Effect of erosion on soil carbon stocks
Since the development of the Soil Carbon Monitoring System (Soil CMS) to predict soil carbon stocks, it has been recognised that mass movement erosion potentially has a major influence on these stocks.
The Ministry for the Environment has recently funded a pilot project to collect field data to address this issue. Soils were sampled at a series of plots with extensive landslide and gully erosion (Figure 1), and the extent of changes in erosion was mapped to determine the effect of erosion on soil carbon over the last 40 years.
Landsliding had a significant effect on soil properties and soil carbon stocks; since the late 1930s it had affected about 50% of a hillslope area at two study sites – Te Whanga in the Wairarapa (Figure 1) and Tutira in Hawke’s Bay hill country. Soil carbon stocks (0–30 cm) were highest in uneroded soils (about 100 t/ha) and averaged 60–65 t/ha on eroded sites. There were clear differences in soil carbon stocks between landslide scars and debris tails, and a trend for soil carbon to increase as landslide age increased. After about 70 years, soil carbon stocks were still well below the value measured for uneroded plots (by about 40% for scars and 20–30% for debris tails), indicating the effect of erosion is very persistent. When the effect of landslide erosion is incorporated into the calculation of soil carbon stocks both sites had about 90 t/ha (Table 1), compared with an estimate of 104 t/ha from the Soil CMS model.
Table 1. Comparison of measured soil C stocks with Soil CMS model prediction
|Erosion type||Site||Measured soil C stocks (t C ha–1)|
|Average uneroded plots||Average eroded plots||Net stocks corrected for erosion||Soil CMS model (t C/ha)|
|Gully||Te Weraroa forest||65||19||57||98|
By contrast, gully erosion had a minor influence on soil carbon stocks. While active gullies and reforested gullies have much lower soil carbon stocks (<30 t/ha), they are minor components of the landscape (generally <5% of the area) and therefore have little influence at whole-landscape scale.
At the study sites there was little change in estimated soil carbon stocks between 1990 (the baseline year for Kyoto Protocol reporting) and 2008 because most of the landsliding had occurred before 1990. However, of the total soil carbon stocks at Te Whanga, 41% are stored on 1977 landslides, while at Tutira, 52% of the total stocks are stored on 1988 landslides, implying major storm events in affected areas have the potential to reduce soil carbon stocks significantly. The results indicate that the Soil CMS model overestimates soil carbon stocks for steep eroding hill country where landsliding is extensive. At Te Whanga the net reduction in soil carbon stocks due to landsliding was about 13% and at Tutira it was about 21%.