Ebb and flow of riverside fortunes

Floods are an ever-present hazard dating back to the beginnings of civilisation, and communities living in river valleys have always sought to protect buildings and cultivated land against periodic inundation. For many centuries, embankments, or levees, were almost the sole form of flood defence. They were sometimes deployed in combination with improvements to the maximum carrying capacity of the natural channel, a strategy that aided other river uses, such as navigation. As the principles of hydraulic engineering became better understood, other structural measures - such as the diversion of excess water into emergency floodways, which bypassed the most valuable land, and the short-term storage of peak flows in large upstream reservoirs - became common. These traditional methods of river regulation, which aim to confront and contain damaging flows, remain dominant today despite the fact that they have been complemented in many countries by non-structural measures, such as emergency warning systems and floodplain hazard zoning.

Despite this complex array of responses, rivers still regularly overspill, creating surprise and loss for millions of people who live on flood-prone land and who believe that flooding is something that should not happen in a modern society. One reason for this paradox is that, as the demand for development land increases through the growth of population and economic prosperity, so flat river floodplains appear to provide attractive building sites and good agricultural opportunities. All too often, the scenic attraction of a river frontage is perceived as an asset rather than a hazard when financial investment and land development interests are at stake. Most of these land conversions also create more rapid run-off. This water then typically pours into a decaying infrastructure of urban drainage that is suffering from prolonged underinvestment and cannot cope with the increased flow. On top of these issues, the significantly increased flood frequencies envisaged in many climate-change scenarios indicate that much greater floodplain protection will be necessary even to maintain the flood hazard at present-day levels.

There are no easy solutions. About 400,000 houses in England and Wales, containing almost 2 per cent of the population, are potentially at risk from floods. Such figures are small compared with, for example, the 10 per cent of the population at risk in the United States. There, floods account for more than 80 per cent of all presidential disaster declarations. On a global scale, floods are the most common type of environmental hazard and they account for about one third of recorded natural disasters.

It is particularly unfortunate that just as a clear need for greater flood protection emerges, there is rising concern about some of the negative consequences of the traditionally engineered defences. For example, levees designed to protect local facilities from flooding simply pass on excess water, which would previously have been stored nearby on the flood plain, to create increased flood peaks downstream. Conversely, flood control reservoirs mean a sacrifice of land upstream for the protection of areas downstream. The increase of channel capacity to contain flood peaks requires not only complex hydraulic solutions but also maintenance of the modified river profile, which will be difficult unless the new channel conforms to a natural geomorphological equilibrium. Many structures are aesthetically displeasing and are sometimes opposed by local residents. All are expensive and all have environmental consequences. But the technology has been adopted so universally that there are very few "wild" or untamed rivers left. In the past century alone, large dams - many built for flood-control purposes - have burdened some nations with debt and have disrupted the ecology of half the world's rivers. It is only within the past few decades that the key role of floods in the healthy functioning of natural hydrological systems has been fully recognised. All human-induced changes, either to the river channel itself or to the wider land-use patterns within the drainage basin that alter the flood regime have adverse consequences for the wellbeing of the whole river corridor.

The underlying strength of this book is that it is one of the first to set river floods firmly within the overall context of dynamic physical, chemical and biological drainage systems and to document their influences on the biodiversity of aquatic and riparian environments.

Such an ambitious approach requires a genuinely interdisciplinary effort and there are 22 authors, all based at institutions in the US, with backgrounds in civil engineering, geography, geology, hydrology, hydraulics and ecology. Such a project also requires firm editorial control. The book contains 19 chapters, which are rather confusingly organised into nine sections, an arrangement for which the editor provides no rationale. It is true that there is a general sequence of material from the physical causes and effects of floods through to flood mitigation strategies, but various anomalies exist between title and content. For example, the section entitled "Responses to flooding" is really concerned with the standard flood prediction and estimation techniques familiar to most hydrologists. The penultimate section "Societal controls on human responses to flood hazards" is a curious mix of rather broad social science perspectives on flooding, replete with basic definitions, which - if included at all - should have appeared earlier in the work. There is little attempt to link chapters, although authors have been encouraged to use three rivers - the Colorado (United States and Mexico), the Tone (Japan) and the lower reaches of the Ganges-Brahmaputra (Bangladesh) - as integrating case studies.

As already indicated, this book considers the "hazards" to flora and fauna - as well as to humans - that are presented both by the presence of and the absence of floods. The flood threats to humans and the behavioural responses that ensue are treated relatively lightly. Only chapters one, 14, 15, 17 and 18 are directly concerned with what many readers would conventionally identify as flood "disasters" and their mitigation. Certainly, the book does not provide a detailed picture of the death and destruction created by floods, especially within third world countries. Given that the objective is in part to provide river floods with a better image, it is even more surprising that the benefits of floods to humans go unmentioned. Even in a disaster-prone country such as Bangladesh, where floods cause so much misery and underdevelopment, a basic distinction is made by the population between the regular barsha flood, which is necessary for food production, and the less frequent bonna years, when large seasonal flood losses outweigh the normal economic stimulus. Throughout the tropics, the moist soil that is left after flood recession is planted with subsistence crops on which millions of people depend. Some of the other human adjustments to floods are handled more securely and there are two notable chapters. One compares flood management strategies, the other discusses the use of non-structural measures to reduce flood losses.

Potential readers should understand that the real thrust of this book, as hinted in the blurb, is concentrated on demonstrating "how the non-occurrence (my italics) of floods, in association with flow regulation and other human manipulation of river systems, may create hazards for aquatic and riparian communities". The natural-science approach is taken by most authors, and their basic premise, that the most complex and diverse freshwater ecosystems flourish near river environments that fluctuate because of flooding, is generally well made. The reader is informed that more than 60 per cent of stream flow in the world is already regulated, much of it in the northern one third of the globe, and that about 500 new dams more than 15m high are constructed each year. These regulated rivers tend to lose not only their peak flows but also the hydrological connectivity that exists between the channel and the flood plain. As a result of this disruption to the natural movement of water, sediments and nutrients, many plant and animal communities in the river corridor suffer a loss of habitat.

For example, the freshwater ecology of the Colorado River is handicapped by a combination of diminished in-channel flow, lack of a seasonal flood pulse, reduced sediment loads and increased salinity. Among several contributions dealing with physical and biological processes, the chapter on inundation hydrology, which deals with floodwater sources at the local scale of a river reach, and those on bottom-land vegetation and aquatic ecosystems respectively are especially relevant.

Other authors focus on more applied issues, notably the urgent need for the rehabilitation of regulated rivers. This involves planning for river-flow regimes that sustain stream biota, as well as human populations, in a situation where the "green" rivers of the future will increasingly be managed through the artificial release of controlled floods. It is a challenging task to provide a balanced treatment of the interaction of river floods with human, riparian and aquatic communities. In emphasising the positive ecological aspects of floods and the damaging effects of river engineering schemes, this book provides a welcome, fresh outlook. It is well illustrated with diagrams and black-and-white photographs. Each chapter lists references and there is an index.

If read selectively, this book can be recommended as an up-to-date account that should establish itself as a useful volume for students in a wide range of field science disciplines. For the more anthropocentric readers, who have perhaps been slightly misled by the title, and who then have difficulty in viewing non-flood hazards to biota in quite the same way as flood hazards to human safety and welfare, other books exist. For those looking to the future, the underlying message is that flood policy is at a crossroads. Despite public expectations to the contrary, river-control engineering will always be a limited solution. Over the coming decades, governments must embrace genuinely comprehensive flood-plain planning and place greater reliance on non-structural measures. This means that although some new and more environmentally sensitive structures will have to be built to protect key urban centres, some frequently flooded areas will become river parks and recreation zones as occupants are relocated to safer sites. Above all, new flood-plain invasion must be halted. For the public and politicians alike, this more strategic and sustainable approach will involve the explicit, and no doubt unpalatable acceptance that floodplain communities will always face some degree of flood risk.

Keith Smith is emeritus professor of environmental science, University of Stirling, and author of Environmental Hazards: Assessing Risk and Reducing Disaster .