Olsgard, F. & Somerfield, P.J. Journal of Aquatic Ecosystem Stress and Recovery (2000) 7: 25. doi:10.1023/A:1009967313147
Although the identification of organisms to the levelof species is the ideal in studies of marinemacrobenthos, there are situations where such a finelevel of taxonomic discrimination may be eitherimpossible or unwarranted, for example when much thefauna is undescribed, or if the task for which samplesare collected does not require them to be identifiedto the species level. The idea that abundances ofhigher taxa, or particular groups of organisms, may beused as surrogates for the total fauna in such studiesis explored in this paper using data from theNorwegian sector of the North Sea. The generalconclusion is that, in surveys of soft sedimentmacrofauna in disturbed areas of the North Sea wherepollution imposes simple spatial gradients on thebenthic communities, little information aboutinter-sample relationships is lost using data based onfamily, polychaete species, or polychaete familyabundances, rather than species abundances. In morepristine areas where spatial patterns are determinedby a number of processes, correlations betweencalculated diversity indices and similarity in faunalpatterns between species and family abundances arestill very high, but less so for polychaete species orpolychaete family abundances. This suggests thatidentification to the level of family may besatisfactory in many routine monitoring surveys,andidentification of only the polychaetes, either to thelevel of species or family, may also be a possiblealternative if there are clear disturbance gradientsin the survey area. Polychaetes are of importance indisturbed areas because the group contains tolerantand intolerant species, and in undisturbed areasbecause within the taxon species have a greater rangeof trophic and reproductive strategies than withinother taxa. Ultimately it is the distribution ofspecies, their identities, and their interactions witheach other and with the environment, that are ofinterest. The use of surrogates is likely to be mostadvantageous if it is only the extent of pollutioneffects from a discrete source that matters, andspecies level baseline studies have already beencompleted.