Search

For the study of field wind erosion, detailed observations of windblown sediment transport in the field are needed. The objective of this study was to determine the best method to quantify the mass of wind-blown material moving past a fixed point during four storms. Twenty-one Modified Wilson and Cooke (MWAC) sediment catchers were installed in a pearl millet [Pennisetum glaucum (L.) R. Br.] field in the Sahelian zone of Niger, on a sandy, siliceous, isohyperthermic Psammentic Paleustalf. Each catcher trapped materials at seven heights between 0.05 and 1.00 m. The vertical profiles of measured horizontal mass fluxes were described by two different models, a three-parameter power function and a live-parameter combined model, which is combination of an exponential function and a power function. For all four storms, both models described accurately the mass fluxes between 0.05 and 0.26 m, but fitted mass fluxes at 0.50, 0.75, and 1.00 m deviated from measured fluxes. Deviations were 21.1, 45.2, and 60.6% for the power function and 12.4, 18.5, and 38.0% for the combined model. Mass transport rates were calculated by integrating the mass flux profiles across height. The differences in calculated mass transport rates were small, but because of the better fit, the combined model was preferred. Correcting for the trapping efficiency of the MWAC catchers (0.49) and multiplying by the storm duration resulted in total mass transport values, which are equal to the mass of soil passing a strip of 1-m width perpendicular to the mean wind direction. The average mass transport values were 102.7, 15.5, 31.8, and 149.8 kg m -1 , respectively, for the four storms.

In northern Burkina Faso (West Africa), a study was undertaken to explore the possibilities of restoring the infiltration capacity of crusted soils through the stimulation of termite activity. Treatments consisted of the application of a mulch of a mixture of wood and straw without insecticides (resulting in “termite plots”) and the application of the same mulch and an insecticide (Dursban with chloropyrifos as the active ingredient) to prevent termite activity (resulting in “non-termite plots”). Three rainfall simulations of 60 minutes duration with an intensity of 50 mm/h at an interval of 24 hours between the first and the second and 72 hours between the second and third simulation were applied, to study the effect of consecutive showers on termite-modified soil characteristics. Cumulative infiltration amounts, final infiltration rates, soil water content and porosity were larger and bulk density was smaller on termite plots as compared to non-termite plots. This suggests that termites may be an important agent in soil-crust control and in the improvement of soil physical properties in Sahelian ecosystems.

The contribution of Cassia siamea hedgerows and mulch to erosion control was evaluated on a 14% slope of a Lixisol/Alfisol at Machakos, Kenya. The four treatments, in 400 m 2 runoff-plots were: hedgerows with prunings applied as mulch to the crop, hedgerows and crops with prunings removed, mulch only applied to the crop and a control. The hedgerows were planted on the contour, 4 m apart and 0.25 m between plants. Maize and cowpea were planted in sequence. The control plot sustained an average annual water loss over 3 y of 31 mm runoff and soil loss of 19 t ha −1 . The best treatment, hedgerows with mulch, reduced losses to 13% and 2% of the control. Hedgerows without mulch reduced losses respectively to 23% and 7%, while mulch without hedgerows reduced losses to 41% and 17%. Soil loss was considerably influenced by one single storm in April 1990 due to the nature of that storm and to the susceptibility of the soil to erosion at that particular time. Differences in crop yield between treatments were small. The hedgerow treatments depressed cowpea yield slightly in less than normal rainy seasons but improved cowpea yields in wet seasons.

This paper deals with the effect of rainfall on the process of wind erosion of beach sands and presents results from both field and wind tunnel experiments. Although sediment transport by splash is of secondary importance on coastal dunes, splash-saltation processes can move sediments in conditions where no motion is predicted by aeolian processes. The effect of raindrop impact on the movement of soil particles by wind was measured on a sand beach plain using an acoustic sediment sampler. In general, an increase of particle movement by wind at the sensor heights was observed during rainfall. Rainfall also affected the wind erosion process during and after rain by changing the cohesive conditions of the surface. The influence of the surface moisture content on the initiation of wind erosion and on the vertical distribution of transported sand particles was studied in a wind tunnel. Moisture significantly increased threshold wind velocities for the initiation of sediment transport and modified vertical sediment profiles.

This paper sets out that the dynamics of the Sahelian rangeland vegetation can be interpreted as a cusp catastrophe and that this interpretation offers a promising basis for the description and analysis of tbis ecosystem. Firstly, an existing scheme of the dynamics of Sahelian herbaceous vegetation is translated into the state-and-transition formulation. Secondly, the application of the cusp catastrophe is explored by studying the behaviour of the Sahelian rangeland ecosystem under changing effective rainfall and grazing inteusity, using the transitions from the state-and-transition formulation as vectors along the cusp manifold. This conceptual cusp catastrophe model subsequently results in the identification of hypotheses and the detection of 5 catastrophic properties of this ecosystem (biiodality, inaccessibiity, sudden jumps, divergence and hysteresis) that have important management implications. The continuous and the discontinuous processes occurring in the Sahelian rangeland ecosystem can both be captured in a tified conceptual model by applying the cusp catastrophe theory. Testing the hypotheses generated by the conceptual model and searching for additional catastrophic properties, such as divergence of linear response and critical slowing down, is a useful direction for future research

Hedgerows of Cassia siamea were planted 4 m apart in a maize/cowpea rotation on a 14% slope of a Lixisol/Alfisol in 1988 at Machakos, Kenya. Infiltration rates were measured in situ across hedgerow barriers with a drip infiltrometer. During the dry season the steady infiltration rate under the hedgerows was 135 mm h −1 while on the alleys in between it was 41–49 mm h −1 and on the control plot 39–48 mm h −1 . Corresponding values for the wet season were 69 mm h −1 for the hedge and 8–11 mm h −1 for the alley. Removal of the hedge plants revealed that much of the increase is due to significantly more macropores in the topsoil beneath the hedgerow than in the topsoil beneath the alley. Differences in pore size distribution in the subsoil were small and not significant. Total increase in average infiltration rate of the hedgerow barrier system in the dry season is 30% and in the wet season is 94% as compared to the alley. The combined effect of above- and below-ground changes makes hedgerow barriers valuable for runoff control.

A method to calculate the statistical distribution of a scaling factor from soil-water retention curves consists of measuring the particle-size distribution by laser scattering, computing soil water retention curves using the Arya and Paris model, and calculating a scale factor for each function. The method was tested by comparing measured and predicted soil-water retention curves. The advantage of this method is in its simplicity and ease of use. The applicability of the method to a given soil texture depends on the verification of the calculated soil-water retention functions.

In a recent article (Diemont et al., 1991) about erosion on Java, it has been postulated that low inputs, not surface erosion, is the main cause of low productivity of upland food crops on this island. In this article it is argued that this hypothesis is too simple. An analysis of empirical field data about the effect of erosion on crop productivity indicates that the effects of erosion and agronomic inputs are confounded. Erosion causes farmers to apply low inputs because only limited returns can be expected on eroded lands. The relations between erosion and crop productivity are site specific; the extent and quality of conservation measures and intensity of agronomic inputs varies per soil type.

As hardly any data can be found on the water requirement and yield response to water of Aloe vera, an irrigation experiment on transplanted aloe poles was conducted on the Caribbean island of Aruba. Compared to non-irrigated treatments irrigation shows a positive influence on growth, number of developed leaves, raw gel figure and gel production per plant. The amount of irrigation water to apply can be best calculated based on crop coefficients (Kc) between 0.20 and 0.30. Treatment based on a Kc value higher than 0.30 did not further increase production. A lysimeter study conducted over a relatively short period revealed that the crop coefficient for Aloe vera varied between 0.25 and 0.40 which is characteristic of the so-called agave/pineapple group.

The infiltration curve is an important characteristic in research on soil erosion in particular on surface erosion. A high infiltration rate, which prevents surface flow, reduces or can prevent soil erosion. We compared infiltration rates measured with a Double Ring Infiltrometer (infiltrometer), a Portable Rainfall Simulator (rainulator) and a Constant Head Permeameter (permeameter). The experiments were conducted in the EEC-sponsored Erosion Research Project in Owerri, South Eastern Nigeria. Based on our measurements we conclude that for research on soil erosion the rainulator is the most suitable method of the three, in terms of realistic results. The infiltrometer yielded too high infiltration rates. The permeameter gives a spatial average measurement and has the advantage that no water is required in the field, and one person can take the samples, which makes the method practical in remote regions.

This paper deals with the effect of the Thomson weir and weirbox on measured recession curves. Small discharges can be measured accurately with a Thomson weir. The variation of the flow rate entering the weirbox changes the storage in the box, which thus gives a time lag in the measured discharge. Three cases of inflow were investigated: zero inflow, exponential recession inflow, and linear recession inflow, as predicted by the Zarmi model. Exponential and linear recession were described by differential equations, solved numerically through a predictor-corrector method. In most cases the time lag is less than 1 min but at very low discharges the time lag can become important. The results presented indicate that the modifying effects of the weirboxes on the outflows from these catchments are not sufficient to explain the differences between the observed exponential recessions and the linear recessions predicted by the Zarmi model.

The irrigation schemes of the villages Gande and Gallade in the Bakel Department in Senegal are managed by a farmer association and an agency. The agency had to install, repair and maintain motorpumps, assist in the lay-out of canals, structures and parcels, and to provide fuel and other inputs on credit. Associations had to manage the irrigation systems and repay the agency in time and in proportion to amounts of inputs received.

This article discusses irrigation technology in relation to the concept of sustainabi‐lity. For sustainable agricultural development, farmers should have control over irrigation technology; they have to be social carriers of technology. Five social variables of technology are defined: product, organization, financial capacity, access and knowledge. Farmers will act as social carriers if these variables match their social situation. The contribution made to sustainable agriculture by flood irrigation and introduced small‐scale pump irrigation technology is evaluated using these five variables.

IntroductionIrrigation schemes all over the world are often marked by a large number of related problems that have an important human dimension and are too complex to be straightforwardly solved. A starting point of this thesis is that these problems have to be dealt with in a learning process that involves all groups and organizations that are relevant to the irrigation scheme. This thesis explores communication processes between irrigation design engineers and farmers in North Senegal and aims at finding out how they can learn from each other.A closer look at the literature about the subject shows that there are two blind spots in the knowledge of design engineers. Both stand in the way of learning. The first concerns a lack of consciousness about the crucial importance of farmers' knowledge of physical phenomena in irrigation schemes such as water flow, soils, topography, etc. (i.e. technical knowledge of farmers). The second concerns a lack of knowledge about the procedures and methods that design engineers could use to improve farmers' participation in the design process.Consequently, the following questions for research arise:1 What is the difference between the technical knowledge of design engineers and that of farmers?2 To what extent do engineers and farmers learn through exchange of technical knowledge, why and how does this exchange take place, and if not why not?3 What is the effect, of the exchange or non-exchange, on the design?4 How can the exchange of technical knowledge be optimized?The research was exploratory and qualitative in nature. The research material was based on detailed observations, informal interviews, semi-structured interviews and group interviews and discussions. Reflection on the collected material led to new perspectives that were useful for experiments at a later stage. Some experiments were part of the development and implementation of a canal maintenance programme for the Ile à Morphil small-scale irrigation project. This required intensive interaction between farmers and myself as an irrigation engineer. The experiments were closely monitored and became, in due course, new research material, which provided in turn a base for new perspectives. In this way several learning cycles were completed during the field research. In the final stage I broadened my perspective and visited several project areas in northern Senegal, where I conducted semi-structured interviews with irrigation engineers as well as groups of farmers.In this thesis, I approach the research material from different angles. I use Bourdieu's concept of habitus, 'a set of dispositions which incline people to act and react in certain ways', in order to explain why design engineers and farmers do not learn from each other. 'Re habitus can be approached indirectly by studying the environments where it developed, as well as by studying peoples' practices, their visible actions. The social interface concept of Long provides clues for what may happen when people who belong to a certain group or category have to deal with 'strangers'. The SoftSystems Methodology (SSM) of Checkland indicates how a learning process can evolve when one faces problems in complex human situations.Practices and environments of farmers and design engineersThe climate in northern Senegal makes it difficult for the Haalpulaar farmers to make a living out of the natural environment. Agriculture often did not provide them with enough to live on and already for some generations, migration work has become an important source of extra income. During a particularly dry period at the beginning of the seventies, farmers were eager to benefit from the extra support of government and donors and managed to integrate irrigated agriculture into their farming system. The Haalpulaar farmers are keen to spread risk and most often divide their efforts between irrigated as well as traditional agriculture and migration work. Although farmers are dependent on the government for the construction of their irrigation schemes and the repair of their pumps, they manage their own schemes and have developed their own technical knowledge, based on some simple initial rules of design engineers. The learning process was facilitated by the existing traditional organization and was adapted to the specific characteristics of the first village schemes.The water potential of the Senegal river, the dry climate and the policy of government and donors to stimulate the Senegalese rice production meant that there were many Senegalese and foreign irrigation engineers in the valley. Irrigation design engineers usually act as natural allies of the government and donors, not only because they depend on them, but also because it is conform their solution-orientated education. This does not alter the fact that they may try to find ways of thinking for the farmers. Several design concepts have evolved since the early seventies. When old concepts seemed to fail or did not satisfy the planners, new concepts were designed. This meant that increasingly more sophisticated and more expensive irrigation concepts evolved. In this process, the gap between the technical design of the engineers and the technical knowledge of farmers gets wider and wider.Differences in technical knowledgeExamples in this thesis make it clear that farmers and design engineers have very different perspectives on irrigation. Irrigation phenomena or design elements are given other priorities, are described differently and are arranged in other ways. They are also embodied at different levels of abstraction and may be or may not be split up into smaller parts.The technical knowledge of design engineers is based on a scientific logic. Generally applicable rules regarding phenomena such as water flow and topography are used in order to be able to design in different localities. To this end, engineers frequently work with abstract models (maps, plans) and are orientated towards generating ideas for future situations. Many technical design elements and physical characteristics are considered separately and may be combined in a design later on. Despite their ability to combine these elements into the design, engineers often attach too much value to discerning these elements. Consequently they may lose sight of the fruitful ideas resulting from an orientation towards interrelationships between the elements. This is illustrated with examples of: water distribution and maintenance, of irrigation, drainage and soil characteristics and of water flow and topography.Compared to engineers the technical knowledge of farmers is closer to physical phenomena in irrigation schemes. It is highly adapted to the specific qualities of the environment (soils, topography) and the simple concept of the village irrigation schemes ( PIVs ). Farmers sometimes use trial and error methods in the field to improve their scheme. In this way they have direct feedback to their design actions. However, in the case of an entirely new design, farmers lack a general overview, not only because their knowledge is bound to a locality, but also because they are so clearly focused on their own plot that most of them do not bother to look at an entire irrigation scheme. Farmers regard physical phenomena and elements as closely connected which often permits them to respond accurately when problematic situations like canal breaching or water scarcity occur.(Non) exchange of technical knowledgeIn northern Senegal communication between design engineers and farmers is limited. With regard to the few situations where communication takes place beyond a superficial level, technical issues receive little attention and the design engineer remains in control of the technical information. At best, a design engineer thinks for the farmers and the irrigation scheme itself often turns out to be the only 'message' of design engineers. One explanation for this is that their employers rarely stimulate and most often discourage communication with farmers. But the lack of communication can also be traced back to design engineers, of whom the majority are not interested in communicating with the fanners. Likewise farmers are not inclined to communicate beyond a superficial level. They prefer not to ask questions because they reason that they may lose the entire project if they do. Besides, their attitude is often a dependent one, as they try to attract new irrigation projects. It is shown that this attitude of farmers strengthens the attitude of planners and design engineers. The reverse is also true. In other words, the habitus of the one triggers and reinforces the habitus of the other.Misunderstandings between design engineers and farmers about technical subjects occur frequently. It has been shown that these most often concern topography, soil suitability, irrigation and drainage requirements, water flow, structures, water distribution and maintenance. The misunderstandings have many dimensions and are difficult to unravel. The explanation for these misunderstandings can be found in the mechanisms of habitus, causing the technical knowledge to change into a technical image that reconfirms itself without engineers or farmers being conscious of it. This implies that design engineers and farmers do not learn from each other, even worse, both draw the conclusion that the technical knowledge of the other should not be taken seriously. They feel justified to be reticent towards the idea of communication about technical issues. A vicious circle occurs: in further design processes communication between design engineers and farmers will only be superficial.Result of the non-exchangeIn many ways the technical knowledge of design engineers is complementary to that of the farmers. Therefore, both are losers with regard to the quality of the technical design. Although new practices could be useful for them, farmers continue with old practices and are not open to suggestions from the design engineer. The new technical designs of engineers are not adapted to the practices of farmers, although these could certainly be useful in a new locality. This thesis shows that the mutual lack of adaptation is costly and has a negative impact on sustainability. It provides examples of farmers who destroy structures or have to adapt the lay out considerably, examples of deterioration of new schemes due to old practices, as well as of designs that are not adapted to the soils and topography of a site. Of course farmers and design engineers blame each other for the resulting problems.How can the exchange of knowledge be optimized?The Soft Systems Methodology (SSM) of Checkland is useful to achieve the shift that is required to deal with the lack of exchange of technical knowledge. It can be used to bring about improvement by activating in design engineers and farmers a learning cycle which ideally is never ending. Learning takes place by means of the iterative process of reflection, discussion, action and again reflection. The reflection and discussion are structured by a number of system models, which may represent desirable future situations. Because of their explicit character the models invite to discuss.This thesis treats a number of experiments with these models. Especially a threedimensional scale- model of a village irrigation scheme that allowed for the imitation of irrigation practices served beyond expectations. It facilitated the exchange of technical knowledge, covering a broad range of technical issues such as water distribution, maintenance, water flow, structures and topography. The scale model bypasses language problems because it is so tangible that it allows both farmers and design engineers to explain their points of view: they just demonstrate what they mean. Other useful explicit models like adapted maps and plans, combinations of drawings, a simple levelling instrument, as well as field visits to other irrigation schemes, may structure a discussion about change. In general, it appears that these models have unfreezing effects that facilitate the communication between design engineers and farmers.Model of a learning system of engineers and farmersThe emerging perspective of my thesis is condensed in a model. The (diagrammatic) model represents a learning system that may eventually lead to the implementation of an irrigation system that is feasible and desirable. The system is meant to avoid the technical misunderstandings as much as possible.The model makes explicit several stages of the learning process. Tile stage during which the irrigation system, or parts of it, is discussed by means of system models is crucial, because it connects the separate learning cycles of design engineers and farmers. During this stage, the learning experiences of both sides are shared, providing a basis for the joint technical knowledge that is required for quality design. Applied research is another stage of the learning system. For this stage, the research methods and the concepts that I presented in this thesis may be useful.In the view of Checkland, models should be tentative, can never replace reality and should not be followed rigidly. The emerging model in this thesis should therefore be seen as a preliminary model that, in the first place, serves to continue a discussion about how to proceed in the context of complex situations in irrigation schemes.

The control, allocation and distribution, of water is the core process of an irrigation system. It is the process by which the available water is divided and distributed to the smaller irrigation units within the system, which in turn is distributed further down to the individual water user who must control it to place it in the crop root zones in particular fields. This study is based on the thesis that water control is a function of both technical/engineering, and organizational and institutional arrangements (social) controls. The combination of the physical system and organizational arrangements in an irrigation system determines how much water goes where. An analysis of both aspects in the context of the local farming system should lead to the better understanding of the more general issues and complexities of water control in irrigation systems.This study presents three cases studies on water control in government-built and managed irrigation systems in the tarai Nepal. Water allocation and distribution practices during the consecutive crop seasons in the three irrigation systems located in three different places in the tarai region were studied from October 1991 to May 1993. There were two principal reasons for focusing research on the tarai irrigation systems. The first is that this region has the greatest number of and the largest area covered by government-built and managed large to medium scale irrigation systems. The second is the challenges and complexities in water control in the government-built and managed irrigation systems. An integrated approach which combined the study of the technical design and of organizational arrangements was followed. The historical development of the irrigation systems has also been traced to see the change or evolution of technical pattern and management practices over time. In addition to the study of the technical design of the irrigation systems, the concept of practice, which in this study refers to the visible actions and undertakings of people, has been used to understand how a certain type of water control technology provokes particular activities of water allocation and distribution. Both quantitative and qualitative data have been used.There is historical evidence that canal irrigation in Nepal has been practised by farmers for centuries, but the development of modern canal irrigation systems in the country began only in the 1920s. The Department of Irrigation, established in 1951, is responsible for the planning, development and management of government irrigation schemes in the country. It consists mostly of civil engineers and is largely construction-oriented.Government investment in irrigation development - especially in the large irrigation systems in the tarai - has increased tremendously since 1970, with the increase in the borrowing of international capital in the form of loans and grants. The development of irrigation facilities has been focused in this region because it has the largest area of irrigable land and the greatest potential for agricultural output using the river water resources. Until the mid-1980s, irrigation development by the government was largely confined to the construction of physical infrastructures of canals and structures, and very little attention was given to the effective management of the completed systems. From 1985, attention began to be paid to the improved management of government-operated irrigation systems. This move was also influenced by the external donor agencies involved in the irrigation development. Further, the Irrigation Regulations promulgated in 1989 emphasized the participation of farmers in all phases of irrigation projects, the operation and maintenance of the government-built and controlled irrigation systems in particular. The premise of increasing farmer participation was that government resources alone were inadequate to meet the country's irrigation development objectives and sustain the management of government irrigation systems after their completion. The same year, action plans for the turnover of small irrigation systems and the participatory management of large irrigation systems were formulated. Subsequently, in 1992, the government adopted a policy under which the management responsibility of the government-built and operated irrigation systems are to be gradually transferred to farmers who are organized into water users' organizations. Under this policy, the irrigation systems serving command areas over 2,000 hectares (in the tarai ) are to be jointly managed by the government and the water users' organization, and the management of the irrigation systems commanding less than this area are to be totally turned over to the water users' organizations.Two modes of irrigation system development can be distinguished in Nepal: extensive and intensive development. In extensive development only basic minimum infrastructure, such as main and distribution canals, are provided and irrigation blocks are typically large, from 250-500 hectares. In intensive irrigation development, physical facilities are extended further down to the tertiary or farm level. The size of an irrigation block is reduced to 30-50 hectares. In recent years, emphasis has been placed on intensive development with farmer participation, to achieve the irrigation potential as envisaged in the design. In Nepal, these two development modes have been influenced by the resources available.Irrigated agriculture in the tarai is basically rice based. Paddy, the staple crop, is cultivated during the monsoon season, followed by the cultivation of mainly oilseeds and wheat in the winter season. Whereas paddy covers almost all the land area, the coverage by winter crops is less intense. However, the cropping intensity varies from one place to another within the region. Farming is labour intensive and traditional methods are generally employed. Landholdings are fragmented and farmers practise intensive fanning on smallholdings. Where irrigation water is available throughout the year, the most intensive fanning is the cultivation of two crops of paddy followed by a winter crop of wheat.The irrigation systems in the tarai region are run-of-the-river systems and have been built primarily to irrigate rice in the monsoon season. They are therefore subjected to inter and intra-seasonal variations in the water supply. Their operation is upstream-controlled and supply oriented. The canals in most irrigation systems are unlined.However, the three irrigation systems studied show that the government-built and managed and controlled large canal irrigation systems in the tarai have varied technical design and operational features for water control. The difference in the water control technology is mainly because of the different design approaches followed by the design engineers, especially foreign consultant engineers. The involvement of different groups of design engineers is largely because of the government's position in irrigation development in the country vis-à-vis different external financing or donor agencies, such as ADB, World Bank, USAID and so forth. The government has to commit to involve foreign design engineers at the time of loan or grant agreements. The consultants are placed in an authoritative position. This situation is favoured by the fact that the Planning and Design Division of the Irrigation Department is not strong enough to discuss, check and question the compatibility and appropriateness of the designs of the consultants. The consequence of the involvement of foreign consultants on the choice of technology could not be foreseen.The first case study, the Banganga Irrigation System, comprises a fully regulated water control technology with manually adjustable gates down to the tertiary outlets. The design engineers emphasized the efficient use of the available water vis-à-vis flexibility in the system operation, given the water constraint. Whereas this fully regulated technology potentially enables an irrigation system to deliver and distribute differential water requirements to the canals and outlets under the conditions of fluctuations in water availability, it considerably increases the operation and maintenance burden of the agency operating the system. It requires frequent resetting of gates at each variation of flow and is therefore complicated and the control structures become the targets of vandalism.The second irrigation system, the Pithuwa Irrigation System, which was designed initially by the engineers of the Irrigation Department and later remodelled by the farmers, is a nonflexible system. It does not have adjustable control structures except for the head regulator of the main canal. The available flow in the main canal is distributed to all branch canals through open pipe outlets whose sizes are proportional to land area served. In this technology, the management input in terms of the operating staff is reduced to the minimum: the operation of the main canal headgate. Interference by people (including farmers) in the distribution of water along the main canal is reduced.In the third case study, the Mahakali Irrigation Project (Stage I), the water control technology is based on the "structured" design concept. In the "structured" system design, regulation of flow is only required above the structured level. The level of structuring can be fixed at any point above the chak outlet up to the headgate at the diversion point. The "structured" system design thus makes the management of the irrigation system relatively easier, as compared to the fully gated controlled technology as in Banganga. But it reduces the flexibility in water distribution in the irrigation system. The Mahakali Irrigation System (Stage I) is "structured" such that flow regulation in the main canal is only required to operate secondary canals at full supply flow, and the flow in secondary canals is divided through the fixed weirs with proportional openings and all the tertiary outlets are set for a fixed discharge.Compared in terms of the "structured" system concept, the three irrigation systems differ in their level of structuring. Whereas Banganga has the lowest level of structuring, at the tertiary outlet, Pithuwa has the highest, at the headgate of the main canal. Accordingly, they present different operational flexibility.Many engineers have a bias towards "modern" technology and operational flexibility in irrigation systems. From the case studies presented in this thesis, it is concluded that the extensively gated technology for operational flexibility needs to be reconsidered, if effective water control is to be achieved. The semi-flexible technology based on the "structured" concept, as in Mahakali, might be appropriate for government controlled and managed large irrigation systems, given the operating staff and budget limitations of the Irrigation Department. The simple water control technology as that in Pithuwa appears to be appropriate for complete turnover of the irrigation system to the farmers. This water control technology needs to be reconsidered and utilized if the investments in the future irrigation development in Nepal are to realize their potential.Water users' organizations are created for farmers' participation in the operation and maintenance of the systems. The reliability and adequate delivery of water to farmers greatly influence on the continuity or stability of the water users' organizations. The legal recognition and institutionalization of water users' organizations enhance the continuity of the organizations, but they do not necessarily motivate farmers' participation in government controlled irrigation systems. This study suggests that the fully gated water control technology does not guarantee reliable and adequate water delivery because of the complexity in operation, especially in government-operated irrigation systems having the limitation of an inadequate skilled manpower resource. The non- flexible technology of fixed weir check structures and open pipe outlets, as in Pithuwa, allows the available water to be distributed fairly and proportionally down the system. However, the limitation of this technology is that it is not possible to alter the distribution ratios of flow, and it inhibits the effective use of rainfall, especially where rotational water distribution is applied. By comparison, the semiflexible water control technology, as in Mahakali, allows water supply to part of the system to be cut off if necessary, because of rainfall. However, this technology should be looked upon as undesirable where there is large spatial variation in soils and crops.Farmers' participation enhances effective water control in government-operated irrigation systems. Farmers will have little incentive to participate in collective efforts in the operation and maintenance activities if they have no confidence in the capability of the technology to deliver reliable and adequate water. The empirical evidence from Pithuwa show that simple inflexible proportional division technology complemented with strong organization for social control is effective in ensuring reliable and fair water distribution.Water control is a complex process requiring intensive organizational effort. Social control strongly influences the water allocation and distribution in irrigation systems. It minimizes water related conflicts and vandalism of physical control structures and enhances fair water distribution. As the case study of Pithuwa shows, social control is more effectively developed by fanners themselves than imposed by an external agency like the Irrigation Department.Finally this study has shown how operational realities can be different from the design engineers' assumption. Water allocation and distribution differ from that intended in the design because of management decisions and the capability of the operating agency to execute the decisions. The availability and value of water also has strong influence upon this. In Mahakali the attempt to introduce rotational water distribution for efficient water management has had little impact because the water supply is abundant.

To develop adequate and sustainable land use plans, a clear assessment of the potential of the land and of the existing farming systems, as well as an identification of ways to attain these potentials is required. The agricultural system in the limestone area south of Malang is classified into different land units, land uses and livestock units. Land uses include a further distinction into agro-forestry system and annual crop. Simulation models are used as a tool to explore the production potential of crops and livestock. Existing crop simulation models are used and adjusted to the situation at hand. A new livestock simulation model is developed that includes major farm household influences, particularly the influence of land and land use on the feed resource and on livestock production. The model considers also the seasonality of feed availability; feed is quantified on a monthly basis. Two production orientations, with their specific technologies, were distinguished; low external-input agriculture (relying on manure only) and yield-oriented agriculture (relying on inorganic fertilizers only). The present crop and livestock productions are far below the potentials, as were explored with the simulation models. Constraints to agricultural production in the research area are high erosion, a low soil organic matter and a declining natural soil fertility. Hence options for land use are explored which focus on biophysical sustainability. The latter is defined here as avoiding excessive erosion, at least maintaining the present soil organic matter content of 2% and maintenance of a closed balance of the most limiting resource.Systems with cattle, more perennials and yield-oriented agricultural systems leave more biomass than current cropping thus creating more cover, litter and mulch. An increase in the use of inorganic fertilizers seems the only viable direction towards more biomass production. Increased fertilizer use will improve the quantity and quality of crop yields as well as of crop residues for animal husbandry. Since 100% of crop residues never will be used in animal husbandry, effort should be made to return the left-overs to the soil in order to maintain or increase the soil organic matter content, i.e. better residue management. The extra manure both in terms of quantity and quality will also contribute to the latter goal. Finally, the higher soil organic matter content will improve the physical state of the heavy clay soils thus improving their water holding capacity, it will lower the soil's erodibility and improve nutrient use efficiency. So, use of inorganic fertilizers will trigger a cascade of effects that all point in the direction of a more sustainable use of the natural resources.

This study was conducted to quantify nutrient losses by saltation and suspension transport. During two convective storms, mass fluxes of wind-blown particles were measured in a pearl millet (Pennisetum glaucum) field in southwest Niger, on a sandy, siliceous, isohyperthermic Psammentic Paleustalf. The trapped material at three heights (0·05, 0·26 and 0·50 m) and a sample of vertically deposited dust were analyzed for total element contents of K, C, N and P. The nutrient content of the material at 0·05 m was similar to the nutrient content of the topsoil. At 0·50 m, the material was three times richer in nutrients than the topsoil, whereas the deposited dust, trapped at 2·00 m, was 17 times richer. For all four elements, a total element (TE) mass flux profile was fitted throughout the observations. From the TE profiles, the following nutrient losses from the experimental plot were estimated: 57·1 kg ha−1 K, 79·6 kg ha−1 C, 18·3 kg ha−1 N, and 6·1 kg ha−1 P. The TE profiles showed a maximum value in the saltation layer. The suspended TE mass fluxes above the saltation layer were an order of magnitude lower than the saltation fluxes, but extended to greater heights. Therefore, saltation and suspension are both able to transport significant quantities of nutrients. While saltation results in only a local redistribution of nutrients, suspension may transport dust over thousands of kilometers, resulting in a regional loss of nutrients.