open access

https://hdl.handle.net/2115/38223

Quantifying sediment production in steepland environments

英語

Sediment budgets

2009-03

Hokkaido University Forests, EFRC

Eurasian Journal of Forest Research

12(1)

9-46

Five published contributions to our understanding of the impacts of erosion processes on sustainable land management are reviewed and discussed. These focus on rapid shallow landsliding and gully erosion which are among the most prevalent forms of environmental degradation in New Zealand's hill country. The over-arching goal of this research has been to quantify the on-site (e.g., soil erosion, land productivity) impacts of these processes. Rather than measure erosion rates over long periods of time, geomorphic techniques such as 'space-for-time substitution' and paired catchment approaches have been employed to overcome the naturally high spatial and temporal variability in erosion processes. Digital elevation models (DEMs) have proven invaluable as a means of measuring rates of gully erosion and of quantifying landform properties at small and large catchment scales. The spatial variability in shallow landslide erosion, recovery of soil on scar surfaces, and long-term evolution of hillslopes is investigated in an area of sandstone hill country in the North Island of New Zealand, to help elucidate the role that vegetation plays in maintaining slope stability. Hillslope evolution is primarily by landsliding on steep slopes and by diffuse creep processes on gentle slopes, punctuated by periods of slope instability related to climatic and vegetative variability. Variation in slope form near to channels, with over-steepened sideslopes and higher benches, implies a history of fluctuating erosion rates, driven by changes in stream base level. Systematic variation in soil depth and slope angle measured at the hillslope scale implies spatial variability in erosion rates and a landscape that is not in morphologic equilibrium. There is about an order of magnitude difference in long-term erosion rates between relatively steep (> 30°) and gentle (< 30°) hillslopes. Steep slopes are located at the head of 1st-order drainage basins and are more closely coupled to base level changes in the drainage network and erode at about the rate of tectonic uplift. A novel 'paired hillslope' approach is used to directly measure the net loss of soil caused by post-deforestation landslide erosion and conclusively show that, had the forest remained, the landslides would not have occurred. Over an 85-year period since deforestation, average net soil loss per unit hillslope area was 0.15 ± 0.04 m, equivalent to a depletion rate of 1.8 ± 0.5 mm yr^[-1]. The rate was higher at 2.7 ± 0.8 mm yr^[-1] on slopes above 28° where most landslides are concentrated. Contemporary erosion rates are at least 3 to 10 times, or more, the long-term rates of erosion depending on hillslope location. The distribution and average depths of soils to bedrock are measured and used to derive a logarithmic function of increasing soil depth with landslide scar age. The rate of soil recovery is found to diminish with time: from 3.5 mm yr^[-1] over the first 40 years after slipping to 1.2 mm yr^[-1] over the following 50 years. The functional form of the relationship implies continued decrease in the accumulation rate with time. When averaged at the scale of hillslopes, accumulation rates are much less than erosion rates implying that there are semipermanent losses in soil depth and associated soil properties, something that has important consequences for sustainable land management of this landslide prone hill country. A further study investigates the methodology and errors involved in determining the amount of sediment produced from gully complexes in catchments at the headwaters of the Waipaoa River basin. Multi-date DEMs constructed from historical aerial photography are used to determine the volume and mass of sediment produced from gullies of varying size over timescales of decades. Results show that the average denudation rate of gullies is proportional to the square root of gully area: large gullies are producing a disproportionately large amount of sediment, not only because of their larger area, but also because of their higher denudation rates. This simple power-law relationship provides an efficient and quick means of estimating the total amount of sediment producted from gullies (of similar morphology) at the catchment scale, from the measurement of gully area alone, and this has direct application to catchment scale sediment budgets. Many of the smaller gullies were found to have stabilized, became smaller, and reduced in sediment yield after tree plantations were established, demonstrating that reforestation should be a major conservation strategy wherever gully erosion is encountered. Large gullies produce a disproportionately large amount of sediment and are less affected by reforestation. Measured sediment yields from the two largest gullies when compared to the basin yield, demonstrate that these gullies have neither individually nor collectively dominated the sediment budget of the Waipaoa River. This is because greater contributions are made by numerous contemporary gullies and/or by diverse sources in other parts of the Waipaoa River basin. Research findings presented in this paper provide information that is valuable for the sustainable land management of areas subject to accelerated erosion processes. Being able to identify and map areas of land that are sensitive to disturbance, and evaluate the on-site and off-site impacts, is an important part of assessing soil sustainability in these steep topographic settings.

紀要論文

Version of Record

1345-8221