Computational protocol: Insights into the abundance and diversity of abyssal megafauna in a polymetallic-nodule region in the eastern Clarion-Clipperton Zone

Similar protocols

Protocol publication

[…] All video from both cameras on the ROV were viewed multiple times and frames archived of each identifiable megafaunal morphotype. The criteria used for selection of megafaunal morphotypes was that individuals were greater than 2 cm in maximum dimension and that there was sufficient detail to identify them to a putative ‘species-level’ morphotype (see ). Morphotypes that could not be identified to species but appeared morphologically distinct were assigned a unique informal species name (e.g. Polynoidae sp. 1). Both metazoans and protists (xenophyophores) were included in this analysis. These were sorted into taxa, identified by taxonomic experts or by using the “Atlas of Abyssal Megafauna Morphotypes of the Clarion-Clipperton Fracture Zone” created for the ISA (, and then used to create an ABYSSLINE megafaunal morphotype atlas for the UK-1 contract area (Amon et al., in prep). This UK-1 atlas was crucial for our quantitative faunal analyses and will also be useful for subsequent megafaunal analyses within the CCZ. This process provided an estimate of the number of megafaunal species in UK-1 Stratum A and the UK-1 contract area, and will aid in delimiting species ranges. However, since the majority of the morphotypes were not collected, it is impossible to confirm species identities in most cases or undertake systematic studies on this fauna. These issues are explored further in the discussion.During surveys, the vehicle had substantial difficulty maintaining constant altitude, direction and velocity over the seabed, thereby limiting the availability of usable imagery. To facilitate quantitative analyses of abundance, the high-definition video from each of the four quantitative transects was split into frames taken every two seconds, using the software Quicktime Pro 7. This yielded approximately 13,700 frames, from which blurry images, images at altitudes >3.2 m and <1.2 m, and overlapping images were removed. The remaining 2458 frames were color-corrected, scaled, and analyzed in ImageJ® (see ). In order to maximize the spatial coverage of our study, we enumerated and identified megafauna (using the criteria defined above) within the maximum usable area within each randomly-selected frame (quadrat sizes varied from 0.2 to 4.5 m2) (see ).Because of the enormous effort required to count and measure the nodules in quantitative surveys (an important environmental variable in our study), exposed nodule abundance and size was evaluated in a subset (10%) of the frames used in the quantitative megafaunal analyses, referred to as quantitative nodule analysis from here forward. These frames were selected at random and then colour-corrected, scaled, and overlain with a randomly-placed quadrat of 0.66 m2 using ImageJ® software. Within each seafloor quadrat, each nodule was counted and then manually outlined to measure plan area. From these values, percent nodule cover was calculated for each quadrat. All megafauna within the quadrat were counted, identified to morphotype using the criteria above, and their underlying substrate recorded. Although many of the nodules in the UK-1 contract area are partially or fully buried in sediment, our nodule counts could only record exposed nodule surfaces. These data reflect the quantity of sediment-free hard substrate available to megabenthos, e.g., for attachment sites. [...] Species accumulation curves and richness estimates were made using Primer v.6. Since Ugland species accumulation curves indicated that species inventories had not yet reached asymptotes and were continuing to accumulate, the recommendations of Magurran were followed and the Chao 1, Jacknife 2 and Bootstrap estimators were used to estimate total species richness.Community similarities were evaluated using a cluster analysis based on Bray-Curtis similarities of square-root transformed abundance data (to allow contributions from both common and rare species) using Primer v.6 software. For these analyses, a large proportion of individuals including (for example) many ophiuroids, could not be confidently assigned to a species-level morphotype (e.g. Ophiomusium cf. glabrum and Amphioplus daleus). For similarity analyses, we assigned unidentified individuals to morphotypes in proportion to morphotype occurrence for confidently identified individuals from the same taxon. This method effectively split the unresolvable individuals into morphotypes in a proportion that was consistent with our confident observations. This approach likely underestimated the number of species present. The contributions to diversity of obligate hard-substrate-dwelling fauna, as well as fauna not requiring hard substrate (i.e., “facultative” species seen on both nodules and sediment, soft-sediment obligates, bentho-pelagic species), were assessed by determining the number of species unique to hard substrate (nodules), and by estimating total species richness by site using Chao-1 species-richness estimators.For quantitative surveys, non-parametric Kruskal-Wallis comparisons were used to test for significant differences between megafaunal abundances and nodule parameters by image at site and transect levels. Correlations between faunal abundance and nodule parameters were also explored using Pearson’s product-moment correlation coefficient. Chi-squared tests were used to measure significant differences between faunal abundances, biodiversity indices (Shannon’s H’, Hulbert rarefaction Es(100) and Pielou’s Evenness J’) and nodule parameters at the site and transect level. A p-level = 0.05 was used throughout as the criterion for statistical significance. These analyses were conducted in SPSS Statistics v. 22.0 . […]

Pipeline specifications