Investigating and modelling the body size structure of benthic communities
Benthic communities were investigated in terms of their body size distributions at three environmentally contrasting study sites: (i) a shallow-water location on the Fladen Ground, North Sea, (ii) a deep-water location in the Faroe-Shetland Channel and (iii) and a mid-slope oxygen minimum zone location on the Oman Margin, Arabian Sea. The construction of body size spectra formed a central component of this analysis and it served as a foundation for further investigations into the functioning and dynamics of these communities. The shape of the biomass size spectra at all three locations could best be described by biomass increasing as a function of body size. In contrast to earlier studies, the biomass distribution patterns did not display distinct evidence of bimodality, implying that biomass size spectra do not distinguish meio- and macro-fauna as two functionally distinct groups of benthic organisms. The body size spectra were found to vary in different environmental conditions. Comparisons of the two NE Atlantic locations revealed that the deeper Faroe-Shetland Channel site (1600 m) was dominated by smaller individuals than the shallower Fladen Ground site (150 m) hence conforming to the deep-sea size miniaturisation hypothesis as suggested by Thiel (1975). The size distribution patterns at the Arabian Sea site also differed significantly from the other two locations. Two taxonomic units (nematodes and polychaetes) overwhelmingly dominated the fauna in the low oxygen environment and this was reflected in the shape of the size spectra. The empirical results formed a basis for a benthic simulation model that attempted to reproduce the trends observed in the field data. The size-based approach was observed to be successful in modelling the benthic biomass distributions. The results suggested that defecation and mortality imposed a strong influence on community size structure. Production and energy flow were also estimated at community level by utilising the empirical size distribution data and the previously established allometric relations.