ICARDA Caravan 6

hanasser valley lies about 40 km east of Aleppo in north-west Syria. It's a wide valley between two low hills--Jebel El Has, about 550 m, to the west and Jebel Shebibe, about 400 m,  to the  east. The valley floor is used for the cultivation of barley, some wheat with supplemental irrigation, and a very little irrigated cotton. On the slopes of the hills, farmers graze their sheep.
        It is a harsh environment. It is hot and exposed. In many parts of the area, the ground water is now too saline to drink, and drinking water must be brought from Aleppo. And there is a real problem with grazing. Beyond Jebel Shebibe lies the steppe, on which farmers have traditionally grazed their sheep. They are no longer permitted to do so. This is a conservation measure; large areas of the steppe in the region has become overgrazed, and the authorities have been forced to close off parts of it to allow the natural vegetation to regenerate.
        In the case of Khanasser, however, the problem has simply been moved elsewhere. Now the sheep which grazed east of the area also graze on the valley slopes, leading to even more overgrazing of the valley. The slopes of the two jebels have thus been badly affected by water erosion.
        Water erosion is a real threat in the region. This may seem curious; surely the problem in the dry areas is too little water, not too much? The problem is that such rainfall as there is, is poorly distributed in time and space. The Mediterranean has winter rainfall, and for much of the year it does not rain at all, but at other times there may be violent precipitation for short periods. This is particularly harmful if the soil surface is not protected by vegetation, such as directly after the dry season. The impact of the rain on the surface may itself loosen the surface, rendering the soil vulnerable to removal by water flow.  Where the flow concentrates in channels, this may lead to the formation of erosion rills and gullies, down which further soil will be pushed by the flow. Heavy runoff down bare overgrazed areas may remove so much soil that the grazing will never regenerate.
        If researchers are to find ways of halting this, they must first understand exactly what is happening. We decided to model the major biophysical factors involved by collecting data on topography, rainfall, soil-nutrient and organic carbon content, and the soil's capacity to hold water. Having done this in a number of locations, one can then enter this data and, using a Geographical Information System (GIS), create a graphical model from which the scientist can see at once the state of the soil in given locations, and the factors that might have influenced it.
        Before describing what we did, it is important to state what we did not do. Water erosion, like other natural phenomena related to agriculture, is not solely a matter of biophysical factors.  Socioeconomic factors are crucial. They were in this case; we guess that the problem has been worsened by overgrazing. It follows that the short answer is to get the sheep away from the affected area. But this

vegetation and soil and their underlying details. This was the first start of the descriptive water erosion model. Out of the model, the assessment criteria and the extent to which they could be measured were defined. Out of these data were selected which could be collected in the time available to construct a more detailed model.
        Back in the field again, the data collection began. The area chosen was a square kilometre; within it, we took four transects--these being lines from the high slopes down to the valley floor below, each a kilometer long, and 250 meters apart. Measurements were taken at ten 100-meter steps, called intersects. Given four transects, there were therefore 40 intersects. The exact location of these intersects was then fixed for mapping purposes with the use of a GPS (Global Positioning System) receiver. The degrees of slope between the intersects were also established using an inclinometer.
        There were two basic types of sample taken at the intersects: disturbed and undisturbed. A disturbed sample is simply a sample of earth removed from the ground; an undisturbed sample is one lifted as a core within a cylinder, so that soil structure is preserved in the sample. The latter method is necessary to measure the water-holding capacity.
        The disturbed samples, however, are quite adequate for chemical analysis. We were looking for the main nutritive elements contained in the soil: nitrogen, phosphorus and potassium (NPK).We were also testing for organic matter content. Variations in this data are not necessarily because of water erosion, but they will help us to assess to what extent the area of a given intersect has been affected by nutrient depletion--which may be caused by water erosion--and how much potential there is for rehabilitation.
        The undisturbed samples helped  us check the water-infiltration rate. Soils with a low infiltration rate will generate more runoff and are more vulnerable to erosion than soils which can take up water quickly.  All this data was entered using a GIS software called ArcInfo, which allowed us to see a picture of the area in terms of soil condition while viewing, for instance, the degree of slope at the same time. The results were revealing. Soil nutrients and organic matter were low everywhere in the sample area. But they were especially so in areas of ancient, abandoned terracing.
        Thus, within this square kilometer, an area was found with a very low degree of fertility where further agriculture is almost impossible--on the transect which ran through this terracing. It might be argued that, okay, this is obvious, people used those lands for years, so they are worn out. This might be true, but the real point is that the modeling method allowed us to quickly  identify the areas where agriculture is not possible any more. With the help of GIS, these areas can be subjected to fast visual comparison. It can help give an idea of which sites are worthy of further attention, and which are a waste of time.
        To test this model properly requires multidisciplinary work; it is also necessary to place what is found within its socioeconomic context, which we did not attempt to do. But the experience in Khanasser suggests that it is worthwhile. GIS is an appropriate tool for natural-resource monitoring and management.
        There is something else that must not be forgotten, though. A GIS model is only as good as the data the scientist gave it. She must still go out into the field, and get her boots dirty.


Ms Judith van Daalen is completing her MSc work at the Agricultural University of Wageningen, The Netherlands. The resources monitoring and land management research in the Khanasser Valley is implemented by the Natural Resources Management Program (NRMP) of ICARDA. This work, carried out in May-July 1997, was part of this research.

Zuhair Arous shows an example of gully erosion.

may not be possible; after all, they had already been moved on from somewhere else! Agricultural research has to be holistic. We confined ourselves to biophysical factors because we only had three months in which to carry out this work and our resources were limited; moreover the objective was to construct and test a GIS model, no more. This is worthwhile in itself, provided one does not imagine that it will give all the answers.
        First came a reconnaissance field trip to get an impression of the area. Before one can decide which measurements can be taken, it is necessary to have a good look around. With the help of local farmers, and a colleague (Zuhair Arous of ICARDA's Natural Resources Management Program, who knows the area well), problem areas were identified and an outline made of the study area. Without this local knowledge, the first stage would have taken far longer.
        At this stage, the first overall information model could be made. A selection was made of the most  important biophysical factors of water erosion: climate, topography,