As far as I am aware the ways to achieve sustainable soil management worldwide are sufficiently outlined and the key technical elements to achieve sustainable soil management are included and take account the great variety of ecosystem services provided by soils. However, I do not think that with our present state of knowledge implementing the guidelines will be sufficient to achieve the Sustainable Development Goals. The guidelines do identify activities that should be avoided but it seems to me, it will be very difficult without addressing social, economic and political factors to bring them about, for example reducing the number of animals grazing the land to stabilise the land surface or addressing the demand for land, especially land to be used sustainably, as population continues to grow. My main concern is with the section on Surface Soil Stability.
Although we know in general terms the causes and impacts of land degradation it seems to me from reading the FAO’s ‘Status of the World’s Soil Resources’ we do not know in many countries what is the prime cause of degradation because the land can be degraded in one, or more ways not often adequately identified. I note that erosion, especially water erosion, is the main cause of degradation in many localities, but other than plot studies what is the evidence for that? It may well be that wash erosion does remove a veneer of soil without leaving any trace, but in my experience that is rare. Indeed, when I have seen wash in action, and that can be turbid flow over a fully vegetated but saturated soil surface, the amounts moved are small and where water has ponded and then evaporated only a very thin (< 1 mm) veneer of fine textured deposits can be seen. Only where rills and gullies form is there significant loss of soil from one location to another and deposits of coarse silt and sand particles seen. I question the assumption that wash erosion strips off a veneer of soil without leaving any evidence.
Rainsplash erosion may move large amounts of soil short distances, but it is flowing water which transports soil particles longer distances, so redistributing soil rapidly within a field/landscape/catchment. I note that in the GLASOD assessment of degradation wash erosion (Wt) is distinguished from the rills and gullies of terrain deformation (Wd). Moreover, from the images that often accompany assessments of what is termed wash erosion as assessed by the Universal Soil Loss Equation, or some form of it, small rills can be seen, but because of the size of the plot, not large ones. A key question it seems to me is how large and how long does a rill have to be before it is put in the wash or the deformation/rill class? As there is also in plot experiments a driver, or pull factor in the experiments, i.e. an increase in velocity of flow at the base of the plot where runoff is collected, and that it is accepted that plot experiments often overestimate erosion at lower rates, how much does the USLE overestimate erosion rates? This is important when it appears that most figures of erosion loss (rates of erosion) are based on the equation. In other words, is the importance of erosion over-stated and that the factor that is most degrading the soil is, for example, loss of organic matter, loss of fertility or an increase in acidity?
The above comments relate to cultivated land. Bare soil can be initiated by animals on very gentle slopes. Erosion of bare soil initiated and maintained by grazing animals is more difficult to assess than water erosion of cultivated land, especially if the edges of the bare soil are not sharply defined. Very often, in my admittedly limited experience of semiarid land, where sheet flow from a compacted grazed surface meets bare soil initiated in some way by animals, flow is concentrated often forming a very low ‘micro-cliff’ there. Very often however bare soil once initiated by animals is also eroded by the hooves of animals, by wind, and in colder climates by frost action, and the constant disturbance of the bare soil surface stops colonisation of the surface by vegetation. It seems to me estimates of erosion of grazing land using the USLE are not appropriate. Indeed, the equation was not devised to do that, it was devised to estimate water erosion of cultivated land.
Some form of field assessment and monitoring of erosion is needed. Much fieldwork - mapping and monitoring - is needed to gain knowledge of soils, their distribution and properties, identify where soils are compacted, identify contaminated sites and sites susceptible to contamination, and assess erosion. It is unlikely that remote sensing techniques will do the job adequately. The designs of the mapping and monitoring schemes are topics for discussion, but the schemes will need many workers/researchers going into the field.
Two other points come to mind. Firstly, and it is rightly noted in the draft document, local users of the soil are important. That cannot be stressed enough. Very often changes in how the land is managed appear to be imposed top down without the views of the indigenous farmer being taken into account. Any changes will have to be done with great care and sensitivity and with the cooperation of the local farmer. It seems to me we need to know what the indigenous farmer thinks, s/he may not think there is a problem, especially if soils are inherently infertile and yields are not trending in a markedly downward direction. Secondly, I support disturbing the soil as little as possible when drilling crops, but only if herbicides are not key to controlling weeds, there are too many other environmental factors at risk, for example weed resistance and pollution of water courses if herbicides are to be used to control weeds.
As far as I am aware the ways to achieve sustainable soil management worldwide are sufficiently outlined and the key technical elements to achieve sustainable soil management are included and take account the great variety of ecosystem services provided by soils. However, I do not think that with our present state of knowledge implementing the guidelines will be sufficient to achieve the Sustainable Development Goals. The guidelines do identify activities that should be avoided but it seems to me, it will be very difficult without addressing social, economic and political factors to bring them about, for example reducing the number of animals grazing the land to stabilise the land surface or addressing the demand for land, especially land to be used sustainably, as population continues to grow. My main concern is with the section on Surface Soil Stability.
Although we know in general terms the causes and impacts of land degradation it seems to me from reading the FAO’s ‘Status of the World’s Soil Resources’ we do not know in many countries what is the prime cause of degradation because the land can be degraded in one, or more ways not often adequately identified. I note that erosion, especially water erosion, is the main cause of degradation in many localities, but other than plot studies what is the evidence for that? It may well be that wash erosion does remove a veneer of soil without leaving any trace, but in my experience that is rare. Indeed, when I have seen wash in action, and that can be turbid flow over a fully vegetated but saturated soil surface, the amounts moved are small and where water has ponded and then evaporated only a very thin (< 1 mm) veneer of fine textured deposits can be seen. Only where rills and gullies form is there significant loss of soil from one location to another and deposits of coarse silt and sand particles seen. I question the assumption that wash erosion strips off a veneer of soil without leaving any evidence.
Rainsplash erosion may move large amounts of soil short distances, but it is flowing water which transports soil particles longer distances, so redistributing soil rapidly within a field/landscape/catchment. I note that in the GLASOD assessment of degradation wash erosion (Wt) is distinguished from the rills and gullies of terrain deformation (Wd). Moreover, from the images that often accompany assessments of what is termed wash erosion as assessed by the Universal Soil Loss Equation, or some form of it, small rills can be seen, but because of the size of the plot, not large ones. A key question it seems to me is how large and how long does a rill have to be before it is put in the wash or the deformation/rill class? As there is also in plot experiments a driver, or pull factor in the experiments, i.e. an increase in velocity of flow at the base of the plot where runoff is collected, and that it is accepted that plot experiments often overestimate erosion at lower rates, how much does the USLE overestimate erosion rates? This is important when it appears that most figures of erosion loss (rates of erosion) are based on the equation. In other words, is the importance of erosion over-stated and that the factor that is most degrading the soil is, for example, loss of organic matter, loss of fertility or an increase in acidity?
The above comments relate to cultivated land. Bare soil can be initiated by animals on very gentle slopes. Erosion of bare soil initiated and maintained by grazing animals is more difficult to assess than water erosion of cultivated land, especially if the edges of the bare soil are not sharply defined. Very often, in my admittedly limited experience of semiarid land, where sheet flow from a compacted grazed surface meets bare soil initiated in some way by animals, flow is concentrated often forming a very low ‘micro-cliff’ there. Very often however bare soil once initiated by animals is also eroded by the hooves of animals, by wind, and in colder climates by frost action, and the constant disturbance of the bare soil surface stops colonisation of the surface by vegetation. It seems to me estimates of erosion of grazing land using the USLE are not appropriate. Indeed, the equation was not devised to do that, it was devised to estimate water erosion of cultivated land.
Some form of field assessment and monitoring of erosion is needed. Much fieldwork - mapping and monitoring - is needed to gain knowledge of soils, their distribution and properties, identify where soils are compacted, identify contaminated sites and sites susceptible to contamination, and assess erosion. It is unlikely that remote sensing techniques will do the job adequately. The designs of the mapping and monitoring schemes are topics for discussion, but the schemes will need many workers/researchers going into the field.
Two other points come to mind. Firstly, and it is rightly noted in the draft document, local users of the soil are important. That cannot be stressed enough. Very often changes in how the land is managed appear to be imposed top down without the views of the indigenous farmer being taken into account. Any changes will have to be done with great care and sensitivity and with the cooperation of the local farmer. It seems to me we need to know what the indigenous farmer thinks, s/he may not think there is a problem, especially if soils are inherently infertile and yields are not trending in a markedly downward direction. Secondly, I support disturbing the soil as little as possible when drilling crops, but only if herbicides are not key to controlling weeds, there are too many other environmental factors at risk, for example weed resistance and pollution of water courses if herbicides are to be used to control weeds.
Dr R Evans
Visiting fellow,
Global Sustainability Institute,
Anglia Ruskin University,
East Road,
Cambridge
UK