Skip to content

Soil carbons under pressure

Well-managed perennial pastures are thought to represent the maximum opportunity for agricultural soils to accumulate C; but what represents a well-managed pasture? Over the past 6 years, extensive field surveys and monitoring have been trying to identify land management practices to increase soil carbon under perennial pastures in the Monaro region of southern NSW. While there are factors that make this region unique, the key messages that are emerging from this work are applicable to pastures in the tableland regions of NSW, as well as mixed farming systems. The management practices investigated in the Monaro region include: liming, nutrient management, introduced perennial pastures, pasture age and occasional minimum disturbance cropping.

Firstly, back to basics; to increase soil carbon, organic matter supply (biomass grown in situ or grown elsewhere and brought to the site, e.g. compost) needs to be greater than the loss of organic matter through decomposition and erosion. Therefore, from a farmer's point of view, pasture management that increases herbage mass for livestock may also increase soil carbon.

In the Monaro region, as is generally the case, soil carbon is influenced by landscape attributes, including: soil type (particularly, soil depth and clay content), aspect and topography. Depending on the extent of these controlling factors within a paddock, carbon variation across a paddock may be greater than the expected short term, detectable change in soil carbon. Additionally, our research has shown that variability is greatest at sites with high soil carbon contents and particularly in pastures where plant growth can cause large variations in soil carbon.

So what can we influence to increase soil carbon?

  1. Address critical nutrient levels. There is a strong positive correlation between soil carbon and total nitrogen, available phosphorus and extractable sulfur. This works in two ways, firstly, organic matter supply; GrassGro modelled outputs for these pasture based systems suggest that addressing critical soil nutrient levels (phosphorus and sulfur) can considerably increase herbage mass production. Secondly, organic matter stabilisation; the formation of stable forms of carbon (for example, humus) requires nutrients such as nitrogen, phosphorus and sulfur.
  2. Address soil limitations to plant growth. If soil pH is limiting plant growth, then liming may increase biomass production, and therefore organic matter supply to soil.
  3. The next obvious question, what is the effect of grazing? The relationship between soil carbon and grazing management is complex, however, if grazing management can achieve increased biomass production, return of senescent vegetation to the soil, increased root turnover and increased groundcover, then soil carbon may increase.

Whether or not any of these changes in soil carbon are detected as significant is context specific due to the influence of a large background variation in carbon content under pastures.

Lastly, what about trading soil carbon? Economically, there are production gains associated with increasing soil fertility, but only when the extra pasture produced is utilised. This means increasing stocking rates, and with that, livestock emissions. Once these calculations are considered, it may be that on farm decisions to increase pasture productivity will be made principally to increase livestock production and any value coming from soil carbon will be largely related to the important soil health benefits, particularly given the reported variability of soil carbon in these pasture based systems.

Table 1 provides some examples of the influence of land management and soil type on soil carbon. Importantly, this table highlights the natural variation in soil carbon within the monitoring plots as well as natural variability that occurs over time

These findings come from NSW Department of Primary Industries research, as well as Australian Government the Action on the Ground funding awarded to Monaro Farming Systems.

Further information regarding the field research presented can be found as a free ibook titled 'Soil Carbon in the Monaro Region – A report from Action on the Ground'.


Table 1. Carbon stocks (tonnes C/ha/ 0-30cm) from the Monaro Farming Systems Action on the Ground project.
Key: Δ indicates the 'difference between' one management factor and another.

 

ManagementCarbon stock tonnes C/ha/0 - 30cm
20122014Difference
Basalt derived soil   
Crop46.3950.123.74
Native pasture59.9864.944.97
Difference - crop/native pasture13.5914.821.23
Low P61.3877.115.72
High P78.8273.48-5.34
Difference - Low/High17.44-3.62-21.06
Low grade metamorphics   
North aspect - intro perennial pasture46.9749.382.41
East aspect - intro perennial pasture42.3748.095.72
Difference North/East aspect4.61.29-3.31
Crop47.6545.36-2.29
Native45.1555.9110.76
New intro perennial pasture68.9669.910.95
Difference - Native pasture/new intro pasture-23.81-13.999.81
Difference - native pasture/crop-2.510.5513.05
Granite derived soil   
Unlimed perenial pasture45.8752.296.42
Lime perennial pasture47.347.430.13
Difference limed/unlimed1.43-4.86-6.29
Crop40.2153.1912.98
Native pasture39.9446.86.86
New intro perennial pasture36.6642.796.13
Difference - native pasture/new intro pasture3.284.010.73
Difference - native pasture/crop-0.27-6.39-6.12
Old intro perennial pasture46.6644.99-1.67
New intro perennial pasture50.840.79-10.02
Difference old intro pasture/new intro pasture-4.154.28.35

Further information:

Susan Orgill
0428 424 566
susan.orgill@dpi.nsw.gov.au