|Home | About | Journals | Submit | Contact Us | Français|
Wetlands are species-rich habitats performing valuable ecosystem services such as flood protection, water quality enhancement, food chain support and carbon sequestration. Worldwide, wetlands have been drained to convert them into agricultural land or industrial and urban areas. A realistic estimate is that 50 % of the world's wetlands have been lost.
This paper reviews the relationship between wetlands and agriculture with the aim to identify the successes and failures of agricultural use in different types of wetlands, with reference to short-term and long-term benefits and issues of sustainability. It also addresses a number of recent developments which will lead to pressure to reclaim and destroy natural wetlands, i.e. the continuous need for higher production to feed an increasing world population and the increasing cultivation of energy crops. Finally, attention is paid to the development of more flood-tolerant crop cultivars.
Agriculture has been carried out in several types of (former) wetlands for millennia, with crop fields on river floodplain soils and rice fields as major examples. However, intensive agricultural use of drained/reclaimed peatlands has been shown to lead to major problems because of the oxidation and subsidence of the peat soil. This does not only lead to severe carbon dioxide emissions, but also results in low-lying land which needs to be protected against flooding. Developments in South-East Asia, where vast areas of tropical peatlands are being converted into oil palm plantations, are of great concern in this respect. Although more flood-tolerant cultivars of commercial crop species are being developed, these are certainly not suitable for cultivation in wetlands with prolonged flooding periods, but rather will survive relatively short periods of waterlogging in normally improved agricultural soils. From a sustainability perspective, reclamation of peatlands for agriculture should be strongly discouraged. The opportunities for agriculture in naturally functioning floodplains should be further investigated. The development and use of crop cultivars with an even stronger flood tolerance could form part of the sustainable use of such floodplain systems. Extensive use of wetlands without drastic reclamation measures and without fertilizer and pesticides might result in combinations of food production with other wetland services, with biodiversity remaining more or less intact. There is a need for research by agronomists and environmental scientists to optimize such solutions.
For many millennia, humans have been cultivating land for food production. Initially, human settlements primarily occurred in fertile areas along rivers. In the floodplains of Mesopotamia, such settlements were the very cradle of human civilization 6000 years ago. From the early beginning of agricultural activities, such riverine wetlands have been recognized as valuable land areas for food and fodder production, because they have fertile soils as a result of regular sediment deposition during flood events. Access to waterways for transport was a major additional advantage. In the course of history, wetlands have been reclaimed for agriculture in many parts of the world with ever more effective drainage and land amelioration measures. The natural wetland ecosystems reclaimed in this way have lost much of their original character, leading to reduced biodiversity and reduced performance of functions other than crop productivity (Hassan et al., 2005). For the global resource of freshwater wetlands, it is certain that substantial wetland areas have been lost because of drainage and development, although quantitative estimates are only available for a number of regions; more than 50 % of the area of peatlands, depressional wetlands, riparian zones, lake littoral zones and floodplains has been lost, mostly through conversion to intense agricultural use, in North America, Europe and Australia (Millennium Ecosystem Assessment, 2005).
Although wetland protection is officially a priority for the 159 nations (as of 2009) that have ratified the Ramsar Convention (www.ramsar.org), wetlands continue to be under threat of being drained and reclaimed. Based on the expected growth of the world population in the next 25 years, the need for food products will increase 50 % by 2030 (Hassan et al., 2005). In addition, there is a growing trend to grow energy crops for use in biofuel production (Smeets et al., 2007). At the same time, measures to enhance ‘climate-neutral’ economic activities will result in initiatives to plant forests in open areas, including non-forested wetlands. All these developments will lead to a greater pressure to reclaim still remaining natural areas for agricultural purposes. This could mean that wetlands run an increasingly higher risk of being drained and destroyed. Another consequence may be the active search for more flood-tolerant and salt-tolerant crop varieties that may grow successfully under limited periods of waterlogging or drought-associated salt stress. This may lead to agricultural activities in wetlands that leave the water regime of the wetland intact but still disturb the wetland ecosystem by adding fertilizer or pesticides. Important aspects of a wetland's character will therefore be harmed and functions other than productivity may still be diminished or destroyed.
The aim of this review is to evaluate the impacts of the agricultural use of wetlands from different perspectives, with special attention to the consequences of past and current developments of land-use dynamics and new agricultural approaches for wetland functions and their benefits worldwide. We first briefly review the history of wetland use, and address to what extent agricultural productivity went together with high biodiversity and other valuable functions. Subsequently, we illustrate these developments with case studies where wetland reclamation has (had) dramatic consequences and give an overview of the latest developments in crop science with respect to flood tolerance. Finally, we summarize the information and give an opinion on the degree of sustainability of various types of agricultural use in wetlands. Sustainability has ecological, economic and social dimensions (Falvey, 2004). We refer to sustainable agriculture in wetlands as the ability to produce food indefinitely, without causing severe or irreversible damage to the wetland ecosystem character.
Wetland definitions often include three main components. Wetlands are distinguished by the presence of water, either at the surface or within the root zone, seasonally or permanent, they often have unique soil conditions that differ from adjacent uplands, and they support vegetation adapted to the wet conditions (hydrophytes) and, conversely, are characterized by an absence of flooding-intolerant vegetation (Mitsch and Gosselink 2007). Wetlands encompass a wide range of different types, characterized by hydrological and biological characteristics: for example, fens, bogs, marshes, swamps, shallow lakes, mangroves and saltmarshes. According to the Ramsar definition, wetlands also include areas of marine water the depth of which at low tide does not exceed 6 m; however, for the present analysis we will not deal with these coastal areas.
Wetland ecosystems have been recognized to provide various services (de Groot et al., 2002; Zedler and Kercher, 2005; Verhoeven et al., 2006). Services often performed/provided by wetlands (although not all at the same time or to the same degree) include storm water detention, flood protection, water quality enhancement, freshwater fisheries, food chain support, feeding grounds for juvenile marine fish, biodiversity, carbon storage and climate regulation (Hassan et al., 2005). These services are brought about by wetland ecosystem functioning, and are therefore also indicated as ‘wetland functions’ (Maltby et al., 2009). If functions provide a benefit for human societies, they are indicated as ‘benefits’ or ‘services’. If wetlands are to be reclaimed, it is important to realize that at least some of their services will no longer be provided, so that the economic profit of the reclamation is actually less than indicated by the developers. For the present review, we attempt to evaluate the extent to which agricultural activities in wetlands in the past and present have been carried out at the expense of the ecological character of the wetlands and the services that they provided. We attempt to distinguish between intensive agriculture, i.e. high degree of soil amelioration, drainage and fertilizer use, and extensive agriculture, i.e. activities in a more or less intact ecosystem without the use of machinery and chemicals. The Ramsar Convention has defined the ‘Ecological Character’ of wetlands and the ‘Wise Use’ concept, implying that wetland ecosystem services may be exploited to a certain extent, as long as the integrity and health of the wetland system remain intact. [The Ramsar Convention defines ‘Ecological Character’ as ‘the combination of the ecosystem components, processes and benefits/services that characterise the wetland at a given point in time.’ Within this context, ecosystem benefits are defined in accordance with the MA (Millennium Ecosystem Assessment) definition of ecosystem services as ‘the benefits that people receive from ecosystems’.]
Reclamation of wetlands (e.g. floodplains, peatlands, wet depressions) for agriculture has mostly involved drainage and soil improvement, and these measures have often totally destroyed their ecological character and the ecosystem services that go with it. That does not imply, however, that low-intensity agricultural activities in a large wetland area could not go together with a high biodiversity and other ecosystem services. It has even been demonstrated that many traditional agricultural systems have resulted in a very diverse landscape with high species densities in a human-created setting. Examples are the moist grasslands and herbaceous fens in Europe, which have been used for grazing and hay-making in the era prior to the use of artificial fertilizers and deep drainage. These meadows once covered large areas and had a very diverse vegetation with up to 30 species m−2 (Delpech, 1980; Vermeer and Berendse, 1983; Daniels et al., 1987; Pegtel, 1987; Bakker, 1989; Heery, 1991). Meadow birds and waterfowl, as well as other faunal elements, were also rich in species and numbers. In other parts of the world (see Fig. 1), rice has traditionally been cultivated in strongly modified landscapes with rice paddies on hill slopes or lowlands with a human-controlled water regime. Until the early 20th Century, rice cultivation has permitted a landscape very rich in diversity, especially in macro-invertebrates, fish and waterfowl (Kawano, 2000; Shimoda, 2007). However, with the intensification of agricultural practices associated with the ‘green revolution’ in the second half of the 20th Century, fertilizer and pesticide use have increased strongly and have led to a dramatic decrease of rice-field biodiversity (Bambaradeniya, 2003).
Summarizing, intensive agricultural use of wetlands has definitely altered their ecological character, because the growth of crops or raising of livestock necessitate reclamation measures such as drainage or tillage. In such wetland areas, biodiversity has often been impacted severely and large parts may no longer even qualify as wetlands. Where, however, low-intensity agriculture takes place in wetlands, involving a regime of extensive use without fertilizers or pesticides, the diversity of the wetland landscape may be high, although the species composition and setting differs strongly from that in its pristine situation. This ‘secondary biodiversity’ is often worthy of protection because it includes many rare and characteristic species. The agricultural intensification of the 20th Century, with its fertilizer and pesticide use, has destroyed many of these values, so that many crop fields, pastures and rice fields are very species-poor. Nevertheless, there are wetland areas in many parts of the world where low-intensity agriculture is combined with ecosystem services other than food, including biodiversity and flood detention (Millennium Ecosystem Assessment, 2005).
Floodplains in river basins in many parts of the world have been used for agriculture because of their natural fertility. Examples are the lower reaches of the Euphrates and Tigris, the Rhine, the Mississippi, the Danube, the Po, the Yangtze and the Ganges. Floodplain sediments are regularly deposited by flooding with river water in very wide, flat areas, with subtle height gradients from natural levees with their relatively coarse sediments to the more clayey backswamps. The higher parts of these floodplains are highly suitable for growing crops, while the lower parts are wetter but are often suitable for grazing. Floodplain soils are nutrient-rich and are naturally ‘fertilized’ as a result of flooding events. In the long history of floodplain use for agriculture, the floodplain systems of north-west Europe remained intact as wetlands with regular flooding until major engineering operations started about 200 years ago to protect crop fields and human settlements from flooding (e.g. Nienhuis, 2008). Dikes were constructed to cut off floodplain areas from the river channel, soils were drained, and canals were dug to drain water in wet periods and bring it to the field in dry periods. In particular, these large ‘land amelioration’ works in floodplains have deprived them of their wetland character in many parts of the world, particularly in the temperate zone, for example the Rhine (Nienhuis, 2008), Danube and Mississippi basins (Mitsch and Gosselink, 2000). In (semi-)arid regions, the use of freshwater for irrigating crops has also created major problems for wetland conservation. A description of the adverse effects of such developments in the Tarim River Basin of China since the 1960s is given by Chew (2003). Here, the Euphrates poplar (Populus euphratica) floodplain forest has declined strongly and many other (semi-)natural floodplain habitats have disappeared or have deteriorated. As in many other river basins, poor water resource management at the catchment scale and constructions of dams and dikes have severely impacted on environmental quality in the entire lower basin of this river.
Apart from agricultural production, river floodplains also provide a major benefit to river fisheries. Many river-dwelling fish species spawn in aquatic vegetation on floodplains and the fish larvae feed on the floodplains, so that fisheries can be strongly enhanced by water management allowing regular river flooding (Welcomme et al., 2006b). Given the severe environmental deterioration of many river basins, with increasing water use for urban development and irrigated agriculture, there are many initiatives calling for a more sustainable water resource management to restore natural flooding and to reserve so-called ‘environmental flows’ to enhance floodplain fertility and river fisheries and at the same time protect river floodplain biodiversity (Coops et al., 2006; Welcomme et al., 2006a). River floodplains remain very attractive for agricultural use, but a less stringent flood control and partial restoration of river functioning will make this use far more sustainable, better reconcilable with river fisheries and even more secure, given the threats posed by global climate change (Rijsberman and De Silva, 2006).
A classic example of the reclamation of a major wetland area for agriculture is the coastal deltaic plain of the Netherlands (Fig. 2). The interaction of the rivers Rhine, Meuse and Scheldt with the North Sea formed this plain behind a sandy coastal ridge. It consisted of a dynamic landscape of meandering river channels, extensive floodplains and large complexes of fens and bogs in those areas that were a little more remote from the direct river and sea influence. Sediment deposition and peat formation kept pace with the gradually rising sea level of the Holocene, so that the level of the land remained well above the water for most of the time (Pons, 1992). Since the 11th Century, inhabitants of the Low Countries have started to modify the hydrology of their surroundings to create protection from flooding for their dwellings and agricultural fields. They built dikes and started to manipulate the water level in the so-called ‘polders’. From the 14th Century, larger-scale reclamations started in the peatlands of the coastal plain, financed by local leaders such as the Bishop of Utrecht. Typically, in addition to enclosing a peatland area with a dike, numerous parallel drainage ditches were dug, leaving long rectangular lots for crop cultivation (Borger, 1992). These were farmed with great success: the mineral content of the peat provided high fertility. However, peat oxidation also resulted in subsidence, and the farmers had to turn to livestock and grassland because their land became too wet for crops. In the 17th Century, windmills were erected and enabled deeper drainage. The farmers were again able to grow crops, until further subsidence resulted in wet meadows. In the 20th Century, water level control became highly sophisticated with powered pumping stations, and levels of fertilizer and pesticide use increased. Figure 3 plots subsidence of the average land level in the Dutch peat meadow district (see Fig. 2), together with data for overall sea-level rise. It is striking that 4 m of peat has disappeared through drainage and oxidation in the last millennium (Schothorst, 1977). The deepest areas are now more than 7 m below mean sea level. The rates of subsidence as well as of sea-level rise have increased recently and are expected to increase further as a result of climate change. The Dutch Government has launched a new national water policy and has called for innovative solutions to reverse this dangerous trend.
The total reconstruction of the coastal plain to make it suitable for agriculture effectively meant the total destruction of the vast wetland area of approximately 8000 km2, leaving most river channels confined to their summer beds, with floodplains cut off from the rivers, and a complicated system of connected polder areas with highly controlled water levels and water flow directions. However, these landscapes have been admired by many because of their cultural authenticity and sheer beauty. Until the early 20th Century, the agricultural activities in these landscapes went together with a high level of biodiversity, with species that partly belonged to this wetland landscape and partly came in from drier areas. The grasslands and drainage ditches had very species-rich plant and macrofauna communities, which differed according to soil, land use and water-level regime. The plant communities of these areas have been described in detail (Westhoff and Den Held, 1969; Schaminée et al., 1995, 1996). The peat meadows had a dense, species-rich and characteristic fauna of meadow birds such as Lapwing, Ruff and Black-tailed Godwit. Reed marshes, small alder woodlands, and fen pools with abundant aquatic and terrestrializing vegetation each had their characteristic plant and animal communities, including waterfowl and birds of prey. This wealth of biodiversity has, however, been decimated by the intensification of land use, with high levels of fertilizer and pesticide use, further rationalizing of the landscape and water pollution. Many of the characteristic species of the peat meadows have become threatened, extremely rare or have disappeared (Vermeer, 1986; Vermeer and Joosten, 1992). Hence, in the end, and particularly in the past century, agriculture in the Dutch wetlands has been shown to be non-sustainable (subsidence) and devastating for wetland biodiversity.
Approximately 45 % of the vast area of tropical peatlands in South-East Asia (270 000 km2, representing 12 % of the total land area in this region) has been deforested in the past 50 years (Maltby, 1986; Maltby and Immirzi, 1993; Hooijer et al., 2006). Most of the deforested peatlands have been drained and are in use for agriculture. Although the more extensive forms of agriculture still permit a landscape with relatively high biodiversity (Rieley and Page, 1996; Phillips, 1998; Page et al., 1999), there is an increasing trend to intensify agricultural production by growing crops that need deeper drainage and fertilizer use. One important crop in these drained peatlands is palm oil, which is increasingly used as a biofuel. The resulting decrease in CO2 emissions from fossil fuel combustion is by far outweighed by the extra CO2 released by the oxidizing and subsiding peat soils used for growing oil palms (Immirzi et al., 1992; Maltby and Immirzi, 1993; Safford et al., 1998; Furukawa et al., 2005). Moreover, after evaluating the quantitative aspects of CO2 emissions from South-East Asian drained peatlands, it was estimated that emissions from Indonesian peatlands alone amount to 516 megatonnes of carbon per year (fires excluded). This equals 82 % of peatland emissions in South-East Asia and 58 % of global peatland emissions. These emissions from drained peatlands in Indonesia also amount to almost twice the emissions resulting from burning fossil fuels in that country, and have moved Indonesia from 21st to 3rd place with regard to global CO2 emissions, behind the USA and China (Hooijer et al., 2006).
This shows that agriculture on these tropical peat soils is not sustainable and has resulted not only in irreversible biodiversity losses but also has turned a valuable ecosystem function (carbon storage) into an environmental hazard (rapid carbon release) in very large areas. Information on concessions for logging pristine peatlands in South-East Asia shows that the major part of the still remaining peatlands can be expected to be logged and drained in the next few decades (Fig. 4). The area used for oil palm plantations, currently estimated at 20 000 km2, is expected to more than double to a surface area of 50 000 km2 by 2050 (Hooijer et al., 2006).
The world's rice fields (1·5 million km2, see Fig. 1) are a very important global food production resource. More than six millennia of rice cultivation have spread the crop from its South-East Asian roots into tropical and subtropical areas around the world. Traditionally, rice cultivation has been a labour-intensive activity in fields that were kept under water for a major part of the year. Rice fields can therefore be classified as ‘agronomically managed temporary wetland ecosystems’. Because rice fields generally have mineral soils, they are hardly subject to subsidence and in many parts of the world they have been farmed for centuries in a sustainable way in landscapes with high biodiversity. Rice fields are traditionally characterized by a rich flora and fauna, with aquatic plants and invertebrates in the aqueous phase of the system and (semi-)terrestrial species in the vegetation. Fish, amphibians, reptiles and birds are often also quite abundant and rich in species; there are many examples of combined cultures of rice production and fishing (Fernando, 1993). Typically, plant and animal species in rice fields include potential crop pests as well as natural enemies of these pests. After the major changes in rice cultivation brought about by the Green Revolution, crop yields have been increased but the degree of sustainability of rice cultivation has decreased due to problems with eutrophication, fish kills caused by toxic effects of pesticides and loss of biodiversity (Bambaradeniya, 2003). Rice field habitats rich in biodiversity remain in almost all rice cultivation regions, especially in areas where intensification of rice growing is not practical or even impossible. Abandoned rice fields are often jewels of species richness, as exemplified by the Nakaikemi wetland, Tsuruga, Fukui Prefecture, Japan, which harbours more than 2000 species in a basin of about 20 ha (Shimoda, 2007). Such refugia of rice field biodiversity are quite important and could serve as starting points for recolonization of rice fields that have a reduced biodiversity.
Rice cultivation is perhaps the best known example of irrigated agriculture. Large amounts of freshwater are required to keep the rice fields flooded in the typically warm climatic conditions where they occur. Evapotranspiration by the rice crop and evaporation from the open-water fields result in high water losses which are often replenished by diverting river water. The scale of this water use has been increasing, particularly in the past century, as will be discussed further in the next section. As a result, wetlands have disappeared or degraded in these river basins due to lack of water.
Recently, the Millennium Ecosystem Assessment has shown that global food production has doubled in the past 40 years, and has been able to keep pace with the increasing human population (Hassan et al., 2005). However, the assessment also showed that this major accomplishment has been realized at the expense of major losses in biodiversity, disruption of global element cycles, problematic eutrophication and toxification of our freshwater resources, and loss of regulating ecosystem functions. The challenge for the next 25 years will be that food production will have to increase by another 50 % merely to match the projected growth of the population. Given that at present there are still food shortages, and that the food habits of large parts of the human population are starting to shift to be more animal-based, the pressure to produce more food per area, as well as to reclaim more land for agriculture, is expected to increase strongly (FAO, 2003). Another trend that will result in additional demands for agricultural land and increasing production is the increasing use of first-generation biofuels as an alternative energy source to fossil fuels (Smeets et al., 2007). All these developments together will inevitably lead to reclamation of natural or marginally used land for intensive crop production. There will be increasing pressure to use wetlands for growing crops, so that an evaluation of the feasibility of such a use is urgently needed. The use of new, more flood-tolerant crop varieties may help to find sustainable solutions where agriculture, wetland ecosystem services and biodiversity can all benefit. In addition, it should be evaluated whether less intensive forms of agriculture could be used in (semi-)natural wetlands and lead to higher food production in a sustainable way, leaving intact species-rich wetland landscapes with additional benefits.
It remains questionable whether major crop species could be made suitable for growth in wetland environments. Research in crop science has shown a range of crop varieties that have better waterlogging tolerance than the regular cultivars. Table 1 provides an overview of the range of waterlogging tolerances of cultivars of wheat, barley, oats, triticale, maize and rice, which had been subject to waterlogging, at the vegetative stage, in the field or in waterlogging-prone soils from target environments. The timing, duration and intensity of the waterlogging events clearly affected plant responses in these preliminary experiments. It should be stressed here that the circumstances investigated are very close to current commercially used agricultural environments and are by no means representative for wetlands. The soils were well drained and ameliorated and the periods of waterlogging were very short in comparison with those of natural wetland environments. Yet the majority of these crop varieties do have some degree of flood tolerance. Further research in crop breeding could result in new cultivars in which waterlogging tolerance is further enhanced.
To our knowledge, no systematic research has been done to evaluate whether new germplasm of commercial crop species other than rice can be grown in wetland environments. Plant breeding and genetic modification is ongoing to develop cultivars that are more flood-tolerant and salt-tolerant. It would be worthwhile to evaluate the overall success and further prospects of these developments and specifically look for opportunities where such new cultivars could be used in selected wetland environments.
In view of the importance of wetland ecosystem services, it would be preferable to practise agriculture in wetland environments without the need of forced drainage measures that basically transform wetlands to dry soils. Moreover, these activities should never be carried out in peatlands, because these cannot be farmed without drainage and are prone to land subsidence and emissions of greenhouse gases. River floodplain systems are more suitable for experimental use of flood-tolerant crops.
Apart from intensive agricultural uses, which have destroyed wetland functions and services across the world, many wetlands are currently subject to extensive land uses, in which food production is often combined with other functions such as water quality enhancement, flood detention or biodiversity. Examples of such land uses are traditional crop cultivation methods without chemical fertilizers or pesticides, grazing schemes involving livestock, or traditional water management schemes to stimulate fish production and to improve fish catches.
In the past few centuries, as shown for the examples of the European wet grasslands and Asian rice fields, such extensive land uses have been shown to result in sustainable wetland landscapes, where local communities have combined food production with other ecosystem benefits. At present, such extensive land uses can be found in regions with subsistence agriculture where local communities produce food on a small scale, mainly for their own family or village, for example in sub-Saharan Africa (Waters, 2007). Combinations of local crop growing, fish production and grazing are being practised in a semi-natural setting. In such cases, natural wetlands are used for agricultural production without complete reclamation, leaving the natural hydrological processes partly intact. Such systems do not necessarily lead to complete loss of the other regulating and supporting wetland functions and services (including biodiversity).
These systems could be optimized to produce more food per unit of (wet)land surface area while conserving the wetland, leaving its hydrology intact as much as possible and protecting its functions, including its biodiversity. Research is needed to explore which combinations of land use and water management are optimal for supporting local communities in a sustainable way. It is of vital importance that agronomists, environmental scientists and local stakeholder groups cooperate to strive for the best combinations of land uses and other measures and for their actual implementation.
With the current trends of increasing agricultural activities worldwide, it becomes crucial to protect our remaining natural ecosystems from non-sustainable forms of human use. As far as wetlands are concerned, we have demonstrated clearly how historical as well as modern uses of peatlands for cultivation of crops or oil palms has created problems beyond the loss of biodiversity. The drainage of peatlands results in peat oxidation, causing major subsidence as well as the switch of peatland systems from carbon sinks to major carbon sources. The use of floodplains and rice fields has proven to be sustainable through millennia, as long as the degree of intensification and fertilizer and pesticide use remain within limits. Floodplain systems could be considered for growing flood-tolerant crop varieties. In fact, the varieties of wheat, oats, barley, soybean and maize reviewed here were successful only if flooding remained relatively short-term and took place relatively early in the growing season. These are just the conditions often found in river floodplains. Where these flooding periods can be predicted, the cultivation period could be geared towards the best performance given their waterlogging tolerance properties. Such agricultural activities in floodplains could also be tested in floodplain restoration projects, where the growth of these varieties could coincide with targets for a more natural flooding regime of the floodplain–river system by removing the structures separating the floodplain from the river channel. In all circumstances, it remains vital to test the performance of the crop varieties locally and to engage with local stakeholders to implement such new land-use strategies.
Another direction for enhancing the provision of food in wetland ecosystems is the further optimization of extensive agricultural uses, which can be combined with other wetland services. If more research could be done (by agronomists and ecologists jointly) on improving traditional wetland agriculture systems and increasing the awareness among local communities and policy-makers of the importance of wetlands for provisioning food as well as other services (flood protection, water purification, biodiversity, etc.), considerable progress could be made.
We gratefully acknowledge the invitation and financial support provided by the International Society for Plant Anaerobiosis (ISPA) to contribute this paper to the 9th ISPA Conference in Matsushima, Sendai, Japan (November 18–23, 2007). Thanks are also due to Professor Kimiharu Ishizawa (Miyagi University of Education, Japan) and Dr Alois Hooijer (Deltares, Delft, The Netherlands) for useful discussions.