Epilithic diatom community response to years of P04 fertilization: Kuparuk River,Alaska (68 N Lat.)

An arctic river was fertilized continuously through the ice-free season with phosphoric acid beginning in 1983. The epilithic diatom community increased in biomass in the first two years in response to the added limiting nutrient (Peterson et al., 1983). The diatom community switched from one dominated by Hannea arcus to one dominated by species of Achnanthes and Cymbella. The immediate responses to the P-addition were decreases in both the Shannon diversity and evenness indices. By the second year, the community diversity increased downriver reaching maximal species richness (110–127 spp). In 1985–1987, the epilithic algal biomass decreased an order of magnitude with both whole-river PO₄ (1985, 1987) and PO₄ + NH₄ addition (1986). In the 5th summer of fertilization, the reduction in biomass was clearly caused by a numerical increase of grazing, refugia-building chironomids (Orthocladiinae, primarily) (Gibeau, 1991; Gibeau, Miller, Hershey, in prep.). We assume the algal biomass reduction in the 3rd and 4th years was similarly caused by grazers with a two year time lag in the numerical response of these monovoltine species. The evenness of the community increased in 1986 as if it might have been grazed; however the number of immigrants was reduced. The community became dominated by Eunotia, Cymbella and Achnanthes, species either fast growing or more prostrate, as the erect species of Hannea Diatoma, and Fragillaria declined. A detrended correspondence analysis of the temporal and spatial diatom samples in species space (186 spp.) showed that the largest variation in the community was between years and less variation was associated with river fertilization. Samples from bioassay tubes run by Peterson et al. (1983) in the Kuparuk River showed P and N + P limitation as found in the river in 1983–84. Like the river samples, the largest change in the diatom community occurred between 15 and 25 day samples, more than that induced by fertilization. Diatoms sampled from all treatments taken at day 25 were more similar to one another than those sampled at day 15. Diatoms colonizing glass slides used in the bioassay tubes were dominated by Achnanthes linearis and Cymbella minuta. Of the 84 species found in bioassays, 26 species were present in all river samples for 4 years. Differences in the communities discriminated by multivariate methods were cause by changes in rare species and abundance patterns of common species.