Assessing lake eutrophication using chironomids: understanding the nature of community response in different lake types

1. Total phosphorus (TP) and chlorophyll a (Chl a) chironomid inference models (Brodersen & Lindegaard, 1999; Brooks, Bennion & Birks, 2001) were used in an attempt to reconstruct changes in nutrients from three very different lake types. Both training sets were expanded, particularly at the low end of the nutrient gradient, using contemporary chironomid assemblages and environmental parameters from 12 British lakes, although this had little improvement on the model performances. 2. Dissimilarity analyses showed that the historic chironomid assemblages did not have good analogues in the original calibration or extended datasets. However, since the transfer functions are based on weighted averages of the trophic optima for the taxa present and not on community similarities, reasonable downcore inferences were produced. Ordination analyses also showed that the lakes retain their 'identity' over time, as the sample dissimilarities within lakes were less than the dissimilarities between lakes. 3. Analysis of the three historic lake profiles showed a range of chironomid community responses to lake development. Chironomids from a shallow lake, Slapton Ley, responded indirectly to nutrient enrichment (TP), probably through altered substrate, macrophyte and fish conditions, rather than directly to primary productivity (Chl a). A stratified lake, Old Mill Reservoir, showed a loss of the profundal chironomid fauna due to increasing primary productivity (Chl a) coupled with increasing hypoxia. A response to nutrients (TP or total nitrogen (TN)) at this site is also indirect, and the TP reconstruction therefore cannot be reliably interpreted. The third lake, March Ghyll Reservoir has little change in historic chironomid communities, suggesting that this well mixed, relatively unproductive lake has changed less than the other lakes. 4. Using chironomids to reconstruct nutrient histories does not follow a simple scheme. The response to changes in nutrients may be direct, but mediated through other ecosystem components. As alternative stable states are possible at a given level of TP it is also likely that alternative chironomid communities exist under similar nutrient conditions. Changes in biological communities can thus occur over thresholds, and it is only biological proxies that can reflect such ecosystem switches within palaeoenvironmental investigations