A comparison of relative abundance versus class data in diatom-based quantitative reconstructions

Area of Science:

• Paleolimnology
• Diatom Ecology
• Water Quality Assessment

Background:

• Diatom relative abundances are standard in paleolimnological studies.
• Plant ecology often uses faster ordinal scale data (class data) for species abundance.
• This study investigates the efficacy of ordinal diatom data in paleolimnological reconstructions.

Purpose of the Study:

• To compare the performance of models using ordinal diatom species class data versus relative abundance data.
• To assess the reliability of reconstructing past total phosphorus (TP) concentrations using both data types.
• To evaluate the cost-effectiveness and time-efficiency of using ordinal class data.

Main Methods:

• Converted relative diatom abundances into ordinal classes (sporadic to dominant).
• Modeled species response to TP using both relative abundance and ordinal class data.
• Reconstructed TP for six sediment cores and validated with an independent test set of recounted samples.

Main Results:

• No significant differences were found between relative abundance and class data in explained species variance.
• TP reconstructions using ordinal class data showed no significant difference compared to relative abundance data.
• Inferred TP values from recounted samples also showed no significant differences between data types.

Conclusions:

• Ordinal diatom class data provides a reliable and cost-effective alternative to relative abundance data for TP reconstruction.
• This method can yield more long-term water quality records more quickly.
• Findings support the use of ordinal data in automated diatom identification and may reduce the need for identifying every valve.