An ecosystem services perspective directly links ecosystem structures and functions to human well-being , providing a tangible rationale for more holistic environmental management and protection . Ecosystem services can be difficult to quantify, especially in marine environments where interactions and boundaries can be dynamic and loosely-coupled . However, technological developments have greatly aided deep-sea scientific research . In particular, imaging of the deep sea provides useful information on physical and biological characteristics of underwater habitats . Imagery can be collected via underwater observatories , drop cameras , landers , autonomous underwater vehicles , remotely-operated vehicles , and human-occupied vehicles . Deep-sea expeditions routinely collect imagery for scientific , outreach , and industry purposes. As a result, there is a wealth of imagery that continues to grow over time as interest in deep-sea exploration and resources expands. Imagery has been instrumental to advances in our knowledge of deep-sea habitats and our ability to properly protect them. Amon et al. characterized the diversity and abundance of megafauna in a polymetallic nodule claim within the Clarion-Clipperton Fracture Zone, providing important baseline information for assessing impacts from potential mining. Another application of deep-sea imagery is evaluating vulnerable marine ecosystems,vertical farming aeroponics such as sponge gardens that enhance local biodiversity and impact biogeochemical cycling .

Additionally, images and videos provide an opportunity to visualize organisms in situ, which can be important for behavioral observations as well as for observing taxa that avoid nets . Application of deep-sea imagery to characterizing ecosystem services explicitly has been limited. Grupe et al. used ROV dive video to characterize the Del Mar methane seep and its megafaunal community. The authors found elevated densities of commercially valuable Sebastolobus spp. at the seep relative to background areas. Other deep-sea studies that utilize imagery often discuss implications for ecosystem services , but do not explicitly aim to do so. With the multitude of deep-sea imagery being collected, there is opportunity to leverage existing data in order to characterize, and ideally quantify, ecosystem services. In the summer of 2015, Ocean Exploration Trust completed an expedition to explore methane seeps and other deep-sea habitats along the southern California continental margin . Methane seeps are found in every ocean from shallow to deep water depths and are still being discovered today . Geological processes lead to seepage of methane and sulfur-rich fluids from the seabed , which fuel chemoautotrophic microbial communities that act as the base of a food web for distinct biological communities . Many “background” species can also be found at methane seeps , aggregating around authigenic carbonates , snail egg towers , or other structures that increase habitat heterogeneity. An additional layer of complexity exists along the northeastern Pacific continental margin in the form of an oxygen minimum zone , which is a midwater feature of naturally-occurring low oxygen . The OMZ can intersect benthic environments to shape local biological communities , and resulting ecosystem services, such as fish catch . Fisheries species have been previously found at methane seeps. Examples include Tanner crabs on the Cascadia margin , and Patagonian toothfish in Chile .

Southern California has four commercial deep-sea fisheries: shortspine thornyhead , longspine thornyhead , sablefish , and Dover sole . Several of these species have previously been found on methane seeps , but it is unclear how these habitats are utilized. Hypotheses include feeding in localized, high productivity areas ; breeding and laying eggs ; avoiding predators ; or removing parasites . These species also interact with seep environments through bioturbation and transporting chemosynthetic production to adjacent environments . As a result of these interactions, the Pacific Fishery Management Council considers methane seeps essential fish habitat , which are all habitats necessary for fish feeding, growth, and reproduction. Climate-regulating services related to carbon may also be provided by methane seeps. Continental margins contribute disproportionately to global carbon and nutrient cycling : although they comprise approximately 20% of global ocean surface area , continental margins have been estimated to sequester more than 40% of carbon in the ocean . Additionally, the coupling of anaerobic oxidation of methane and sulfate reduction by seep microbes serves as a methane sink through carbonate precipitation . OMZs, like the one off the coast of southern California, can also contribute to regulating services through their influence on nitrogen and sulfur cycling . The objective of this paper is to develop an approach that characterizes deep-sea ecosystem services at and around methane seeps using deep-sea imagery. We adapt biological trait analysis to target ecosystem services and focus specifically on fisheries and carbon services. These services are likely mediated, in part, by megafauna whereas services such as element cycling are facilitated by microbes which cannot be observed with imagery. However, we do discuss visual indicators of microbially-driven services where relevant. We use examples from three southern California, upper slope, methane seeps : Point Dume , Palos Verdes , and Del Mar . For two of these , we provide the first detailed characterization of megafauna. Key questions addressed are: Which megafaunal taxa are present at a given site? What functional traits or behaviors do the community exhibit? How might these traits promote ecosystem services?

And How can deep-sea exploration and observing be conducted in ways that facilitate quantification of ecosystem services? We examine the hypothesis that habitat heterogeneity at methane seeps provide more fisheries and carbon services than adjacent non-seep areas by testing for differences among active seep sites, transition areas, and non-seep background areas. We also hypothesize that fisheries and carbon services increase with diversity, which has been shown to increase ecological function, such as benthic fluxes of nutrients , that can contribute to ecosystem services. Additionally, we investigate how these services relate to depth, dissolved oxygen concentrations, and temperature.Exploratory dives were conducted by ROV Hercules in July and August 2015 as part of the OET southern California borderlands expedition NA066. High-definition video was taken continuously during each dive which ranged between 7-20 hours duration. The ROV recorded location, depth, temperature, conductivity, sound velocity, and oxygen concentrations. Because OET is focused on ocean exploration and telecommunication, we were not able to extract quantitative data from the dive videos due to changes in altitude, zoom, and non-visible laser references. However, qualitative descriptions based on presence-absence and frequency of occurrence are still useful, especially in deep-sea systems that are rarely visualized. Metadata from each dive are summarized in Table 4.1. Video from each dive was segmented into five-minute clips that were each treated as a “sample” and annotated by-hand in MS Excel . Information regarding the ROV setting and megafauna encountered was collected. Seep activity is separated into three categories: active seep sites with visual indicators of active seepage , transition areas with visual indicators of sparse or prior seepage , and non-seep background areas generally associated with soft sediment habitats. For the first minute of each video,vertical indoor hydroponic system animals were counted and identified to the highest possible taxonomic resolution. For the remaining four minutes, a full list of morphotypes was generated.Observable traits that support fisheries or carbon services were chosen , and each morphotype was assigned a score for selected traits . We used fuzzy coding to capture the extent to which trait modalities contribute to each service . Each video was scored for the morphotypes present that demonstrate the traits chosen. We were not able to calculate faunal densities from the videos due to unknown and variable camera field-of-view so we used presence-absence data. Scores were standardized by the number of morphotypes found in each clip. The minimum fisheries score a morphotype could have was zero whereas the maximum score was four. For carbon services, the minimum score was two and the maximum score was seventeen.All statistical analyses were done in R , using the base package unless otherwise noted. Data were tested for normality using a Shapiro-Wilk test. Because data did not meet normality conditions, non-parametric tests were used. The Kruskal-Wallis test-by-ranks was used to test for significant differences among groups , and a post hoc Dunn test with a Bonferroni correction was used to identify which groups were different. Correlations were tested using Spearman’s rank coefficient. All ecosystem services score analyses were done for the first-minute subset as well as for the whole video clip in efforts to decrease temporal dependence among samples. Species accumulation curves were plotted using R package ‘vegan’. Deep-sea imagery is often used to describe biological communities . When pictures and videos are collected in a systematic way , faunal densities, distributions, and biomass estimates can be produced, and compared over space and time .

Qualitative data can also be helpful. In our study, number of morphotypes increases with oxygenation among sites: the Palos Verdes dive had the highest number of morphotypes and had the highest mean overlying oxygen of 20.24 µmol/kg . This could be an artefact of the larger distance and wider depth range covered by the dive. However, oxygen has been shown to influence biodiversity on Pacific continental margins with a strong threshold effect . This threshold effect has been shown to decrease diversity beginning at approximately 22 µmol/kg , a higher oxygen concentration than the dive mean, but the highest number of morphotypes observed in a single video at Palos Verdes was associated with oxygen levels ranging from 2.18-2.26 µmol/kg. Within the Palos Verdes dive, the number of morphotypes observed in each video was negatively correlated with oxygen.In some cases, hypoxic conditions can exert selective pressure that increases specialization of taxa for increased diversity . Gallo & Levin found diverse assemblages of fish in the Pacific, Atlantic, and Indian Oceans with physiological, morphological, and behavioral adaptations for life in OMZs. Additionally, increased biodiversity with water depth to peak from 2000-3000 m has been documented in several taxa , such as demersal fish in the northeast Atlantic and cnidarians, echinoderms, and gastropods in the Caribbean . Because oxygen and depth covaried , it is not possible to separate their effect on number of morphotypes during the Palos Verdes dive. In contrast to Palos Verdes, Point Dume lies within the core of the California OMZ with mean oxygen levels of 2.76 µmol/kg. Here, number of morphotypes was significantly correlated with oxygen and depth . Because the Point Dume seep is in suboxic water, further decreases in dissolved oxygen may surpass physiological tolerances of some taxa . This may provide some evidence for loss of available habitat and shifting faunal distribution due to climate change . As deoxygenation continues to expand and intensify the OMZ , animals that cannot tolerate low oxygen conditions willlose available habitat while those that can will distribute accordingly . The decrease in number of morphotypes with depth observed in Point Dume videos may be driven by the significant negative correlation between oxygen and depth . The correlation between number of morphotypes and depth here is the opposite of that observed during the Palos Verdes dive, highlighting differential effects of environmental parameters at different locations along the continental margin.All three sites had some visual evidence of bacterial mats during the dive , likely indicating microbial sulfide oxidation and possibly some methane oxidation. Sulfideoxidizing microbes can detoxify sulfidic water , which could facilitate occurrence of morphotypes that contribute to ecosystem services. Sulfide oxidation also consumes oxygen with active seep areas, consuming two orders of magnitude more oxygen than non-seep areas . However, seep influence on sediment macrofauna communities, on which megafauna could be feeding, seems to be limited . Only 25% of morphotypes occurred on the bacterial mat, most frequently Liponema anemones , P. rufescens , and galatheid crabs . One morphotype of polychaete was found exclusively on bacterial mats with two occurrences. Our results suggest that the active seep areas of Point Dume have lower fisheries and carbon scores than transition and background areas . Intense seepage with hydrogen sulfide and exceptionally low oxygen may act synergistically to reduce the occurrence of functional traits that generate ecosystem services. During the Point Dume dive, we did not find significant correlations between ecosystem services scores and oxygen , which ranged from 2.01-4.73 µmol/kg. However, scores were significantly correlated with water depth which negatively covaried with oxygen . Unfortunately, separating these effects with our dataset is not possible. Palos Verdes transition areas provided significantly higher fisheries services than active and background areas . As mentioned before, this is likely driven by the large aggregations of fish found on carbonate rocks in transition areas .