|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Page last updated: 17 June, 2004 |
| Dryland myths and preconceptions |
| Objectives | Myths list | References |
|
|
| Objectives | |
|
It is my experience that the accepted so-called 'knowledge' of dryland river behaviour, the character of related environments, and the sediments they leave behind, is riddled with misunderstanding and presumption. Whilst the whole of this website is dedicated to correcting this situation, and advancing true understanding, this particular section is intended to highlight some of the worst problems I frequently encounter. After summarizing the myth, I attempt to show how it is false. The first category of problem is the presumption that dryland rivers are fundamentally no different to rivers elsewhere. Related to this is misunderstanding about the nature of drylands themselves. That dryland rivers are distinctive should no longer be in doubt. I have summarized on a separate web page the main points of difference. See also the excellent review by Knighton & Nanson (1997). The second category of problem is the tendency for unsubstantiated guesswork about what dryland rivers are like, based on what people think they know. This includes extrapolation from other settings, and the creation of purely hypothetical models unanchored in observations of modern systems. It also ends up oversimplifying the nature of dryland environments. Many of the myths in this second category relate to topics I believe are, or should be, at the leading edge of dryland rivers research. In these cases, you will find further discussion of the subject through the Hot topics menu option at the top-left of this page. The mistakes resulting from these myths are particularly important for those involved with interpretation of the rock record, especially from the subsurface. Perhaps the biggest impact is on palaeoclimate interpretation and depositional setting. Do you have your own favourite dryland myth? Please . |
 Not built by Hollywood ... click for more information |
| Myths list | Top  |
|
"Drylands are dry, aren't they? The dominant physical processes are those driven by the wind. Water action is trivial to non-existent." This is a common misconception that comes about because precipitation in drylands is infrequent. Because vegetation is limited or absent in drylands, the wind takes on a geomorphological role to a much larger degree than in other environments. But this does not mean water action is absent. Indeed, it is one of the startling paradoxes of drylands that although they are regions of little rain, the details of their surfaces are mostly the products of the action of rivers (Graf, 1988, p.3). Surface runoff is of considerable importance, though usually occurring episodically. Even in the driest areas, high-magnitude sheet floods can have significant geomorphic effects, though they have rarely been observed or recorded (because they are generally unpredictable, but see McGee 1897 and Rahn 1967 for exceptions). Rather special to drylands is that low-intensity rainfall will produce significant runoff, because infiltration rates are low (partly as a result of the lack of vegetation). Aeolian sand-seas (the general impression the lay-person has of
deserts, thanks to films such as Lawrence of Arabia) generally form
less than 30% of the dryland landscape, rising to just 38% for Australia
(see Thomas 1997, p.9, table 1.5). |
|
|
An example of how water action dominates over wind action comes from the north end of the Namib Sand Sea, in Namibia, Africa (Fig. 1). This example is just one of many that could be selected to make this point. The general direction of sand movement due to wind action is roughly to the north (Lancaster 1989). Although the dunes encroach into the course of the ephemeral River Kuiseb between floods, the irregular floods are still capable of preventing the dunes from advancing further north. The Kuiseb catchment is large enough to supply adequate water to achieve this, but not all Namibian rivers are so lucky. See Fig. 1 for more details. |
 Fig. 1 |
|
This myth is common amongst European-trained geologists attempting subsurface sedimentological interpretation. Their experience of the world is mostly (exclusively?) of temperate humid conditions, and their knowledge of drylands comes solely from the news and entertainment media, or from brief vacations to drylands (when of course they see no rainfall). |
To myths list |
|
"All stream flow is in the form of flash floods each lasting just a few hours." If a person manages not to succumb to the "It never rains"
myth, then usually their next statement is to perpetuate the myth
that all dryland floods are flash floods. I suspect this arises
because most of the (few) detailed surveys of flood behaviour in
drylands have been done in small, high-gradient drainage catchments
where flows are dominated by convective storms and flash floods
truly are the norm. |
|
|
In reality, there is a great temporal and spatial variability in
precipitation, and consequently a wide range in flood magnitude
and duration. Loosely, flood duration correlates directly with catchment
area. The literature is biased towards rivers with small catchments.
This is clear from the examples shown in Fig.
2. |
 Fig. 2 |
|
This topic is covered in greater detail on the page dealing with the Distinctiveness of Dryland Rivers. |
To myths list |
|
All dryland rivers are ephemeral "Dryland rivers are all ephemeral, flowing for just a few hours at a time, then drying up completely." This myth arises partly out of the previous one "All floods are flash floods", partly from a lack of precision with the term 'ephemeral', and partly because of a lack of thought about dryland geography. Etymologically, the word "ephemeral" derives from the Greek meaning "lasting only a day". It is used colloquially to mean "lasting a very short time". But it should not be used synonymously with 'dryland river', which commonly is what happens. As shown above (Fig. 2), floods vary greatly in duration. Two aspects of dryland rivers often get overlooked. The first is groundwater contribution to flow. The second is sourcing of flow outside the dryland region. A common classification of streams is into:
These terms are in common use but lack rigid definition. Also, there is a continuous progression from perennial streams through intermittent streams to ephemeral streams. Whereas the middle and lower reaches of streams in humid regions rarely or never cease flowing and can properly be called perennial, almost every year many of their upstream feeders run dry where they are not fed by springs. In basins cut in impermeable bedrock, prolonged droughts can halt flow in most channel reaches. Karst (limestone country) that has some surface drainage often includes streams that are spatially intermittent; frequently it also contains temporally intermittent streams that flow only when heavy rain raises the groundwater table and reactivates outlets above the usual level. Temporally intermittent streams also occur in dry areas where, at low stage, only some channel reaches contain flowing water. Groundwater contribution There is a general presumption that groundwater contribution is negligible in drylands. The reasoning usually goes like this: because precipitation is low and infrequent, and evaporation rates are high, water-table height will be depressed, and usually will be considerably below the base of channels. Though this is certainly true in many places, it cannot be taken as an all-pervading rule, not in time nor in space. For example, a common phenomenon in dryland rivers is the occurrence
of rapid and deep incision, producing what in the western USA are
referred to as arroyos. This can occur without any change in base-level,
the dominant control probably being climatic (Waters & Haynes,
2001). |
|
|
Such incision can easily take the channel base down to or below the prevailing water-table, allowing significant groundwater contribution to the river, prolonging flood flow and resulting in semi-permanent surface water. It is unclear exactly what impact this might have on sediment transport and internal structures. In the example documented from the Santa Cruz River, Arizona, USA (Waters 1988), the groundwater flow led to the creation of permanently wet marshes and the accumulation of organic-rich clays within the channel. If such a sedimentary record was intersected in a borehole, it is likely it would be interpreted as the product of a lake rather than an in-channel deposit of restricted extent. Similar examples exist on other dryland rivers, such as the middle reaches of the Huab River in northwest Namibia (Figs 3 & 4), but I know of none that have been documented in any detail. Externally-sourced rivers Rivers which depend on catchments outside dryland regions may display very complex behaviour. For example, the Nile River is sourced from three principle headwaters: the White Nile drains large lakes in the southern headwaters of Uganda, and is clearly sourced from a tropical humid region; the Blue Nile and the Atbara, its other two sources, drain northern and central parts of Ethiopia, a dryland region. The Blue Nile and Atbara supply the bulk of the sediment to the lower Nile and Nile Delta, an interesting observation in the context of which climatic regions produce the greatest amount of sediment (see the Distinctiveness page Sediment Production item). But the lower Nile is dependent for water volume on the White Nile. The net result is the Quaternary sedimentological history of the Nile is highly complex (see overview in Williams et al., 1998, p.153-155). |
|
| To myths list | |
|
"It is too dry to support vegetation except in trivial amounts." Certainly vegetation is much more limited than in other regions, and this has a profound impact on infiltration, surface runoff and sediment availability. But it is definitely not true that vegetation is totally absent. What does happen is that vegetation tends to concentrate along, and often in, the channel, and is more restricted on the floodplain. See for example Fig. 4 above, the Huab River in Nambia, where the modern course of the river is picked out by vegetation. This topic is covered in greater detail on the page dealing with the Distinctiveness of Dryland Rivers. |
To myths list |
|
"Because it is so dry, sediments deposited in drylands are not disturbed by bioturbation - it is too dry for bugs. If you see bioturbation (in any significant amount), then it these must be lake sediments." Occasionally this myth is modified by the concession there may be small amounts of rooting, though of course plant material is not preserved because the environment is too strongly oxidising. This myth reflects again a profound misunderstanding of the amount of biological activity in drylands. Drylands are host to an extremely wide range of insects (especially flies and beetles), arachnids (spiders and scorpions), reptiles (lizards and snakes), and mammals (gerbils, moles, rats), that dwell in, feed in, or deposit their larvae in, dryland alluvium and aeolian sediment. Granted, the particular species are typically highly evolved to deal with the conditions (e.g. the tenebrionid beetles in the Namib that get water by standing upside down at the crest of dunes when it is foggy). Bioturbation intensity reflects sedimentation rate and moisture availability. Recognition of terrestrial trace fossils is a specialised skill. Distinction of dryland traces from those of other climates is even more difficult. But this is no excuse for presuming drylands are devoid of trace fossils. Steve Hasiotis (University of Kansas) is in the process of finishing a new book on trace fossils and ichnology that will be of considerable help in this regard. Look for it in 2002, to be published by AAPG. |
To myths list |
|
"Because it is dry, there will be no plants or bugs, and thus little organic matter, so soils do not form." This is an example of trying to guess what happens based on an incomplete understanding of drylands. It is partly based on a misunderstanding of the nature of soil processes - few sedimentologists get good training in soil formation and palaeosol recognition. It also comes about because dryland soils are often relatively immature, so are much harder to spot in the rock record. And, as with so many other aspects of drylands, soils in such settings are much more poorly understood and documented than those in temperate and humid regions. Most of the studies claiming success in using soils in interpreting the architecture of dryland region sediments are premised on the soils of those regions developing in exactly similar ways to those in other climate regions. Careful analysis of the literature on modern soils shows that dryland region soils are distinctly different to the soils of other regions. Furthermore, they may be developed to such a minimal degree that they show negligible variation across the floodplain, even though they may represent similar amounts of time to well-developed soils in other climates. Application of temperate region models in these circumstances would lead to the incorrect interpretation that sedimentation rate was high, and the river systems combed the floodplain at a rapid rate. The published sedimentological literature on dryland soils is biased almost exclusively to two 'types' of soil: Vertisols and Calcretes (horizons enriched in calcium carbonate, not strictly a soil type). Anyone taking the time to read relevant literature carefully would discover that this gives a misleading view of dryland alluvial systems, as it overlooks that:
|
To myths list |
|
Bijou Creek is a valid sedimentological model "The Bijou Creek lithofacies model gives the diagnostic appearance of dryland rivers." When it comes to interpreting the rock record, especially in the subsurface, we are dependent on having knowledge of the range of possibilities - we see what we expect to see. Lithofacies models can be extremely helpful in untangling the complexities of nature. But this is a case in point where the models can be positively dangerous. Given all the statements made above, and in the page on the Distinctiveness of Dryland Rivers, it is in any case unreasonable to expect one model to represent the wide range of complexity. Of 16 fluvial facies models recently summarised by Miall (1996, fig. 8.8, p.203-205), three in particular are considered to be relevant to drylands: he calls them "ephemeral sandy meandering", "sheetflood distal braided", and "flashy, ephemeral sheetflood". The first two of these bear considerable similarity to each other, and to others of the fluvial profiles. There is of course the problem about what ephemeral should be taken to mean (see above), so maybe some of the other models may be relevant, but that has yet to be shown in the literature. The "ephemeral sandy meandering" model seems to be needed because of a single description in the literature (Shepherd, 1987) of a single occurrence of lateral accretion surfaces at a bend in the Rio Puerco River in Mexico. Whilst it is perfectly conceivable that dryland rivers can meander, and leave behind point bars, this instance has now been brought into doubt in the course of new detailed studies by Schlumberger Research to build an outcrops database: the Shepherd example is believed now to be downstream accretion at a section cut at an unusual angle (Ian D. Bryant, personal communication 1999). The last model is also based on a single instance in the literature, this time on the description of Bijou Creek by McKee et al. (1967), and is dominated by upper flow-regime plane-bed lithofacies, a dominance which Miall admits to be misguided (Miall, 1996 p. 54 paragraph 3), though he hasn't modified the model or the descriptions! Such plane-bed lithofacies occur in just one of the logged Bijou Creek sections and form only about 5% of the deposits observed by Williams (1966, 1970, 1971) from the large floods in the Lake Eyre basin of central Australia. My own studies have led me to the general conclusion that the so-called
Bijou Creek facies model may be relevant solely for high-gradient
rivers in small catchments. We have no reliable models for other
dryland rivers. Please don't get me wrong here, I am not accusing Andrew Miall of creating this myth. Indeed, we should all be grateful to him for the exhaustive compilations he has produced, which I am sure we all turn to at some time or other. He might have perpetuated the myth a bit, but it is the general users of the literature who must take ultimate responsibility for their own conclusions. The main problem is that our overall understanding of dryland river systems is poor. Sedimentological interpretations continue to be rooted in the 1970s and early 1980s, and have not kept pace with new geomorphological knowledge. The important advances made in geomorphology in the last decade have had little impact on sedimentological thinking, and have not been matched with complementary sedimentology research. The little new work there is, such as on the Plio-Pleistocene of the Rio Grande (e.g. Mack et al., 1997; Mack & Leeder, 1998; Mack & Leeder, 1999; Pérez-Arlucea et al., 2000), the role of river capture (Mather, 2000), and the way climate can induce phases of incision or construction without requiring a base level change (Mack et al., 1998; Harvey et al., 1999), have yet to influence mainstream writing. The existing sedimentological literature is premised on a tiny number of original field studies of modern dryland systems. To make matters worse, derivative papers that use these primary results, to interpret the rock record, get cited as support for subsequent interpretations as though they themselves are primary sources. Concepts and models have been built, and spread through the literature, that are based on extremely weak science and a very limited appreciation of dryland systems. This limited view is not necessarily a fault of the authors, just a reflection of the lack of data on modern drylands. One of the most common examples of this phenomenon is the way the work of Ian Tunbridge on interpretation of the Devonian-age Trentishoe formation in North Devon (an extension of the Old Red Sandstone of S. Wales) is quoted as justifying interpretation of other units as dryland "ephemeral" rivers (citing in support either Tunbridge, 1981; or Tunbridge, 1984). Tunbridge did a good job interpreting his outcrops in the light of the then-current knowledge. But this work was solely on rock record examples, not modern rivers, and it in turn draws on just a tiny number of primary studies of ephemeral rivers. I see this phenomenon several times a year in papers I am asked to review and in graduate student dissertations. |
To myths list |
|
|
| Objectives | Myths list | References |