Scholarlink

Purpose <p>Littoral zones (aquatic-terrestrial ecotones, ATEs) of lakes are critical interfaces for global nutrient cycling and ecosystem stability, yet they are highly vulnerable under climate change and intensified land use. Imbalances in soil nitrogen (NS) and phosphorus (PS) storage can exacerbate eutrophication risks, but the underlying mechanisms remain inadequately understood.</p> Methods <p>An integrated approach combining machine learning and entropy-TOPSIS modeling was applied to analyze the distribution patterns and driving factors of NS and PS in the soils of the ATE of a plateau lake in southeastern Tibet.</p> Results <p>The results revealed a pronounced spatial decoupling between NS and PS across the ATE. NS showed significant lake-ward accumulation up to 1.532&#xa0;kg/m², whereas PS remained relatively stable. This led to an elevated N: P ratio, indicating a potential nitrogen enrichment relative to phosphorus. This pattern was also observed in cropland and grassland. Machine learning identified air temperature as the primary driver of this spatial decoupling, with its effect amplified by land use–climate interactions. Entropy-TOPSIS evaluation further indicated that natural wetlands effectively regulate soil N-P balance and reduce nutrient stocks through vegetation-mediated processes.</p> Conclusions <p>Based on these findings, we propose a land-optimization strategy involving an artificial wetland–grassland mosaic to help balance nitrogen and phosphorus dynamics, providing a low-cost and sustainable approach to manage eutrophication in plateau lakes. The integrated framework offers a transferable methodology for diagnosing nutrient dynamics and guiding conservation efforts in global lake ecosystems.</p> Graphical Abstract <p></p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Machine learning and entropy-TOPSIS model reveal soil nitrogen-phosphorus dynamics in plateau lake littoral zones: Implications for land-use optimization

  • 期刊论文

  • 发表日期：

    2026年7月8日

查看全文
  • Yunxia Ma,
  • Jiqing Yin,
  • Jing Xiong,
  • Pengfei Yin,
  • Yuxin Jia,
  • Wenxiang Zhang,
  • Taohui Li

摘要

Purpose

Littoral zones (aquatic-terrestrial ecotones, ATEs) of lakes are critical interfaces for global nutrient cycling and ecosystem stability, yet they are highly vulnerable under climate change and intensified land use. Imbalances in soil nitrogen (NS) and phosphorus (PS) storage can exacerbate eutrophication risks, but the underlying mechanisms remain inadequately understood.

Methods

An integrated approach combining machine learning and entropy-TOPSIS modeling was applied to analyze the distribution patterns and driving factors of NS and PS in the soils of the ATE of a plateau lake in southeastern Tibet.

Results

The results revealed a pronounced spatial decoupling between NS and PS across the ATE. NS showed significant lake-ward accumulation up to 1.532 kg/m², whereas PS remained relatively stable. This led to an elevated N: P ratio, indicating a potential nitrogen enrichment relative to phosphorus. This pattern was also observed in cropland and grassland. Machine learning identified air temperature as the primary driver of this spatial decoupling, with its effect amplified by land use–climate interactions. Entropy-TOPSIS evaluation further indicated that natural wetlands effectively regulate soil N-P balance and reduce nutrient stocks through vegetation-mediated processes.

Conclusions

Based on these findings, we propose a land-optimization strategy involving an artificial wetland–grassland mosaic to help balance nitrogen and phosphorus dynamics, providing a low-cost and sustainable approach to manage eutrophication in plateau lakes. The integrated framework offers a transferable methodology for diagnosing nutrient dynamics and guiding conservation efforts in global lake ecosystems.

Graphical Abstract