Anthony J. Parsons

Geomorphology of Desert Environments

In popular concept, a desert should be hot, barren, and, preferably, sandy. In reality, many deserts are few, or none of these things. Most deserts are, however, areas of aridity, and it is upon this property that the scientific definition of deserts has generally hinged. Even so, providing an acceptable measure of aridity upon which to base a definition of desert areas has not been straightforward, and several attempts based upon a variety of geomorphic, climatic, and/or vegetational indices of aridity have been made to identify the world distribution of deserts.

Plot-scale studies of vegetation, overland flow and erosion interactions: case studies from Arizona and New Mexico

Rainfall-simulation experiments have been carried out on a series of plots ranging in size from 1 m2 to c 500 m2 in order to observe process and flux-rate changes resulting from the replacement of the dominant vegetation type from grassland to shrubland in the American South-west. Results have demonstrated variations in infiltration rates, flow hydraulics, splash and interrill erosion rates and nutrient transport rates. Furthermore, the shrubland areas develop rills, which are responsible for significant increases in overall erosion rates. The small-plot experiments allow the definition of controlling factors on the processes, and highlight the importance of vegetation controls. Although the small-plot approach has a number of significant advantages, it also has a number of disadvantages, which are discussed in detail. Some of these problems can be overcome with a careful consideration of experimental design. It is argued that plot-scale studies play an important part in improving our understanding of complex, open systems, but need to be integrated with other approaches such as the monitoring of natural events and computer modelling so that mutually consistent understandings of complex ecohydrological systems can be achieved. Copyright

Effects of vegetation change on interrill runoff and erosion, Walnut Gulch, southern Arizona

During the past 100 years grassland has been replaced by shrubland in many parts of the American Southwest. The effects of this vegetation change on interrill runoff and erosion are investigated by performing field experiments on small and large runoff plots located on contemporary grassland and shrubland hillslopes in Walnut Gulch Experimental Watershed. The experiments indicate that the vegetation change causes runoff and erosion to increase in interrill areas by decreasing resistance to overland flow, decreasing runon infiltration, increasing the spatial heterogeneity of the plant canopy, and possibly increasing the susceptibility of the soil to frost action. Increased runoff and erosion result in stripping of the soil A horizon, the formation of desert pavement in intershrub areas, and the development of rills. Fines remain and accumulate only under shrubs, where they form islands of fertility. Shrubs are able to regenerate in these islands, but it is difficult for any plants to become established in intershrub areas. Consequently, there is an increase in the spatial heterogeneity of soil resources, and this spatial heterogeneity is self-perpetuating. Increasing spatial heterogeneity of soil resources is considered to be characteristic of the process of desertification.

Do Changes in Connectivity Explain Desertification

Arid and semiarid regions cover more than 40% of Earths land surface. Desertification, or broadscale land degradation in drylands, is a major environmental hazard facing inhabitants of the worlds deserts as well as an important component of global change. There is no unifying framework that simply and effectively explains different forms of desertification. In this article, we argue for the unifying concept that diverse forms of desertification, and its remediation, are driven by changes in the length of connected pathways for the movement of fire, water, and soil resources. Biophysical feedbacks increase the length of connected pathways, explaining the persistence of desertified landscapes around the globe. Management of connectivity in the context of environmental and socioeconomic change is essential to understanding, and potentially reversing, the harmful effects of desertification.

Nutrient losses in runoff from grassland and shrubland habitats in Southern New Mexico: I. rainfall simulation experiments

Rainfall simulation experiments were performed in areas of semiarid grassland (Bouteloua eriopoda) and arid shrubland (Larrea tridentata) in the Chihuahuan desert of New Mexico. The objective was to compare the runoff of nitrogen (N) and phosphorus (P) from these habitats to assess whether losses of soil nutrients are associated with the invasion of grasslands by shrubs. Runoff losses from grass- and shrub-dominated plots were similar, and much less than from bare plots located in the shrubland. Weighted average concentrations of total dissolved N compounds in runoff were greatest in the grassland (1.72 mg/1) and lowest in bare plots in the shrubland (0.55 mg/1). More than half of the N transported in runoff was carried in dissolved organic compounds. In grassland and shrub plots, the total N loss was highly correlated to the total volume of discharge. We estimate that the total annual loss of N in runoff is 0.25 kg/ha/yr in grasslands and 0.43 kg/ha/yr in shrublands — consistent with the depletion of soil N during desertification of these habitats. Losses of P from both habitats were very small.

Resistance to overland flow on desert hillslopes

Abstract This study examines the relation between the Darcy-Weisbach friction factor, f , and the Reynolds number, Re, for overland flow on six runoff plots in semiarid southern Arizona. The surfaces of these plots are irregular and covered with stones. As overland flow increases, the stones and microtopographic protuberances, which constitute the major roughness elements, are progressively inundated, thereby altering the flow resistance. Analyses of 14 cross-sections reveal that the resulting f -Re relations have two basic shapes: convex-upward and negatively sloping. These shapes bear little resemblance to the conventional f -Re relation for shallow flow over a plane bed, whose shape is a function of the state of flow. Rather, they are explained in terms of the simultaneous operation of two processes. The first is the progressive inundation of roughness elements and increase in their wetted upstream-projected area as discharge increases. This process causes f to increase. The second is the progressive increase in the depth of flow over already inundated parts of the bed as discharge increases. This process causes f to decline. Whether the f -Re relation has a positive or a negative slope depends on whether the first or second process dominates, and this depends on the configuration of the bed and level of discharge. These findings have profound implications for the mathematical modeling of overland flow on desert hillslopes. Whether a model is based on the Saint-Venant equations or makes use of the kinematic-wave approximation, it requires the specification of a relation between f and Re (or surrogates thereof), and the computed hydrograph is very sensitive to the form of this relation. Hitherto, the most widely used relation has been the conventional one for shallow flow over a plane bed. However, the present study suggests that this relation does not apply to desert hillslopes. Additional field and laboratory studies are needed to learn more of the behavior and controls of f -Re relations on such hillslopes so that models of overland flow can employ more realistic relations and better simulate runoff hydrographs.

Relation between infiltration and stone cover on a semiarid hillslope, southern Arizona

Abstract A knowledge of the spatial distribution of infiltration rates is essential to the application of realistic, distributed rainfall-runoff models in semiarid areas. One method of determining this distribution is to develop a predictive equation based on a readily observable surface property, such as stone cover. Previous studies of the relation between infiltration and stone cover on shrub-covered semiarid hillslopes have yielded both positive and negative correlations. It is suggested that positive correlations result where infiltration measurements are confined to areas between large shrubs. Negative correlations, on the other hand, reflect pronounced shrub-intershrub differences in infiltration and stone cover and are found where both shrub and intershrub areas are sampled. In this study of a semiarid hillslope in southern Arizona both shrub and intershrub areas are sampled, and a negative correlation is obtained between infiltration and stone cover. This correlation arises from the fact that under shrubs fine sediments have accumulated primarily as a result of differential splash, whereas between shrubs such sediments have been selectively removed by a combination of rainsplash and overland flow, leaving behind a gravel lag. Consequently, infiltration rates are higher under shrubs than between them, owing to the higher percentage of sand in the surface soil, the larger quantity of organic matter both within the soil and on its surface and the greater digging and burrowing by animals. In addition, simulated rainfall experiments imply that shrubs with moderately dense canopies are more effective than adjacent surface gravels in dissipating the kinetic energy of raindrops and thereby reducing surface sealing and promoting infiltration. Inasmuch as the field site is typical of gentle, shrub-covered piedmont hillslopes over much of the Sonoran Desert, inverse relations between infiltration rate and stone cover are probably characteristic of such hillslopes and may be used to predict the spatial distribution of infiltration rates for inclusion in rainfall-runoff models.

Resistance to overland flow on semiarid grassland and shrubland hillslopes, Walnut Gulch, southern Arizona

One hundred and thirty six experiments were conducted on runoff plots on semiarid grassland and shrubland hillslopes at Walnut Gulch, Arizona. Graphs of Darcy-Weisbach friction factor f against Reynolds number Re are positively sloping or convex-upward for the grassland and predominantly negatively sloping for the shrubland. These trends are attributed to the progressive inundation of the roughness elements, with the submergence of the gravel on the shrubland being greater than the submergence of the plants on the grassland. The f-Re relations are of little value as models for predicting flow resistance at locations other than where they were developed because each location has its own unique relation which is a function of the surface properties at that location. Consequently, multivariate models that include surface properties among their predictive variables were derived for the grassland and shrubland hillslopes using multiple regression analysis. On the grassland 69.5% of the variation in f is accounted for by basal plant stem and litter cover, whereas on the shrubland 56.3% of the variation is explained by gravel cover and gravel size. The inclusion of Re improves the explained variation by 5.4% on the grassland and 7.6% on the shrubland. These results suggest that for most purposes the extra effort involved in measuring or modeling flow rate is not worth the small improvement in predictive accuracy and, therefore, that the practice of estimating resistance to overland flow solely on the basis of surface properties is entirely reasonable. A comparison of the two sets of experiments reveals that both infiltration and resistance to flow are higher on grassland than shrubland. Consequently, where shrubland has replaced grassland during the past century, in addition to the ground surface becoming more exposed to raindrop impact, overland flow has increased in volume and velocity. Together these changes have resulted in accelerated erosion in the form of rill development and stripping of the soil A horizon.

Development of badlands and gullies in the Sneeuberg, Great Karoo, South Africa

Abstract The study aims to examine the origin and development of land degradation with particular emphasis on badland and gully systems in the Sneeuberg uplands of the Great Karoo. This is an area of semiarid extensive stock farming where land degradation in the form of rill and gully erosion has accompanied the replacement of grassland by shrub vegetation. Species diversity has declined and ground cover has been reduced, leading to a positive feedback loop which exacerbates the degradation. Many foot slopes developed in shales, clays and colluvium have extensive, incipient badland development with closely spaced gullying up to 1.5 m deep. In valley-bottom and valley-side depression locations gullies up to 8 m deep have developed, usually cut to bedrock through valley fills of mainly Holocene colluvium. Both badlands and gullies appear to have developed since European settlement and to be part of the same hydrological system with extensive areas of bare ground (badlands) feeding water to incising gullies. Experiments using simulated rainfall throw some light on current processes. Badland areas are active under high-frequency, low-magnitude rainfall events. Major gullies are likely to be the result of occasional, high-magnitude events, but these have not been observed. Overgrazing in the past is the most likely cause of the degradation.

Emergence and erosion : a model for rill initiation and development

Soil erosion by overland flow, resulting from infiltration-excess rainfall, generates rill networks on hillslope areas. The way in which these networks emerge and develop suggests that hillslope erosion functions as a self-organizing dynamic system. Based upon this argument, a model for soil erosion (RillGrow 1) has been developed: this operates at the spatial scale of raindrops and microtopography. In this paper the second generation of the model (RillGrow 2) is described and applied to four different soil surfaces. Results suggest that, even at this early stage in its development, RillGrow 2 is capable of replicating the success of the earlier model and in some cases of extending them. The success of both models suggests that this self-organizing view of rill generation may capture some fundamental aspects of the operation of real erosional systems. Copyright