Work

Modern surface sediment diatom assemblages and conductivity modeling in northern China

Year: 2025

Type: article

Abstract: • Established a robust diatom-based transfer function across 56 lakes in northern China. • Identified electrical conductivity as the principal ecological filter shaping diatom communities. • Identified regional diatom diversity patterns and ecological thresholds to salinity across climatic zones. • Offers a robust tool for reconstructing past salinity dynamics in the lakes of Northern China. Despite intensifying climate and anthropogenic pressures on freshwater ecosystems, northern China remains without a regionally calibrated diatom-based electrical conductivity (EC) model hindering efforts to quantify ecological responses across its hydroclimatically diverse lake systems. We analyzed diatom assemblages from 86 lakes across northeast, north, and northwest (NW) of China to establish the first regionally calibrated diatom-based EC inference model spanning this broad hydroclimatic gradient. Diatom composition varied significantly among regions, but the most widespread taxa Cyclotella meneghiniana , Stephanodiscus minutulus , and Aulacoseira granulata occurred consistently across the full study domain. Richness and Shannon diversity peaked under intermediate EC conditions (0.26–2.95 mS/cm) and declined sharply at high EC (>20 mS/cm), particularly in the arid NW, reflecting ecological thresholds to salinity stress. Canonical correspondence analysis (CCA) identified EC and water depth as primary environmental drivers, and a weighted averaging partial least squares (WA-PLS) model achieved strong predictive performance ( R j a c k 2 = 0.74; RMSEP = 0.14). The modern calibration set offers a new framework for assessing ecological sensitivity across hydroclimatically diverse and data-scarce continental interior lake systems, incorporating paleosalinity reconstructions to advance paleoecological research and climate change evaluations.

Source: Ecological Indicators

DOI: 10.1016/j.ecolind.2025.114229