A ‘debt’ based approach to land degradation as an indicator of global change

We propose a way to synthesize different approaches to globally map land degradation by combining vegetation and soil indicators into a consistent framework for assessing land degradation as an environmental 'debt'. our combined approach reveals a broader lens for land degradation through global change, in particular, identifying hot-spots for the different kinds of land degradation.

ing populations and affluence, agriculture will need to considerably expand its capacity to meet demand (Hunter et al., 2017). At the same time, the most productive soils are already in use (Ramankutty et al., 2008). Currently (2018), FAO estimates ~12% of land is cultivated and ~25% used as meadow or pasture, and only ~21% of world soils are without major soil constraints for cultivation (Bot et al., 2000). This nexus brings the issue of land into sharp policy focus; more so given an estimated third of land was already degraded at the beginning of the 1990s (Oldeman et al., 1991) and an estimated 1.3 billion people lived on degraded land in 2010 (Barbier & Hochard, 2018), and both numbers keep growing. The challenge and opportunity for policy makers is now to solve land scarcity issues by halting and reversing global land degradation, to meet the increasing global demand for agriculture, provide food security and mitigate the environmental impact that follows land degradation, such as biodiversity loss and climate change (Smith et al., 2016).
Major global initiatives recognize the prescience of the issue.
Most prominently, the United Nations (U.N.) sustainable development goals (SDGs) have land degradation neutrality by 2030 as a target (indicator 15.3.1; Sachs et al., 2019). Yet, what exactly is land degradation remains an unresolved issue. To date, indicators of degradation have tended to focus on either above-or belowground characteristics of the biosphere. Here we propose a way to synthesize these approaches combining vegetation and soil indicators into a consistent framework for assessing land degradation as an environmental 'debt'. Single indicators can give a limited lens through which we see an issue and underestimate the level of debt, whereas our combined approach reveals a broader lens for land degradation through global change, in particular, identifying hot-spots for the different kinds of land degradation. We propose that this approach advances our ability to assess degradation and should serve to provide important measures by which success in reversing it can be better quantified. Some years ago, Gibbs and Salmon (2015) already raised this issue, by comparing maps of the world's degraded lands using different approaches, including expert opinion on soil status (e.g. GLASOD), satellite observation of plant productivity (e.g. GLADA), biophysical models (e.g. potential vegetative productivity) and remote sensing classifications of land abandonment. As to be expected, there are many regions in the world that are degraded, or are being degraded, according to some measure but not another.
The United Nations is a good example of an organization using multiple definitions (see discussion in Supporting Information). The different definitions provide valuable, but divergent, spatial assessments of land degradation. It is apparent, that combining them would build upon their individual strengths and compensate their individual limitations. We demonstrate how this can be achieved using a natural resource 'debt' approach to capture global change. In particular, what we propose: 1. Combining the current SDG indicators-land use, carbon stocks above-and below-ground-with globally modelled soil erosion by water using the Global Soil Erosion Modelling platform approach already used by IPCC, 2. Capturing land degradation more comprehensively, as status and trend, and 3. Building on the idea of a 'land, soil and carbon debt', defined as the difference in each land degradation indicator's current value and what it would be without human intervention, or in a native condition. This utilizes recent advances in remote sensing, machine learning, and computational resources, and can now be implemented in a straightforward manner at the global scale. This considerably extends, for example, the utility of the 15.3.1 indicators in an important way: It lets us distinguish between the natural and the man-made change. As a framework, it can also, easily be augmented to include additional physical processes, such as phosphorus depletion, acidification, diffuse contamination, and others.
This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. F I G U R E 1 Global Land degradation as 'debts'. (a), the difference between natural forest potential and actual tree cover globally, and aggregated by latitude and by world region (see Figure  S1), (b), the difference between natural soil erosion and actual soil erosion globally, and aggregated by latitude and by world region, (c), the difference between natural above-ground carbon and actual above-ground carbon globally, and aggregated by latitude and by world region, (d), the difference between natural below-ground carbon and current below-ground (0-30 cm) carbon globally, and aggregated by latitude and by world region. Regarding our global tree cover debt, naturally, there could be 4.6 Gha of tree cover but currently there are only 3.2 Gha, so our global tree cover debt is 1.4 Gha (correspondingly, if we define forests as areas with >10% tree cover, naturally there could be 8.8 Gha, currently there are 5.9 Gha, and the implied forest debt is then 2.9 Gha; a). The natural rate of soil erosion would be 10 Gt per year, but currently, it is 36 Gt. Thus, our global soil erosion debt is 26 Gt -and rising (b). The above-ground biomass would naturally be 871 Gt C, but currently, it is only is 601 Gt C (c). This means our global above-ground carbon debt is 270 Gt C. Below-ground, naturally, there would be 899 Gt C, but currently, there are only 863 Gt C, which means our global below-ground carbon debt is 36 Gt C (d). The maps (a)-(d), and maps of each indicator's current and natural condition can be found in high resolution in the Supporting Information (Figures S2-S6  can also be seen that at a more disaggregated level, sub-regional areas differ substantially in their debt profiles and this shows especially clearly in the latitudinal profiles of each map. In particular, the above-and below-ground carbon debts (c and d) are to some extent inversely distributed and the soil erosion debt is globally the most concentrated indicator (b), much of it in southeast Asia, South America and Africa, while the forest debt is globally the least concentrated indicator (a).
This shows that only by considering land degradation through this broader lens, are we able to fully capture its global distribution.
If we omit soil erosion as an indicator (which it currently is for the progress assessment of SDG 15.3.1.), land degradation in southeast Asia, Africa and South America is overall underestimated. On the other hand, by only relying on soil erosion as an indicator (as is currently done by the IPCC), land degradation is underestimated in North America, Eurasia and Oceania.
Our analysis also reveals areas subject to historic degradation (e.g. arable lands in Europe and North America, the Ethiopian highlands, the Chinese plateau), and areas where degradation is a severe threat (e.g. Amazonia, Madagascar). As a result, our proposed approach both compliments, and strengthens, efforts such as the UN SDG 15.3.1 approach, providing more information to support policy development aimed at achieving land degradation neutrality and reversing it. In addition, our analysis suggests that global land use has so far decreased global tree cover by 30%, carbon stored in biomass by 20% (average for above-and below-ground carbon), and increased soil erosion almost fourfold, suggesting that our global soil erosion debt is especially large and deserving of special attention.
Our research has implications both for policy and research. For future research, our land degradation maps might be useful, for example, to understand the role of land degradation in the global distribution of agricultural yield gaps or the relationship between land degradation and socio-economic outcomes.
Policy wise, we clearly see the disadvantage of the different UN working groups operationalizing land degradation in different ways.
Some ignore soil erosion and others use only soil erosion. Both approaches do not give the full picture. In addition, no current approach quantifies overall potentials but instead only measures contemporary trends and outcomes. Yet, it is valuable to understand overall capacities to inform potential action. We propose that the methodology presented improves on the efforts of the UN working groups, consolidating their efforts into a consistent and more encompassing approach. If current policy development is to learn anything from history, the very rise and fall of entire civilizations has been determined by their ability, or inability, to manage the land (Hillel, 1992). Given the global nature of the problem, reversing degradation is a clear and present societal challenge we must address for the wellbeing of future generations. A holistic assessment is a pivotal stepping stone in this direction. Capability.

CO N FLI C T O F I NTE R E S T
All authors declare that they have no conflicts of interest. Panos Panagos 4