CSIRO report on sampling soil organic carbon released

8 November 2013
News article

The Australian Government has partnered with the Grains Research and Development Corporation and commissioned the CSIRO to develop a statistically efficient approach for land managers to sample soil carbon at the paddock scale.

Working with the CSIRO, the Department of the Environment is using the findings of this report to develop both a user-friendly method for designing project‑scale soil sampling programs and also standards for analysing soil carbon.

The resulting report ‘Sampling soil organic carbon to detect change over time’ sets out a soil carbon sampling strategy that will enable farmers and land managers to sample soil carbon as efficiently as possible, maximising accuracy while minimising costs.

A key feature of CSIRO’s innovative approach is to divide the relevant carbon estimation area into even sized sub-plots (strata) to ensure even sampling coverage and reduce bias. Random samples are then taken and combined into composites in order to minimise laboratory costs. This process is then replicated in subsequent sampling rounds to determine change in soil carbon levels over time.

The Report makes a range of assumptions to estimate potential sampling costs and measurement accuracy based on Australian and international data.

  • Costs for two rounds of sampling and analysis are likely to fall within the range of $2-3 per tonne of carbon dioxide equivalent.
  • It is anticipated that the costs of soil sampling and analysis will continue to decline as new technologies, such as magnetic infrared spectroscopy, are proven to be effective.
  • Regardless in some situations, particularly where soil carbon levels vary greatly cross a given area, it may not currently be cost effective to sample intensively enough to detect statistically significant changes in soil carbon.

The Report sets out a strategy for soil carbon sampling that maximises accuracy while minimising sampling costs.  To maximise applicability, this approach accommodates proponents who have no initial data on soil carbon distribution across their land which could be used to better target sampling efforts.