- Current intensive agricultural practices are unsustainable.
- Regenerative farming practices, however, can sequester atmospheric carbon into the soil, reduce soil loss and build new soil stocks.
- Measures must be enacted to bring down the cost of measuring soil organic carbon using satellite observation systems.
- Doing so could improve agricultural productivity, profitability and sustainability and facilitate increased volumes of atmospheric carbon to be sequestered in soil.
This paper is the first in a two-part series that examines the role soil organic carbon (SOC) plays in maintaining healthy soils that support productive, profitable and sustainable agriculture.
The first question: Are our agricultural practices productive, profitable and sustainable?
There are strong differences of opinion here, both among farmers and researchers. Some point to the ongoing contribution that agriculture makes to the Australian economy and Australia’s leading position based on international productivity comparisons. Others suggest that modern intensive agricultural practices (with an over reliance on chemical inputs and fungicides/pesticides) are destroying the very fabric of the soil and that the loss of SOC and associated soil biology must be addressed. This is not a recent argument but is a long-standing debate between intensive agriculture and so-called “regenerative or conservation farming” techniques. A recent piece in The Australian Financial Review gives a good summary of the debate.
My conclusion is that current intensive practices are not sustainable and will lead to losses in productivity and profitability. Moving towards regenerative farming can stem this undesirable trend with an increase in productivity, profitability and sustainability in the long term.
The second question: Are changed agricultural practices capable of sequestering large amounts of carbon from the atmosphere?
Again, there are strong differences of opinion. Answers range from relatively minor to quite massive negative emissions. Debate centres on several points, including the inherent difficulty and costs of obtaining reliable measurements of SOC, the longevity of different carbon compounds and just how much could be sequestered before further uptake was minimal.
My conclusion is that moving towards regenerative farming practices can sequester very large amounts of atmospheric carbon into the soil, can reduce soil loss and even build new soil stocks.
Soil organic carbon is important
Plant carbon comes from photosynthesis, not directly from the soil, unlike nitrogen which comes from the soil in the form of nitrates. Microbes in the soil thrive in higher levels of SOC and can convert organic nitrogen into available nitrate. Farmers in the past have focused more on Nitrogen-Phosphorous-Potassium (NPK) and Calcium, Sulfur, Zinc (et al) but not Carbon (C) (or the associated soil biology) as this is not taken up. But SOC feeds the microbiome. Nitrates can also be available directly from inorganic fertilizer sources without any intervention of soil microbes.
Conventional soil chemistry, eg the ability to hold K+ and Ca++, misses the microbial action such as the mycorrhizal fungi (MRF) that feed on decaying organic matter and provide root extensions. Without adequate C in soil, the MRF cannot function. There are many interactions, eg quorum sensing where on reaching a critical mass of bacteria present, growth stimulants are produced. The plant then produces even more of the extrudates on which the bacteria feed. The case is made that soil life (and not just microbial) is the main determinant of soil health. Thus the paradox, that while SOC is not a plant nutrient, yields and soil health depend on SOC and the interdependent soil biology.
The recent Decadal Plan for agriculture by the Australian Academy of Science suggests that:
soils are the most complicated biomaterial on the planet. It is not surprising that, in contrast to the huge amounts of information available with regard to above-ground plant performance, knowledge concerning the physical and biological soil–plant interface is still very patchy.
In the past, the study of soils (and their management) was largely a physical and chemical one. Now the importance of biology is recognised. Soils and their interaction with plants need to be recognised and approached as an ecological science. As such, it is now more complex; hence, the seemingly variable and sometimes contradictory findings. With the right scientific approaches, applied in the context of Australia’s unique (agro)ecosystems, however, we can cut through this complexity to understand how best to manage the diversity of soils (and their ecosystems) for long-term productivity, profitability and sustainability, an outcome which is undeniably dependent on increasing and maintaining SOC levels.
Without maintaining SOC, our soils will lose the ability to sustain crop yields, and the massive R&D invested in genetic et al improvement in agriculture will be wasted if Australian farmers are unable to grow the new varieties and breeds to their full potential.
On soils, and plant soil interaction, we are rather ignorant
The Australian Academy of Science goes on to say:
Major questions remain around many issues including levels of redundancy among soil organisms; functional links between below-ground processes and plant performance; second-order interactions with animals; the ways in which micro-organisms affect the availability of nutrients; the role soils play in sequestering carbon; and the contribution of soil biology to agro-ecosystem sustainability.
This is all salutary in terms of not being able to quantify the benefits of increasing SOC with certainty. It also helps explain why opinions are so divergent. Nonetheless, the United Nations Food and Agriculture Organization is clear cut in its message:
Increases in SOC and biodiversity are generally beneficial for crop production, and decreases in both are equally deleterious for crops; however, providing evidence for these qualitative statements and establishing predictive relationships has been difficult because crop growth is dependent on a suite of interacting factors.
Questions on measurement
Arguments to defend different practices need to be informed by reliable data. There are challenges with SOC: there is much variability, both spatially (in three dimensions) and temporally. One answer to this problem would be to simply increase the number of samples taken until the variability was reduced to an acceptable level. The challenge here is that sampling and measurement are expensive. Furthermore, there is argument as to whether the statistics of normal distributions is relevant to soils. In any case, farmers can generally not afford to take the large number of samples needed to give highly accurate estimates of SOC in particular fields. They can and do, however, take systematic measurements from closely controlled specific locations and return to these locations each year. Providing the actual sampling and measurement are technically reliable, and they generally are, such “non-spatially representative” sampling is, however, highly indicative of temporal trends.
Let us for now consider the country-wide soil and landscape grid as a starting point, but it is hard to pick significant trends.
Trends, however, are perhaps better seen in direct data from specific regions. This is the basis of the five-yearly State of the Environment report, which assesses the state and trends of soil carbon for 39 areas.
The data suggests that we have a long way to go, although the overview in the report gives words of comfort: ‘Land management practices are improving, particularly in relation to soil management and soil conservation measures. Pesticide and nutrient run-off is also being significantly reduced in some industries, although increasing in others.’
Consider the areas where there is sampling. The summary comments are telling:
State and trends of soil carbon – 2016
|1 Paroo Plain and Warwick Lowland||Rangelands with extensive grazing and wind-borne soil erosion, particularly on sandplains. In western New South Wales, 74% of soil monitoring units report SOC reduction as an issue||Deteriorating|
|2 Warrego Plains||Rangelands with minor opportunity cropping. In western NSW, 74% of soil monitoring units report SOC reduction as an issue||Unclear|
|3 Tenterfield Plateau||Grazing of modified and natural pastures, and nature conservation are major land uses. In the Central Plateau of NSW, 33% of soil monitoring units report SOC reduction as an issue||Deteriorating|
|5 Cobar Plains||Historically poor management has depleted SOC. Overgrazing by feral goats is causing further decline, despite improving land management. In western New South Wales, 74% of soil monitoring units report SOC reduction as an issue||Deteriorating|
|6 Barrier Ranges||Surface SOC is low as a result of grazing and prior clearing. In the Central Tablelands region of New South Wales, 40% of soil monitoring units report SOC reduction as an issue||Deteriorating|
|7 Gunnedah Lowlands||Declining trend due to intensification of cropping. In the north-west region of New South Wales, 18% of soil monitoring units report SOC reduction as an issue||Unclear|
|8 Macleay Barrington Fall||Area used for nature conservation and production forestry, with some grazing. Possible decline in SOC due to logging. In the Hunter region of New South Wales, 50% of soil monitoring units report SOC reduction as an issue||Unclear|
|9 Merriwa Plateau||Mixed farming on naturally fertile Ferrosols and Vertosols. In the Hunter region of New South Wales, 50% of soil monitoring units report SOC reduction as an issue||Unclear|
|10 Condobolin Plains||Soils are Sodosols and Vertosols, used for cropping and grazing. In the Central West region of New South Wales, 19% of soil monitoring units report SOC reduction as an issue||Unclear|
|11 Bathurst Tablelands||Grazing of modified and natural pastures dominates. In the Central Tablelands region of New South Wales, 40% of soil monitoring units report SOC reduction as an issue||Unclear|
|12 Hawkesbury Shoalhaven Plateaus||Diverse landscape with natural conservation, forestry, grazing, horticulture and urban land uses. Fire regime and land management practices are most likely causing a decline in SOC. In the Greater Sydney region of New South Wales, 57% of soil monitoring units report SOC reduction as an issue||Deteriorating|
|13 Cumberland Lowland||Mostly urban and industrial land use. In the greater Sydney region of New South Wales, 57% of soil monitoring units report SOC reduction as an issue||Unclear|
|14 Werriwa Tablelands||In the South East region of New South Wales, 59% of soil monitoring units report SOC reduction as an issue. No data from the Australian Capital Territory||Deteriorating|
|15 Monaro Fall||Good levels of SOC under nature conservation, forestry, and grazing. Land management is improving||Unclear|
|16 Australian Alps||Mostly used for nature conservation. Controls on grazing and reduced erosion stabilised early losses, but the increased intensity and extent of fires are likely to be causing a decrease, particularly in Organosols||Unclear|
|17 Mallee Dunefield||Cropping, grazing and nature conservation with irrigated agriculture along the Murray River. Improved farming practices have improved soil condition in some areas||Unclear|
|18 Wimmera Plain||Mainly cropping and grazing. Former grazing lands now used for nature conservation may still be experiencing declining carbon content. Changing farming practices to no-till may be increasing SOC content in some areas, especially on heavier soils||Unclear|
|19 Riverine Plains||Dryland cropping and irrigated agriculture, with grazing in the west. None of the soil monitoring units in the Riverina region of New South Wales reported SOC as an issue||Unclear|
|20 West Victorian Plains||Grazing, cropping and expanding plantation forestry. Areas converted from pasture to cropland are probably declining, as are soils used for continuous cropping||Unclear|
|21 Midlands Plain||Dryland cropping, grazing and increasing irrigated cropping. Intensification of cropping is probably causing a decline in SOC||Unclear|
|22 Lakes Plateau||Nature conservation reserves. Where wildfire and grazing have resulted in sheet erosion over large areas, SOC has been lost, with limited potential for recovery||Unclear|
|23 West Tasmanian Ridges||More frequent and/or hotter fires in conservation reserves are causing losses, especially in Organosols. Production forestry in the north suggests little potential for increase in SOC sequestration||Unclear|
|24 East Tasmanian Hills||Production and plantation forestry, with minor decline due to erosion. Irrigated cropping in the south-east and north-east is causing a decline in SOC||Unclear|
|25 North West Ramp||Decline in SOC is associated with irrigated cropping||Unclear|
|26 Roe and Carlisle Plains, Coonana–Ragged and Bunda Plateaus||Mainly grazing of native vegetation. Shift from perennials to annuals and possible increase in fire frequency may lead to decline in SOC||Unclear|
|27 Southern Goldfields Plateau||SOC decline is restricted to pastoral areas||Unclear|
|28 Swan Plain||Urban areas and intensive agriculture. High levels of SOC are often associated with irrigated pasture. Decline in SOC is likely under intensive horticultural systems||Unclear|
|29 Woodramung Hills||Low input cropping and grazing. Drying trends have compounded effects of clearing and cropping on SOC loss||Unclear|
|30 Murchison Plateau, Leemans and Yaringa Sandplains, Carnegie and Glengarry Hills, Augustus Ranges||Areas with extensive grazing of native vegetation, with declines in SOC in more heavily grazed areas. Few data in driest areas||Unclear|
|31 Carnarvon Plain||Nature conservation, extensive grazing; small areas of intensive irrigated horticulture likely to have a decline in SOC||Unclear|
|32 Fitzroy Plains||Extensive grazing of native vegetation||Unclear|
|33 Daly Basin||Small areas of intensive agriculture are likely to have declining SOC. Remainder is used for extensive grazing||Unclear|
|34 Whirlwind Plain and Birrundudu Plain||Extensive grazing, with small areas of more intensive development on better soils. Possible minor decreases in SOC due to high seasonal stocking rates||Unclear|
|35 Barkly Tablelands||Extensive grazing on clay plains, with decline in SOC likely||Unclear|
|36 Toowoomba Plateau||Ferrosols used for cropping and pasture, with increasing agroforestry. Slow recovery after large historical loss of SOC||Unclear|
|37 Central Uplands||Partially cleared grazing country. SOC is likely to be declining in recently cleared areas; otherwise stable||Unclear|
|38 Atherton Tableland||Fertile land with high rainfall. Diverse land uses, with SOC now recovering under pastures and tree crops; it is likely still decreasing under small-grain and horticultural crops||Unclear|
|39 Garnet Uplands||Recently intensified land use after clearing; therefore, SOC is likely declining||Unclear|
This is hardly a picture of long-term sustainable agricultural practices.
It is consistent, however, with recent data for Australian wheat crops with yields post 2000 being highly variable.
On the measurement of SOC
Building on earlier reviews by Sanderman et al, CSIRO led a collaborative project to produce a soil carbon atlas of Australia. The Soil Carbon Research Program (SCaRP) cost $23m and is ‘the first nationally coordinated program of soil carbon research [which] has gathered a wealth of information on soil carbon stocks that will underpin Australia’s greenhouse gas accounting, carbon farming and sustainable agriculture.’
There is little doubt that at an individual sample level, the processes used to produce the SOC results are technically correct and precise. That said, however, the fundamental approach of the sampling methodology, and thus the veracity of the SCaRP data, was suspect due to the extreme sampling variability. In fact, where a population does not fit a normal curve but exhibits extreme variability due to other causal factors, like in a soil, it may not be mathematically, and certainly not scientifically, valid to assume that it complies with a normal curve. It is, therefore, difficult to validly sample this population to get a mean value and even more difficult to obtain a time-based increment based on sequential mean values. While the spatial projections from the SCaRP can be queried, any temporal conclusions, including the potential to predict the capacity of changed practices to deliver SOC improvements, is suspect.
One can argue that the only method capable of producing veracious results for whole field temporal changes in SOC is a method that involves whole field measurement. This is now possible using satellite measurement.
On the nature of SOC
The nature of SOC and its permanence is contentious, which is strange as there was widespread agreement on the efficacy of growing forests as a way of sequestering carbon from the atmosphere, yet the questions of the different form and permanence of carbon are similar. The amount of carbon in the soil is three times greater than that in foliage above the ground and two times larger than the amount in the atmosphere.
What is clear is that the nature of the SOC varies with depth and time. This has led some people to argue that only the deeper, long-lived humates should be considered in terms of sequestration or monitoring of soil health. Indeed, complex systems of categorisation and tracking SOC at different depths are a hallmark of the Australian Soil and Landscape Grid (derived from the SCaRP). The approach of Lehmann and Kleber suggests a more holistic approach of looking at the totality of the SOC and how the sum varies with time – just as we do with carbon credit for forests. Carbon tends to remain in the soil for extended periods when it is trapped in the humus and this is a binder of carbon – biologically, chemically and physically. The labile carbon (the part of the SOC that decomposes relatively quickly) tends to be nearer to the surface and rises and falls according to the season and moisture levels. Soil carbon oxidizes in times of drought and also escapes in wet events; however, soils that are healthy tend to replace the lost carbon quite quickly.
We should move quickly to get the costs of measuring SOC down using satellite observations tied to ground proofing. This will be a game changer that will enable more targeted improvements in agricultural productivity, profitability and sustainability. It will also facilitate a much enhanced uptake of carbon from the atmosphere to be sequestered in our soils at depth, with an associated increase in micro and macro biological diversity and populations.
Many people have contributed to the arguments and data presented in this paper, including many of the authors cited such as Jeff Baldock, Deli Chen and Kadambot Siddique. Others (in no particular order) are all senior figures in agriculture and groups focussed on the long-term health of our soils, eg Michael Jeffery, Walter Jehne, Deane Belfield, Snow Barlow, Michael Crawford and John White. All have been most helpful on my journey to date in coming to grips with the potential of SOC.
Note, however, that the views expressed are mine and do not necessarily coincide with the views of those listed above.
 “Conservation agriculture” refers to simultaneous use of three steps: minimum soil disturbance, soil always covered by cover crops or residues and diverse crop rotations. It includes both conventional and “organic” farming practices. “Regenerative farming” follows the same principles and adds various forms of grazing and off-farm organic inputs such as compost, humates, leanordite. Most significantly, regenerative farming puts soil health at the centre of decision making rather than productivity.