The effect of land uses and rainfall regimes on runoff and soil erosion in the semi-arid loess hilly area, China

Wei Wei, Liding Chen, Bojie Fu, Zhilin Huang, Dongping Wu, Lida GuiJournal of Hydrology 2007 · DOI

TL;DR

This study investigates how different land uses and rainfall patterns interact to influence runoff and soil erosion in China's semi-arid loess hilly region. By analyzing 14 years of field data, the researchers identified three distinct rainfall regimes and found that high-intensity rainfall events, coupled with certain land uses like cropland, pose the greatest erosion risk. The findings highlight the importance of considering both rainfall characteristics and land management strategies for effective soil conservation.

What Problem Does This Paper Address?

Soil erosion is a significant environmental problem, particularly in semi-arid regions like China's Loess Plateau, where it leads to land degradation and impacts agricultural productivity. While land use and rainfall are known to be major drivers of erosion, their complex interactions, especially under varying rainfall patterns, are not fully understood. Previous research often relied on simulated rainfall, which may not accurately reflect natural conditions. This paper aims to bridge this gap by analyzing long-term, real-world field measurements to understand how different land cover types respond to distinct rainfall regimes, providing crucial insights for land management and conservation strategies in this vulnerable region.

Methods

The study employed a combination of field measurements and statistical analysis. Over 14 years (1986-1999), researchers collected data on rainfall, surface runoff, and soil loss from experimental plots representing five different land use types (cropland, pastureland, shrubland, woodland, and grassland) situated on varying slopes.

A key methodological contribution is the classification of rainfall events into distinct "rainfall regimes" using K-means clustering. This approach, applied to rainfall depth, duration, and maximum 30-minute intensity, allowed for a more nuanced understanding of rainfall's erosive potential compared to treating all rainfall events uniformly.

Runoff was quantified using the runoff coefficient (ratio of surface runoff to rainfall), and soil erosion was measured by the erosion modulus. Statistical analysis, including ANOVA, was used to determine the significance of observed differences.

Datasets

The study utilized a dataset collected over 14 years (1986-1999) from 15 experimental plots in the Anjiapo Catchment, Dingxi, Gansu province, China. This dataset included:

  • Rainfall Data: Depth, duration, and maximum 30-minute intensity for 131 rainfall events that produced runoff.
  • Runoff Data: Surface runoff volume for each rainfall event.
  • Soil Erosion Data: Sediment loss for each rainfall event.
  • Land Use Data: Information on five land use types (cropland, pastureland, shrubland, woodland, grassland) and their vegetation coverage over the years.

No new benchmark datasets were introduced; the study relies on the long-term field measurements collected by the research team.

Key Insights and Ideas

  • Rainfall Regimes Matter: Not all rainfall is equal in its erosive power. The study successfully categorized rainfall into three regimes, revealing that high-intensity, short-duration, and frequent events (Regime II) are significantly more erosive than others.
  • Land Use Hierarchy for Erosion Control: There's a clear hierarchy in how well different land uses protect against erosion. Shrubland emerged as the most effective, followed by grassland and woodland, while cropland and pastureland were the least effective.
  • Alfalfa's Double-Edged Sword: While pastureland (specifically alfalfa) is often promoted, this study found it to be less effective for soil conservation than shrubland, grassland, or woodland. This is attributed to its harvesting cycle and water absorption, which can degrade soil.
  • Plant Succession is Crucial: The effectiveness of land use in controlling erosion changes over time. Newly established vegetation (saplings) is more vulnerable to erosion, but as plants mature and vegetation cover increases, erosion rates significantly decrease, especially under high-intensity rainfall.
  • Interaction is Complex: The relationship between land use and erosion is not static; it's dynamic and influenced by the specific rainfall regime and the stage of vegetation development.

Key Results

  • Erosion Ranking by Land Use: The average runoff coefficient and erosion modulus followed this order: cropland > pastureland > woodland > grassland > shrubland.
  • Erosion Ranking by Rainfall Regime: Rainfall Regime II (high intensity, short duration, high frequency) caused the most runoff and soil loss, followed by Regime I (medium intensity/duration) and then Regime III (low intensity, long duration).
  • Shrubland's Superiority: Shrubland exhibited the lowest erosion modulus (131 t km⁻² a⁻¹), often considered negligible. Cropland, in contrast, had a very high erosion modulus (8599 t km⁻² a⁻¹), exceeding the threshold for "highly intense" erosion.
  • Vegetation Recovery: Under Rainfall Regime II, shrubland and woodland showed a significant decrease in runoff and erosion after the initial 3-4 years of plantation, stabilizing at lower levels as vegetation matured.

Limitations and Future Work

  • Alfalfa's Water Use: The paper suggests further research into alfalfa's water exhaustion and its impact on erosion control, as this was identified as a reason for its poor performance.
  • Plant Succession Stages: The study highlights the vulnerability of initial sapling stages. Future work could focus on effective interim measures (e.g., straw mulch, plastic film) during these critical periods.
  • Generalizability: While the study provides valuable insights for the semi-arid loess hilly area, the specific rainfall regimes and land use responses might vary in other regions. Further studies could explore the transferability of these findings.
  • Seasonal Distribution: The authors recommend more attention to the seasonal distribution of the most erosive rainfall types to inform more targeted conservation efforts.

Summary updated

BibTeX
@article{wei2007effect,
  author = {Wei Wei and Liding Chen and Bojie Fu and Zhilin Huang and Dongping Wu and Lida Gui},
  title = {{The effect of land uses and rainfall regimes on runoff and soil erosion in the semi-arid loess hilly area, China}},
  journal = {Journal of Hydrology},
  year = {2007},
  doi = {10.1016/j.jhydrol.2006.11.016},
}
Figure