Multiple Scale Approach to Understand Connectivity and Fragmentation of Seamount Megafaunal Assemblages in the North Pacific

In the North Pacific, seamounts are important ecological hotspots for deep-sea species found below 200 m. Connectivity of deep-sea fauna in this area is of great interest to fisheries and conservation management for possible restoration and mitigation of anthropogenic impacts. This work examines how seamount fauna are connected at multiple scales to further understand how the variety of seamount habitats can impact the distribution and connectivity of those assemblages. To look at connectivity at a fine scale, environmental data and assemblage structure was analyzed from image data between three sides of Mokumanamana seamount between 200 and 700 m of depth. The assemblages found at the three sides are significantly different (global R=0.807, p=0.001), though shallower assemblages are more similar to each other at any side than deeper assemblages. Assemblages were also different between shallow and deep samples within any one side of Mokumanamana (global R=0.80, p=0.001), and all assemblage differences were influenced by environmental variables of sediment levels, substrate variability and roughness, oxygen, particulate organic carbon, and surface currents. The microhabitats caused by substrate changes and changes in depth allow for unique assemblages to be found within a feature, and greatly increase diversity of the area. To understand connectivity at a much larger geographic scale, populations of the precious red coral Hemicorallium laauense were sampled at 16 locations across the Hawaiian Archipelago, and seven microsatellite loci were amplified in 270 individuals to compare genetic connectivity within and between seamounts. Corals on these locations show significant population differentiation at a global scale (G'ST = 0.17, p = 0.01), as well as differentiation within features. DAPC showed a pattern of separation of the Main Hawaiian Islands from the Northwest Hawaiian Islands, though samples from Ka'ena Point appear to form their own cluster and Bank 8 clusters with the Main Hawaiian Islands. Membership assignments showed moderate admixture between some locations, while East and West Northampton and Raita Bank showed almost no admixture. K-means cluster assignments instead show five populations that are well distributed throughout the Hawaiian Archipelago, and some pairwise G'ST values are nearly zero for very distant populations. The unexpected pattern of more differentiation at smaller spatial scales compared to wider spatial scales is likely tied to extremely patchy distributions, episodic recruitment, and overlapping generations of long-lived corals that creates a sweepstakes effect on allele frequencies. Also an important location in the North Pacific, the Emperor Seamount Chain and North Hawaiian Ridge (ES-NHR) have been heavily fished by deep-sea trawling vessels since the 1950's, but this high-seas region has recently come under management by the North Pacific Fisheries Commission (NPFC). With an understanding that the affected seamount communities can have very patchy distributions, and be dominated by local recruitment for larvae, a fine-scale approach was taken to better define the fishing footprint of trawling vessels using the open-source satellite Automated Identification System data provided by Global Fishing Watch. Bottom-fishing activity was quantified from 2012 – 2018 at a 0.01 by 0.01-degree scale, and much of the fishing was focused on just four seamounts of 14 features managed by the NPFC. Another four seamounts show minor bottom fishing activity, while no activity was identified on the remaining six features. Voluntary closures on Koko Seamount and C-H seamount to protect known spawning grounds and vulnerable ecosystems appear to be observed by fishing vessels. In addition to satellite data, Autonomous Underwater Vehicle (AUV) surveys from 2014 and 2015 provide visual data on the fishing footprint through trawl scars and locations of abundant megafauna where fishing has likely been avoided. AUV data suggests the fishing footprint was either not fully encompassed by satellite data, or it has moved through time. Locations of remnant or recovering coral and sponge populations found in AUV data can provide starting points for further small-scale closures that will likely not impact fishing in the area but will protect vulnerable ecosystems in this area. These studies provide further understanding on the connectivity of seamount assemblages, the potential human impacts on connectivity, and can aid in developing science-based management plans that work with the heterogeneity and patchiness that have been found for seamount ecosystems.