Liu, Yingying; Jin, Xiaowei
ORCID: https://orcid.org/0000-0002-0478-5306; Fang, Jing; Huang, Lingjie; Hou, Lin; Xie, Huiyu; Luo, Ying; Chen, Miao; Johnson, Andrew C.
ORCID: https://orcid.org/0000-0003-1570-3764; Xu, Jian
ORCID: https://orcid.org/0000-0001-5287-1009.
2026
River diversity outcomes at different trophic levels can be greater than expected when land use and physical and toxic stressors combine.
Environmental Science & Technology, 60 (22).
15641-15655.
10.1021/acs.est.6c02614
Disentangling the impact of toxic mixtures amidst multiple stressors remains a grand challenge in riverine ecology. Using the Yangtze River Basin as a model system, we integrated concentrations of 107 toxic substances with multidimensional biodiversity data across three trophic levels (macroinvertebrates, phytoplankton, and periphyton) from 89 sites sampled in March–April 2023 to quantify chemical impacts along a land-use gradient. We reveal a distinct regime shift in ecological drivers: while communities in low-human-impact zones are associated with climate and nutrients, toxic mixtures emerge as a dominant driver in high-impact zones, explaining up to 25% of macroinvertebrate multidiversity. Crucially, responses diverged across trophic levels: macroinvertebrates suffered taxonomic loss (sensitive species elimination), whereas algae maintained apparent stability through functional reorganization (redundancy compensation). The response patterns suggested interactions included synergistic or antagonistic nonlinear patterns. Threshold Indicator Taxa Analysis identified community-level impacts for pesticides (e.g., carbamates) at surprisingly low ng/L levels, which are orders of magnitude lower than current Predicted No-Effect Concentrations. These findings imply “invisible” ecological risks from micropollutants can be masked by functional compensation. This finding highlights the risks that traditional single-stressor assessments may underestimate ecological harm in complex, human-dominated landscapes.
Altmetric Badge
Dimensions Badge
![]() |
