The answer to this question depends on perspective. To answer this question well and understand the different points of view let us consider the following sub-questions…

Can rain have negative impacts in drylands?

Yes it can. This is an apparent paradox resulting from the often labile dryland ecosystem structures together with sometimes excessive rain intensities. Because dryland ecosystems are often poorly vegetated having large areas of exposed crusted soil, rain water from such intensive events starts running of the surface and rapidly acquires high energy that causes rills, gullies and finally massive erosion structures. The Wadi Attir site displayed massive soil erosion and collapses before project start, caused by runoff water encountering and dissolving deep layers of exposed loess soils, excavating thousands of cubic meters of valuable soil (Figs. 1 and 2).

Fig. 1: The run-off area: a moderate slope of de-vegetated crusted loess soil provides ideal conditions for the formation of large amounts of runoff water. Out of 10 mm precipitation, over half, or more than 50000 liters of water can run off per hectare (the area shown here is several hectares big. With rain intensities of over 10 mm per hour occurring regularly, the destructiveness of such events can be easily imagined.

Fig. 2: where the water masses (see fig 1) encounter flat land with deep soil, the water enters cracks typical for degraded loess soils, and initiates a below ground erosion process creating large caverns that ultimately collapse. The collapsed structure shown is 3 – 5 m deep and documents the loss of over 1000 m3 of valuable topsoil.

Consequently such rain, rather than feeding valuable vegetation and ecosystems, induces a vicious cycle resulting in aggravated soil degradation and erosion. Water, temporarily available in excess, runs off unused, carrying with it valuable topsoil, organic matter, seeds and nutrients, and causing massive erosion and damaging floods further downstream. The fraction of moisture maintained in the soil subsequently evaporates rapidly as the exposed soil surfaces efficiently absorb the intensive sunrays, heat up and cause rapid evaporation to no benefit to vegetation.

In conclusion, any exposed crusted soil surfaces in drylands are causing increased water runoff reducing biological productivity and ecosystem resilience. The causing agents are crusted soil, lacking vegetation or plant litter cover, low soil water infiltration, and long exposed slopes allowing water running off without interference. All those are characteristics of degraded dryland ecosystems, which are being maintained in their degraded state because of the small proportion of precipitation made available to vegetation. Biological productivity therefore is too small to significantly improve soil quality or vegetation cover, and degradation can remain persistent.

Can this vicious cycle be broken by capturing water in dams and ditches?

It cannot. A technology widely used in Israel and elsewhere called contour trenching includes the large scale remodeling of topography to create a maximum of water catchments to stop water runoff, and collect the runoff water for growing local vegetation, mostly trees planted at the bottom of the ditches. However, this technology has failed in many instances and caused more serious erosion and runoff problems, with many dams collapsing under the onslaught of water masses created even by moderate precipitation events (Fig. 3).

Consequently such remodeled topographies have been recorded as the poorest productive ecosystem in the area, with the poorest soil quality and soil organic matter recorded, indicating a large negative impact of such ‘contour trenching’ efforts, see publication below.

Click icon to view a scientific article about the topic above.

Against better knowledge and expert advice, such large topographic remodeling was attempted also within the Wadi Attir project, attempting to block and harvest water from a 5 ha untreated area. Initial results indicate that such efforts are futile, with dams being breached every year since by flood waters (fig 4). Loess built damming structures are too unstable to hold more than 20 – 30 cm of floodwaters or more than 1.5 m of height difference, the rest will have to be drained by well-designed drainage structures.

Fig. 3: Chiran forest is an area north of project Wadi Attir where KKL-JNF is attempting unsuccessfully to stabilize and reverse desertification by tree planting using the so called contour trenching method.

A further problem of such large liman structures is evident as well in Fig. 4. The large amounts of water captured contain on one hand large amounts of seeds and nutrients that would otherwise be lost from the ecosystem, but also import a high load of silt that leads to soil sealing and crusting once the water dries out. Such rapid drying can cause soil salinization and will reduce water infiltration, two effects countering our restoration efforts. Therefore planting of dense tree populations resistant to water logging is the ideal complementation to stabilizing such runoff catchment areas (Fig. 5). In this sense the very sparse planting of contours by KKL-JNF using trees that suppress growth of herbaceous vegetation is clearly ill advised (Fig. 3) and scientifically unjustified. Much denser planting of adequately selected species would likely provide far better results (see next chapter).

Fig. 4: Fate of the large, deep limans created at Project Wadi Attir. Those limans are collecting the runoff from 5 hectares of undeveloped land outside the project (top right, upper picture), and have a height of about 5 meters (top). The massive inflow of water from occasional floods has overwhelmed the loess dams in all three years since completion (lower picture). Such structures require carefully engineered exits for excess water and stabilization by dense tree planting to eventually fulfill their expected function.

Fig. 5: Densely planted woodland of fast growing Prosopis juliflora, a tree having no problems with waterlogging (top), or a mix of A. victoriae, Balanites aegyptiaca and Morus (bottom), have within three years covered these large limans with their canopy and litter, and will promote their resistance against further floods, while providing litter, and improving soil stability and productivity.

A further problem of such large scale soil movement is the final destruction of any soil structure, soil microbial life, and soil nutrient and soil organic matter inventories due to oxidative processes. This loss of soil fertility and stability has been well documented, in the Chiran area the contour trenched area was the poorest in terms of soil quality and plant productivity among all systems analyzed. Immediate dense planting of soil improving tree species (Fig. 5) could alleviate such problems within a few years as demonstrated successfully in Fig 5. Based on our experience, no more than three years are required to stabilize such critical locations by means of dense vegetation and the locations in Chiran area shown could be stabilized and improved within a view years by very dense planting of the major water collection structures.

Are there methods to optimize the design and construction of water catchment structures?

There certainly are. Dependent on circumstances water catchment can be highly beneficial to avoid water runoff and erosion, while restoring local islands of productivity to the benefit of wildlife and biodiversity. Especially in the more arid areas of the Negev, such liman and terrace systems have been maintained for thousands of years to create agroforestry islands even under arid conditions. Recently restored liman systems restore tree populations, as well as soil and productive vegetation by damming of small creeks, when adequately engineered (Fig. 6).

Fig. 6: Large JNF liman in the arid Negev, capturing water from a local wadi and releasing excess via a well-designed stabilized drainage structure. Such limans are expensive to build, but can provide water for up to 1 ha of highly productive woodland.

Those limans pose significant engineering challenges, and interestingly seem not to be required, as the same trees freely spaced along the dry riverbeds will grow equally well, providing a more ecologically relevant context. Project Wadi Attir has concentrated on capturing the water in very small cheap catchments, closely spaced and with an average height drop of less than 1.5 m (Fig. 7). Those approaches were highly successful, even where water overflows, the low height differences limit massive erosion damage and allow distributing water evenly between the different catchments.

Fig 7: large numbers of smaller harvesting structures are far cheaper and more efficient than few very large and expensive ones.

Fig. 8: Restored gallery vegetation along of one of Wadi Attir’s minor wadis. Capturing water from the very top of the watershed by densely spaced dams allows for major improvement of water retention that is exploited to establish this gallery agroforestry woodland within areas of recovering farmland.

Immediate irrigation assisted planting of such liman systems very rapidly restored the original gallery forest structure expected in this arid environment (Fig. 8), with trees spaced along the wadis and stream stabilizing soil and vegetation while recovering and redistributing resources contributed by flood waters. Those trees will also provide wind break, and recover and redistribute nutrients to the surrounding farmland.

Can a combination of water harvesting, re-vegetation and soil restoration succeed in restoring original ecosystem function?

Correct. There are a number of factors affecting dryland degradation, and consequently combinations of measures are required to restore fully functional ecosystems. Fig. 9 schematically explains the various interactions possible when precipitation reached dry ecosystems of various composition and complexity. In short, if more vegetation and plant litter is available more water is captured and kept on site, while denuded areas display high water runoff leading to soil erosion. Furthermore perennial plants provide shade reducing evaporation, and leaf litter providing nutrients and soil cover.

Fig. 9: Interaction of precipitation with different dryland surface structures. Trees reduce the impact energy of raindrops, and provide litter reducing runoff and enhancing infiltration. They also provide shade to reduce evaporation so that in woodlands soil moisture remains higher than in exposed soils (Fig. 10). Shriubs provide similar services, but normally achieve much smaller canopy areas and thus provide less service.

Fig. 10: Soil water content of three different ecosystem structures analyzed in detailed research projects in the course of a full year starting in May: Woodland soil maintains double soil moisture throughout the year because of litter cover and shading, allowing establishment and growth of fresh seedlings and large amounts of herbaceous vegetation. Both degraded and conserved, exposed soils capture less of the precipitation and dry out significantly faster, resulting in much lower plant productivity. Interestingly there is little moisture difference between the degraded and conserved open shrubland soils.

Try this quiz and test your knowledge of the question:

Is water the decisive factor determining productivity in drylands?

Utilizing the information presented in this section, do you best to answer the following questions to test your knowledge of this topic question.