The vast majority of marine species produce young that are dispersed by ocean currents. Therefore, replenishment of local populations is dependent on both the transport of young produced elsewhere, and the retention of young in their native location. The relative contribution of these sources of young to replenishment of local populations has profound consequences for the ecology, evolution and management of marine species. Nonetheless it remains one of the most challenging questions in marine ecology and evolution. We explored this question with the kelp rockfish, Sebastes atrovirens, which is one of the most common fishes in kelp forests along the central coast of California. This is the first study to our knowledge that has identified and measured realized dispersal of a species with a substantial larval duration (two months) along the open coast of North America. As such, the study provides a critical proof-of-concept for our approach to measuring dispersal patterns for coastal marine fishes along continental coastlines. 

We sampled recruiting offspring (4,269) and reproductive adults (1,887) and, because kelp rockfish are largely sedentary following larval settlement, the distance between a reproductive adult and its settled offspring is a close approximation of its larval dispersal distance. We sampled in four areas (“local populations”) in southern Monterey Bay and Carmel Bay. We estimated the density of adults at each population from visual diver surveys each year and multiplied those densities by the area of rocky reef from high resolution seafloor maps to estimate the size of each adult population.  The resulting estimates of adult population size relative to the numbers of adults and young sampled, and parent-offspring matches detected in each year and each population allowed us to determine the contribution each population made to replenishing itself and other populations.

We identified a total of nine parent-offspring pairs, which were distributed throughout the study area. Dispersal trajectories included north to south, south to north and short distance events. Across the four-year study, we estimated that 8% of larvae recruiting within the study were produced by populations within the study region. However, this contribution varied markedly among the four years (0 to 13%). At the scale of subregions, approximating local populations, we detected self-replenishment at only one population (North Carmel Bay) in only one year (2014); however, the level of self-replenishment (8%) was surprisingly high for a species of such long larval duration along the open coast.

We also found 31 pairs of recently settled full siblings that were similarly spread throughout the study area. As with parent-offspring pairs, frequency of these sibling matches varied interannually with a corresponding pattern among the four years of the study. In addition, we identified a pair of adult siblings, and two pairs of siblings born in consecutive years, indicating either monogamy or long term sperm storage.

Together, the parent offspring matches and distribution of siblings indicate that kelp rockfish populations can be self-replenishing at a scale corresponding to an individual marine protected area (MPA), while simultaneously contributing to the replenishment of other MPAs and populations outside the MPA.  

The high dispersal potential and gene flow of the study species made highly accurate pedigree reconstruction more difficult and therefore required us to develop of an entirely novel set of genetic markers, methods and analytical framework. We leveraged the power of next generation DNA sequencing to develop a novel type of population genetic marker, the ‘microhaplotype’, that utilizes the same short read sequence data as traditional single nucleotide polymorphism (SNP) analyses, but has orders of magnitude more power for pedigree reconstruction. Our set of 96 microhaplotypes not only allowed us to identify parents and offspring and full siblings with high accuracy, but also to distinguish kelp rockfish from 48 other common co-occuring rockfish species. Many juvenile rockfish are visually indistinguishable, so this allowed us to identify the ~45% of sampled juveniles that were kelp rockfish and not other closely related species. This microhaplotype method is already being widely adopted in molecular ecology.

The project also required development of a high resolution nearshore ocean circulation model across the study region, achieved through grid nests from coarse resolution models resolving the California Current System coast-wide down to the Carmel and Monterey Bay regions, all using Regional Ocean Modeling System. The model allowed us to evaluate whether circulation patterns might explain the patterns of larval dispersal revealed by genetic sampling including higher levels of larval retention and sibling cohesion seen in one year of our study (2014). Highly variable ocean circulation enabled both local retention within and transient two-way exchange between Carmel Bay and southern Monterey Bay, consistent with genetic results. The nesting approach used in this study is now being applied to other projects examining nearshore coastal processes at very high resolution.

Last Modified: 03/05/2018
Modified by: Mark H Carr