28 August 2026 ยท White paper
What the Status of the World's Soil Resources 2026 means for grazing livestock
The FAO published its second global soil assessment in eleven years last week, and the picture it paints is worse than the first. What is more interesting, for anyone working with grazing systems, is the company the report was launched in, and what the governments in the room decided to do about it.
On 27 August 2026 the Food and Agriculture Organization released the Status of the World's Soil Resources 2026, the second edition of an assessment first attempted in 2015. It was produced by the Global Soil Partnership and its Intergovernmental Technical Panel on Soils, drawing on literature published between 2015 and 2025 and on expert input from 75 countries, and it was launched at the seventeenth Conference of the Parties to the UN Convention to Combat Desertification in Ulaanbaatar, Mongolia. The timing was not incidental, because the same fortnight produced a set of intergovernmental decisions on rangelands, grasslands and pastoralism that read very differently from the way livestock are usually discussed in environmental policy.
What the assessment found
The headline findings are unambiguous and they are not encouraging. Soil management was rated poor or very poor across 52.5 percent of the regions and subregions assessed, fair across roughly a third, and good or very good across only 14 percent. Fifty eight percent of areas report a deteriorating trend since the 2015 baseline, while only around ten percent report improvement. Approximately a third of land globally is now classified as moderately to highly degraded, and erosion is identified as the single largest threat, with soil being washed or blown away considerably faster than it can form under prevailing management.
The framing the FAO chose for all of this is worth quoting, because it sets up everything that follows: the organization argued that globally we must shift from a mindset of soil exploitation to one of soil stewardship. That is not a technical recommendation, it is a statement about who is expected to do the work, and it points at land managers rather than at land use categories.
The second, and in some ways more damning, finding is about adoption rather than about soil itself. The report is explicit that proven practices to protect and restore soils exist and are simply not being taken up at the pace or the scale required. More than half of the regional assessments reviewed indicate low or very low adoption of sustainable soil management. The problem being described is therefore not primarily a knowledge gap, but an implementation gap, which shifts the policy question away from "what should be done" and toward "why is it not being done, and what would change that".
How the report is put together
The assessment runs in two parts. The first eight chapters deal with global soil issues, moving from the principles of the World Soil Charter and the guidelines it sets for stakeholders, through an overview of global soil resources and how they have been assessed historically, into the anthropogenic and natural pressures that drive chemical, physical and biological change in soils. Land use and soil management sit at the centre of that pressures analysis. Subsequent chapters take in the socio-economic conditions that shape land management decisions, including population dynamics, market access, education and cultural values, alongside climate change and its anticipated effects on soils, before arriving at the status and trends of the major soil processes threatening ecosystem services, and finally at the policy, institutional and land use management options available in response.
The threat chapter is the analytical core, and it provides a description and a ranking of the major soil threats that endanger ecosystem functions and services, both globally and separately for each region. Those threats include erosion, loss of soil organic carbon, nutrient imbalance, salinization and alkalinization, contamination, acidification, sealing, compaction, poor water status, and loss of soil biodiversity. The second part of the report, chapters nine to sixteen, then works through regional assessments covering Africa south of the Sahara, Asia, Europe, Latin America and the Caribbean, the Near East and North Africa, North America, the Southwest Pacific and Antarctica, which is what allows the headline percentages to be broken down by region rather than presented as a single global average that conceals enormous variation.
Soil biodiversity, and where nematodes sit in it
Loss of soil biodiversity is treated as a degradation threat in its own right rather than as a pleasant secondary consequence of good management, which matters, because it puts biology alongside erosion and compaction in the ranking rather than beneath them. Soils are estimated to hold in the region of 59 percent of all species on Earth, and that biodiversity is what underpins nutrient cycling, decomposition, water regulation and, ultimately, the stability of the carbon that everybody is trying to count. Roughly 70 percent of the world's soil biodiversity hotspots currently fall outside any protected area, which is a striking figure when set against the amount of conservation attention directed at aboveground habitats.
Nematodes feature specifically within that account, and for reasons that will be familiar to anyone who has read the rest of this site. They are the most abundant animals on Earth, and, unusually, they occupy essentially every trophic level of the soil food web simultaneously, from bacterial and fungal feeders through to omnivores and predators. Together with springtails, mites, earthworms and other soil arthropods they drive decomposition, regulate microbial community composition, and govern the rate at which nutrients are made available to plants. The practical consequence, which recent global analyses have reinforced, is that soil biodiversity cannot be understood by counting species alone, because the feeding roles of soil animals and the way those roles shift across land use systems and climate regions are what actually predict whether a soil system is stable and functioning. That is precisely the argument for using nematode trophic structure as an indicator rather than treating soil biology as an undifferentiated mass.
What was decided at COP17
The conference ran from 17 to 28 August under the theme "Restoring Land. Restoring Hope.", with delegates from all 197 Parties to the convention, and it was convened in alignment with the UN International Year of Rangelands and Pastoralists. Two outcomes matter here.
The first is that governments adopted decisions aimed at protecting, sustainably managing and restoring rangelands and grasslands, at supporting pastoralists and local communities, and at promoting a transition toward nature positive agriculture. This was backed by a Rangelands Flagship Initiative bringing together approximately 1.2 to 1.3 billion United States dollars across 45 projects in 23 countries on five continents, of which 644.5 million dollars is new finance and 216.4 million dollars is already confirmed and moving toward implementation. It is the largest investment in the rangelands sector in the history of the convention.
The second is a failure. Parties could not agree a global protocol or framework on drought preparedness and resilience, an issue that had already been deferred once at COP16 in 2024, and negotiations have been pushed to COP18 in 2028. WWF and other observers described the conference as laying a foundation on land restoration while falling short on drought, which is a fair summary.
Why this pushes against the binary on livestock
The prevailing public account of livestock and land is close to binary: animals are a problem, grazing degrades soil, and the environmental gain lies in reducing both. The COP17 materials do something different, and they do it explicitly rather than by implication.
The rangelands assessment presented at the conference directly challenges what it calls the long held assumption that pastoralism is primarily a driver of land degradation. It identifies the actual drivers of rangeland degradation as drought, cropland expansion, urbanisation, and the fragmentation of traditional grazing routes, which is a materially different diagnosis, because three of those four are about grazing being displaced, constrained or broken up rather than about grazing being too intense. Set against that, the conference text recognises the essential contribution of pastoralism and rangeland stewardship to climate resilience, biodiversity conservation, food security, social stability and peace. Rangelands cover more than half of the Earth's land surface, support the direct livelihoods of around 500 million people, provide roughly a sixth of global nutrition, and were valued in the COP17 reporting at somewhere between 21 and 47 trillion dollars a year in benefits, with individual hectares reaching up to 5,000 dollars a year.
The underlying ecology supports the reframing. Grasslands co-evolved with large herbivores over geological time as coupled systems of grasses, soil biota, grazers and predators, and livestock grazing is used deliberately in conservation practice to mimic the grazing regimes that wild herbivores once provided, because it is a primary determinant of plant community composition in natural and semi-natural grassland. Removing herbivory from a system that evolved with it is an intervention, not a neutral baseline. India's Guide to Grasslands and Other Open Natural Ecosystems, launched at the same conference, makes the same move in national policy terms, naming sustainable grazing as one of the core levers for restoration alongside watershed development and fodder development, and positioning pastoralists as custodians rather than as a threat.
Why this is a positive development
It is positive for three reasons, and none of them is that livestock have been exonerated.
First, it moves the policy question from land use to land management. If the driver of degradation is how grazing is done rather than whether grazing happens, then the target for intervention becomes stocking method, rest period, timing and infrastructure, all of which a farmer can actually change, rather than the existence of the enterprise itself. That is a question with a workable answer.
Second, it attaches money to that reframing. A decision text without finance is a statement of intent, whereas 45 funded projects across 23 countries is a delivery mechanism, and the fact that 216.4 million dollars is already confirmed and moving means some of this is not contingent on future pledges materialising.
Third, it puts soil biology into the same conversation as soil carbon. The assessment treats biodiversity loss as a ranked threat and identifies the soil food web as the machinery behind nutrient cycling and carbon stability, which is the necessary precondition for any serious argument that grazing management should be judged on its biological outcomes rather than on a single bulk carbon number.
What it does not settle
Two honest caveats. The first is that "well managed grazing supports rangeland function" is not the same claim as "grazing is always beneficial", and the evidence is genuinely conditional. Work on heavily degraded desert steppe, for instance, has found grazing exclusion to be more effective than continued grazing for the initial phase of vegetation recovery and nutrient transfer. Degradation state, climate and system history all condition the answer, and a framing that recognises pastoralism as essential to functioning rangelands is not a licence to ignore that.
The second is that regenerative agriculture appeared throughout COP17, on a dedicated Food Systems and Soil Health Day, in the business and finance tracks as an investable category, and across national side events, but it did so without the conference tightening its definition or attaching an evidence standard to it. The term is being funded faster than it is being specified. That gap between a practice being promoted and a practice being demonstrated is the space this doctorate is working in, and this report has made the case for the work considerably easier to state.