Seascape genomics reveals metapopulation connectivity network of Paramuricea biscaya in the northern Gulf of Mexico

Abstract

The degree of connectivity among populations influences their ability to respond to natural and anthropogenic stressors. In marine systems, determining the scale, rate, and directionality of larval dispersal is therefore central to understanding how coral metapopulations are interconnected and the degree of resiliency in the event of a localized disturbance. Understanding these source-sink dynamics is essential to guide restoration efforts and for the study of ecology and evolution in the ocean. The patterns and mechanisms of connectivity in the deep-sea (> 200 meters deep) are largely understudied. In this study, we investigated the spatial diversity patterns and metapopulation connectivity of the octocoral Paramuricea biscaya throughout the northern Gulf of Mexico (GoM). Paramuricea biscaya is one of the most abundant corals on the lower continental slope (between 1200 and 2500 m) in the GoM. The 2010 Deepwater Horizon oil spill (DWH) directly impacted populations of this species and thus are considered primary targets for restoration. We used a combination of seascape genomic analyses, high-resolution ocean circulation modeling, and larval dispersal simulations to quantify the degree of population structuring and connectivity among P. biscaya populations. Evidence supports the hypotheses that the genetic diversity of P. biscaya is predominantly structured by depth, and that larval dispersal among connected populations is asymmetric due to dominant ocean circulation patterns. Our results suggest that there are intermediate unsampled populations in the central GoM that serve as stepping stones for dispersal. The data suggest that the DeSoto Canyon area, and possibly the West Florida Escarpment, critically act as sources of larvae for areas impacted by the DWH oil spill in the Mississippi Canyon. This work illustrates that the management of deep-sea marine protected areas should incorporate knowledge of connectivity networks and depth-dependent processes throughout the water column.

Kilometer-Scale Larval Dispersal Processes Predict Metapopulation Connectivity Pathways for Paramuricea biscaya in the Northern Gulf of Mexico

“…The output of our biophysical model showed a mostly congruent agreement with the estimated genetic connectivity patterns (Galaska et al, 2021). In CROCO we found a prevailing westward pathway following the ∼1,000-2,000 m isobath along the continental slope of the northern GoM regardless of seasons in 2015.…”

“…Considering the relative locations of sampled sites and the magnitude of the connectivity network, we can infer that the three sites in the Mississippi Canyon (i.e., MC344, 1,853 m; MC297, 1,571 m; and MC294, 1,371 m) are not well-connected to each other (m < 1.5%). This is consistent with the depth-differentiation hypothesis that posits that genetic differentiation is greater across depth than geographic distance, even between sites relatively close by (Quattrini et al, 2015;Bracco et al, 2019;Galaska et al, 2021). Secondly, gene flow is predominantly westward.…”

“…To evaluate the performance of our models in predicting connectivity, we compared our results with the genetic connectivity estimates (migration rates m) by Galaska et al (2021). Briefly, Galaska et al (2021) produced single nucleotide polymorphisms (SNPs) data from individuals collected at the six main sites (DC673, MC344, MC297, MC294, GC852, and KC405) using the reduced representation DNA sequencing (RAD-seq) method (Baird et al, 2008;Reitzel et al, 2013). Migration rates (m), defined as the proportion of immigrant individuals in the last two generations, were estimated using BAYESASS v3.0.4.2 (Wilson and Rannala, 2003).…”

“…The genetic analysis of the Paramuricea samples by Galaska et al (2021) shows that migration rates are generally low (average m = 0.011), with a few exceptions (Figure 4). Considering the relative locations of sampled sites and the magnitude of the connectivity network, we can infer that the three sites in the Mississippi Canyon (i.e., MC344, 1,853 m; MC297, 1,571 m; and MC294, 1,371 m) are not well-connected to each other (m < 1.5%).…”

“…This paper is a companion to the paper by Galaska et al (2021) that describes the analyses of genetic connectivity and seascape genomics. Here, we compare the modeled current velocities to those of mesoscale resolving HYCOM-NCODA reanalysis.…”