Ostracod and chironomid transfer functions for inferring past lake water conductivity in neotropical aquatic ecosystems

MASSAFERRO, J., PÉREZ, L., BRENNER, M., SCHARF, B., SCHWALB, A

Konstanz

Congreso; International Limnogeological Congress; 2011

TUI

Chironomids (Insecta: Chironomidae) and ostracodes (Crustacea: Ostracoda) are two of the most abundant, diverse and environmentally sensitive groups of organisms in northern neotropical aquatic ecosystems. This study shows the potential of using different organisms (ostracods, chironomids) as environmental indicators for quantitative reconstructions We identified 66 chironomid morphospecies and 29 non-marine ostracode taxa on the Yucatan Peninsula and in surrounding areas. Lack of limnological and autecological data for these lacustrine organisms, however, had precluded their use in neotropical paleoenvironmental studies. We developed a training set with 63 waterbodies containing environmental data and information on species tolerances and optima. Multivariate canonical analysis indicated that lake water conductivity controls ostracode assemblages and was one of the three significant variables (conduct, TOC and HCO3) explaining chironomid distribution. Next, we developed chironomid- and ostracode-based transfer functions using Weighted Average Partial Least Squares (WA-PLS) regression and a leave-one-out cross validation method to infer past water conductivity in lakes. Conductivity ranged from 127 to 5,960 S cm-1. The chironomid-based transfer function based on a five-component WA-PLS regression provided a coefficient of determination (r2jack) of 0.66 and a root mean square error of prediction (RMSEPjack) of 686.7 S cm-1. The two-component WA-PLS model based on ostracode data displayed better predicting performance (r2jack=0.79, RMSEPjack = 159.18 S cm-1). Our results indicate that both taxonomic groups can provide accurate quantitative estimates of past environmental changes. Transfer functions will be applied to samples from long cores taken on the Yucatán Peninsula and in surrounding areas to infer late Quaternary environmental conditions.