Environmental quality measurements for the St. Lawrence River are relatively scarce for periods prior to the 1970s, yet long-term data are often required for effective ecosystem management. For example, in the St. Lawrence system, important questions remain, such as how have shifts in macrophyte growth responded to past changes in nutrient loading. In this study, we develop a paleolimnological transfer function to provide information on past aquatic habitat structure. Using diatom algae collected from the three dominant habitats in the river (rocks, macrophytes, filamentous algae), habitat preferences of the diatom taxa were identified. Inference models were then developed to reconstruct past habitat conditions from fossil diatom assemblages using logistic statistical techniques. Logistic equations were developed by performing correspondence analysis (CA) on diatom assemblage data, and fitting logistic regressions to the sample ordinations. The three models appeared to reliably predict habitat types from the field data (observed versus inferred r2 values: rock = 0.85, Cladophora = 0.94, macrophyte = 0.97). These diatom-based transfer functions can now be used on fossil diatoms preserved in dated sediment cores to infer past environmental changes, which can assist the development of rehabilitation efforts in the St. Lawrence River.
Bibliographical noteFunding Information:
Funding for this research was provided by the Tri-Council and by an NSERC grant to J. Smol. The map of sites was provided by J. R. Glew. The manuscript has benefitted from critical reviews by J. Johansen, R. P. Richards, T. J. Cuffney, and colleagues at PEARL. We would also like to thank R. Montgomerie for help with logistic statistics, and J. O'Connell for helping with fieldwork and diatom analysis of the Cladophora samples.
- Correspondence analysis
- Logistic model
- St. Lawrence River