Use of functional units of periphytic protozoa for monitoring water quality in marine ecosystems: bioindicator redundancy.

Although periphytic protozoan communities have long been used for the bioassessment of water quality, their utility is hampered by functional redundancy, leading to high "signal-to-noise" ratios. In this study, a 1-year baseline survey of periphytic protozoan communities was carried out in coastal waters of the Yellow Sea, northern China, in order to determine redundancy levels in conditions of differing water quality. Samples were collected at four sampling sites along a pollution gradient. Environmental variables such as salinity, chemical oxygen demand (COD), and concentrations of dissolved oxygen (DO), soluble reactive phosphates (SRP), ammonium nitrogen (NH-N), and nitrate nitrogen (NO-N) were measured to compare with biotic factors. A total of 53 functional units (FUs) were identified from 144 observed protozoan species based on four biological traits, i.e., feeding type, body size, movement type, and source of food supply. For reducing the "signal-to-noise" ratios of species-abundance/biomass data, the peeling procedure was used to identify the bioindicator redundancy levels based on these FUs. Three consecutive subsets of response units (RU1-RU3) with correlation coefficients > 0.75 of the full FU dataset were identified, comprising 12 FUs, 21 FUs, and 9 FUs, respectively. Algivores and bacterivores were dominant in RU1 and RU2 among the polluted sites, whereas raptors were dominant in RU3 at the unpolluted site. In terms of relative abundance, RU1 was the primary contributor to the protozoan communities during the 1-year cycle and its relative abundance increased with the increasing pollution, whereas RU2 and RU3, with complementary temporal distributions, generally decreased with increasing pollution. Ordinations based on bootstrapped average analyses revealed a significant variation in the functional pattern of all three RUs among the four sampling sites. Biological-environmental match analysis demonstrated that the variability was driven by the increasing concentrations of nutrients (e.g., NH-N, NO-N, and PO-P) and decreasing concentrations of DO (P < 0.05). There were high levels of functional redundancy among periphytic protozoan communities which could be used as bioindicators of marine water quality.