On critical levels of ecological status for macrozoobenthic communities in Peter the Great Bay of the Japan Sea: model studies

Abstract

Representative sets of biotic and abiotic parameters were taken from maps of their distribution at the sea bottom, using interpolation. Good correspondence between real and modeled data was noted that allowed to engage the latter for correct estimation of the mean explained variance (MEV) by regression analysis. The magnitude and structure of MEV variability depended on external influence that proves possibility of its application as a measure for total effect of abiotic factors. Among many variants of the ecological stress index (PES), the more convenient for modeling was that one with the maximum correlation between distribution density or biomass of certain taxa and the total density or biomass of macrozoobenthos at stations (in samples). This approach provides better biological plausibility, as well, so as the benthic dwellers, like other animals, are very diverse in size and interspecific relations, that’s why the distribution density of a species correlates well with either total density or total biomass, and the same for biomass of a species. Results of in situ studies of PES dependence on MEV were confirmed by modeling with virtual MEV values. The difference between real and modeled PES values was statistically insignificant, but the MEV ranged wider in the models with virtual values and the dependence curve had weaker inflection when reached a plateau. The final dependence is S-shaped (r2 = 0.939), with the greatest curvature at 85.4 % of MEV (PES of 15 %), and the point of minimum or beginning of growth (beginning of the model curve plateauing) at 94.0 % of MEV (PES of 36 %). These levels are critical for macrozoobenthos communities. The PES level for the 2nd critical level was previously estimated as 30 %, but this value was corrected to 36 % by calculation with biologically more reasonable method.