Interrelationships between rural land-use, sediment transport, overland flow and soil depth variability

Abstract

The loss of fertile topsoil over time, due to erosive processes, could impact New Zealand’s economy and be devastating to individual land owners.

Soil degradation can be assessed through the application of landscape models, however applicability at different sites varies due to uniqueness created by limited data collection and availability. Use of unique models can be broadened with the addition of data sets from differing localities, land types, land use and at different scales. Additional data sets can be used to inform model representation and enhance parameterisation.

The aim of this study was to measure fine scale background (day to day) sediment transport rates, soil depth and soil characteristic information for different rural land uses. This would provide a data set for a steep land area where little information was currently available that could potentially be used to inform model representation.

The sediment transport model developed in this study estimated 5t/km²/yr hillslope erosion for the catchment. A total yield was estimated by combining the modeled estimate and an estimate of long term redistributed landslide debris material (estimated using satellite imagery). This gave a final estimate of 114.3t/km2yr for the study year. The sediment model was based on individual event sediment and overland flow captured in Gerlach traps over a year. The model is a lower estimate of erosion when rainfall events not exceeding 47mm occur in a given year. Higher volume events are known to occur with very high rainfall events occurring approximately every 10 years. Stream flow suspended sediment was also analysed to provide an estimate of sediment leaving the catchment, producing an estimate of 97.6t/km2/yr, this is comparable to the hillside sediment transport estimate.

Additionally, a fine scale empirical soil depth predictor was developed with the intent of using that predictor to enhance environmental and hydrological models that have a need to incorporate soil depth at varying scales.

The empirical depth predictor was based on 200 point measurements from locations of varying slope, elevation, soil type and curvature. Over 70% of observed soil depths were estimated within 30cm by the depth predictor, with the remaining points within 75cm of observed values. The depth predictor is limited to areas where soil depths are below 150cm, additionally it is recommended for areas where slope is greater than 15° due to limited sampling below this.

Where data is available, the empirical soil depth predictor has the potential to estimate soil depths at a much finer scale than 15m² due to being heavily influenced by the local slope.

Understanding soil depth and the ability to predict variability at multiple scales is extremely important for agriculture, erosion mitigation, hazard management and hydrological modelling.