Repository logo

First network analysis of interspecific associations of abyssal benthic megafauna reveals potential vulnerability of abyssal hill community

Accepted version

No Thumbnail Available



Change log


Mitchell, Emily G 
Durden, Jennifer M 
Ruhl, Henry A 


The distribution of organisms is related to both environmental factors and interactions between organisms. However, such associations between organisms across an abyssal megafaunal community have not previously been investigated at landscape scale because of a lack of positional data on specimens over such scales. We quantified spatial distributions and investigated interspecific associations in benthic megafaunal communities in three contrasting habitats on the Porcupine Abyssal Plain, two on the abyssal plain and one on the flank of a modest abyssal hill (~50 m above the plain). We used a Bayesian Network Inference Algorithm approach, which considers the ecosystem as a network, facilitated by robust positioning of specimens determined through seabed photography captured with an autonomous underwater vehicle. We found non-random interspecific distributions of most morphotypes in all areas. The organisms in two interspecific networks on the abyssal plain had high connectance and link density, while the network at the Hill site was notable in the lack of inter-dependencies and highly dependent on one morphotype – Ophiuroidea. The reduced connectance of the hill community suggests that it is operating under a different regime and potentially more vulnerable to perturbation than those on the plain. Interspecific dependencies on the abyssal plain occurred across broad taxonomic groupings, and were thought to be a result of similar relationships between pairs of organisms and the substrate, and competition for detrital resource. In addition, some intraspecific pairs changed dependency direction at different scales. Our results suggest that the scales of inter- and intraspecific aggregation will be important considerations in the design of community assessments, and in spatial planning for their conservation.



Invertebrate, Interaction, Bayesian, Species distribution, Environmental filtering, Porcupine Abyssal Plain, Aggregation, Non-random distribution

Journal Title


Conference Name

Journal ISSN


Volume Title



Elsevier BV
Natural Environment Research Council (NE/P002412/1)
Natural Environment Research Council (NE/S014756/1)