Agricultural Systems 123 (2014) 1–11

Contents lists available at ScienceDirect

Agricultural Systems

journal homepage: www.elsevier .com/locate /agsy

Review

Sustainable rice production in African inland valleys: Seizing regionalpotentials through local approaches

0308-521X/$ - see front matter � 2013 Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.agsy.2013.09.004

⇑ Corresponding author. Tel.: +255 222780768, 688425335.E-mail address: j.rodenburg@cgiar.org (J. Rodenburg).

Jonne Rodenburg a,⇑, Sander J. Zwart b, Paul Kiepe a, Lawrence T. Narteh c, Wilson Dogbe d,Marco C.S. Wopereis b

a Africa Rice Center (AfricaRice), East and Southern Africa, P.O. Box 33581, Dar es Salaam, Tanzaniab Africa Rice Center (AfricaRice), 01 BP 2031, Cotonou, Beninc Food and Agriculture Organization of the United Nations (FAO), Viale delle Terme di Caracalla 0153, Rome, Italyd Savanna Agricultural Research Institute (SARI), P.O. Box TL 52, Tamale, Ghana

a r t i c l e i n f o a b s t r a c t

Article history:Received 21 December 2012Received in revised form 18 September 2013Accepted 19 September 2013Available online 29 October 2013

Keywords:LowlandsWetlandsIntegrated crop managementWater managementBiodiversityParticipatory approaches

With an estimated surface area of 190 M ha, inland valleys are common landscapes in Africa. Due to theirgeneral high agricultural production potential, based on relatively high and secure water availability andhigh soil fertility levels compared to the surrounding uplands, these landscapes could play a pivotal rolein attaining the regional objectives of food security and poverty alleviation. Besides agricultural produc-tion, i.e. mainly rice-based systems including fish-, vegetable- fruit- and livestock production, inland val-leys provide local communities with forest, forage, hunting and fishing resources and they are importantas water buffer and biodiversity hot spots. Degradation of natural resources in these vulnerable ecosys-tems, caused by indiscriminate development for the sole purpose of agricultural production, should beavoided. We estimate that, following improved water and weed management, production derived fromless than 10% of the total inland valley area could equal the total current demand for rice in Africa. A sig-nificant part of the inland valley area in Africa could hence be safeguarded for other purposes.

The objective of this paper is to provide a methodology to facilitate fulfilment of the regional agricul-tural potential of inland valleys in sub-Saharan Africa (SSA) such that local rural livelihoods are benefitedand regional objectives of reducing poverty and increasing food safety are met, while safeguarding otherinland-valley ecosystem services of local and regional importance. High-potential inland valleys shouldbe carefully selected and developed and highly productive and resource-efficient crop production meth-ods should be applied. This paper describes a participatory, holistic and localized approach to seize theregional potential of inland valleys to contribute to food security and poverty alleviation in sub-SaharanAfrica. We analyzed over a 100 papers, reference works and databases and synthesized this with insightsobtained from nearly two decades of research carried out by the Africa Rice Center and partners. We con-clude that sustainable rice production in inland valleys requires a step-wise approach including: (1) theselection of ‘best-bet’ inland valleys, either new or already used ones, based on spatial modelling and adetailed feasibility study, (2) a stakeholder-participatory land use planning within the inland valley basedon multi-criteria decision making (MCDM) methods and using multi-stakeholder platforms (MSP), (3)participatory inland-valley development, and (4) identification of local production constraints combiningmodel simulations and farmer participatory priority exercises to select and adapt appropriate practicesand technologies following integrated management principles.

� 2013 Elsevier Ltd. All rights reserved.

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22. Current inland valley use. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

2.1. Drivers for inland valley use. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32.2. Multi-functional character of inland valleys . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

3. Constraints to development and use of inland valleys in Africa . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4

Fig. 1.uplandWindm

2 J. Rodenburg et al. / Agricultural Systems 123 (2014) 1–11

3.1. Development constraints . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43.2. Production constraints . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4

4. Towards sustainable inland valley development and exploitation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

4.1. Selecting suitable inland valleys. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64.2. Participatory land use planning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74.3. Designing, implementing and evaluating best-fit water management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74.4. Optimizing productivity and profitability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8

5. Conclusions – regional potentials, local approaches . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

1. Introduction

Inland valleys can be defined as seasonally flooded wetlandscomprising valley bottoms (fluxial) and hydromorphic fringes(phreatic) but excluding river flood plains (Fig. 1; Table 1). Withan estimated land area of 190 M ha (FAO, 2003) inland valleysare abundantly available in Africa and serve a multitude of ecosys-tem functions. Inland valleys, in particular the valley bottoms –bas-fonds, fadamas, inland swamps in West Africa; mbuga in EastAfrica and vleis, dambos, mapani, matoro, inuta or amaxhaphozi inSouthern Africa according to Acres et al. (1985) – generally havea high agricultural production potential due to their relative highand secure water availability and soil fertility (Andriesse et al.,1994). The hydromorphic slopes of the inland valleys are oftenused for dryland rice and cash crops like cotton, while the upperslopes, with lower groundwater levels (Fig. 1), are often grownby high value fruit trees, like mangos and cashew nut, and foddercrops (Balasubramanian et al., 2007), and the crests by maize orsorghum (e.g. Lawrence et al., 1997). The ground cover providedby these trees and crops on higher parts of the slope reduces soilrun-off towards the hydromorphic slopes and valley bottom (e.g.de Ridder et al., 1997; Rodenburg et al., 2003). The only major foodcrop that can be grown under the temporary flooded conditions ofthese valley bottoms is rice (e.g. Andriesse and Fresco, 1991).Depending on the species (Oryza sativa or Oryza glaberrima),

Schematic landscape presentation of rice production environments along the– lowland continuum, and their hydrological regimes (Adapted from:

eijer and Andriesse, 1993).

sub-species (japonica or indica) and cultivar, this crop can be grownalong the upland – lowland continuum (e.g. Saito et al., 2010). Thedevelopment of inland valleys into rice-based production systems,can be accomplished with relatively small-scale technologies thatwould require moderate investments (Roberts, 1988). For this rea-son, inland valleys, comprising such huge and yet largely unex-ploited area, are strategically important for the development ofthe African rice sector (e.g. Sakurai, 2006; Balasubramanian et al.,2007).

Wetlands, such as inland valleys, are particularly important as-sets for the rural poor as they can fulfil many services (Turner et al.,2000). Apart from agricultural production, these ecosystems sup-ply local communities with a range of goods, including hunting,fishing, forest and forage resources (e.g. Roberts, 1988; Scoones,1991; Adams, 1993) and they are local hot-spots for biodiversity(Chapman et al., 2001). As different inland-valley ecosystem func-tions may conflict with agricultural objectives, and because thereare large area-specific differences in development suitability andrisks, indiscriminate development should be avoided (McCartneyand Houghton-Carr, 2009). Ecosystem functions of inland valleys,such as biodiversity and water buffering, are affected when inlandvalleys are used for agriculture. Where developments are imple-mented without proper impact assessments, they can negativelyaffect local livelihoods and environments (e.g. Whitlow, 1983). In-deed, aligning food production with biodiversity conservation is animportant future challenge for agronomic and environmental re-search (Verhoeven and Setter, 2010). Following the above, the cen-tral aim of this paper is to develop an approach to fulfil the regionalagricultural potential of inland valleys in sub-Saharan Africa (SSA)such that local rural livelihoods are benefited and regional objec-tives of reducing poverty and increasing food safety are met, whilesafeguarding other inland-valley ecosystem services of local andregional importance.

A number of useful frameworks have recently been proposed tocharacterize wetlands for their agricultural and ecological poten-tials in order to make informed decisions on their use (e.g.McCartney and Houghton-Carr, 2009; Kotze, 2011; Sakané et al.,2011). As a step forward compared to earlier methods specificallytargeted to inland valleys, such as the ones proposed by Andriesseand Fresco (1991) and Andriesse et al. (1994) that were primarilybased on biophysical and land use characterizations, these ap-proaches combine biophysical with socio-economic characteris-tics. The next step forward is to integrate these characterizationsin a comprehensive methodology, supported by appropriate tools,that runs from selection of the most suitable inland valley for agri-cultural production to the actual development and eventually tosustainable management practices. Such methodology should alsoprovide guidelines on how to ensure participation of local stake-holders in all these stages. The current paper, focussing specificallyon sustainable realization of the inland-valley potential for rice-based production systems, attempts to do just that, as we believethat for the sustainable development of these ecosystems, the siteselection, land use planning and design, development and resource

Table 1Rice growing ecosystem characterization (water supply, agro-ecological zone and main biophysical production constraints); inland valleys may cover the whole range fromhydromorphic fringes to irrigated lowlands. Sources: Andriesse et al. (1994), Kiepe (2006), Thiombiano et al. (1996), Wopereis et al.(2007).

Ecosystem Upland Hydromorphic fringes Rain-fed lowland Intensified lowland Irrigated lowlands

Main water supply Rainfall Rainfall + water table Rainfall + watertable + unregulatedfloods

Regulated floods Full irrigation

Agro – ecological zone Guinea savannah –humid forest

Guinea savannah – humidforest

Sudan savannah tohumid forest

Sudan savannah tohumid forest

Sahel to humid forest

Main biophysicalproductionconstraints

Drought, Weeds, Pest &Diseases, P and Ndeficiency, Soil erosion,Soil acidity

Drought, Weeds, Pest &Diseases, P and N deficiency,Soil erosion, Soil acidity, Irontoxicity

Drought/flooding,Weeds, Pest & Diseases,P and N deficiency, Irontoxicity

Drought/flooding,Weeds, Pest & Diseases,P and N deficiency, Irontoxicity

Weeds, Pest & Diseases,Salinity/Alkalinity, P andN deficiency, Irontoxicity

J. Rodenburg et al. / Agricultural Systems 123 (2014) 1–11 3

management should follow a participatory, integrated and system-atic approach. We aim to provide a framework for such an ap-proach based on a review of the literature and insights obtainedfrom recent research carried out by the Consortium for the Sustain-able Use of Inland Valley Agro-Ecosystems in sub-Saharan Africa(short: Inland Valley Consortium, IVC) and its convening organiza-tion, the Africa Rice Center (AfricaRice). The IVC, composed oftwelve West-African national agricultural research sytems and anumber of international (AfricaRice, IITA, ILRI, IWMI, FAO, World-fish and CORAF) and advanced research institutes (CIRAD, Wagen-ingen University), was founded in 1993 with the objective todevelop, in concerted and coordinated action, technologies andoperational support systems for the intensified but sustainableuse of inland valleys in sub-Saharan Africa.

2. Current inland valley use

2.1. Drivers for inland valley use

There are no reliable figures about the percentage of the totalinland valley area (190 M ha) currently under rice production insub-Saharan Africa. Andriesse et al. (1994) were only able to pro-vide a rough estimate for this area in West Africa (10–25%) and thisestimate includes inland valleys in peri-urban areas that aremainly used for vegetable production due to proximity of markets(e.g. Erenstein, 2006; Erenstein et al., 2006). The share of inlandvalley area under rice or rice-based production systems in thewhole of Africa, hence including the central, eastern and southernparts, is expected to be much lower. Inland valleys are, however,increasingly used for agricultural production, partly driven by thedrought spells in the 1970s (e.g. Niasse et al., 2004), and followingdeclining soil fertility in the uplands due to unsustainable farmingpractices (Windmeijer and Andriesse, 1993). Valley bottoms andhydromorphic fringes generally have higher water availabilityand higher soil fertility levels compared to upland soils (e.g.Andriesse et al., 1994; van der Heyden and New, 2003), eventhough soil fertility is still often suboptimal to sustain high cropproductivity. Rice yields in rain-fed upland rice systems in SSAare currently around 1 t ha�1 (e.g. Rodenburg and Demont, 2009)and production years should be followed by 3–7 years of fallowto maintain soil fertility and control pests, diseases and weeds(e.g. Becker and Johnson, 2001a). With good management, inlandvalley rice can produce 5–6 t ha�1 without the need for suchunproductive fallow periods required in the uplands (Wakatsukiand Masunaga, 2005).

Global changes have also given a new impetus to inland valleydevelopment. While there are a number of conflicting projectionswith respect to the severity, timing and geographic distributionof future wetting and drying (e.g. Cook and Vizy, 2006; Hoerlinget al., 2006; Biasutti et al., 2008), model forecasts suggest changingand increasingly variable precipitation patterns in Africa resulting

in less rain in the Sahel (Giannini et al., 2008) and more in theequatorial zones (Christensen et al., 2007). A secure harvest froma wetland produced crop becomes of invaluable importance inthe increasingly dry and unreliable agricultural environments(e.g. Scoones, 1991; Sakané et al., 2011). However, because of thesensitivity of inland valley systems to changes in quantity, qualityand frequency of water supply, climate change also poses anhydrological threat to these ecosystems, requiring adaptive man-agement strategies (e.g. Erwin, 2009).

Besides the aforementioned biophysical assets, inland valleydevelopment also has a clear economic driver. About 10 milliontonnes of milled rice, approximately 40% of the annual regionalconsumption, is imported into Africa (mainly from Asian countries)each year, worth about US $5 billion (Seck et al., 2010, 2012). Re-gional production has, however, increased steeply since the early2000s due to a declining availability of global rice stocks for export,and consequently an increase in regional farm-gate prices from anestimated average US $285 per tonne in 1999 to US $564 per tonnein 2009 (Based on available data from 20 rice-producing countriesin sub-Saharan Africa; FAO, 2010). These significant price changeshave encouraged many small-scale farmers to take up rice produc-tion, as reflected in an increased inland valley use (e.g. Sakurai,2006).

2.2. Multi-functional character of inland valleys

Apart from their importance for agriculture, mainly rice andmaize production and horticulture (Sakané et al., 2011), inland val-leys have essential ecosystem functions such as biodiversity con-servation, water storage, local flood and erosion control, nutrientretention and stabilization of the micro-climate (Adams, 1993;Wood et al., 2013). These environments are also used for recreationand tourism and for retrieving clay and sand for crafts and con-struction, and for collection and use of forest, wildlife, fisheriesand forage resources and they contribute to local cultural heritage(Dugan, 1990; Adams, 1993). Inland valleys are important loca-tions for local communities to collect non-agricultural plant re-sources, and rural people generally recognize useful plant speciesand dispose knowledge on their use, abundance and collectionplaces (Rodenburg et al., 2012).

Due to their multifunctional character, inland valleys are attrac-tive for exploitation and therefore vulnerable to degradation. Theeconomic opportunities of inland valleys have been widely recog-nized and investments have indeed been made to make these areasbetter accessible and profitable. Indiscriminate development ofthese vulnerable environments will however lead to degradationof the natural resources they harbor, and thereby jeopardize theirunique and divers ecosystem functions (e.g. Dixon and Wood,2003). The trade-off between conservation of natural resourcesand agricultural land use is particularly critical in African wetlands(e.g. McCartney and Houghton-Carr, 2009; Wood et al., 2013) and

4 J. Rodenburg et al. / Agricultural Systems 123 (2014) 1–11

therefore, any development activities in such ecosystems need tobe planned with care and should only be implemented when par-ticipation of the local users is guaranteed. Until recently, theimportance of wetland functions for local communities have, how-ever, often been ignored in policy planning (Silvius et al., 2000;Wood et al., 2013). Understanding the use and management of eco-system functions by local communities would be the first neces-sary step to generate recommendations for their sustainable use(Rodenburg et al., 2012). Different ecosystem services do not nec-essarily conflict. For instance, agricultural fields can be consideredas important locations to find useful non-cultivated plants too(Rodenburg et al., 2012). Farmers recognize the useful weed spe-cies during weeding and leave them untouched or keep them apartafter uprooting (see references in: Rodenburg and Johnson, 2009)and at field clearing useful species (predominantly trees) are oftenmaintained (e.g. Leach, 1991; Madge, 1995; Kristensen and Lykke,2003). In fact this is a common strategy to cope with declining for-ests (Shepherd, 1992). Other strategies, observed by Rodenburget al. (2012) around inland valleys in Togo and Benin, include theestablishment of a community garden with useful species andthe conservation of a small community forest. These observationsshow that local communities depending on natural resources inand around inland valleys are able to exploit these landscapes syn-ergistically, balancing agricultural production with biodiversityconservation, use and management.

3. Constraints to development and use of inland valleys inAfrica

3.1. Development constraints

Inland-valley utilization efforts are driven by the aforemen-tioned environmental and economic motives but mired by healthand cultural constraints. Traditionally, inland valleys were not of-ten used for agricultural production purposes in Africa (Adams,1993; Verhoeven and Setter, 2010). This is partly because inlandvalley bottoms are difficult to manage and they are also oftenassociated with water-borne diseases such as malaria (e.g. Plasmo-dium falciparum, Plasmodium malariae, and Plasmodium ovale), riverblindness (onchocerciasis; vector: Onchocerca volvulus, source:Wolbachia pipientis), bilharzia (schistosomiasis; Schistosomahaematobium and Schistosoma mansoni) and sleeping sickness(trypanosomiasis; Trypanosoma brucei or Trypanosoma brucei rho-desiense) (e.g. Gbakima, 1994; McMillan et al., 1998; Yapi et al.,2005). And even controlling such diseases is no guarantee for prof-itable (agricultural) exploitation of these environment, due tocounteractive policies (e.g. unfavourable tax regulations, cheapfood imports) and the lack of suitable technologies (McMillanet al., 1998).

Many water management infrastructures built in the 1970shave been abandoned. Such failures are thought to have resultedfrom the lack of local community participation during selection,design and planning of the developments (e.g. Dries, 1991;Maconachie, 2008), or because traditional local land-tenurearrangements were overlooked (Brautigam, 1992). In the 1960sand 1970s many irrigation scheme developments in West Africawere funded by public investment corporations and developmentprojects. The Benin–China Cooperation, for instance, developed atotal of 1400 ha inland valleys and flood plains into medium-sized(25–150 ha) irrigation schemes by equipping them with waterretention structures, irrigation and drainage canals, inlets and out-lets, cofferdams and small bridges. Most of these irrigationschemes are currently under-utilized or abandoned. One of theexceptions is the irrigation scheme of Koussin-Lélé, where farmersgrow rice on 106 ha developed land using gravity irrigation. A

comparison between this scheme and the nearby mostly unutilizedschemes of Bamè and Zonmon (33 and 84 ha respectively) showedthat careful selection of the valley and local stakeholderparticipation in planning, design, implementation and use of thedevelopments are prerequisites for successful development efforts(Djagba et al., 2013).

The socio-economic environment in inland valleys is highly var-iable and complex. Inland valley exploitation is also often compli-cated by unfavourable land tenure arrangements (e.g. Thiombianoet al., 1996; Fu et al., 2010; Oladele et al., 2010) or prohibiting cus-tomary beliefs. Land tenure arrangements vary between locationsand range from ownership by states to ownerships by individuals.The most common ways in which farmers in inland valleys in Afri-ca acquire ownership over land is through inheritance, marriage, orthrough renting, lease or sharecropping (Oladele et al., 2011).When farmers lack stable land ownership the incentive for long-er-term investments is usually low and this in turn is likely to havea negative effect on productivity and sustainable resource manage-ment. Farmers working in the inland valleys do often not possessthe rights over the land and are therefore not always benefitingfrom inland-valley development investments. Land tenurearrangements also often affect gender relations. Land is mostlyowned by men but cultivated by women, in particular when thevalue of the land is low due to, for example, low soil fertility or lackof control over water. Upon development, when the value of theland is raised, men can claim their rights again. Such social con-structions should be considered when inland-valley developmentprojects are designed that aim at benefiting the poor and empow-ering disadvantaged groups like women. In the Pegnasso inlandvalley in south-east Mali, for instance, the French DevelopmentCooperation deliberately opted for a partial rather than a completedevelopment (Abdoulaye Hamadoun, personal communication).The logic was that large investments would increase the land valuewhereby women, using the land prior to the development, wouldrisk losing access, and would not benefit from the project. Themodest improvements (a small water retention structure, one cen-tral inlet and bunded plots) enabled farmers to make better use ofthe available water for a prolonged period of time and thereby in-crease rice productivity. As rice production in this inland valley islocally mainly the responsibility of women the project succeededin its mutual goal to benefit the community while strengtheningthe position of women.

As inland valleys in Africa are socio-economically and biophys-ically diverse and complex (Sakané et al., 2011) the development ofthese landscapes for crop production requires a flexible and carefulapproach (e.g. Andriesse et al., 1994), with actively participatingstakeholders.

3.2. Production constraints

Estimated actual rice yields in inland valleys across Africa(1.4 t ha�1 according to Rodenburg and Demont, 2009) are muchlower than the attainable yield, i.e. the potential yield limited bythe available water and nutrients in a given environment(Rabbinge, 1993), achieved under optimal management conditions(5–6 t ha�1 according to Wakatsuki and Masunaga, 2005). Based ona survey among rice scientists in eight countries in West Africaweed competition, poor soil fertility and diseases, were classifiedas the three most important biophysical production constraints,responsible for these low actual yields in inland valleys(Thiombiano et al., 1996; Table 1).

Dominant weeds in inland valley rice are grasses like Echino-chloa spp. and Oryza spp. (wild rice), sedges such as Cyperus spp.and a variety of broad-leaved weeds like Sphenoclea zeylanica,Ludwigia spp. and Heteranthera callifolia (Rodenburg and Johnson,2009). Another emerging problem in inland valleys across Africa,

J. Rodenburg et al. / Agricultural Systems 123 (2014) 1–11 5

in particular the ones with poor water control, is the parasitic weedRhamphicarpa fistulosa (Rodenburg et al., 2010) causing crop yieldlosses in infested farmers’ fields to exceed 60% (Rodenburg et al.,2011). Other important biotic production constraints in inland val-leys are diseases such as Rice Yellow Mottle Virus (RYMV), leaf blast,bacterial leaf blight and brownspot. RYMV, endemic to Africa, istransmitted by beetles (order Coleoptera, family Chrysomelidae)and can lead to total yield losses ranging from 5% to 100% inrain-fed lowland rice in Africa (Kouassi et al., 2005). Pests suchas insects (e.g. African rice gall midge, stem borers and rice bugs)and rodents and birds can also cause significant yield losses(Balasubramanian et al., 2007). African rice gall midge, commonin both West and East Africa, damages rice tillers causing up to65% yield loss (Nacro et al., 1996). It should be emphasized thatnone of these pests and diseases is restricted to inland valleysalone. If these ecosystems are to be put under production, how-ever, one should find effective ways to deal with them.

Low soil fertility is a general production constraint in inland val-leys despite enrichment caused by soil deposition of silt and fineclay factions through runoff from the surrounding uplands (Ogbanand Babalola, 2003) increasing the exchangeable bases (Kyuma,1985), calcium and magnesium (Fagbami et al., 1985) and phospho-rus contents (Ogban and Babalola, 2003). Soil fertility in these envi-ronments is often far from optimal for sustainable and profitablecrop production. While soil fertility varies across agro-ecologicalzones (Issaka et al., 1997), studies on soil fertility in inland valleysacross West Africa revealed low to very low levels of nitrogen, avail-able phosphorus, pH, CEC and total carbon (Issaka et al., 1996), defi-ciencies in micro-nutrients like sulfur and zinc (Buri et al., 2000)and poor clay mineralogy (Abe et al., 2006). A commonly associatedproblem with low soil fertility in inland valleys is iron toxicity(Becker and Asch, 2005; Audebert and Fofana, 2009). This is a com-plex nutrient disorder caused by excessive iron in the soil solutionunder the specific but typical water-logged conditions of rain-fedand irrigated lowlands, in particular inland valleys (Narteh andSahrawat, 1999). Direct and indirect effects of iron toxicity can leadto 40–45% rice yield reductions in lowlands but this can be miti-gated by effective water and soil fertility management and byselecting tolerant cultivars (Audebert and Fofana, 2009).

Lack of inputs, credits, water control and labor, were the mostimportant institutional and socio-economic constraints mentionedin the survey of Thiombiano et al. (1996). These constraints are allinter-related and also closely related to the main biophysical con-straints. Most farmers working in inland valleys in Africa are re-source-poor subsistence farmers (Balasubramanian et al., 2007).Such farmers generally have limited financial means and monetarysurpluses (e.g. Ismaila et al., 2010) and they would need credits topurchase inputs. Indeed, for resource-poor rice farmers, the finan-cial means or level of credits often determines the level of inputs,such as fertilizer (Donovan et al., 1999), necessary to alleviate

Table 2Estimated inland valley area (and percentage) needed to cover the total paddy demand of Aassumed that mean productivity can be increased by at least 1 t ha�1 following improved wRodenburg and Demont (2009), Seck et al. (2010).

Milled rice import in Africa (2010)Conversion rate paddy-milled riceEquivalent paddy quantity of imported rice in Africa (2010)

Paddy production in Africa (2010)Total paddy demand in Africa (2010)Current productivityProductivity (paddy) increase with improved water and weed managementInland valley area needed for total paddy demand in Africa

Total inland valley area in AfricaShare of total inland valley area in Africa needed to cover total paddy demand

the negative effects of biophysical production constraints. Withoutaccess to credits, inputs are difficult to obtain and labor inputs willhave to increase to avoid crop losses. This labor trade-off is mostobvious in weed control. Weeding is by far the most time-consum-ing practice in rice cultivation in inland valleys and the time can besignificantly reduced when farmers have access to herbicides (e.g.Lawrence and Dijkman, 1997) or, presumably, when they havecontrol over water, as flooding is one of the most effective weedcontrol practices (e.g. Rodenburg and Johnson, 2009).

4. Towards sustainable inland valley development andexploitation

While regional food security is an important goal, and agricul-tural production from already exploited as well as new inland val-leys can contribute to this, the selection, development andexploitation of these environments should be approached withcare. Not all inland valleys are necessarily suitable for crop produc-tion (e.g. Kotze, 2011; Sakané et al., 2011). Crop production derivedfrom the valleys that are suitable, would be enough to contributesignificantly to the region’s food security, in particular when waysare employed to increase productivity and resource-use efficiency.The remainder of the valleys should be safeguarded for non-agri-cultural ecosystem services such as pastoralism, biodiversity andwildlife sanctuaries and natural (excess) water buffers. To illus-trate this, if the regional average rice yields in inland valleys couldbe raised by only 1 t ha�1 through improved water and weed man-agement (as shown by Becker and Johnson, 2001b), the current to-tal continental rice production and imports could hypothetically becovered entirely by production from only 9.1% of the estimated to-tal land area of these landscapes in Africa (Table 2), and hencemore than 90% could be conserved. Future increases in rice de-mand should be accounted for by a further increase in productivity.In reality significant rice production is also derived from other ricegrowing environments, such as uplands and river flood plains(Balasubramanian et al., 2007; Seck et al., 2012), hence the 90%of inland valleys that could be saved for other purposes, can beconsidered as a lower limit.

Indeed, such a strategy requires systematic approaches andmethodologies for (1) characterizing and selecting the ‘best bet’ –most suitable and least-risk – inland valleys for agricultural devel-opment to avoid investment failures or unnecessary destruction ofwetlands, (2) participatory multi-stakeholder land use planningwithin the inland valley, (3) designing, implementing and evaluat-ing the ‘best-fit’ – locally adapted and communally managed –water management development infrastructure and (4) optimizing– and locally adjusting – crop management practices for increased(rice) productivity. Proposed tools for each step are summarized inTable 3.

frica, calculated as a sum of total rice import and production (based on 2010 figures),ater and weed management. Sources: Becker and Johnson (2001b), FAO (2003, 2012),

Estimated figure Calculation

10 M t0.616.7 M t 10� 1

0:6

24.7 M t41.4 M t 16.7 + 24.71.4 t ha�1

2.4 t ha�1

17.25 M ha 41:42:4

190 M ha9.1% 100� 17:25

190

Table 3Essential steps and tools for selection, planning and implementation of sustainable inland valley development.

Step Proposed tools Purpose Key references

1. Selecting suitable inlandvalleys

Remote sensing/GIS analysis/spatialmodelling

To identify inland valleys and assess development potentialat national level

Thenkabail and Nolte (1996),Thenkabail et al. (2000)

Random Forest method To assess agricultural potential Laborte et al. (2012)Detailed agro-ecological survey To determine the suitability for agricultural development Andriesse et al. (1994)Working Wetland Potential index(WWP)

To determine the suitability for agricultural development McCartney and Houghton-Carr(2009)

2. Participatory land useplanning

Multi-criteria decision making(MCDM) such as the WWP

To collect decision criteria from key stakeholders Raj (1995), McCartney andHoughton-Carr (2009)

Multi-stakeholder platform (MSP) To reach a workable compromise between stakeholders’interests

Warner (2006)

3. Designing, implementingand evaluating best-fitwater management

Rapid pre-development diagnostics(DIARPA)

To assess inland valley water dynamics to select the best-fitintervention

Lidon et al. (1998)

Participatory Learning and ActionResearch for Integrated RiceManagement (PLAR-IRM)

To guide farmer participatory inland valley developmentfor improved water management to acquire ownership overthe development structures

Wopereis et al. (2007)

Simple, Participatory Inland ValleyDevelopment Approach

To guide farmer participatory inland valley developmentfor improved water management

Worou (2013)

Irrigation performance assessment To monitor and improve productivity and sustainability ofirrigations systems

Dembélé et al. (2012)

4. Optimizing cropmanagement practices

Sawah system development (SSD) To improve water, nutrient and weed management to raisecrop productivity

Abe and Wakatsuki (2011)

Participatory Learning and ActionResearch for Integrated RiceManagement (PLAR-IRM)

To create farmers’ awareness, identify local productionconstraints and locally available solutions to solve them

Wopereis et al. (2007)

Modelling (e.g. ORYZA, EPIC, farmhousehold models)

To prioritize production constraints, analyze managementoptions and guide decision making

Lansigan et al. (1997), Boling et al.(2007), Laborte et al. (2009),Worou et al. (2012)

6 J. Rodenburg et al. / Agricultural Systems 123 (2014) 1–11

4.1. Selecting suitable inland valleys

The selection and characterization of suitable inland valleys isof vital importance for successful agricultural development inter-ventions. This accounts both for new, unexploited valleys as wellas for valleys that are already (partly) used for crop production.Suitability depends on a wide range of socio-economic and bio-physical factors that require investigation before proceeding to fol-lowing steps. Pioneering work on assessing inland valley systemsand their potential for development has been conducted byAndriesse et al. (1994) who proposed a comprehensive agro-ecological characterization on four levels: ‘macro’ (1:1,000,000–1:5,000,000), ‘reconnaissance’ (1:100,000–1:250,000), ‘semi-detailed’ (1:25,000–1:50,000) and ‘detailed’ (1:5,000–10,000). Atthe first level (macro), the major agro-ecological zones are distin-guished based on the length of the growing period. They are thensub-divided into agro-ecological units and sub-units at the ‘recon-naissance’ level. This subdivision is based on information on lithol-ogy, hydrology, soils and climate using Geographic InformationSystem (GIS) tools (e.g. Narteh et al., 2007) and land use statisticsretrieved from national sources and rapid rural appraisals. Thismethodology was further elaborated by Thenkabail et al. (2000)who used various spatial data sets of soil, cropping seasons, roads,population density, discharge and rainfall data in combinationwith maps on land use and inland valleys classified from satelliteimages. Each of the spatial layers was given a certain weight basedon expert knowledge, and by summing these weights inland val-leys with high potentials were identified. This approach of spatialmodelling was later repeated for a research area in Ghana (Gummaet al., 2009). Experts outlined bio-physical, technical, socio-economic and eco-environmental indicators that affect the poten-tial for development. But data scarcity for the indicators as well asthe debatable justification for the quantification of the weightsmake the methodology inaccurate and subjective. Random Forests(RF) procedures, a classification method based on a decision-tree,may provide an alternative for the aforementioned methods as itdoes not require prior knowledge or assumptions about the

relationships of, or interactions between, the variables and distri-bution of the data (Breiman, 2001). Random Forests procedureswere successfully applied to assess the potential for paddy rice cul-tivation in Laos using predictors on topography, climate, accessibil-ity and demography and poverty (Laborte et al., 2012).

Remote sensing or remote-sensing derived products have beenused to map inland valleys. Simple image classification has beenused in Benin (Thenkabail and Nolte, 1996), Ghana (Gummaet al., 2009) and Cote d’Ivoire (Thenkabail et al. 2000) to classifyimages with good results. Thenkabail and Nolte (1996), Gummaet al. (2009), and Chabi et al. (2010), identified inland valleys usingthe normalized difference vegetation index (NDVI) determinedfrom satellite images, and a slope map generated by GIS softwareusing a digital elevation map. Cloudy conditions are, however, pre-valent in most regions and this inhibits the implementation of suchmethodologies on national or regional scales. Recently, AfricaRicedeveloped an automated mapping procedure based only on infor-mation from a digital elevation model, which is globally availableat a spatial resolution of 30 m. This standardized methodology iscurrently being implemented and validated for the entire West-African region (Zwart and Linsoussi, personal communication).Such spatial modelling tools are helpful in the first necessaryassessment of availability, suitability and locations of inland val-leys and will save project developers or policymakers valuabletime and resources.

However, the application of spatial modelling, using GIS and re-mote sensing, can only provide an indication for the developmentpotential of inland valleys. Soil fertility and soil depth are of greatimportance for this purpose as well, but information on such soilcharacteristics are not widely available in maps with sufficient de-tail and cannot be derived with remote sensing techniques. Along-side biophysical and agronomic assessments, socio-economicvariables such as availability of markets, extension services or so-cial customs are important to assess which valleys can be devel-oped for agriculture (e.g. Narteh et al., 2007) and many of thesecan simply not be mapped and must therefore be assessed usingterrain surveys and feasibility studies.

J. Rodenburg et al. / Agricultural Systems 123 (2014) 1–11 7

The Working Wetland Potential (WWP) index (McCartney andHoughton-Carr, 2009) is a comprehensive assessment tool thatcan be implemented at the valley scale to assess its suitabilityfor agricultural development. It consists of assessments on: (1)the ecological potential, (2) the social and economic importance,(3) the agricultural suitability and (4) the environmental and so-cio-economic risks involved in the actual development. The prod-uct of ‘suitability’ and ‘risk’ scores based on these assessmentsresults in the WWP index which provides an indication of the agri-cultural use-potential of an ecosystem. The ‘best-bet’ inland valleysfor rice production should score high on (agricultural) productionand marketing potential and low on environmental and socialrisks, and preferably also low on other ecosystem functions suchas biodiversity. Hence a thorough assessment of the valley’s eco-nomic value for local communities, including direct, indirect andnon-use benefits is required (Scoones, 1991) as well as an environ-mental systems analysis of upstream – downstream impacts fromdevelopments.

In conclusion, a two-step approach is advocated. First, spatialmodelling and analysis at national or sub-national level shouldprovide the location of the inland valleys and a first indication ofthe potential for development. Second, a detailed feasibility study(we propose the aforementioned Working Wetland Potential– WWP – index) must be implemented using field and farmersurveys to assess the true potential for development of the selectedinland valleys.

4.2. Participatory land use planning

After site selection, an inventory of land uses should be fol-lowed by the actual land use planning within a valley. For this stepto be successful, active participation of all important stakeholdersis required (e.g. Perfecto and Vandermeer, 2008; Del Amo-Rodriguez et al., 2010). Here, multi-criteria decision making(MCDM) methods could be practised in recognition of the insightthat stakeholders are likely to base decisions on more than justone criterion (e.g. Raj, 1995). The earlier mentioned WWP index,which principally is an MCDM method, can again be used to eval-uate the potential of different agricultural activities within the val-ley and to find synergies, or at least compromises, betweendifferent ecosystem services (McCartney and Houghton-Carr,2009). Preferably all local stakeholders, including local authoritiesand politicians, should play a role in the land use planning within aselected inland valley and find consensus on the directions to take.To facilitate this process so-called multi-stakeholder platforms(MSP) could be created. This is perhaps the most challenging partof the approach, as getting all stakeholders around the table hasproven difficult in some situations (Warner, 2006) and reachingconsensus among a wide ranging group of stakeholders with differ-ent interests, might be another considerable hurdle to take. Theoutcome such MSP processes should aim for would be to reach acompromise between economic, social and environmental gainsand risks and the identification of hot-spots for specific ecosystemservices within the inland valley under consideration. The make-up of such a compromise would of course depend on the specificvalues local stakeholders place on the services that the ecosystemprovide them with, relative to alternative income sources andinterest (McCartney and Houghton-Carr, 2009). The MSP processshould lead to a detailed strategic plan for the local land use of aspecific inland valley, with areas designated for agricultural pro-duction and areas that are maintained or managed to fulfil otherecosystem functions. This should result in the sort of spatiallyand temporally mixed, land use forms that were earlier proposedby Dixon and Wood (2003) and assumed to be inherently environ-mentally and socially sustainable.

4.3. Designing, implementing and evaluating best-fit watermanagement

After these first two, rather difficult and time-consuming, stepsof inland-valley selection and participatory planning, the actualdevelopment, mainly consisting of clearing of vegetation and con-struction of irrigation and drainage structures to increase watercontrol, can start. As much as possible this work should involvethe future beneficiaries of the development to enable them to ac-quire ownership. Information on the extent of participation inthe development work provided by an individual stakeholder canbe used to guide plot partitioning once the inland-valley develop-ment is finalized. Personal time investments at these stages willalso ensure user-commitment to future management and mainte-nance of the infrastructure and thereby benefit the sustainability ofthe inland-valley production system. In the Blétou valley in south-west Burkina Faso, for instance, an IVC project installed a contour-bund system consisting of small water retention bunds along thecontour lines. A community-participatory development approachwas used, whereby plots were distributed among farmers,according to their participation in the construction of thesewater-management structures (Youssouf Dembélé, personal com-munication). The active participation of farmers in the constructionreduced the costs of the investment and, more importantly, pro-vided them with ownership. The plots were assigned in a participa-tory manner following each individual’s contribution to the work.This resulted in fair distribution of the plots based on group con-sensus, respecting individual time investments and disregardinggender or age.

Clearly, every inland valley is unique in terms of biophysicaland socio-economic characteristics and there is no ‘off-the-shelf’technology with a broad and indiscriminate application range.Technologies for inland valleys must be locally adapted. For in-stance, the regulation of water – e.g. to control flooding, optimizeirrigation, conserve water for late-season use – requires a designthat takes local conditions into account. Various physical factors,such as the size of the catchment area, the valley morphologyand soil texture (determining hydrological behavior) need to beconsidered for the design of the most suitable water managementsystem. This requires a thorough diagnostic study on the spatialand dynamic water movements within the valley (Wopereiset al., 2007) followed by in-depth discussions with stakeholders(preferably through a multi-stakeholder platform) to ensure fullbuy-in of the community and to make sure that land tenure issuesare identified and solved beforehand. Inviting local politicians andvillages chiefs in such multi-stakeholder platforms is imperative inthis respect as they often have the responsibility or power to allo-cate land to land users. Discussions on the land and water develop-ment options that could be put in place should also considerconsequences for water availability for downstream users.

The first steps towards improved water management in inlandvalleys in Africa will be the construction of main and secondarydrainage pathways and outlining, bunding and levelling of individ-ual fields with minimal soil movement. A valley with a slight slopewill result in fairly large bunded fields, whereas a valley with asteep slope will lead to more ‘terraced’, bunded fields. The intro-duction of such very simple water management structures will al-ready lead to substantial yield gains (1–2 t ha�1), especially ifaccompanied with good crop management practices (e.g. Beckerand Johnson, 2001b; Toure et al., 2009). For this type of partialwater control structures, that can be developed entirely by farmersthemselves, Worou (2013) developed a useful guide.

To raise rice productivity with improved water control, relevantmodules of the curriculum for Participatory Learning and ActionResearch (PLAR) for Integrated Rice Management (IRM) can beimplemented. PLAR-IRM was developed by AfricaRice and

8 J. Rodenburg et al. / Agricultural Systems 123 (2014) 1–11

partners, based on the insight that a locally adapted and integratedapproach is required to increase rice productivity in inland-valleyproduction systems in Africa (Wopereis and Defoer, 2007). It isessentially a farmer participatory, step-wise approach to put in-land valleys under rice production using good agricultural prac-tices (Defoer et al., 2004; Wopereis et al., 2007). Farmersinvolved in the Japanese funded SMART-IV project implementedby AfricaRice and partners in Togo and Benin obtain very good re-sults in inland valley settings, introducing relatively simple, low-cost water management structures (drainage canal development,bunding, levelling) that can be constructed and maintained totallyby farmers themselves. Use of power tillers, while not essential atthe first stage, can substantially speed up land development oncefarmers are familiar with the technique.

If full water control is targeted, five main water managementsystems for inland valleys can be envisaged: (1) the traditionalrandom-basin system, (2) the central-drain system, (3) the inter-ceptor-canal system, (4) the head-bund system and (5) the con-tour-bund system (for technical details see: Oosterbaan et al.,1987; Windmeijer and Andriesse, 1993; Windmeijer et al., 2002).Lidon et al. (1998) developed a diagnostic interactive tool calledDIARPA (‘diagnostic rapide de pré-aménagement’; Eng.: rapid pre-development diagnostics) which works as a decision tree and helpsto assess the ‘best-fit’ type of intervention at a given location and agiven level of investment, with the purpose to optimize agricul-tural production with limited hydraulic risks. While the technical-ities of the choice of the water-control design and parts of theactual implementation of the development may be beyond the re-source-poor farmers’ capacities and expertise, we again stress theimportance of involvement of farmers in the activities that canbe carried out by them to gain the aforementioned necessary ‘own-ership’ over the water-control structures. Moreover, the properfunctioning, management and maintenance of the water-controlstructures requires the actual users (i.e. the farmers) to understandthe basic principles. After completion of the water managementstructures, a performance evaluation, similar to the one suggestedby Dembélé et al. (2012), should be carried out on a seasonal basisto enable farmers to make necessary adjustments and thereby fur-ther improve water productivity. This requires farmer training andfacilitation.

4.4. Optimizing productivity and profitability

Following the selection of the inland valley, the land use plan-ning within the selected inland valley and the development of thatpart of the inland valley designated for agricultural production, thelast step is to establish and optimize crop production within thesedesignated areas. At this stage, a set of local constraints need to betackled in order to benefit from the inherent inland-valley produc-tion potential. Simulations with physiological crop models (e.g.ORYZA), and multi-criteria models (e.g. EPIC, farm household mod-els) can be used to analyze productivity and sustainability of crop-ping systems, or to quantify effects of different stresses on cropyield, and as such they can provide a decision support tool to im-prove management on the crop, farm or inland valley level (e.g.Lansigan et al., 1997; Boling et al., 2007; Laborte et al., 2009;Worou et al., 2012).

Following water management, key factors for raising produc-tivity in inland valleys are weed and soil fertility management(Wopereis and Defoer, 2007) and pest and disease control(Table 1). However the order of importance of production con-straints needs to be locally assessed for each inland valley. Thedata collected during the aforementioned detailed characteriza-tion should be helpful in this respect. Such characterizations inturn can guide the selection of technology interventions (e.g.Sakané et al., 2011). The aforementioned PLAR-IRM curriculum

also provides a very useful method to identify key productionconstraints, as well as locally researchable issues. PLAR-IRM fur-ther stimulates farmer experimentation in order to test ‘whatworks best’ under the given local (biophysical and socio-eco-nomic) conditions, using an integrated management approach.The available modules of the PLAR curriculum provide guidelinesfor such approaches from which facilitators and farmers can tapto improve the local productivity. Through integrated water, soilfertility and weed management in inland-valleys, rice yields canincrease considerably. Bunding, puddling (if possible) and level-ling for instance, facilitates water management and decreasesweed competition – as many weed species are not well adaptedto permanently flooded conditions (e.g. Kent and Johnson, 2001)– and generally increases nutrient use efficiencies. These rela-tively simple technologies have shown to increase rice yields by40%, and reduce weed infestation by 25% across agro-ecologicalzones (Becker and Johnson, 2001b; Toure et al., 2009). Bunding,levelling and puddling (mainly to improve water management)is also proposed through Sawah system development (Abe andWakatsuki, 2011) a labor-intensive approach which basically pro-motes the aforementioned contour-bund system combined withgood agricultural practices. Further yield improvements can thenbe attained by using improved rice cultivars. For instance, somecultivars of NERICA (New Rice for Africa) adapted to lowland con-ditions have an inherent high weed competitiveness (Rodenburget al., 2009) and a high yield potential (Sié et al., 2008). Followingsuch investments in crop productivity increases, one should alsoaim at reducing post-harvest losses due to birds, insects and ro-dents, mainly by shortening the time between crop maturityand harvest and storage and by improving transport and grainstorage facilities (e.g. Yusuf and He, 2011). Finally, conducivepolicies ensuring good prices for producers on the local market,are imperative for profitable agricultural exploitation of inlandvalleys (e.g. McMillan et al., 1998).

5. Conclusions – regional potentials, local approaches

Tapping the regional potential of inland valleys in sub-SaharanAfrica requires development of high-potential and low-risk areasthat are yet unexploited, as well as improvement of already usedareas, through development of water management structures andthe use of improved crop management technologies. We proposea step-wise and locally-adaptable stakeholder-participatory ap-proach for site-selection, land-use planning, water managementdesign and implementation and crop management, to realize this,while maintaining other important ecosystem functions of theselandscapes. A number of approaches, tools and technologies havebeen developed over the past three decades that contribute toachieving this. We propose: (1) the selection of ‘best-bet’ inlandvalleys, whether unexploited or already used, based on spatialmodelling and analysis at national or sub-national level (usingGIS and remote sensing tools) followed by more detailed typolo-gies using the Working Wetland Potential (WWP) index, (2) astakeholder-participatory land use planning within the inland val-ley based on the earlier characterizations (notably the WWP index)and using multi-stakeholder platforms (MSP), (3) participatory in-land-valley development (e.g. clearing, levelling and constructionof water management structures) following guidelines developedby Worou (2013), relevant modules of the Participatory Learningand Action Research (PLAR) curriculum and the pre-developmentdiagnostic tool DIARPA followed by regular performance assess-ments of the water-control system, and (4) identification of localproduction constraints combining model simulations and farmerparticipatory priority exercises (e.g. PLAR), to select and adaptappropriate management practices and technologies following

J. Rodenburg et al. / Agricultural Systems 123 (2014) 1–11 9

principles of Integrated Rice Management (IRM). While there issome experience with the first and the last two steps of this ap-proach no published evidence exists yet showing that the secondstep (e.g. formation of MSPs for detailed strategic planning) indeedresults in the desired sustainable use of inland valleys in Africa. Fu-ture research and development projects, like some of the currentones carried out by AfricaRice and partners, should test and fine-tune such approaches.

We conclude that it is essential to use systematic analyses ap-proaches for the selection of ‘best-bet’ inland valleys for rice pro-duction as only a fraction of the available inland valleys in Africawould need to be used for agricultural production in order to attainregional self-sufficiency in rice. The remaining inland valleys couldthen be safeguarded to fulfil other ecosystem services. However,for this strategy to be effective, an environmental impact assess-ment should be compulsory before any development takes place.Conservation regulations and monitoring and evaluation mecha-nisms need to be established to help protecting those inland val-leys that are either too vulnerable (e.g. to soil and waterdegradation or social conflicts) or too valuable (because of otherecosystem functions such as biodiversity) for agricultural develop-ment. Selection of ‘best-bet’ production valleys should be based onboth biophysical and socio-economic criteria and be broadly sup-ported by the local communities depending on them. The same ap-proach is proposed for the identification of locations within theinland valley that should be used for crop production and thosethat should continue to fulfil any other ecosystem function. Thisagain requires involvement of local stakeholders. Following thesesteps, the actual development is the next challenge. The rightchoice of water-management system is of pivotal importance andthis depends largely on the valley morphology and the local soiland hydrological characteristics. Development and implementa-tion of such water management systems and the agricultural pro-duction practices following such development should notnegatively impact the water quality and availability downstream.For the actual crop production, high-yielding and stress-resilientlowland rice cultivars and locally adapted and integrated cropmanagement practices are required. Harvesting technologies andpost-harvest facilities, for drying, threshing, milling, storage andtransport should also be included in inland-valley developmentplans.

For the sustainable realization of the regional potentials offeredby inland valleys in Africa, full local stakeholder participation is re-quired in all stages, ranging from decision-making to developmentand implementation. This should result in consensus on the selec-tion and land use plans of inland valleys and the implementation ofbroadly supported interventions and flexible, locally adaptable,and cultural and socio-economical acceptable solutions to thenumerous constraints encountered in inland valleys in this region.

Acknowledgements

The insights herewith presented have resulted from work of theInland Valley Consortium (IVC) and the Africa Rice Center (Africa-Rice) and partners. We are indebted to all those who have contrib-uted to this work over the past decades. This paper is dedicated toour respected colleague Youssouf Dembélé, who passed away sountimely and unexpectedly. For many years he contributed invalu-ably to research and development of inland valleys in West Africa.

References

Abe, S.S., Wakatsuki, T., 2011. Sawah ecotechnology – a trigger for a rice greenrevolution in Sub-Saharan Africa Basic concept and policy implications. OutlookAgr. 40, 221–227.

Abe, S.S., Masunaga, T., Yamamoto, S., Honna, T., Wakatsuki, T., 2006.Comprehensive assessment of the clay mineralogical composition of lowlandsoils in West Africa. Soil Sci. Plant Nutr. 52, 479–488.

Acres, B.D., Rains, A.B., King, R.B., Lawton, R.M., Mitchell, A.J.B., Rackham, L.J., 1985.African dambos: their distribution, characteristics and use. Z. Geomorphol., 63–86.

Adams, W.M., 1993. Indigenous use of wetlands and sustainable development inWest-Africa. Geogr. J. 159, 209–218.

Andriesse, W., Fresco, L.O., 1991. A characterization of rice-growing environmentsin West Africa. Agric. Ecosyst. Environ. 33, 377–395.

Andriesse, W., Fresco, L.O., Duivenbooden, N., Windmeijer, P.N., 1994. Multi-scalecharacterization of inland valley agro-ecosystems in West Africa. Neth. J. Agric.Sci. 42, 159–179.

Audebert, A., Fofana, M., 2009. Rice yield gap due to iron toxicity in West Africa. J.Agron. Crop Sci. 195, 66–76.

Balasubramanian, V., Sie, M., Hijmans, R.J., Otsuka, K., 2007. Increasing riceproduction in sub-Saharan Africa: challenges and opportunities. Adv. Agron.94, 55–133.

Becker, M., Asch, F., 2005. Iron toxicity in rice-conditions and managementconcepts. J. Plant Nutr. Soil Sci. 168, 558–573.

Becker, M., Johnson, D.E., 2001a. Cropping intensity effects on upland rice yield andsustainability in West Africa. Nutr. Cycl. Agroecosyst. 59, 107–117.

Becker, M., Johnson, D.E., 2001b. Improved water control and crop managementeffects on lowland rice productivity in West Africa. Nutr. Cycl. Agroecosyst. 59,119–127.

Biasutti, M., Held, I.M., Sobel, A.H., Giannini, A., 2008. SST forcings and Sahel rainfallvariability in simulations of the twentieth and twenty-first centuries. J. Climate21, 3471–3486.

Boling, A.A., Bouman, B.A.M., Tuong, T.P., Murty, M.V.R., Jatmiko, S.Y., 2007.Modelling the effect of groundwater depth on yield-increasing interventionsin rainfed lowland rice in Central Java, Indonesia. Agric. Syst. 92, 115–139.

Brautigam, D., 1992. Land rights and agricultural development in West Africa – acase study of 2 Chinese projects. J. Dev. Areas 27, 21–32.

Breiman, L., 2001. Random forests. Mach. Learn. 45, 5–32.Buri, M.M., Masunaga, T., Wakatsuki, T., 2000. Sulfur and zinc levels as limiting

factors to rice production in West Africa lowlands. Geoderma 94, 23–42.Chabi, A., Oloukoi, J., Mama, V.J., Kiepe, P., 2010. Inventory by remote sensing of

inland valleys agro-ecosystems in central Benin. Cah. Agric. 19, 446–453.Chapman, L.J., Balirwa, J., Bugenyi, F.W.B., Chapman, C., Crisman, T.L., 2001.

Wetlands of East Africa: biodiversity, exploitation and policy perspectives. In:Gopal, B. (Ed.), Biodiversity in wetlands: assessment function and conservation.Blackhuys, Leiden, pp. 101–132.

Christensen, J.H., Hewitson, B., Busuioc, A., Chen, A., Gao, X., Held, I., Jones, R., Kolli,R.K., Kwon, W.T., Laprise, R., Magaa Rueda, V., Mearns, L., Menndez, C.G., Risnen,J., Rinke, A., Sarr, A., Whetton, P., 2007. Regional climate projections. In:Solomon, S., Qin, D., Manning, M., Chen, Z., Marquis, M., Averyt, K.B., Tignor, M.,Miller, H.L. (Eds.), pp. 847–940.

Cook, K.H., Vizy, E.K., 2006. Coupled model simulations of the West Africanmonsoon system: twentieth- and twenty-first-century simulations. J. Climate19, 3681–3703.

de Ridder, N., Stomph, T.J., Fresco, L.O., 1997. Effects of land-use changes on waterand nitrogen flows at the scale of West African inland valleys: a conceptualmodel. In: Teng, P.S., Kropff, M.J., ten Berge, H.F.M., Dent, J.B., Lansigan, F.P., vanLaar, H.H. (Eds.), Applications of systems approaches at the farm and regionallevels: proceedings of the second international symposium on systemsapproaches for agricultural, development. vol. 1, pp. 367–381.

Defoer, T., Wopereis, M.C.S., Idinoba, P.A., Kadisha, T.K.L., Diack, S., Gaye, M., 2004.Curriculum for Participatory Learning and Action Research (PLAR) for IntegratedRice Management (IRM) in Inland Valleys of Sub-Saharan Africa: Facilitators’Manual. WARDA, Cotonou, Benin/IFDC, Muscle Shoals, USA.

Del Amo-Rodriguez, S., Vergara-Tenorio, M.D., Ramos-Prado, J.M., Porter-Bolland, L.,2010. Community landscape planning for rural areas: a model for bioculturalresource management. Soc. Natur. Resour. 23, 436–450.

Dembélé, Y., Yacouba, H., Keita, A., Sally, H., 2012. Assessment of irrigation systemperformance in South-Western Burkina Faso. Irrig. Drain. 61, 306–315.

Dixon, A.B., Wood, A.P., 2003. Wetland cultivation and hydrological management ineastern Africa: matching community and hydrological needs throughsustainable wetland use. Nat. Resour. Forum 27, 117–129.

Djagba, J.F., Rodenburg, J., Zwart, S.J., Houndagba, C.J., Kiepe, P., 2013. Failure andsuccess factors of irrigation system developments in West Africa – a case studyfrom the Ouémé valley in Benin. Irrig. Drain. http://dx.doi.org/10.1002/ird.1794.

Donovan, C., Wopereis, M.C.S., Guindo, D., Nebie, B., 1999. Soil fertility managementin irrigated rice systems in the Sahel and Savanna regions of West Africa. Part II.Profitability and risk analysis. Field Crop. Res. 61, 147–162.

Dries, I., 1991. Development of wetlands in Sierra-Leone – Farmers rationalityopposed to government policy. Landscape Urban Plan. 20, 223–229.

Dugan, P.J., 1990. Wetland Conservation: A Review of Current Issues and Action.IUCN, Gland, Switzerland.

Erenstein, O., 2006. Intensification or extensification? Factors affecting technologyuse in peri-urban lowlands along an agro-ecological gradient in West Africa.Agric. Syst. 90, 132–158.

Erenstein, O., Oswald, A., Mahaman, M., 2006. Determinants of lowland use close tourban markets along an agro-ecological gradient in West Africa. Agr. Ecosyst.Environ. 117, 205–217.

Erwin, K., 2009. Wetlands and global climate change: the role of wetlandrestoration in a changing world. Wetl. Ecol. Manage. 17, 71–84.

10 J. Rodenburg et al. / Agricultural Systems 123 (2014) 1–11

Fagbami, A., Ajayi, S.O., Ali, E.M., 1985. Nutrient distribution in the basement-complex soils of the tropical, dry rainforest of southwestern Nigeria 1.Macronutrients – calcium, magnesium, and potassium. Soil Sci. 139, 431–436.

FAO, 2003. TERRASTAT. <http://www.fao.org/ag/agl/agll/terrastat/> (Date accessed14.02.03).

FAO, 2010. FAOSTAT. <http://faostat.fao.org/> (Date accessed 07.10.10).FAO, 2012. Rice Market Monitor. Food and Agriculture Organization of the United

Nations, Trade and Markets Division, Rome, p. 33.Fu, R.H.Y., Abe, S.S., Wakatsuki, T., Maruyama, M., 2010. Traditional farmer-

managed irrigation system in central Nigeria. Jarq-Jpn. Agric. Res. Quart. 44, 53–60.

Gbakima, A.A., 1994. Inland valley swamp rice development – malaria,schistosomiasis, onchocerciasis in south central Sierra-Leone. Public Health108, 149–157.

Giannini, A., Biasutti, M., Held, I.M., Sobel, A.H., 2008. A global perspective onAfrican climate. Climat. Change 90, 359–383.

Gumma, M., Thenkabail, P.S., Fujii, H., Namara, R., 2009. Spatial models for selectingthe most suitable areas of rice cultivation in the Inland Valley Wetlands ofGhana using remote sensing and geographic information systems. J. Appl.Remote Sens. 3.

Hoerling, M., Hurrell, J., Eischeid, J., Phillips, A., 2006. Detection and attribution oftwentieth-century northern and southern African rainfall change. J. Climate 19,3989–4008.

Ismaila, U., Gana, A.S., Tswanya, N.M., Dogara, D., 2010. Cereals production inNigeria: problems, constraints and opportunities for betterment. Afr. J. Agric.Res. 5, 1341–1350.

Issaka, R.N., Masunaga, T., Kosaki, T., Wakatsuki, T., 1996. Soils of inland valleys ofWest Africa general fertility parameters. Soil Sci. Plant Nutr. 42, 71–80.

Issaka, R.N., Ishida, F., Kubota, D., Wakatsuki, T., 1997. Geographical distribution ofselected soil fertility parameters of inland valleys in West Africa. Geoderma 75,99–116.

Kent, R.J., Johnson, D.E., 2001. Influence of flood depth and duration on growth oflowland rice weeds, Côte d’Ivoire. Crop Prot. 20, 691–694.

Kiepe, P., 2006. Characterization of Three Key Environments for IntegratedIrrigation-Aquaculture and their Local Names, Integrated Irrigation andAquaculture in West Africa, Concepts, Practices and Potential. FAO, Rome, pp.1–5.

Kotze, D.C., 2011. The application of a framework for assessing ecological conditionand sustainability of use to three wetlands in Malawi. Wetl. Ecol. Manage. 19,507–520.

Kouassi, N.K., N’Guessan, P., Albar, L., Fauquet, C.M., Brugidou, C., 2005. Distributionand characterization of Rice yellow mottle virus: a threat to African farmers.Plant Dis. 89, 124–133.

Kristensen, M.K., Lykke, A.M., 2003. Informant-based valuation of use andconservation preferences of savanna trees in Burkina Faso. Econ. Bot. 57, 203–217.

Kyuma, K., 1985. Fundamental Characteristics of Wetland Soils, Wetland Soils:Characterization, Classification and Utilization. International Rice ResearchInstitute (IRRI), Los Banos, Philippines, pp. 193–205.

Laborte, A.G., Schipper, R.A., Van Ittersum, M.K., Van den Berg, M.M., Van Keulen, H.,Prins, A.G., Hossain, M., 2009. Farmers’ welfare, food production and theenvironment: a model-based assessment of the effects of new technologies inthe northern Philippines. Njas-Wageningen J. Life Sci. 56, 345–373.

Laborte, A.G., Maunahan, A.A., Hijmans, R.J., 2012. Opportunities for expandingpaddy rice production in Laos: spatial predictive modeling using RandomForest. J. Land Use Sci. 7, 21–33.

Lansigan, F.P., Pandey, S., Bouman, B.A.M., 1997. Combining crop modelling witheconomic risk-analysis for the evaluation of crop management strategies. FieldCrop. Res. 51, 133–145.

Lawrence, P.R., Dijkman, J.T., 1997. The introduction of animal traction into inlandvalley regions. 2. Dry season cultivation and the use of herbicides in rice. J.Agric. Sci. 129, 71–75.

Lawrence, P.R., Dijkman, J.T., Jansen, H.G.P., 1997. The introduction of animaltraction into inland valley regions. 1. Manual labour and animal traction inthe cultivation of rice and maize: a comparison. J. Agric. Sci. 129,65–70.

Leach, M., 1991. Engendered environments: understanding natural resourcemanagement in the West African forest zone. IDS Bull. 22, 17–24.

Lidon, B., Legoupil, J.C., Blanchet, F., Simpara, M., Sanogo, I., 1998. Le diagnosticrapide de pré-aménagement (DIARPA). Agric. Dévelop. 20, 61–80.

Maconachie, R., 2008. New agricultural frontiers in post-conflict Sierra Leone?Exploring institutional challenges for wetland management in the EasternProvince. J. Mod. Afr. Stud. 46, 235–266.

Madge, C., 1995. Ethnography and agroforestry research: a case study from theGambia. Agroforest. Syst. 32, 127–146.

McCartney, M.P., Houghton-Carr, H.A., 2009. Working Wetland Potential: an indexto guide the sustainable development of African wetlands. Nat. Resour. Forum33, 99–110.

McMillan, D.E., Sanders, J.H., Koenig, D., Akwabi-Ameyaw, K., Painter, T.M., 1998.New land is not enough: agricultural performance of new lands settlement inWest Africa. World Dev. 26, 187–211.

Nacro, S., Heinrichs, E.A., Dakouo, D., 1996. Estimation of rice yield losses due to theAfrican rice gall midge, Orseolia oryzivora Harris and Gagne. Int. J. Pest Manage.42, 331–334.

Narteh, L.T., Sahrawat, K.L., 1999. Influence of flooding on electrochemical andchemical properties of West African soils. Geoderma 87, 179–207.

Narteh, L.T., Moussa, M., Otoo, E., Andah, W.E.I., Asubonteng, K.O., 2007. Evaluatinginland valley agro-ecosystems in Ghana using a multi-scale characterizationapproach. Ghana J. Agric. Sci. 40, 141–157.

Niasse, M., Afounda, A., Amani, A., 2004. Reducing West Africa’s vulnerability toclimate impacts on water resources, wetlands and desertification. Elements fora regional strategy for preparedness and adaptation. The World ConservationUnion (IUCN), Gland, Switzerland and Cambridge, UK.

Ogban, P.I., Babalola, O., 2003. Soil characteristics and constraints to cropproduction in inland valley bottoms in southwestern Nigeria. Agric. Water.Manage. 61, 13–28.

Oladele, O.I., Bam, R.K., Buri, M.M., Wakatsuki, T., 2010. Missing prerequisites forGreen Revolution in Africa: lessons and challenges of Sawah rice eco-technology development and dissemination in Nigeria and Ghana. J. FoodAgric. Environ. 8, 1014–1018.

Oladele, O.I., Kolawole, A., Wakatsuki, T., 2011. Land tenure, investment andadoption of Sawah rice production technology in Nigeria and Ghana: aqualitative approach. Afr. J. Agric. Res. 6, 1519–1524.

Oosterbaan, R.J., Gunneweg, H.A., Huizing, A., 1987. Water Control for RiceCultivation in Small Valleys of West Africa, Annual Report. ILRI, Wageningen,The Netherlands, pp. 30–49.

Perfecto, I., Vandermeer, J., 2008. Biodiversity conservation in tropicalagroecosystems – a new conservation paradigm. Ann. NY Acad. Sci. 1134,173–200.

Rabbinge, R., 1993. The Ecological Background of Food Production. John Wiley &Sons, Chichester, pp. 2–29.

Raj, P.A., 1995. Multicriteria methods in river basin planning – a case study. WaterSci. Technol. 31, 261–272.

Roberts, N., 1988. Dambos in development – management of a fragile ecologicalresource. J. Biogeogr. 15, 141–148.

Rodenburg, J., Demont, M., 2009. Potential of herbicide resistant rice technologiesfor sub-Saharan Africa. AgBioForum 12, 313–325.

Rodenburg, J., Johnson, D.E., 2009. Weed management in rice-based croppingsystems in Africa. Adv. Agron. 103, 149–218.

Rodenburg, J., Stein, A., Noordwijk, M., Ketterings, Q.M., 2003. Spatial variability ofsoil pH and phosphorus in relation to soil run-off following slash-and-burn landclearing in Sumatra, Indonesia. Soil Tillage Res. 71, 1–14.

Rodenburg, J., Saito, K., Kakai, R.G., Toure, A., Mariko, M., Kiepe, P., 2009. Weedcompetitiveness of the lowland rice varieties of NERICA in the southern GuineaSavanna. Field Crop. Res. 114, 411–418.

Rodenburg, J., Riches, C.R., Kayeke, J.M., 2010. Addressing current and futureproblems of parasitic weeds in rice. Crop Prot. 29, 210–221.

Rodenburg, J., Zossou-Kouderin, N., Gbèhounou, G., Ahanchede, A., Touré, A., Kyalo,G., Kiepe, P., 2011. Rhamphicarpa fistulosa, a parasitic weed threatening rain-fedlowland rice production in sub-Saharan Africa – a case study from Benin. CropProt. 30, 1306–1314.

Rodenburg, J., Both, J., Heitkonig, I., van Koppen, K., Sinsin, B., Van Mele, P., Kiepe, P.,2012. Land-use and biodiversity in unprotected landscapes: the case of non-cultivated plant use and management by rural communities in Benin and Togo.Soc. Natur. Resour. 25, 1221–1240.

Saito, K., Azoma, K., Sokei, Y., 2010. Genotypic adaptation of rice to lowlandhydrology in West Africa. Field Crop. Res. 119, 290–298.

Sakané, N., Alvarez, M., Becker, M., Böhme, B., Handa, C., Kamiri, H.W.,Langensiepen, M., Menz, G., Misana, S., Mogha, N.G., Möseler, B.M., Mwita,E.J., Oyieke, H.A., van Wijk, M.T., 2011. Classification, characterisation, and useof small wetlands in East Africa. Wetlands 31, 1103–1116.

Sakurai, T., 2006. Intensification of rainfed lowland rice production in West Africa:present status and potential Green Revolution. Dev. Econ. 44, 232–251.

Scoones, I., 1991. Wetlands in drylands: key resources for agricultural and pastoralproduction in Africa. Ambio 1991 (20), 366–371.

Seck, P.A., Tollens, E., Wopereis, M.C.S., Diagne, A., Bamba, I., 2010. Rising trends andvariability of rice prices: threats and opportunities for sub-Saharan Africa. FoodPolicy 35, 403–411.

Seck, P.A., Diagne, A., Mohanty, S., Wopereis, M.C.S., 2012. Crops that feed the world7: Rice. Food Security 4, 7–24.

Shepherd, G., 1992. Managing Africa’s Tropical Dry Forests: A Review of IndigenousMethods. Overseas Development Institute, London.

Sié, M., Séré, Y., Sanyang, S., Narteh, L.T., Dogbe, S., Coulibaly, M.M., Sido, A., Cissé, F.,Drammeh, E., Ogunbayo, S.A., Zadji, L., Ndri, B., Toulou, B., 2008. Regional yieldevaluation of the interspecific hybrids (O. glaberrima x O. sativa) andintraspecific (O. sativa x O. sativa) lowland rice. Asian J. Plant Sci. 7,30–139.

Silvius, M.J., Oneka, M., Verhagen, A., 2000. Wetlands: lifeline for people at the edge.Phys. Chem. Earth Pt. B 25, 645–652.

Thenkabail, P.S., Nolte, C., 1996. Capabilities of Landsat-5 Thematic Mapper (TM)data in regional mapping and characterization of inland valley agroecosystemsin West Africa. Int. J. Remote Sens. 17, 1505–1538.

Thenkabail, P.S., Nolte, C., Lyon, J.G., 2000. Remote sensing and GIS modelingfor selection of a benchmark research area in the inland valley agroecosystemsof West and Central Africa. Photogramm. Eng. Rem. Sens. 66, 755–768.

Thiombiano, L., Jamin, J.Y., Windmeijer, P.N., 1996. Inland Valley Development inWest-Africa: Regional State of the Art. WARDA, Bouaké, Cote d’Ivoire, p. 69.

Toure, A., Becker, M., Johnson, D.E., Kone, B., Kossou, D.K., Kiepe, P., 2009. Responseof lowland rice to agronomic management under different hydrologicalregimes in an inland valley of Ivory Coast. Field Crop. Res. 114,304–310.

J. Rodenburg et al. / Agricultural Systems 123 (2014) 1–11 11

Turner, R.K., van den Bergh, J.C.J.M., Soderqvist, T., Barendregt, A., van der Straaten,J., Maltby, E., van Ierland, E.C., 2000. Ecological-economic analysis of wetlands:scientific integration for management and policy. Ecol. Econ. 35, 7–23.

van der Heyden, C.J., New, M.G., 2003. The role of a dambo in the hydrology of acatchment and the river network downstream. Hydrol. Earth Syst. Sci. 7, 339–357.

Verhoeven, J.T.A., Setter, T.L., 2010. Agricultural use of wetlands: opportunities andlimitations. Ann. Bot. 105, 155–163.

Wakatsuki, T., Masunaga, T., 2005. Ecological engineering for sustainable foodproduction and the restoration of degraded watersheds in tropics of low pHsoils: focus on West Africa. Soil Sci. Plant Nutr. 51, 629–636.

Warner, J.F., 2006. More sustainable participation? Multi-stakeholder platforms forintegrated catchment management. Int. J. Water Resour. D 22, 15–35.

Whitlow, J.R., 1983. Vlei cultivation in Zimbabwe: reflections on the past.Zimbabwe Agric. J. 80, 123–135.

Windmeijer, P.N., Andriesse, W., 1993. Inland Valleys in West Africa: An Agro-ecological Characterization of Rice Growing Environments. ILRI, Wageningen.

Windmeijer, P.N., Dugué, M.J., Jamin, J.Y., Van de Giesen, N., 2002. Describinghydrological characteristics for inland valley developments. In: Proceedings ofthe Second Scientific Workshop of the Inland Valley Consortium. WARDA/ADRAO, Bouaké, Cote d’Ivoire.

Wood, A., Dixon, A., McCartney, M., 2013. Wetland Management and SustainableLivelihoods in Africa, first ed. Earthscan, Routledge, Abingdon, Oxon, p. 281.

Wopereis, M.C.S., Defoer, T., 2007. Moving methodologies to enhance agriculturalproductivity of rice-based lowland systems in sub-Saharan Africa. In: Bationo,A., Waswa, B., Kihara, J., Kimetu, J. (Eds.), Advances in Integrated Soil FertilityManagement in Sub-Saharan Africa: Challenges and Opportunities. Springer,Dordrecht, The Netherlands, pp. 1077–1091.

Wopereis, M.C.S., Defoer, T., Idinoba, P., Diack, S., Dugué, M.J., 2007. ParticipatoryLearning and Action Research (PLAR) for Integrated Rice Management (IRM) inInland Valleys of sub-Saharan Africa: Technical Manual. WARDA, Cotonou,Benin/IFDC, Muscle Shoals, USA.

Worou, S., 2013. Approche Simple Et Participative d’aménagement de Bas-Fonds.AfricaRice, Cotonou, Benin.

Worou, O.N., Gaiser, T., Saito, K., Goldbach, H., Ewert, F., 2012. Simulation of soilwater dynamics and rice crop growth as affected by bunding and fertilizerapplication in inland valley systems of West Africa. Agr. Ecosyst. Environ. 162,24–35.

Yapi, Y.G., Briet, O.J.T., Diabate, S., Vounatsou, P., Akodo, E., Tanner, M., Teuscher, T.,2005. Rice irrigation and schistosomiasis in savannah and forest areas of Coted’Ivoire. Acta Tropica 93, 201–211.

Yusuf, B.L., He, Y., 2011. Design, development and techniques for controlling grainspost-harvest losses with metal silo for small and medium scale farmers. Afr. J.Biotechnol. 10, 14552–14561.