Diatom-based water chemistry reconstructions from northern Sweden: a comparison of reconstruction techniques

Abstract

The diatom composition in surface sediments from 119 northern Swedish lakes was studied to examine the relationship with lake-water pH, alkalinity, and colour. Diatom-based predictive models, using weighted-averaging (WA) regression and calibration, partial least squares (PLS) regression and calibration, and weighted-averaging partial least squares (WA-PLS) regression and calibration, were developed for inferences of water chemistry conditions. The non-linear response between the diatom assemblages and pH and alkalinity was best modelled by weighted-averaging methods. The lowest prediction error for pH was obtained using weighted averaging, with or without tolerance downweighting. For alkalinity there was still some information in the residual structure after extracting the first weighted-averaging component, which resulted in a slight improvement of predictions when using a two component WA-PLS model. The best colour predictions were obtained using a two component PLS model. Principal component analysis (PCA) of the prediction errors, with some characteristics of the training set included as passive variables, was performed to compare the results for the different alkalinity predictive models. The results indicate that calibration techniques utilizing more than one component (PLS and WA-PLS) can improve the predictions for lakes with diatom taxa that have broad tolerances. Furthermore, we show that WA-PLS performs best compared with the other techniques for those lakes that have a high relative abundance of the most dominant taxa and a corresponding low sample heterogeneity.

Effects of climate, fire and vegetation development on Holocene changes in total organic carbon concentration in three boreal forest lakes in northern Sweden

“…and the diatom-inferred pH decreases from 6.4-6.6 to 5.5-5.8 and the diatom-inferred color increases and has the highest value during the entire Holocene. Although the maximum value of diatom inferred color corresponds with a period of high TOC values, no clear relationship was observed between these two variables despite previous findings that color can be used as an estimate for TOC concentrations of lake water (Rasmussen et al, 1989) and statistical significant models have been used for both color (Korsman and Birks, 1996) and TOC (Rosén, 2005).…”

“…Diatoms are another paleolimnological tool, which are known to be sensitive indicators to pH (Renberg et al, 1993;Battarbee et al, 1999). Diatom-based models for reconstruction of TOC and color also exist (Korsman and Birks, 1996;Rosén et al, 2000a), but these models may be confounded by the pH signal. Therefore there is need to combine a variety of different proxies to assess carbon cycling based on lake sediments.…”

An expanded calibration model for inferring lakewater and air temperatures from fossil chironomid assemblages in northern Fennoscandia

“…The following criteria were applied to identify unusual samples (e.g., Birks et al, 1990;Korsman and Birks, 1996;Weckström et al, 1997): (i) the sample score fell outside the 95% confidence limits of the sample score means on any of the first four DCA axes; and (ii) the lake had a large (Ͼ 5%) residual distance to the constrained environmental axis in the DCCA using the environmental variable of interest as the sole explanatory variable, respectively. Additional checking of outliers in terms of an unusual combination of environmental variables was carried out using leverage diagnostics (ter Braak, 1990; in redundancy analysis (RDA), as well as principal components analysis (PCA) of the environmental data.…”

A Bayesian hierarchical model for reconstructing relative sea level: from raw data to rates of change

“…Inverse approaches initially gained popularity because of their reduced computational complexity (e.g., Birks, 2010) resulting in quicker processing compared to classical methods. Furthermore, inverse methods often demonstrate equal or superior performance when compared to classical approaches (e.g., Toivonen et al, 2000;ter Braak and Juggins, 1993;Korsman and Birks, 1996). The parameters in transfer functions are estimated using empirical data (a modern training set) from environments likely to be analogous to those encountered in core material (e.g., Juggins and Birks, 2012) and are treated as fixed and known.…”