.008 to 0.032 between clusters), because detection of low FST values can require markers with high power. Also, the Kamin et al. (2014) study treated each haul collection as a sampling unit and conducted tests on the allele frequencies among the hauls, transects, locations, and years, but did not examine genetic clustering based on individual admixture analysis. However, the presence of genetic structure in our study is consistent with Palof et al. (2011), who detected isolation-by-distance population structure in the adults. The complete mixing among the genetically distinct groups of YOYs would be expected to result in a lack of population structure within just a few generations if the mixed fish maintained their grouping through settlement, recruitment and spawning. Our observations are consistent with both the DisMELS (Stockhausen 2009) and Palof et al. (2011) results indicating dispersal is not the primary mechanism by which POP population structure is maintained.
Our study suggests that distinct POP populations that are sympatric during the larval and YOY stage are likely geographically segregated and genetically differentiated during spawning. The presence of genetic clusters in spite of larval stage sympatry may indicate that once the fish settle out in the nearshore rearing habitat, they may be able to home-in to their natal locations over the following few years. If homing to their natal locations begins after fish settle out of the water column into their nearshore rearing habitat, then the mixtures of genotypes would be evident among larvae as they advected towards shore by cross-shelf currents.
The homing behavior in adult Sebastes spp. has been well documented (i.e. Carlson and Haight 1972; Matthews 1990; Carlson at al. 1995). It is unknown, however, when this behavior begins. Schools of age 1+ fish are spatially segregated (Carlson and Haight 1976), although it is unknown if those individuals are from a single or multiple source populations. It may be that these single cohort schools are composed of individuals from multiple sourced populations and like salmon, leave the school when natal location is nearby.
Homing behavior would result in genetic isolation and population structure consistent with our observations. Westrheim (1975) noted that POP schools were separated by bathymetry and would not cross deep trenches once in demersal stage. Withler et al. (2001) also described POP populations that were genetically distinct, yet lived within close proximity of each other, even when sampled in different seasons. Therefore if larvae from discrete nearby parturition locations, separated by bathymetric features such as canyons and ridges, were jointly entrapped in the oceanic currents, these clusters would resemble our observations. If homing to their natal locations begins after fish settle out of the water column into their nearshore rearing habitat, then the mixtures of genotypes would be evident among larvae as they advected towards shore by cross-shelf currents.
Another explanation for the fate of these YOY fish is that they are entrained in the coastal current and mesoscale eddies and fail to find suitable rearing habitat prior to winter settlement and are therefore destined to die, and our sampled fish were already the “swimming dead”. The selection that we observed would then be the sign of various phenotypes dying at different rates, while the unobserved fish, the ones that did not get advected away from natal grounds and mixed with other similar-fated YOYs, are the only ones that successfully reach suitable nearby rearing habitat. Westrheim (1958), and Carlson and Haight (1976) noted the extreme successes and failures among POP year classes, which perhaps may be indicative of advection rate away from the natal grounds or high larval mortality, assuming consistent spawning population.