Desertification: Fragility vs. Resilience


‘Resilience’ is an ecological term, one synonymous with the concept of relative stability, the rate at which a stable ecological system returns to a steady state following a perturbation. Resilience can be more-precisely defined as the time it takes a system to return a specified fraction of the way toward a steady state following a perturbation (DeAngelis, Bartell, and Brenkert, 1989). ‘Fragility,’ on the other hand, is not a quantifiable concept, so is not therefore simply the antonym of resilience. Rather, fragility refers to the idea that some parts, if not the whole, of an ecosystem can be damaged relatively easily. Arguably, the U.S. Midwestern soils swept away by the wind in the Dust Bowl conditions of the 1930s were fragile, since they were easily removed in vast quantity after plowing had destroyed the soil-binding capability of the undisturbed ecosystem (Sears, 1959).

In speaking of the resilience of an arid ecosystem from the perspective of sustainability, we must consider the following components: (1) the capability of soil particles to bind rather than become dustlike (a function of: soil organic matter, clay fraction, vegetation cover, human technological prowess); (2) the ability of the land surface to resist water erosion (a function of: slope, vegetation cover, human technological prowess); (3) the ability of microorganisms and plant propagules to estivate or lay dormant during times of adverse conditions (a function of: species diversity, genetics, vegetation cover, human technological prowess); and (4) human pressure. Periods of drought stress arid ecosystems, but species inhabiting these regions have adapted to this stress. Even the human species has adapted; cultural ecologists have investigated the various coping mechanisms of many arid-lands indigenous peoples, and found that they are ubiquitously well-adapted and resilient to perturbations of various sorts (see, e.g., Netting, 1986).

Dryland ecosystems and production systems use various strategies – especially the spatial and social spread of risk – to track climatic variations. Thus, the resilience of arid ecosystems and production systems is definable simply as the ability to track rainfall, and thus to rebound to full bloom from apparent lifelessness when the rains are adequate. Any less than a return to “full bloom” status implies some fragility within the system, and suggests some degree of irreversibility.

The Microscale Argument

There are at least two spatial scales to consider when we assess how resilient a system is – the micro and the regional. At the microscale, field studies tend to indicate that dryland plant populations are highly responsive to temporal variability in environmental stresses (e.g., Parker, 1993). Anecdotal evidence from Ethiopia, for instance, relates that seedlings sprang from the dry, dusty earth within hours of the first rains following the 1984 drought (Pearce, 1992). Helldén (1991) found no trend in the merging or expansion of desert patches around 103 examined villages and waterholes in the Sudan from 1961 to1983. Conversely, in a 1984 to 1987 study of sand-dune vegetation in the Thar Desert, Kumar and Bhandari (1993) found that “High pressure beyond the carrying capacity of the land has caused the depletion of many pioneer vegetation species. In the face of human activities, rainfall has little impact on the development of vegetation.” Such conflicting evidence suggests that at the microscale, evidence for desertification must be considered on a case-by-case basis. Disclaiming the existence of desertification from a handful of studies is no less a sweeping generalization than statements such as those issued at the 1992 Rio conference: that desertification affects 70 percent of all drylands (3.6 billion ha, about one-fourth the world's land area), and that there exists a cause-effect link between desertification and poverty (UNEP, 1993).

The Regional Scale Argument

At regional scales, evidence is equally conflictual. From a Kordofan Province (Sudan) study, Helldén (1991) was unable to identify any major changes in vegetation cover and crop productivity that could not be explained by variations in rainfall. Indeed, rainfall variations explain 71 percent of variations in crop yield. While the 1965 to 1974 drought severely impacted crop yield, recovery followed as soon as the rains returned. Unfortunately, these time series have not been continued past 1983, and so the question of a similar recovery following the severe drought of the following year remains unresolved. However, consider the case of Ethiopia (FEWS, 1994). Until the extreme drought of 1984-85, long-term food production remained equal to, or a little above, consumption needs. Food production never fully recovered after 1985, although it continues along the same general growth path of the pre-drought years. I can only speculate as to why this is the case. The Eritrean liberation war and the malignant agricultural policies of the Mengistu regime were certainly part of the equation, and the end of both of these in 1990 is perhaps reflected by a small productivity upswing. But the likely answer can be blamed on land degradation. Experts define any soil erosion rate above 50 tons per square kilometer (0.5 tons per ha) as “unusual.” Others say that 10 tons per ha is barely “acceptable.” In some parts of the Ethiopia, soil erosion figures as high as 450 tons per ha per year are not uncommon (Darkoh, 1993).

A Mechanism of Degradation: The “Islands of Fertility” Model

Helldén and his colleagues at Lund University, Sweden, contend that desertification is a myth, but this stance ignores vegetation quality in favor of quantity (see a significant fraction (from 10 to 20 percent) remains unexplained and may be linked to deterioration resulting from poor land management practices.

Schlesinger et al. (1990) suggest how overgrazing can disrupt the pulse-reserve strategy that enables arid ecosystems to resiliently survive extreme variability. In the U.S. Southwest's undisturbed perennial grasslands, the ground surface generally has vegetation cover of some kind, thus slowing runoff from rain events and enhancing infiltration. Relatively uniformly in a spatial sense, soil reserves of water and nutrients slowly build up. Intensive grazing not only removes vegetation selectively, but changes the soil surface: trampling compacts the soil, thus increasing runoff, which in turn results in increased soil erosion and nutrient removal. The net effect is to reduce the spatial homogeneity of soil moisture and nutrients. Shrubs, rather than grasses, are better adapted to exploit spatially heterogeneous moisture and nutrients, which at this stage are often found deeper in the soil profile. The biogeochemical cycling of plant nutrients is subsequently and increasingly restricted to the less-harsh environment directly beneath the shrubs, leading to the development of “islands of fertility” in between which nothing grows. In southern New Mexico, the net primary productivity of desert shrubland is similar to that of grassland.

So Is “Desertification” Real?

Grasslands are generally more fragile than shrublands; shrublands have a built-in resilience, since individual plants are deep-rooted and better able to access moisture and nutrients than the shallow-rooted grasses. Further, the islands of fertility become preferred sites for shrub regeneration, and thus lead to self-perpetuating levels of localized fertility. In terms of ecosystem resilience, therefore, if the Lund University researchers are viewing an already-degraded ecosystem, it comes as no surprise that it should appear quite stable. Such a conclusion in no way addresses how the level of economic potential may have changed as a result of a desertification process. Nor does it address the question of what will happen next. Even relatively resilient systems will have a threshold beyond which marked changes result.

Hutchinson (1996) suggests that as soil moisture and nutrient reserves are drawn down or eliminated in the Sahel, plants and animals (including humans) develop an increasing dependence on the infrequent and sporadic pulses that accompany rainfall. The system becomes ever-more reliant on annual grass species, which appear only after rainfall events. Demands for woody shrubs for browse and fuelwood in the Sahel accelerate the conversion process. Reduction in overall soil organic matter now feeds forward into concomitant decrease in soil water-retention capacity and nutrient recycling, which increases the spatial heterogeneity of these reserves. It is difficult to conceive how to avoid labeling such a downward spiral in the drylands as “desertification.”


Darkoh, M.B.K., 1993. Desertification: the scourge of Africa. Tiempo 8, April. International Institute for Environment and Development (London, UK) and the University of East Anglia (Norwich, UK).

DeAngelis, D.L., S.M. Bartell, and A.L. Brenkert, 1989. Effects of nutrient recycling and food-chain length on resilience. The American Naturalist 134:5, 778-805.

FEWS, 1994. Vulnerability Assessment (July). Washington, D.C.: U.S. Agency for International Development, Bureau for Africa, Disaster Response Coordination, Famine Early Warning System.

Helldén, U., 1991. Desertification—time for an assessment? Ambio 20:8, 372-383.

Hutchinson, C.F., 1996. The Sahelian desertification debate: A view from the U.S. Southwest. Journal of Arid Environments 33:4, 519-524.

Kumar, M. and M.M. Bhandary, 1993. Impact of human activities on the pattern and process of sand dune vegetation in the Rajasthan Desert. Desertification Control Bulletin 22, 45-54.

Netting, R.M., 1986. Cultural Ecology (second edition). Prospect Heights, IL: Waveland Press.

Parker, K.C. 1993. Climatic effects on regeneration trends for two columnar cacti in the northern Sonoran Desert. Annals of the Association of American Geographers 83:3, 452-474.

Pearce, F., 1992. Mirage of shifting sands. New Scientist, 12 December, 38-42.

Schlesinger, W.H., J.F. Reynolds, G.L. Cunningham, L.F. Huenneke, W.M. Jarrell, R.A. Virginia, and W.G. Whitford, 1990. Biological feedbacks in global desertification. Science 247, 1043-1048.

Sears, P.B., 1959. Deserts on the March (third edition, revised). Norman: University of Oklahoma Press.

UNEP, 1993. Agenda 21, Chapter 12: Managing fragile ecosystems: combating desertification and drought. Desertification Control Bulletin 22, 9-18.

The correct citation for this page is:
Milich, L., 1997. Desertification.

The Table of Contents of my work on desertification and food security is available.

Back Onward!

This site last updated August 10, 1997.