Projects
Back to Chemosynthetic Ecosystems

Geographic and Temporal Genetic Structure
Within Rimicaris exoculata Along the Northern Mid-Atlantic Ridge
Poehls and Shank


Abstract
Genetic communication between disjunct communities of hydrothermal vent shrimp, Rimicaris exoculata, on the Mid-Atlantic Ridge (MAR) was investigated using sequence data from the mitochondrial cytochrome b (cyt b).

Adults and juvenile Rimicaris exoculata were collected from four deep MAR vent fields, Rainbow, Broken Spur, TAG and Snake Pit in 2002 and from TAG in 2003 to include a study of temporal genetic variation. Previous allozyme studies suggest genetic homogeneity among TAG, Snake Pit, and Broken Spur; however, prior results have demonstrated that allozymes may not recover existing genetic subdivision. High-relief transform faults, dynamic latitudinal ocean circulation patterns (perhaps associated with transform faults or fracture zones), and/or the relative stability of venting activity along the slow-spreading Mid-Atlantic Ridge present potential mechanisms for genetic isolation between vent fields hundreds of kilometers apart.

Complete panmixia was not observed given haplotype differences between Rainbow and TAG, Broken Spur and Snake Pit, nor between adults and juveniles. However, phylogeographic, coalesence, and AMOVA analysis of the cyt b gene at these four vent fields along the MAR are consistent with allozyme data, suggesting that high levels of gene flow do occur within R. exoculata. Larvae likely overcome the barriers to dispersal by transport in upper ocean currents beyond the direct influence of the ridge topography.

Despite observed geographic homogeneity, temporal variation between years and between adults and juveniles may offer insight into dispersal patterns on an ecological scale. Barriers to dispersal may still exist, driven instead by upper ocean current regimes such as the separation between the North Atlantic Central Gyre and its southern counterpart. In order to accurately address the connectivity within R. exoculata, other ocean basins must be examined.

Hypotheses
  • Transform faults act as barriers to gene flow.
  • Metal concentrations can impose selection.
  • Genetic structure & diversity change over time
Temporal Change
Few studies have considered genetic changes over time or between juveniles and adults of a particular species. A previous allozyme study (Creasey et al 1996) observed a loss of rare alleles from the juvenile to the adult stage in R. exoculata. This loss could be due to selection against the rare genotypes during development. The present study considered both temporal differences between adults and juveniles, as well as change at a vent site, TAG, between the summers of 2002 and 2003.

Adults and juveniles and collections from TAG in 2002 and 2003 were not significantly different from each other based on Fst estimates and diversity estimates. This suggests that high levels of selection are not acting on the shrimp over the year long time interval and that the level of gene flow is similar to the TAG vent site for 2002 and 2003. The time interval may not have been long enough to detect temporal variability. For the ontogenetic changes, it is not certain that the juveniles from 2002 encorporated into the adult population by 2003 because the growth rate of Rimicaris is unknown. The same juveniles may have been sampled both years. This is unlikely though because length frequency data revealed a normally distributed size distribution with the same size mean each year (data not shown). The same juveniles would have been sample only if no growth occured over the course of the year.

Geographic Patterns
Transform faults and chemical differences between vent sites, each separated by 100s of kilometers, have the potential to genetically structure populations. The study sites included pairs of sites that were separated by major transform faults (e.g. Rainbow and Broken Spur) and sites that were within the same first order segment (e.g. Broken Spur and TAG). Additionally, the vent sites differed in chemical composition, especially in metal enrichment. These site characteristics and relationships allow us to test the effects of faults and metals on genetic structure and diversity.

Genetic analyses suggest that there is little to no genetic structure within this geographic region. Most of the pairwise comparisions have a very large number of migrants per generation exchanged and all are well above Nm=1 (the level suggested to counter act genetic drift). However, some comparisons with Rainbow vent had significant Fst values. Rainbow also lacked a large number of vent specific rare alleles. This may be indicative of selection due to the ultramafic host rock or the beginning of a cline to the north. Further study at other ultramafic host rock vents and at other vent sites further north and shallower may reveal genetic structure.

R. exoculata are believed to feed in the upper water column during larval dispersal, allowing shrimp to disperse to vents despite large physical barriers on the seafloor. The present study does not support the hypothesis that physical structures act as barriers to dispersal and gene flow for R. exoculata. The duration and behavior of the larval planktonic life of Rimicaris remains uncharacterized, but has the potential to control the community structure and genetic structure on the Mid-Atlantic Ridge. Time-series studies of genetic structure are needed to understand the processes that maintain populations at MAR vent sites.

Summary
  • Little to no significant population genetic structure (based on mtCytb) was observed among the deep MAR sites- Rainbow to Snake Pit.
  • Haploytipic differences at Rainbow suggest the existence of a allelic cline to the north or an effect of habitat selection (e.g., ultramafic fluid geochemistry).
  • Further investigation in the north near the triple junction and of other vents with ultramafic rock is warranted to understand gene flow in Rimicaris.
  • Genetic structure and diversity were not significantly different among adults and juveniles sampled by only a year a part at TAG.
  • Longer time-series collections are needed to examine the role that ecological processes may play in maintenance of vent species populations.
Acknowledgements
We thank the pilots and crew of the R/V Atlantis and the DSV ALVIN, and the R/V Akademik Keldysh, and Mir 1 & 2 for their invaluable assistance. We especiallly thank R. Reves-Sohn, S. Humphris, J.P. Canales, W. Lange, and A. Sagalevitch for the opportunity to participate on cruises. This work is supported by the Ocean Life Institute and the Deep Ocean Exploration Institute of the WHOI.