9. Wetlands in Drylands: Floodplain Agriculture

Drylands are regions characterized by high (but variable) levels of aridity. Growing crops in drylands generally requires irrigation. Outside the central tropics, much of Africa is either arid or semi-arid dryland (see map below). Therefore expanding irrigation in these regions seems imperative.

Fig. 1 Map showing aridity zones across the African continent (Source: COMPETE Africa website, but map made by World Meteorological Organization (WMO) and United Nations Environment Programme (UNEP) prepared for the IPCC Third Assessment Report, 2001)

However, paradoxically within these drylands there are in fact numerous, small, widely distributed 'wetlands' (see map below). Wetlands are regions characterized by seasonal inundation and are created by localized interactions between highly seasonal rainfall, topography and geology (Tooth & McCarthy, 2007). In fact, the map below only shows broad regions; there are actually many more small wetlands that are not shown at this scale. Lehner & Döll (2004) estimate that between 4.7% and 6% of SSA is wetland. Although still relatively small, this is roughly the same as the area of land currently irrigated in SSA! These wetland areas do not require irrigation and can support a range of agricultural activities without manmade storage infrastructure, as discussed below.

Fig. 2 Distribution of the main wetland and dryland areas in Africa. NB this map only shows broad regions; there are many more small wetlands are not shown at this scale. (Source: Tooth & McCarythy (2007), wetland distribution based on data from Mitsch & Gosselink (2000); dryland distribution based on data from UNEP (1992) and Thomas (1997))

Flood cropping:
Indigenous wetland communities have developed cropping systems that use the seasonal rise and fall of floodwaters to grow crops, particularly rice. Flood rice is planted as the rains begin, grown as the floods rise (a natural paddy field!) and then harvested as the floods recede. 

Fig. 3 Flood rice growing in the Hadejia-Nguru wetlands, NE Nigeria, in the 1990s (Source: Julian Thompson)

Fishing:

Wetlands also support large fisheries. The life cycle of many fish are linked to seasonal flood regimes. As floods rise, fish migrate out of the river channel onto the floodplains to spawn because there is more food and less predation. They then return to the channel when the floods recede (Adams, 1993). This both supports the fish populations and concentrates the fish into certain areas at certain times of year making them easier to catch.

Fig. 4 Local fishermen in the Hadejia-Nguru wetlands, NE Nigeria, in the 1990s (Source: Julian Thompson)

Livestock grazing:
Residual soil moisture means that floodplains can support vegetation during the dry season, even though they are not flooded. Therefore, pastoralists can graze their cattle on them during the dry season when the surrounding grasslands dry up. Thus wetlands indirectly sustain grazing systems over large areas of surrounding drylands. 

Fig. 5 Livestock and temporary pastoralist settlements in the Hadejia-Nguru wetlands during the dry season (Source: Julian Thompson)

Groundwater recharge:
Floodwaters replenish shallow aquifers under and around the wetlands (because these aquifers move seasonally). Although irrigation is not needed on the floodplains themselves, this groundwater can be used to irrigate small-scale farms located outside the floodplain. 

Fig. 6 Groundwater pump extracting water from shallow aquifers near Hadejia-Nguru Wetlands to irrigate surrounding farms (Source: Julian Thompson)

The indirect importance of wetlands in sustaining irrigation and livestock grazing in the surrounding drylands means that the total production area dependent on wetlands tends to be much larger than their actual extent (Adams, 1993). For example, Barbier & Thompson (1998) estimated that the production area dependent on the Hadejia-Nguru wetlands, in NE Nigeria, is roughly 6.5 times greater than the actual area flooded. Consequently, wetlands contribute to the livelihoods of millions of people across Africa, particularly in SSA (Rebelo et al.., 2010) and SA.

Fig. 7 The Hadejia-Nguru wetlands in NE Nigeria (Source: Julian Thompson)

Threats to wetlands: dams

Wetlands clearly have significant agricultural importance, particularly when embedded within dryland areas. Yet, they are one of the fastest disappearing ecosystems in the world (Millennium Ecosystem Assessment, 2005). This is largely due to an increase in upstream damming projects that reduce the seasonality and extent of downstream discharge (check out this link to explore the global distribution and extent of dam capacity!!). Ironically, the purpose of these dams is often to store rainwater to supply upstream irrigation (rainwater storage was advocated in my last post). However, in the context of dryland Africa and in terms of agricultural production the loss of downstream wetlands is generally not outweighed by the gain of upstream irrigation. This is because upstream irrigation projects often use water inefficiently and because the agricultural importance of wetlands includes numerous different types of agriculture and is larger than their physical extent. Furthermore, the wetlands generally support local indigenous people whereas upstream irrigation projects tend to benefit wealthier, often non-local, farmers.

The Hadejia-Nguru wetlands (NE Nigeria): a case study
These wetlands have significantly reduced in size since the construction of two large upstream dams in the 1990s, which supply the Kano River and Hadejia Valley irrigation projects. Thompson & Hollis (1995) and Barbier & Thompson (1998) tried to quantify the economic value of both the wetlands and the irrigation projects to establish which is more valuable in terms of crop agriculture, fishing and fuelwood (3 key direct floodplain benefits). (Although fuelwood is not a food source and therefore less relevant to this blog, I couldn't find a study that solely quantified agricultural value). They concluded that the economic value of the wetlands per unit area and per unit of water is over five times greater than that of the upstream irrigation projects. Furthermore, this may be an underestimate of the true economic value of wetlands because it does not take into account any of their indirect benefits such as groundwater recharge and livestock grazing.

Fig. 8 Challawa Gorge dam that was built from 1990-1992 on a main tributary of the Hadejia River in NE Nigeria. THe construction of the Challawa Gorge dam and the Tiga dam have significantly decreased food extent in the downstream Hadejia-Nguru Wetlands (Source: Julian Thompson)

How can we save the wetlands?

It would be unrealistic to 'de-commission' dams that have already constructed. However, 'controlled flood releases' have been suggested to try to achieve a better compromise between upstream irrigation and downstream flood extent (Scudder, 1980, 1989, 1991; Adams, 1993; Thompson & Hollis 1995). Traditional regimes release a little bit of water from the dam all year round, whereas controlled flood regimes try to mirror the natural hydrological regime by releasing less water during the dry season and more during the wet season. This creates artificial wet season flooding. Although this flooding will never be as great as under natural conditions, it will maintain much more of the wetland than before. Furthermore, Thompson and Hollis (1995) suggest that the loss in floodplain extent under this regime could be outweighed by an increase in productivity per unit area. This is because flood releases would increase the certainty about the timing and extent of flooding, which removes some of the risks associate with flood cropping under natural conditions.