Desert Rivers - Cameron Herrington

Much of the terrestrial landform of the world’s continents lies in either arid or semi-arid zones. Additionally, estimates show that approximately 36% of the entire human population resides in water-limited environments. The average global surface temperature is expected to increase by an amount that is between 1.5 and 7°C by 2050. Droughts will be longer and more intense, rainfall more unpredictable, and snowpacks will decrease through much of the mountain ranges that feed desert river systems. As such, it can be expected that water will become even more limiting in many regions and that this condition will extend into regions that border the arid or semi-arid zones.

Desert river corridors have long been a focus of human manipulation as destructive floods resulting from monsoonal rain events and the need to store water in times of drought have placed an importance on the control and management of the water resources that are transported through these regions. Many of the plant and animal species within desert ecosystems exist on the brink of failure even in their natural states, and anthropogenic influences have further limited species survival as artificial controls have narrowed the extent of floodplain inundation and discharges from urban metropolises have greatly altered the biogeochemistry of the upstream headwaters. In the last twenty years river restoration techniques have become the preferred solution towards repairing the damage done by mega-dam building and extensive basin-wide water extraction. These projects are often accomplished with the premise that their impacts will be beneficial to the ecosystem as well as sustainable. Time will tell if their recent economic expense will be able to offset the ecological one that has been paid for centuries.

1. Introduction

Desert landscapes are some of the most extreme environments on this planet. It is not uncommon in high elevation deserts for the temperature to soar to over 37 °C (100°F) in the day and to subsequently drop to 4 °C (40°F) or below at night. Precipitation varies with the elevation and physical location of desert ranges, and these systems are normally classified according to their lack of rainfall as extremely arid (12 consecutive months without rainfall), arid (less than 250 mm annually) and semi-arid (between 250 and 500 mm annually). In contrast would be a tropical climate in which all twelve months would see an average rainfall of at least 60 mm. Evapotranspiration is a significant consumer of any water ponding on the desert’s surface or occupying the plant material protruding out of it, and in desert climates exceeds any annual rainfall amounts. These drylands are found in regions that are either isolated from ocean moisture patterns, i.e. the Great Basin of North America lying to the east in the rain shadow of the Sierra Nevada chain, or in mid-to-lower-latitudes beneath somewhat permanent high pressure systems caused by the global circulation system, i.e. the Sahara in Northern Africa and Great Victoria Desert in Central Australia [Graf, 1988]. The inhabitants of desert ecosystems thus depend greatly upon the waterways that flow perennially, ephemerally or intermittently through these regions. Desert rivers are the sole lifeblood running through these otherwise desolate landscapes [Poff, 2010].

Rivers that feed dryland systems can either be endogenic, meaning the waters originate from within the desert region, or allogenic and are sourced from mesic regions upstream. Allogenic rivers tend to have perennial flow characteristics whereas endogenic ones are characterized by either intermittent or ephemeral flows [Kingsford and Thompson, 2006]. Arid regions which are primarily fed by endogenic surface runoff tend to completely consume this resource and retain it within its boundaries in waterholes and large floodplain wetlands.


Figure 1- Desert river runs through the Mongolian steppes at sunset.

1.1. Location of the World’s Desert Rivers

Of the total world land area, over one-third of it is occupied by arid or semi-arid landscapes [Kingsford et al., 1998]. Deserts can be found on most every continent on the globe. They vary in temperature from the cold (i.e. the Great Basin Desert in North America with minimum winter temperatures of -6 °C) to the undeniably hot (the Danakil Desert of Ethiopia has an average daily temperature of 34 °C while the Lut Desert in Iran has reached temperatures as high as 70 °C).
Examples of some of the major desert river systems are the Colorado and Rio Grande River in North America, the Nile River in Africa and the Yellow River that flows through the Gobi Desert in Asia. The larger river systems found within these lands are almost always sourced from areas upstream of the desert in mountain highlands fed by winter snowfall and spring melt turning into overland runoff or downward infiltration resurfacing in groundwater springs.


Figure 2- Afar men taking blocks of salt from a depression in the Danakil Desert Depression. Photo from

1.2. Why Are These Systems Important?

As the effects of global climate change (GCC) become more evident and prevalent in today’s society, so too will we see an increase in the extent of dryland coverage worldwide and thusly further global desertification and the predominance of intermittent or ephemeral flow regimes in our river systems [IPCC, 2013; Larned et al., 2010; Sabo et al., 2012; Schlesinger, 1990]. In fact, greater than half the length of all rivers in Greece, the United States, and South Africa can be considered intermittent in nature [Larned et al., 2010]. Headwater streams are not included in this figure due to the difficulty in locating them while they are actively discharging and the total distribution of global intermittent river systems might be greater than 50% of the worldwide river network [Datry et al., 2014].

The more reliable, perennial flowing systems that sustain arid regions also have been experiencing enormous pressures in the face of population growth and agricultural demand for decades. The water limited environment surrounding these rivers leads to a tendency for increased water withdrawals in order to maintain life in areas that evapotranspiration predominates precipitation. The anthropogenic effects of water diversion projects and dam building can have catastrophic impacts on the ecological systems that rely on periodic flooding or the existence of a natural flow regime [Kingsford, 2000; Poff, 2010] as a result of the interruption in the energy balance of the river continuum concept [Vannotte, 1980].

The connectivity of downstream floodplains is often compromised as flood pulses are decreased in magnitude and frequency and riparian corridors that serve as biological refuges and nutrient sinks are further diminished or subjected to species invasions (i.e. salt cedar and Russian olive incursions in the southwestern United States). Even mighty rivers such as the Colorado River in North America are not immune to GCC responses such as increased rainfall variability, longer drought durations, and higher winter mean daily temperatures or atmospheric soil particle deposition on the snow pack leading to early snow melt. In the face of human imposed alterations to the ecosystem and a warmer Earth, the characteristics of desert rivers are vital for the effective management of regional and global water resources.

2. Characteristics of Dryland River Basins

2.1. Hydrology

Landscapes that house desert rivers are alike in that they experience very limited rainfall and generally suffer long periods of dryness between precipitation events. The lack of consistent wetting results in a parched soil column with little to no retained soil moisture and minimal organic material as a top layer. Vegetative cover is sparse in the majority of the desert environments. The peak of the rainfall in these regions occurs either frontally or in short, convective bursts caused by summer monsoonal patterns, often resulting in violent deluges that quickly inundate the soil column with water and creating flash floods that transport large flood pulses over normally dry riverbeds [Junk, 1989]. Perennially flowing rivers will even occasionally top their banks and fill ancient floodplains once channel capacities are exceeded. At one moment the desert is fluidly alive, then the storm moves on and the sun emerges to mop up any saturated remnants. The highly stochastic nature of desert rainfall patterns leads to a system that is extreme in its temporal and spatial variability.

2.2. Geomorphology

Changes to channel morphology within dryland river systems is primarily accomplished naturally through episodic flooding as a result of seasonal precipitation patterns as defined previously. Desert rivers can be generally characterized as possessing high sediment loads as a result of sediment suspension and flotation by localized surface flooding (flash floods) and subsequent drying of the riverbed after flow cessation. Both channel aggradation and degradation are prevalent conditions of these systems as they are often intermittently flowing and perennial reaches are normally not free of human alteration. When these channels are in their natural state they are generally wide and shallow. Flood control practices and water diversion projects often prefer that rivers within these desert reaches are constricted to their banks and tend to minimize spreading of the flow to the floodplain to protect property as well as to reduce the surface area that is available for evaporation. The resulting effect on the river is one of transition and change between various states as the system either craves or deposits its sediment load. A channelized river that is not allowed to overbank because of engineered controls will quickly transport its sediment load downstream and will be in need of a greater supply. This process results in a cutting of the river into the bed material and the deepening or narrowing of the channel from its preferred state if left alone. Desert rivers as a result of flood control methods and surface water withdrawals are often subjected to this type of solution, which is discussed in greater detail in Section 4 of this article below.

2.3. Ecology

The vegetation along the riparian corridor that accommodates dryland rivers tends to be quite distinct from the surrounding areas that are farther from the water source [Jacobson, 1995]. These linear oases offer cover for amphibians, reptiles, desert mammals and migrating birds. Distribution and density of tree cover over desert rivers is largely governed by the flow regime and/or frequency of flooding events with coverage varying from completely sparse to shaded. The amount of light that can penetrate the canopy and the organic loading available for biological processing is hence a function of the permanence of water flow in the region and can also deviate over the course of a few kilometers.

3. System Responses to Global Climate Change Scenarios

It is well-understood that the Earth has naturally entered both periods of colder and warmer than average climatic cycles for millions of years. This process had taken place quite some time before the dawn of mankind and the laws of thermodynamics tell us that nature has an inherent tendency towards maximizing entropy or increasing disorder. In 2013, the Intergovernmental Panel on Climate Change (IPCC) released a document that indicted human beings on the charges of global warming. The panel claimed that over half of the total increase in surface temperatures globally were a direct result of anthropogenic influences, namely greenhouse gas emissions [IPCC, 2013]. This is the same group that despite immediate criticism from naysayers won the Nobel Peace Prize in 2007 on the same workings and with a lower certainty than the 2013 findings. Despite strong scientific evidence that our actions are driving this response, the worldwide political idiosyncrasies have mostly ignored their proclamations and avoided aggressive actions to correct the trends. As such, more of the planet will experience the environment that creates and propagates the dryland river system.

3.1. Shifts in Rainfall Timing and Seasonality

Average daily temperatures and the ratio of warm days to cold days within arid or semi-arid regions are expected to increase moving forward into the 21st century [IPCC, 2013]. Many dryland river ecosystems, i.e. allogenic systems, depend greatly on early spring snowmelt to inundate their floodplains. The process of river overbanking in many desert rivers enables seedlings of native species, such as the cottonwood tree in the Middle Rio Grande Valley in New Mexico, to establish themselves and continue to offer habitat for permanent inhabitants as well as migratory waterfowl. Native vegetation and aquatic species are tuned into the timings of these riparian flood pulses and rely upon them for spawning and species proliferation. GCC threatens to alter regional runoff patterns by shifting winter precipitation from snowfall to rainfall and lessening the snowpack depths that supply the necessary streamflow in the early spring months. These systems tend to experience peak flows around June in the northern hemisphere and due to reduction in snowpack and early melt we can expect this peak to both dampen and to move to a point earlier in the season. Temporal shifts in river hydrographs might result in lowered riparian connectivity in many desert river systems and increased pressure on the native vegetation to offer some resiliency against invasions against exotic species such as the salt cedar or Russian olive. Additionally, fish species such as the Rio Grande Silvery Minnow (RGSM) are thought to spawn in the backwaters created by flood pulses in the spring and can face greater survival challenges as GCC influences historical flow regimes.

3.2. Increased Impact from Fire Loading

Human settlements within water-limited regions generally depend upon a combination of surface and ground water flows as their drinking water source. In fact, over half the world’s population inhabits dryland environments [Thoms et al., 2006]. Ground water withdrawal of ancient water sources in many areas greatly exceeds the ability of the underground aquifer to recharge through infiltration processes because of low annual rainfall amounts. This problem is further exacerbated through urbanization and subsequent transformation of natural landscape to impermeable land-types, i.e. concrete and asphalt. For this reason, many communities have begun to rely more heavily on nearby surface water sources (desert rivers) as a vital mainstay for their drinking water requirements. GCC predictions show a much higher probability of occurrence for dryland wildfires as well as an extension of the fire season because of warmer winter temperatures. Drought severity, insect outbreaks, and dated fire management practices have all contributed to the likelihood of fire in the midst of GCC to greatly impact the water quality within headwater streams that lead to many of these communities’ drinking water sources.


Figure 3- Burn scar material is washed downstream into the Rio Grande river following monsoonal rainfall that occurred the month following the Las Conchas fire in New Mexico.

3.3. Ecological Sensitivities to Temperature Change

The species that inhabit a desert region have long adapted to the historical fluctuations in rainfall and temperature patterns that result as a function of natural global processes. Scientists have shown evidence in the natural record that our world has undergone shifts from warm to colder periods throughout time and that it has an innate ability to self-regulate. However, the human record is much more brief and our massive release of carbon dioxide into the atmosphere since the American and European industrial revolution in the early 1800s has resulted in atmospheric levels that remain unparalleled in the natural system. Elevated carbon dioxide levels in Earth’s atmosphere act to raise the surface temperature of the Earth by absorbing reflected infrared radiation from the sun [Hansen, 2005; Lacis, 1973]. The contribution of carbon dioxide to the Earth’s “Greenhouse Effect” has led to GCC predictions of 1.5-2 °C by the end of the century [IPCC, 2013].

Anthropogenic influence is now thought to be the single largest contributor to elevated surface temperatures globally [IPCC, 2013]. Heat, drought, intermittency of rainfall all act to apply additional stress to arid ecosystems that are already pushed to the edge of their limits. As water becomes even more of a scarcity in these environments we can predict that desert rivers will flow at shallower depths once the snowmelts fade. Human beings will attempt to divert or to store any surplus of the available water resources and contribute to depletions in river depths. As these sources become ever shallower they will have decreased ability to absorb the sun’s radiation and to distribute its energy without raising the temperature of the water column. As water temperature rises, its dissolved oxygen (DO) content plummets. A stream or river that carries a reduced oxygen load below 5 mg/L threatens the health of the pelagic and invertebrate communities it serves.

Another undesirable side-effect of temperature increase and subsequent extension of the growing season is that algal species will have the benefit of a longer period for productivity than is normal. Algal blooms in aquatic systems can have the propensity to result in large diurnal swings in DO concentrations within the water column and can lead to anoxic conditions within the river reach. Species that would inhabit these systems might not be able to survive under future GCC if surface temperatures increase.

3.4 Fluctuations in Biogeochemical Processing

Desert riverine ecosystems are dynamic and hold a large capacity to process nutrients such as nitrates and phosphates when healthy. Our own water treatment facilities and technologies are modeled after the biogeochemical processing that takes place within these natural systems. Studies have shown that even intermittent systems quickly become busy bioreactors once flow is restored to the channel. Given that these processes are carried out at the molecular level, our understanding of them is in its infancy and when coupled with the heterogeneity of groundwater networks that exchange water with the river channel (interaction zone that houses many of these organisms responsible for metabolism) become increasingly complex to identify or model. Microorganisms can be located in some of the most inhospitable locations on the planet and many species are very good at adaptation to a wide range of conditions so it remains difficult to quantify any impacts on the biogeochemical processing capacity of desert river systems in the face of GCC. However, their importance for converting organic matter into useable forms for consumption within the environment is vital for sustaining life. It should be expected that these organisms will be asked to process greater nutrient loads as burn scar material is conveyed through surface runoff and transported downstream.

4. Anthropogenic Influences on Desert River Systems

Carbon emissions are not the only system altering influence humankind has charged the planet with resolving. Drylands are notorious for their extreme lack of moisture, however, when water does return to the system it often does so in monsoonal downpours that have the ability to entirely reshape the desert landscape. Flash floods that characterize the summer storm events in the southwestern U.S. produce regional flooding and typically transport large sediment loads downstream. It is the natural tendency of desert rivers to convey what can be considered catastrophic flows quickly through the channel system and for this reason human populations in these regions have constructed many flood control facilities in the drainage basins to minimize the loss of property and life. Another primary concern of people living in water-limited landscapes is that of water security. In order to irrigate crops and provide drinking water in the summer months once runoff from the spring snowmelt has passed and precipitation is scarce, people have also constructed water storage dams or agricultural diversions. Commonly, not as much consideration is given to the impacts on the river ecosystem when these decisions are made. Consequently, the channel responds to the artificial disturbance in what are often unfavorable ways that require additional economic resources and lead to degradation within the impacted reach.

4.1. Dams and Diversions

There are two primary reasons for the construction of dams or diversions in desert channel systems. One is that of storage and the other for flood protection. A secondary motive for their construction could also be for hydropower production, i.e. the Hoover Dam in North America. In their nature dams are designed to soften the variability of flow in a natural system and they effectively smooth out and diminish the high peak flows in the river’s hydrograph. While this does benefit the human populations who tend to build homes along the river banks during periods of average flooding and river flow (ignorant of limits to the floodplain at long flood return periods), it greatly impacts the biology of the river floodplain. Flows that historically would fill the entire floodplain and inundate riparian corridors are decreased in frequency or removed altogether. Conversely, low-flow periods that would be characteristic of the summer and winter months instead see the river experience slight increases in flow as water storage is released by dam operators for supply to cities or irrigators. The loss of floodplain connectivity caused by reduction in peak flows endangers many native species of flora and fauna who are accustomed to episodic flooding. Greater river flows during drier periods can additionally favor the spread of invasive plant species with longer growing seasons than the native inhabitants.


Figure 4- Hoover Dam regulates the flows from Lake Mead on the Colorado River. The dam sits at the meeting point of the Mojave, Sonoran
and Great Basin deserts in North America.

4.2. Natural Channel Responses

Channel response behavior by biotic organisms in the river system was not well understood until the introduction of the River Continuum Concept (RCC) by Robin Vannotte and others in 1980, decades after major dam construction projects had been carried out worldwide and centuries since smaller dams had been built by the Romans or millennia since their construction in ancient Mesopotamia [Vannote, 1980]. The RCC describes the method in which river systems respond to change through the efficient use and distribution of energy based on a harmonious existence with the physical channel. In the case of dams, the concept holds that through constructing this river feature we should expect that the result from a reduction in sediment loads should elicit a reset response from the biological organisms within the river basin as well as a physical response from the channel itself.

Changes made within and around a river channel can result in aggradation, degradation or a combination of each as the system attempts to reach an equilibrium state that it had possessed prior to the disturbance. Many desert river systems are in various states of equilibrium now due to the presence of dams and other water diversion projects built within their basins. Figure 5 provides an illustration of the Channel Evolution Model (CEM). Type V reaches in the CEM are the most equilibrated (pristine) and represent the point in the RCC that a balance has been achieved between the sediment transport capacity and the sediment supply of the river system. Because the majority of desert river systems within the world have been manipulated in some way to either store or extract water resources they are typically found somewhere from a Type I-IV classification depending on their distance from the impoundment. Those closest to the dam outfall would be sediment deficient and would fall into the Type I or II classification.


Figure 5- Illustrates the various stages of the Channel Evolution Model. Photo from

It all comes down to a fundamental difference in perspectives on whether society should view these channels as innate water delivery pipelines or natural systems that are alive and dynamically evolving. Likely, a compromise of each viewpoint would be preferable if we are to live in harmony with the natural environment. A continuation of this principle is supplied by the concept of environmental flows, by which the anthropogenic effects of dams manifested through natural channel responses could possibly be minimized through ecologically responsible flow management [Poff et al., 2010].

5. Dryland Restoration Practices

Desert river systems within the arid and semi-arid ecosystems have for the most part experienced substantial modifications as a result of human influences on the natural flow regime [Keshtkar et al., 2013; Kingsford, 2000; Palmer et al., 2007; Poff et al., 2010; Sabo et al., 2012]. These modifications are often focused on flood control, water storage, or conversions of land-use for agricultural and urban expansion; traditionally, without regards for the welfare of the natural system and its native inhabitants. Recently, desert river restoration projects have become more numerous and greater efforts have been made to correct the mistakes of the past since governmental regulations such as the Clean Water Act of 1972 was passed in the U.S. and the Water Framework Directive came into effect in Europe in 2000 [Follstad Shah et al., 2007; Palmer et al., 2007].

Restoration techniques in desert river systems, in contrast, are focused primarily on re-establishing floodplain connectivity to the river corridor [Palmer et al., 2007]. This is most often accomplished through determining the minimal river flow necessary for overbanking and to lower the river banks to the required elevation through heavy equipment excavation. Any non-native vegetation such as tamarisk in the southwest U.S. is then removed and replaced with those that originally filled the landscape. This is often done in conjunction with habitat restoration efforts for migratory birds and also to promote slackwater or backwater regions on the river banks to provide higher quality fish habitat [Dudley and Platania, 2007]. A secondary benefit that is produced through riparian recruitment and flow pulsing is the increase in biogeochemical processing of nutrient loads as a function of larger heterogeneity and biodiversity within the river corridor.

In addition to concerns over riparian functionality and the health of the river’s chemical water quality, most desert river systems find themselves in varying degrees of either channel aggradation or degradation depending on the distance of the study reach from its nearest agricultural diversion dam or flood control facility. Those reaches that are closest to the outfall of a flood control facility are sediment deprived and as a result cut deeper into the river bed in an effort to gain the sediment load back into river channel. Reaches that are further downstream often experience the contrasting state in which they are receiving large sediment inputs from intermittent streams that flow during flood events, the degraded upstream channel, or are subjected themselves to periodic drying in which the sediment drops out of the water column once the flow velocity falls below a certain magnitude. Sediment plugs in these systems are normally undesirable to river managers as flows in the river are lost as they tend to spread out and be evaporated at higher rates or as water is drawn into starved sediment layers.

6. Conclusion

Water is the limiting resource to biological growth and proliferation within desert regions. For this reason many of the larger human civilizations have inhabited the lands immediately adjacent to the drainage networks of desert rivers and have relied solely on these systems for their sustenance since ancient times. A characteristic of many of these systems is that they tend to experience episodic flooding within their wide floodplains and as anthropogenic communities have become more concentrated in these corridors it has become necessary for human survival to attempt to tame these rivers through flood control as well as to store their most precious resource for times of drought. These practices have greatly disrupted the natural flow regimes of these systems and have led to unintended consequences and depletion of other ecosystem services. The current solution to this quandary is to spend millions of dollars in restoration efforts, often without much ecological foresight and without any regulated methods of determining restoration effectiveness. As water becomes scarcer in the face of global climate change in these hot and extensively inhabited regions, it should be expected that further anthropogenic manipulation of the desert river will be the norm and that well-managed restoration projects will become vital to human sustainability within these regions.

7. Literature Cited

Datry, T., S. T. Larned, and K. Tockner (2014), Intermittent Rivers: A Challenge for Freshwater Ecology, BioScience, 64(3), 229–235, doi:10.1093/biosci/bit027.

Dudley, R. K., and S. P. Platania (2007), Flow regulation and fragmentation imperil pelagic-spawning riverine fishes, Ecol. Appl., 17(7), 2074–2086.

Follstad Shah, J. J., C. N. Dahm, S. P. Gloss, and E. S. Bernhardt (2007), River and riparian restoration in the Southwest: results of the National River Restoration Science Synthesis Project, Restor. Ecol., 15(3), 550–562.

Graf, W.J. (1988), Fluvial Processes in Dryland Rivers, Berlin: Springer-Verlag.

Hansen, J. (2005), Earth’s Energy Imbalance: Confirmation and Implications, Science, 308(5727), 1431–1435, doi:10.1126/science.1110252.

IPCC (2013), Climate Change 2013: the Physical Science Basis, Cambridge University Press, Cambridge.

Jacobson, P.J., K.M. Jacobson, and M.K. Seely (1995), Ephemeral Rivers and their Catchments: Sustaining People and Development in Western Namibia, Desert Research Foundation of Namibia, Windhoek, pp. 160.

Junk, W.J., P.B. Bayley, and R.E. Sparks (1989), The Flood Pulse Concept in River-Floodplain Systems, Can. Spec. Publ. Fish. Aquat. Sci. 106, 110-127.

Keshtkar, A. R., A. Salajegheh, A. Sadoddin, and M. G. Allan (2013), Application of Bayesian networks for sustainability assessment in catchment modeling and management (Case study: The Hablehrood river catchment), Ecol. Model., 268, 48–54, doi:10.1016/j.ecolmodel.2013.08.003.

Kingsford, R.T. (2006), Ecology of Desert Rivers, Cambridge University Press, Cambridge U.K.

Kingsford, R. T. (2000), Ecological impacts of dams, water diversions and river management on floodplain wetlands in Australia, Austral Ecol., 25(2), 109–127.

Kingsford, R. T., A. J. Boulton, and J. T. Puckridge (1998), Challenges in managing dryland rivers crossing political boundaries: lessons from Cooper Creek and the Paroo River, central Australia, Aquat. Conserv. Mar. Freshw. Ecosyst., 8(3), 361–378.

Lacis, A.A. and J.E. Hansen (1973), A Parameterization for the Absorption of Solar Radiation in the Earth's Atmosphere, Journal of the Atmospheric Sciences,31, 118-133.

Larned, S. T., T. Datry, D. B. Arscott, and K. Tockner (2010), Emerging concepts in temporary-river ecology, Freshw. Biol., 55(4), 717–738, doi:10.1111/j.1365-2427.2009.02322.x.

Palmer, M., J. D. Allan, J. Meyer, and E. S. Bernhardt (2007), River restoration in the twenty-first century: data and experiential knowledge to inform future efforts, Restor. Ecol., 15(3), 472–481.

Poff, N. L. et al. (2010), The ecological limits of hydrologic alteration (ELOHA): a new framework for developing regional environmental flow standards: Ecological limits of hydrologic alteration, Freshw. Biol., 55(1), 147–170, doi:10.1111/j.1365-2427.2009.02204.x.

Sabo, J. L., K. Bestgen, W. Graf, T. Sinha, and E. E. Wohl (2012), Dams in the Cadillac Desert: downstream effects in a geomorphic context: Downstream effects of dams, Ann. N. Y. Acad. Sci., 1249(1), 227–246, doi:10.1111/j.1749-6632.2011.06411.x.

Schlesinger, W.H., J.F. Reynolds, G.L. Cunningham, L.F. Huenneke, W.M. Jarrell, R.A. Virginia, and W.G. Whitford (2006), Biological Feedbacks in Global Desertification, Science, New Series 4946, 1043-1048.

Thoms, M.C., P.J. Beyer, and K.H. Rogers (2006), Variability, Complexity and Diversity: The Geomorphology of River Ecosystems in Dryland Regions, Ecology of Desert Rivers, Cambridge University Press, Cambridge U.K.

Vannote, R.L., G.W. Minshall, K.W. Cummins, J.R. Sedell, and C.E. Cushing (1980), The River Continuum Concept, Can. J. Fish. Aquat. Sci. 37, 130-137.