GOVERNMENT OF ETHIOPIA WATER RESOURCES DEVELOPMENT AUTHORITY MASTER DRAINAGE PLAN FOR MELKA SADI AND AMIBARA AREAS FINAL REPORT July 1985 VOLUME 8 ANNEXE : AGRICULTURE Wjter Resources Development Authority P 0 Box 5673 Addis Abeba Ethiopia Sir William Holcrow & Partners Consulting Engineers Burderop Park Swindon Wiltshire SN4 OQD United KingdomGOVERNMENT OF ETHIOPIA WATER RESOURCES DEVELOPMENT AUTHORITY MASTER DRAINAGE PLAN FOR MELKA SADI AND AMIBARA AREAS FINAL REPORT July 1905 VOLUME 8 ANNEX E : AGRICULTURE Water Resources Development Authority P 0 Box 5673 Addis Ababa Ethiopia Sir William Hal crow & Partnen Consulting Engineers Burd er op Park Swindon Wiltshire SN4 CQD United KingdomMASTER DRAINAGE PLAN FOR MELKA SADI 4 AMI3ARA AREAS FINAL REPORT VOL 1 VOL 2 VOL 3 VOL 4 VOL 5 VOL 6 - - - - - - VOL 7 SUMMARY MAIN REPORT ANNEX AL : SOILS AND DRAINABILITY CLASSES ANNEX A2 : SOILS AND DRAINABILITY CLASS MAPS ANNEX A3 I DESCRIPTION OF AUGER BORES ANNEX B GROUNDWATER AND SALINITY ANNEX C 1 HYDROLOGY ANNEX D : ENGINEERING VOL 8 * ANNEX S 1 AGRICULTURE - VOL 9 - ANNEX F MARKETING AND PRICES - - ANNEX G ANNEX H : PROJECT COST ESTIMATES ; FINANCIAL AND ECONOMIC EVALUATION - ANNEX I EVALUATION METHODOLOGY - GUIDELINES VOL 10 ANNEX J : ENVIRONMENT AND HEALTHANNEX E AGRICULTURE Contents 1. INTRODUCTION 2. METHODOLOGY OF FIELDWORK 3 2.1 Yield-Sample Plots 3 2.2 Monitoring Soil Moisture: Waterlogging Index 5 2.3 Soil Salinity: Salinity Index 6 2.4 Rooting Index 7 2.5 Variation in Microtopography 3 2.6 Random Sample Survey 8 3. CROP YIELDS 10 3.1 Crop Characteristics and Growth Patterns in Cotton 3.2 Crop Characteristics in Bananas 3.3 Review and Analysis of Historical Yields 3.4 Effect of Waterlogging on Yield 3.5 Effect of Salinity on Yield 3.6 Results of the Sample Survey of Cotton 3.7 Other Causes of Yield Loss 10 14 14 17 19 21 23 4. SURFACE WATERLOGGING AND SALINITY 27 4.1 Soil Porosity 27 4.2 Perched Watertables 28 4.3 Microtopography 28 4.4 Irrigation and Leaching 29 5. OPPORTUNITIES FOR IMPROVEMENTS 31 5.1 Optimisation of Drain Spacing 32 5.2 Water Management 34 5.3 Tillage 37 5.4 Other Possibilities for Improvements 41 6. FUTURE LANDUSE AND CROP YIELDS IN STAGE I AREA 47 6.1 Present and Future Landuse 47 6.2 Present and Future Crop Yields 51 6.3 Timing of Drainage Installation 54 REFERENCES TABLES FIGURES APPENDIX 1LIST OF TABLES Table No. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 16A 17 18 Crop Data from Cotton Yield-Sample Plots Soil Data from Cotton Yield-Sample Plots Crop and Soil Data from Banana Sample Plots Data from Random Sample Survey in Cotton Average Cotton Yields for Farms in Project Area Date of Field Operations, Yield and Yield Change to 1983 for Selected Pairs of Fields (1981 to 1983) Cotton Yield Adjusted for Effect of Waterlogging Overall Analysis of Variance of Yield Data from Random Sample Survey Summary of Yield Data from Random Sample Survey Gross Margin at Different Drain Spacings on Soils of Different Drainability Class, and Estimated Optimum Spacing Proposed Leaching/Irrigation Schedules for Cotton Irrigation Parameters for Bananas and Estimated Schedule Irrigation Parameters for Irrigated Pasture and Estimated Schedule Gross Land Areas in Rationalised Groundwater Salinity Classes Estimated Number of Years Before the Schedule-Year Crop Failure Occurs Projected Rate of Abandonment of Land within Stage I Projected Rate of Abandonment of Land Within RRC: Alternative Scenarios 2 and 3. With Project Banana Replanting Schedule and Yields Potential Landuse in RRC Settlement Scheme in Relation to 18A 18B 19 19A 20 21 22 23 24 24A 25 Groundwater Features Potential Land Use in RRC Settlement Scheme in Relation to Groundwater Features: Scenario 2 (all Cotton) Potential Land Use in RRC Settlement Scheme in Relation to Ground water Features : Scenario 3 (50Z Cotton, 25Z Maize and Pasture Stage 1 Area - Projected Land Use With and Without Drainage Projects Projected Lane Use With and Without Drainage Projects for Alternative Cropping Patterns at RRC Percent Loss of Crop in the Years Before Schedule-Year Average Crop Yield in Years Before and After Schedule-Year as Affected by Waterlogging Only Decline in Crop Yields Due to Rising Watertable and Salinity Example of Calculation of Yield Decline of Cotton En terprise at RRC Without Subsurface Drainage Stage 1 Area - Projected Average Crop Yields With and Without Drainage Projects Projected Average Crop Yields With and Without Draina ge Projects for Alternative Cropping Patterns at RRC Expected Yield of Firewood and Poles in RRC Irrigated Forest Area APPENDIX TABLES . J Basic Financial Data Used in Drain Spacing Optimisation ip? Calculation of Gross Margin at Different Drain Spacings. P ’ Example: Model for Coarse-Textured Soil, Drainability Classe 1A ,\p.3 Drain Optimisation. Analysis of Variance and Summary. 'P‘ Example: Model for Coarse-Textured Soil, Drainability Class lA , Estimation of Optimum Drain Spacing for Soil Drainability ClassesI I I I I I I I I I I I I d A I I ■ LIST OF FIGURES Figure No. 1.1 1.2 to 1.9 2.1.2.2 3 4.1 to 4.5 5 6 7 8 9 10 11 12 13 Features and Description of Moisture and Salinity Profiles in Cotton and Banana Yield-Sample Plots: Example Moisture and Salinity Profiles and Estimated Seed Cotton Yield in Yield-Sample Plots Moisture and Salinity Profiles and Average Maximum Height of Pseudostems in Banana Plots Variation of Conversion Factor of ECl;2 to EC Between e 14 15 16 Effect of Soil Solution Concentration on Germination of I Cotton Seed Effect of Salinity Index on Che Estimated Yield of Seed Cotton Effect of Salinity on Height and Vigour of Bananas 17 18 19 20 21 22 23 24 25 26 APPENDIX I Contrasting Soil Types and With Soil Moisture Content Ground Elevation and Crop Height Along Selected Sections Map of AIP Showing Location of Random-Sample Fields Nomogram to Estimate Seed Cotton Yield from Average Crop Height and Colour Index Stylised Growth Pattern of Cotton Planted on 1 June at AIP Map of Average Cotton Yields in Melka Sadi and Melka Warer State Farms - 1983 Map of Yield Decline of Cotton Between 1982 and 1983 Seasons Map of Yield Decline of Cotton Between 1981 and 1983 Seasons Relationship Between Rooting Index and Average Height of Cotton Plants Relationship Between Rooting Index and Colour Index in Cotton at Non-Saline but Chemically Fertile Sites Effect of Rooting Index on Che Estimated Yield of Seed Cotton Relationship Between Average Height of Crop at Maturity and Estimated Yield of Seed Cotton Relationship Between Colour Index of the Leaves at about 145 Days and the Estimated Seed Cotton Yield at Non-Saline Sices Effect of Interdrain Spacing at Different Drain Depth on Marginal Return Effect of In-field Drain Depth on Maximum Marginal Return to Drainage Illustration of the Moisture Profile Following Irrigation in Soils of Contrasting Profile Parameters for Calculation of Cotton Irrigation Schedule Optimisation of Cotton Irrigation Schedule Effect of Age of Crop at Final Irrigation on the Loss of Potential Cotton Yield Indian Raised-Bed System for Minimum Tillage Moisture Profile of Soil after Irrigation when Average Annual Depth of Watertable is Less Than 1.0m Apl Effect of Waterlogging on Harvestable Yield and its Distribution Between First and Second Harvests of Cotton Ap2 Comparison of Curves Fitted to Gross Margins at Different Drain Spacings; Fitted by ’Eye’ and by Quadratic Regression-1- I. INTRODUCTION Agronomy field-work was divided into two stages. The objective of the first stage, of one month duration, was to provide the background to the nature and extent of the problems of salinity and high watertable within the scheme. Data and information were collected from Che various institutions involved in the scheme 3nd these were summarised in the Discussion Document (Ref 1). Stage two of the study, which lasted two months, is described in detail in this Annex. The main objective was to provide a sound basis for the projection of future yields with and without a drainage project. In order to achieve this it was necessary to investigate the independent and quanti tative effects of salinity and waterlogging on the yield of crops under the particular circumstances of Amibara. Field study was restricted to cotton and bananas since other crops were too limited in extent for field investigation. As field work progressed it became increasingly obvious that a rising ground-watertable was not the only reason for surface salinity and drainage-related problems. It is the most visibly devastating feature having caused the abandonment of several hundred hectares of banana plan tation and cotton fields. Furthermore, the area affected clearly will increase in future unless a remedy is found. However, equally apparent were crop symptoms of waterlogging and salinity in areas where the ground watertable is known to be too deep to be causing any damage. With Chis realisation came the need to investigate in greater detail firstly Che reasons for surface waterlogging, and in places, its associated salinity, and secondly their extent and impact on the crops throughout the whole scheme. Tn order to minimise the fieldwork required co establish the quantitative relationship of yield with salinity and all forms of waterlogging, sample plot9 were clustered mainly around the target area. Most were in productive fields of cotton and bananas, yet close to abandoned areas where a high level of ground-water could be expected. Nevertheless, a few observations were made on the Melka Warer State Farm, in Che cotton and maize of the Amibara Settlement Scheme, and in Unit 1 of Melka Sadi State Farm. In order to establish the extent and impact of drainage-related problems in cotton, and to assess how much potential yield is presently being lost, a random sample-survey was carried out. Tn the time available this survey could not be comprehensive but it was sufficiently accurate and repre sentative to make a reasonably confident estimate. Furthermore, the estimate of overall average yield for the 1983/84 season is very close to that made independently by the staff of MAADE. The Annex continues by examining the evidence available to explain the reasons for surface waterlogging and its related salinity, and finally by discussing the opportunities which exist to overcome the problems.-2- A brief explanation is given of some of the principles of plant and crop physiology which underlie the discussion of waterlogging and salinity. Although this study is about the estimation of the benefits of drainage, a few comments are included on more general aspects of husbandry. Based only on casual observations rather than on systematic study, some hopefully constructive recommendations are made which might assist in raising the general level of crop profitability and output. The overall impression of the cotton enterprise at Amibara, is that in most respects the management is good. Some of the problems apparent in 1983/84 have now been overcome and a marked improvement in yield is forecast. While considerable scope remains for yet further achievement, it may be observed that the yield prediction of the Feasibility Study has been exceeded in almost every season since the start of cotton production at Amibara. The situation in the banana plantation is less favourable, but there too are signs that considerable progress has been made in the fields which have been replanted recently. However, it should be recognised that Amibara is an environment of conspiciously lover potential for bananas than for cotton and it is unlikely that many of the inherent problems could be overcome.-3- 2. METHODOLOGY OF FIELDWORK The objective of the Yield-Sample Plots was to make a reasonably accurate estimate of yield at locations quite deliberately chosen on the expectation that the soil would show a measurable degree of waterlogging or salinity. The objective of the Random Sample Points was to make a rapid survey at entirely random locations. Yield was to be estimated using the relation ships established in the Yield-Sample Plots, and the nature of the problem to be assessed by reference to certain observations on the soil. Tn order to verify the importance of variation in microtopography, levels were taken at a selection of the Yield-Sample Plots. 2.1 Yield-Sample Plots 2.1.1 Cotton A total of 54 Yield-Sample Plots was established between 25th September and 22nd October, 1984. Of these, 46 were deliberately chosen in fields which were about to be irrigated so that the moisture characteristics of the profile could be assessed. In the event, only 32 of the plots were immediately irrigated and were completely monitored. In most fields, a cluster of sample points was chosen to provide a range of yield estimates from extremely good (or the best in the immediate area) to very poor, or in some cases, zero yield because plants had died. Once a sample point had been chosen, a sample plot was marked out ensuring, as far as possible, uniformity of crop within it. The plot measured 5m of each of five rows. The width of the plot was measured, recorded and was found to vary slightly around a nominal value of 4.5a (interrow spacing 90cm). Strings were tied around four comer posts, the latter being left in situ for the duration of field work. The average plant height was recorded as the mean of the average measured at six random points in the plot. The fourth, fully-expanded leaf from the terminal bud was sampled systematically from every second to fifth plant in each row of the plot (depending on plant population). Four leaves were sampled in each row, giving a total of twenty per plot. A Redness Index was calculated by classifying the sample leaves into the following classes: Class I Class 2 Class 3 No redness 0-10* of leaf area reddened 10-40* of leaf area reddened Class 4 Class 5 40-30* of leaf area reddened 80* of leaf area reddened The side of the leaf showing the greater amount of redness was nsed in tihe classification. Redness Index is the mean class:-u- where n is the number of leaves in the ith class. A visual 9core was made for yellowness, class 1 being leaves predominantly dark green, to class 5 in which all leaves would be at least pale green and older leaves yellow. A simi . ar score was made of severity of leaf necrosis, with class 1 nil to class 5 severe. (For an explanation of the physiological significance of these scores, see Section 3.1). A visual score of crop vigour was also made, with 1 best and 5 worst. Records were made of the number of bolls and plants in each row of the plot. A check was made initially on accuracy of counting. Each of the two recorders was asked to count the bolls in each row of the plot and later to recount them, without having been forewarned. The differences were less than + 5Z and seemed to be random both between recorders and between suc cessive counts. At the first count only bolls > 15nn in diameter were recorded because at that time it was not clear what proportion of new squares and small bolls would survive to harvest. During the last week in October all plots were revisited in order to count Che number of unsplit bolls. In some plots, particularly those of luxuriant growth and with a high proportion of unsplit bolls, the total number of bolls was recounted and on Chis basis the original yield estimate of some plots was adjusted. There is no reason therefore to suspect bias in the yield estimates. In one plot of particularly luxuriant growth, a substantial number of bolls had been shed due to over-irrigation and the excessive LAI (see Section 3.1) A systematic sample of the seed cotton from 100 bolls was taken from each plot in order to obtain average weight of seed cotton per boll (« ’average boll-weight*). In very short cotton only a single 100-boll sample was taken Co represent all bolls. In most plots separate samples were taken from the upper and lower halves of the plants so that independent estimates of boll weight were possible. There was considerable variation in average boll weight both between plots and between upper and lower regions of the crop. The average top/bottom ratio was approximately 1.25 which was then used to estimate average boll-weight in those plots where only a single sample of lover bolls was possible due to luxuriant growth and delayed maturity. The basic data obtained from cotton Yield-Sample Plots are shown in Tables I and 2. 2.1.2 Bananas Accurate estimation of banana yield is not possible in a short period so no attempt was made to mark out plots. Plots within a radius of about 5m from a selected sample point were assessed for average height of mature pseudo stem and scored for vigour (on a scale of 1 to 10, with 1 best). The average number of hands per bunch was estimated from a sample of about 10 bunches. Visual scores were made of plant population (on a scale of 1 to 5, with 1 highest) and the yellowness of leaves (on a scale of I to 5, with 5 most yellow). Taking into account the above observations, an overall assessment of yield was made by scoring the site on a scale of 1 to 10. with 1 best. Based on present annual production data, a 9core of 1 may be taken to correspond to a yield of 18t/ha and a score of 10 to zero yield. Each unit change in score therefore corresponds to an increment in annual yield of 2t/ha. Banana data from the Yield-Sample plots are given in Table 3.-5- 2.2 Monitoring Soil Moisture: Waterlogging Index Some moisture characteristics of the soils of Amibara have been reported elsewhere (Ref 2). Hydraulic conductivity has been measured in this study and the results are presented in Annex A?. The movement of irrigation water down the profile is difficult to monitor and for accuracy requires the use of complex equipment which was not available in this study. It was considered, however, that variation in the moisture profile is responsible for some of the variation apparent in growth of cotton and bananas. Thirty two of the cotton and five of the banana yield sample plots were selected for monitoring of the moisture profile. Beginning, as far as possible, 24 hours after irrigation had stopped and for nine days thereafter, the soil was augered daily to a depth of 2m at randomly selected points in the plot. By fee! and visual inspection three moisture states were recognised. From previous studies (Ref 2) it is possible to infer their approximate equivalent moisture contents (mm(mm soil) to be: Vertisol Al 1 uv i urn (clay, silty clay) (other textures) Saturated (above FC) 0.64 - 0.74 0.58 - 0.64 < 0.58 0.41 - 0.54 At Field Capacity (FC) 0.37 - 0.41 Moist below FC <0.37 During monitoring further subdivisions were recognised. Soil moist below FC which had been wetted by the irrigation was recorded as distinct from that which was moist but unaffected by the irrigation being monitored. Moist soil was considered to be above permanent wilting point (PWP) and that which was not was recorded as dry. In practice it is impossible to distinguish these subdivisions by feel with any precision in certain soil textures. However, it was unimportant as the emphasis of Che study was to characterise soils in which moisture excess rather than deficit is affecting the growth of crops. Accordingly, the data are presented in terms of the three main divisions, noted above, with the emphasis upon the Saturated class. The most convenient way of presenting the data is in diagramnatic form, which is explained in Figure 1.1. The observed moisture profiles under cotton are shown in Figures 1.2 to 1.9 and under bananas in Figure 2. This format, of a moisture profile which varies with time, is not immed iately useful as a diagnostic criterion for assessing the influence of waterlogging on yield. A single value, to be called Waterlogging Index (Wl), is required which recognises the greater activity of roots in the surface horizons. It also must be sensitive to the presence of a static watertable in that part of the profile normally regarded as rhizosphere. A suitable index for cottop was derived but the corresponding index for bananas was less satisfactory.-6- The Waterlogging Index for cotton involves the calculation of a daily weighted mean depth of saturated 9oil following irrigation. The WI is then the simple mean value over the 10 days of observation. The weighting factors tried were the same as those used in optimising the Salinity Index and those which showed the best fit to yield data were the same (see Section 2.3). This is to be expected since the weighting factors are a simple reflection of the distribution down the profile of vital root activity. Cotton roots were observed below 2m in auger samples from plots where plants had reached more than 2m in height. These plots did not produce the highest yields which confirms that this exceptional root development was in response to opportunity rather than necessity. The daily weighted mean depth of saturated soil is calculated by summing the products, for each horizon of 25cm, of the number of cm of saturated soil and the corresponding weighting factor, (because the sum of the weighting factors is value for each of the 10 day9 of observation follows: Xn-10 Xn-l and is x where f is the weighting factor and d the depth then dividing by 100 ). The simple mean of the the WI. The equation is as 10 -3 of saturated soil in the ith horizon for the nth day. The observation period was chosen to begin approximately one day after irrigation stopped in order to avoid problems of augering through a surface cover of water. In some plots, surface water persisted for several days. The period of 10 days was chosen as a compromise between the irrigation interval of 14 to 21 days, the time required for a saturated horizon to exert a measurable influence on crop growth, the shape of the watertable recession-curve following irrigation, and the amount of work involved. The calculated values of WI in those cotton and banana plots which were monitored are given in Tables 2 and 3. 2.3 Soil Salinity: Salinity Index At the same time that Yield Sample Plots were marked out and measurements made on the crop, samples of soil were taken from an auger hole to 2m depth, at each 25cm interval. Each approximately 200 ml sample of soil was placed in a bottle with an equal volume of water. After shaking thoroughly, the electrical conductivity of this 1:2 suspension, ECl*2« was measured by a submersible probe. All 54 cotton and 15 banana Yield Sample plots had salinity measured in this way and in seven cotton plots, the salinity wa9 remeasured after irrigation in order to observe the short-term effect of irrigation on the salinity profile.-7- The salinity profiLe is best represented diagrammatically and those of the plots in which moisture profiles were monitored are shown in Figures 1.1 to 1.9. On account of the majority of root activity being located in surface horizons, a simple mean of the salinity values down the profile may be expected to be a less-than-9atisfactory statistic by which to assess the influence of salinity on crop yield. A selection of weighting factors was cried before it was concluded that the following provided the best fit to the data: Horizon depth (cm) 0-25 25-50 50-75 75-100 100-125 125-150 7150 Weighting factors (Z) Cotton 50 30 10 5 3 2 0 Banana 5 10 35 50 0 0 0 The weighted-mean salinity (WMS) is calculated: Z* i-8 WMS ’ <1-1 fiSi X 10-2 where fi is the weighting factor and s£ is the EQ. 2 of the ith horizon. For cotton and bananas, WMS is called the Salinity Index (SI). Values of SI are shown in Table 2 and 3 for cotton and bananas respectively. The accuracy of the EC[:2 measurement is less chan that of Che EC of the saturation extract (EC«). The latter is the electrical conductivity of the filtrate of a saturation paste. The accuracy of ECi-2 is dependent on the accuracy with which Che volume of soil is matched by the volume of water added and whether or not there is compensation for the moisture content of the soil sample at the time of making the suspension. Since EQ; 2 values were measured in the field it was not practicable to measure accurately either soil volume or moisture content. Since most samples were taken from dry profiles before irrigation, most estimates of Salinity Index would not be seriously affected by variation in moisture content. The minority of samples at or close to saturation would be somewhat underestimated for salinity, by possibly as much as 50Z compared with a corresponding dry sample The conversion factor from ECi;2 to EC and Che influence of soil moisture e content at Che time of EQ *2 ®^ASurement is shown in Figure 3 for a typical vertisol and alluvium (as described in Ref 2). For soils at field capacity when ECi:2 is measured, Chen EC would be 2.8 and 2.3 times greater in an e alluvium and a vertisol respectively. This is confirmed by the graphical relationship for a number of soil samples described in Annex A. 2.4 Rooting Index The Waterlogging Index (Section 2.2) i9 calculated from the moisture profile of the soil on the assumption that rooting depth is not limited to less than 150cm by any ocher feature chan soil saturation. A very compact and generally saline horizon was observed in the root zone of a few cotton-8- plots. In order to achieve a closer relationship between yield of cotton and the Waterlogging Index it was necessary to treat the presence of any such rooting barrier as a notional watertable. The adjusted index is called the Rooting Index and is identical to the WI in the majority of plots, as shown in Table 2. 2.5 Variation in Microtopography Elevation of the furrow-bottom was measured at intervals along a number of sections. Sections were chosen to pass through or close to cotton Yield- Sample plots and most ran parallel to furrows. At each point the height of the cotton also wa9 recorded. Data are shown diagrammat ically in Figures 4.1 to 4.5. In most cases a field-plan shows rhe location of the sections relative to each other, to the plots and to the furrows and canal. Ground level is shown as elevation in cm below the highest point recorded at the site. Intervals along the sections are 5m (paced) in all cases, with the exception of intervals of lm along a part of the section in Figure 4.5. The location of plots along the sections ia shown together with the corresponding yield and Salinity Index (SI). 2.6 Random Sample Survey Study time was insufficient to locate a large number of Yield-Sample plots in a random manner such that their data would represent the whole estate with a reasonable level of confidence. It was possible, however, in the time that remained after completion of the Yield-Sample plot work, to make a rapid survey using Random Sample points. A total of 145 sample points were measured. Sample points were selected using random numbers in such a way as to ensure that all parts of the estate had an equal opportunity of being selected. All 486 fields were numbered consecutively and twenty random numbers in the range I to 486 were chosen to represent sample fields. Each field notion- ally wa9 subdivided into tenths along the length of the field canal, numbering from the inlet end. This was to allow for variable length of field. For each field a set of ten random numbers in the range 1 to 10 was selected representing the first coordinate of ten random sample points. Few fields are wider chan 250m so for each sample field a second set of ten random numbers in the range 1 to 250 was selected. These were the second coordinates, which corresponded to the first coordinates, and located each of the ten sample points within each selected field. The second coordinate wa9 Che number of paces (m) down the furrow from the field canal. Had a sample point fallen outside a field narrower than 250m, it would have been replaced by a ’spare’ random pair of coordinates, but in the event Chis did not occur. The distribution of sample fields is illustrated in Figure 5.-9- At each sample point a number of characteristics of crop and soil were recorded; these and the data are shown in Table 4. Salinity Index was calculated from the EC measured in a 1:2 suspension of samples taken at 25cm intervals to a depth of 150cm. If the two uppermost horizons were non saline then measurement of ECj 2 was discontinued to save time. In the : calculation of SI (see Section 2.3) these two values heavily weight the mean and high salinity at depth has little impact. In Table 4 a real value of SI indicates the complete measurement of the profile and a higher level of salinity; otherwise a descriptive low (l) or very low (VL) is used. Soil texture and moisture characteristics at sampling time were used to assess the likelihood of waterlogging or chemical infertility being hazards to the crop and responsible for yield being less than the potential. In both cases a descriptive code letter was used in their assessment. The yield was estimated from the crop height and its Colour Index (q.v) using the nomogram featured in Figure 6. The derivation of the nomogram is explained in Section 3.6.-10- 3. CROP YIELDS 3.1 Crop Characteristics and Growth Patterns in Cotton 3.1.1 The Roots The root system continues to extend as long as four basic conditions exist. Firstly, there must be an adequate flow of assimilates from the foliage to the sites of root extension and thickening. Secondly, the resistance of the soil to rupture must be less than the strength of the root to penetrate. Thirdly there must be an adequate and sustained supply of oxygen and water from the root environment. Deficiency of mineral nutrients is mainly an indirect effect, slowing the production of assimilates in the foliage. Fourthly, the root environment must not be toxic, of extreme pH nor of high osmotic pressure. There is perhaps conjecturally a fifth condition, namely, that there must be some advantage to be gained from further extension. Given that the first four conditions exist it is an observable fact that cotton roots will branch and extend to considerable depth. At Amibara, roots were found in auger samples from below 2m in some plots. Yet it is equally observable that small, compact root systems will support productive plants grown hydroponically. Generally the observed root system is the consequence of the conditions which prevail. It is therefore impossible to define an ideal root system or an ideal rooting depth in the soil profile. Equally, it is difficult to describe a root system which is on the threshold of limiting the productive performance of the crop. Cotton is one of the majority of plants which increase the absorptive potential of the root system by cellular extension, the root hairs. These are the major site of absorption of water and mineral nutrients. Any suppression of this development has a significant impact on the growth of the crop. 3.1.2 Ionic Absorption Mineral nutrients are absorbed by a partly inactive process, with certain ions penetrating at a greater rate than others. Cotton ordinarily absorbs a greater stoichiometric quantity of anions chan cations. The predominant anion is nitrate which loses it9 electron during its conversion to an organic radical. Chloride is an anion almost as readily absorbed but which is not anabolised. Potassium is the most readily absorbed cation, but sodium at high concen tration in the soil solution will readily enter because selectivity in absorption L9 mainly on the basis of ionic diameter. Cations are either esterified or remain free in solution and therefore retain their charge. The consequence of Chis process, and the need to buffer the vital enzyme processes at optimal pH, is the creation of organic anions. The plant’s ability to maintain thia balance between cations and anions therefore is a function of the capacity to create or destroy organic anions, the composition of ionic absorption and the ability Co excrete cations.-11- The selective excretion of cations is a separate physiological process. Potassium is released into the soil solution in increasing amounts as the plant ages. Sodium is excreted selectively in order to maintain an acceptable level in the epidermal cells. Cotton, a crop regarded as relatively tolerant of salinity and sodicity, is relatively effective at controlling its internal sodium concentration in this way. The so-called ’essential* elements are known to have specific biochemical functions. There are a number of processes which are not ion-specific, so that one ion may substitute for another. Tn this context sodium can have a beneficial role, or at the least be tolerated, unless the concentration exceeds a toxic threshold. Thus a cotton crop, unhindered by nutrient deficiency, is not merely adequately supplied with essential nutrients but is able to maintain the vital balance between cations and anions. In Chis regard both sodium and chloride may either assist or detract, depending upon the overall nutrient status of the crop. 3.1.3 The Physiology of Waterlogging A permanent watertable prevents root extension. Its effect on crop development depends upon its depth in the profile and the ability of the soil above it to fulfil the demands made on the root-zone. A superficial root system is prone to periodic water deficit and effectively a nutrient deficit, since crop water requirements may be supplied from the watertable. Transitory waterlogging, such as may result from heavy rain, flooding or irrigation producing perched watertables at or below soil surface, has a quite different effect but a similar result. Temporary anaerobic conditions create an oxygen deficit in the roots and certain products of anaerobic decomposition are strikingly toxic. Finer branching of roots and the growth of root hairs is suppressed and the normal rate of nutrient absorption is severely reduced. In either event, the waterlogged plant displays the classic sysmptoms of nutrient deficiency. Growth rate is reduced and the normal maturation process is accelerated. This is characterised by the loss of the green chlorophyll (a) and (b) pigments, to reveal the red carotene and yellow zanthophyll pigments, which are masked in the normal green leaf. In other respects the crop is precocious, flowering earlier on a shorter plant, bolls splitting earlier and at a small size. 3.1.4 The Physiology of Salinity For the reasons given above, the cotton plant is tolerant of a high concentration of salts in the soil solution and even of sodium chloride. If it survives the vulnerable seedling phase Che established crop which is salinity affected, may yield satisfactorily. Its appearance is quite different, however, from the crop which is affected by waterlogging.-12- A high tissue concentration of sodium has the effect in all plants of reducing leaf lamina expansion. The leaf is thicker and differential expansion between upper and lower epidermis causes the cotton leaf to curl upwards in a characteristic way. A second consequence, common to all plants affected by sodicity, is the production of more chlorophyll (a) and (b) pigments. Their concentration in a smaller lamina produces a foliage which i9 markedly darker green than a normal crop. The third charac teristic of salinity is reduced moisture content of the foliage as a result of greater osmotic pressure around the roots. A reduced transpiration stream produces a higher leaf canopy temperature, which is detectable on infrared film or by an infrared-sensing thermometer. A fourth charac teristic of salinity, and in contrast to waterlogging, is that maturity is somewhat delayed. In summary then, both waterlogging and salinity tend to produce a cotton crop shorter than normal in stature. The waterlogged crop early in the season becomes marked by yellow and red foliage, the symptoms of fertiliser deficiency, is weaker yet is strikingly precocious in flowering and boll splitting. The salinity-affected crop is sturdy, dark green, with curling leaves and somewhat delayed in flowering and boll splitting. It is a feature of sites like Amibara that salinisation of the soil almost invariably is associated with some degree of waterlogging. As they are largely opposite in effect, the question arises of what is the dominating influence on the cotton crop. As the fieldwork revealed, the more significant effect at Amibara is waterlogging. In isolation and small areas (excepting the abandoned area near Melka Sadi) the salinity may increase from a low to a high level in a short distance. Surface salinity too high at germination causes seedling death and subsequent bare patches. These are characteristically ringed by 9turdy, dark green cotton showing all the symptoms of salinity. Beyond that generally there is a rapid merging from the dark green to the bright red and yellow foliage of water logged soil. 3.1.5 Growth Patterns, Irrigation and Yield Upland cotton has a somewhat indeterminate growth pattern; that is, its growth pattern is not rigidly predictable but responds to features of the environment. For this reason it is convenient to describe the stylised pattern and then discuss the effect of variations in the environment upon it. The stylised growth pattern is illustrated in Figure 7. The greatest rate of vegetative growth is around 45 days from planting. This point is characterised by the attainment of the most efficient leaf canopy; that is the Leaf Area Index (LAI) is optimal, and the Net Assimilation Rate (NAR) reaches its peak. In cotton, Like most plants, Chis condition stimulates flowering, which in turn makes a fundamental change in physiology. Assimilates progressively are diverted from further vegetative growth, the race of which slows. With good management, vegetative growth will diminish almost to cease by 120 days (approximately October), while flowering reaches a peak at about 75 days.-13- Bolls develop slowly at first, but the rate develops exponentially to reach a peak, some two to three weeks after the point of maximum flower production. First-formed bolls are low down in the canopy and are slow to reach maturity. They generally split from 90 days after planting but boll splitting accelerates and reaches a peak rate around 100 days (late September) enabling a first pick in early October. To some extent this is a self-regulating process yet it can be manipulated by crop management. As indicated above, waterlogging accelerates and salinity delays the process. Another important feature at Amibara, now apparent in the light of experience and from observation of growing crops, is the impact of the irrigation schedule. Like most semi-xeromorphic crops of indeterminate growth pattern, cotton has a survival mechanism for its seeds. The response of the plant to excessive moisture (from rainfall, irrigation or waterlogging) during early and peak flowering, is to abort flowers and squares, to ensure that seeds are not exposed to rotting before being matured and dispersed. Thus excessive moisture between 50 and BO days causes the natural process to reverse; the loss of the ‘metabolic sink’ (the flowers and squares) diverts assimilates back into vegetative growth. Stems elongate and new leaves are formed. The crop grows to an excessive height and, providing there is no impediment, the root system extends deep in the profile. More seriously, the formation of new leaves increases the LAI beyond its optimum and the NAR may fall. That is, assimilates relative to synthesis . Any bolls which survive in a microclimate condusive the leaves become less efficient, respiring more the amount produced in daylight hours by photo the period of shedding develop deep in the canopy to boll-rot. Pest control by aerial spraying becomes less effective as droplets fail to penetrate the canopy. Bollworms survive on lower bolls adding to the loss of crop. New bolls subsequently form high in the overgrown canopy but the peak of boll-splitting is substantially delayed. Harvesting likewise is both delayed and extended, and on account of the added difficulty of moving through the crop and reaching bolls, is slower. Irrigation ceases perhaps before all the bolls are fully developed and either crop eradication is delayed, with serious consequences for pink bollworm control in the following year, or the late crop is destroyed by grazing animals. Thus the apparently luxuriant crop has a lower yield potential and is much more difficult to manage. Water deficit during vegetative growth and boll-splitting causes yield loss, yet during flowering ensures the survival of the early bolls and a manageable crop. The most effective way of idealising the moisture regime is by varying the depletion factor, the proportion of available moisture in the rootzone which is consumed by the crop between irrigations. In effect this means a longer irrigation interval during the 50-80 day period; this is discussed in Section 5.2.1, As illustrated in Section 3.6, the most extensive problem at AIP is temporary waterlogging, which tends to be a feature of most and parti cularly the early irrigations. Its consequence is a crop which generally is shorter, and more precocious than the potential crop. These two features are advantageous but the price is a substantially reduced yield.-14- If surface and subsurface drainage and water control are unproved without the corresponding changes in schedule to avoid moist conditions during flowering, then the resulting crop will be luxuriant and unmanageable. 3.2 Crop Characteristics in Bananas Bananas are a perennial crop with considerable foliage production but a short, compact and adventitious root system. The crop originates from a humid tropical area so it is inevitable that its physiology should contrast with that of cotton. The banana has much less control than cotton over its transpiration stream. On days when evaporative demand is high, particularly during the period March to July at AIP, absorption and translocation of water is likely to be exceeded during part of the day by evaporation loss. The banana, lacking the semi-xeromorphic features of cotton, i9 unable to cope with reduced turgor and growth declines. This is so even at an ideal soil moisture content but would be aggravated by moisture deficit prior to irrigation. Being a forest species it is not adapted to high wind and whereas some leaf shredding is thought to have little effect, the damage, particularly by wind-borne particles, clearly is serious at ATP. The root system of banana is very different in form to that of cotton. It has only limited capability for extension, with roots normally not penetrating deeper than lm even under ideal soil conditions. Roots are succulent and like cotton are very sensitive to temporary waterlogging but unlike cotton are sensitive to moisture deficit. Under ideal conditions, in which the soil remains moist but never waterlogged, the roots form a dense superficial mass extending out in a radius of more than 4m. This does not happen at AIP on account of the surface soil drying out after irrigation, and there is evidence that roots penetrate more deeply, unless impeded by a watertable. Unlike cotton, the banana is exceptionally sensitive to both salinity and sodicity because the crop does not encounter such conditions in its native environment. The osmotic effect, reducing transpiration, has a serious effect on yield. The chloride ion is readily absorbed but the banana expends considerable metabolic energy in excreting 9odium ions, attempting to maintain a low tissue concentration. The plant therefore is not successful in maintaining an optimal balance between cations and anions in order to buffer cellular pH. The usual effect of sodium in inducing dark green foliage, is not observed in bananas. The predominant effects of both waterlogging and salinity are yellowing of the foliage and reduced vigour, symptoms of nutrient deficiency. 3.3 Review and Analysis of Historical Yields 3.3.1 Predicted Cotton Yield Trends In the Italcon9ult 1969 Feasibility Study (Ref 5) for the Melka Sadi - Amibara Irrigation Project, predicted cotton yields for the purpose of project economic evaluation were estimated as 16 quintals/hn seed cotton for the first 10 years of project life and 22 quintals/ha thereafter.-15- The Angelele and Bolhamo Feasibility Study (Halcrow 1975 Ref 6) was more optimistic, and predicted yields on State Farms rising from 20 quintal9/ha initially to 30 quintals/ha in year 6. Following two years cropping in Melka Warer and Melka Sadi State Farms yield predictions for the next 6 years were made in the Interim Report /Halcrow 1982 Ref 3). A yield decline to 32 quintals/ha in the second year was predicted initially, followed by a yield build-up, in subsequent years, to a maximum of 36 quitals/ha in the sixth year. The pattern of initial yield decline followed by yield increase was based on management rather than technical factors. 3.3.2 Recorded Cotton Yield Trends Average cotton yields for the State Farms within the project area have been compiled for the period 1977-1983 from State Farm and AIP records. The data are presented in Table 5. Inspection of these data shows that for Melka Warer and Melka Sadi State Farms, yields have declined as the irrigated area has increased. The most recent average yields are lower than predicted. The reasons for the decline have been examined in some detail in previous reports (Halcrow-ULG 1983, Halcrow 1984, Ref 3 and 2). 3.3.3 Variation in Cotton Yield Within State Farms By 1983 about 602 of the developed area of the State Farms had been planted to cotton (5,252 ha) and that 19 the latest year for which yield data are available. The average yield by field is shown in Figure 8. The data have been classified for mapping but the range between class 1 (7 41 q/ha) and class 5 (<15 q/ha) is considerable. Just as average yield declined as a greater area was planted, it is clear that in 1983 the highest yields generally were achieved on land being planted for the first time (blocks 1, 2 and 11). The exceptions are the centre of block 3 and the western sides of blocks 6 and 11, which had been planted only one season previously. The areas of lowest yield are the southern part of block 3 (near the Pilot Area), the north-western side of block 4 (adjacent to Melka Warer) and most of blocks 5 and 6. Figures 9 and 10 show the yield change (for fields for which a comparison is possible) between the 1983 yields and those of 1982 and 1981 respectively. The change in yield is recorded in terms of the change in the number of yield classes, the yield classes being defined on Figure 8. Most yield changes are in fact yield declines. Comparison of Figures 9 and 10 shows that decline is a fairly steady process with most fields declining by only a single class per year. However, the yield decline in block 5A, since first planted in 1980, has been more than 20 q/ha. The yield of the eastern strip of block 3. adjacent to the main canal, was not good when it was first planted in 1982 so the yield decline to 1983 is relatively modest. 3.3.4 Analysis of Yields and Yield Decline The amount of variation in yield within fields makes a successful analysis of historical yield data on a f ield-by-field basis extremely unlikely.-16- A previous report (Ref 3) had suggested that the amount of cut during levelling operations was influencing yield. Fields adjoining the main canal in block 3 had been deeply cut on account of the steep rise in slope at the eastern end of the fields and yields were poor. A regression analysis of 1983 yields on volume of cut in block 3 fields produced an Re value of only 2.52 which tended to confirm the belief that this kind of analysis would not be worthwhile. Variance of block 4 yields in 1982 and 1983 was analysed with respect to whether or not the fields were pre-irrigated and the number of post-plant irrigations given. Tn both seasons, pre-irrigat ion increased yield substantially (q/ha): (Block Without 4 only) Pre-irrign. With Pre-irrign. SE diff. 1982 30.8 35.7 * 1.53 * (Significant) 1983 22.7 31.0 + 1.10 *** (V.highly significant) It is probably significant that with increasing area of cotton to be planted, a considerable number of fields were dry-planted to avoid excessive delay, and this and delay are likely to have contributed to yield decline. There was some evidence that yield was increased by the number of post planting irrigations which confirms the work from the TAR, Melka Warer (Ref 4, p33). With the objective of attempting to identify one or more factors which could explain the serious yield decline in some fields, four pairs were selected, two each from Melka Warer and Melka Sadi Estates. Each member of a pair has shown little change in yield in the three years to 1983 whereas the yield of the other field has declined seriously. The two fields of a pair were selected to be as close together as possible to minimise the chance of major variation in soil; soil is known to vary greatly over short distance so this criterion is no guarantee. The field numbers and details of their operations and yields are given in Table 6 The greatest yield declines were from 42.7 to 13.5 q/ha between 1981 and 1982 in field 6/28A. and from 44.7 to 15.0 q/ha between 1981 and 1983 in field 4/3/19A. Both fields (and others in their vicinity) are known to have 'serious problems with levels and water distribution' (personal communication by Management Staff). While this is clearly true it is difficult to explain why it should contribute to yield decline, since fields might be expected to have suffered for this reason from the outset. Two possible reasons are that since levelling is an annual operation, in recent years the operators lacked the necessary skills, and secondly, that there may have been some settling of formerly levelled land. Waterlogging was observed to be a particularly serious problem in all fields in which yields have seriously declined. A possible reason for an increase in impermeability of soils is chat subsoiling has not been repeated since the first season and cracks and natural structure may have collapsed. This perhaps is aggravated by the passage of machinery when-17- soil is moist, as is the case for operations following the pre-irrigation and in interrow cultivation. It is possible for clay to be leached by irrigation water and to collect in a compact profile with reduced hydraulic conductivity. A final reason may be the formation of a plough pan, the base of the plough-share smearing and sealing the profile immediately below the plough-1aver. Planting date repeatedly has been shown in work at IAR, Melka Warer, to markedly affect cotton yield (Ref A), with yields falling sharply from an optimal date of mid-May (about 1/9 on the Ethiopian Calendar). Table 6 shows that field 4/2/12, in which yield has slightly increased with time, has been consistently planted close to the optimum date. The latest date in the small sample was of field 6/28A in 1981 planted well into August. There is little evidence however that delay in planting is contributing to the yield decline. Careful study of Table 6 has not revealed any evidence that yield decline can be clearly attributed to the timing of any operation or the number of any of the types of operation. The conclusion must be therefore that the reasons are most likely to be associated with water distribution and water logging as suggested above. 3.4 Effect of Waterlogging on Yield The fieldwork to study this effect is described in Sections 2.1 and 2.2 above, where the Waterlogging Index (Wl) is defined. It was found necessary also to take account of any horizon which 19 impenetrable to roots, and the definition of Rooting Index (Rl), to replace WT, is given in Section 2.4. The data are given in Tables 2 and 3 and in Figures 1 and 2. 3.4.1 Cotton Waterlogging was found to affect plant height; the relationship is illustrated for cotton in Figure 11. A maximum height occurs when the RI has a value between 3 and 5. At dry sites, characterised by a low RI and where the amount of irrigation water applied was less than average (and probably less than the amount scheduled), the plant height was less as a response to water-stress. Above a RT of about 5 the plant height diminished rapidly but was 60 cm at the most seriously waterlogged site (Plot 51), with a RI close to the maximum of 25. Note that since plant height and waterlogging were measured from furrow-bottom, Che real height of plants above ridge-level is some 5 to 25 cm less than the values given. Equally, plant roots had the volume of the ridge in which to survive serious waterlogging of the furrow-bottom, which accounts for the modest growth of Che plants in plot 51. Although plots affected by salinity and chemical infertility and which were judged to have had exceptional irrigation prior to monitoring were omitted, three plots (as indicated) were outlyers to an otherwise good relationship. Plot 53 almost certainly was affected by the leaking canal and otherwise would have qualified for a lower R.I. There is no obvious explanation for the other two points.-18- A9 explained in Section 3.1.3 waterlogging suppresses chlorophyll pigments, producing red. orange and yellow leaves. As an index of this, the product of the Redness Index and Yellowness Score was used as a Colour Index (Cl). Figure 12 shows that waterlogging increased the CI as expected. Saline plots with a high RI are excluded since salinity exerts a dominating influence in increasing greenness of foliage. Saline plots with low RI are included in order to complete the curve, but their inclusion is responsible for the inflexion in the curve. The effect of waterlogging on yield is illustrated in Figure 13. As the Rooting Index increases, the yield falls, but it is noteworthy that the form of the relationship is somewhat different to that with crop height. The significant features are firstly that relatively tall plants under serious waterlogging produce very little yield. Secondly, the very tall plants growing at a RI of 3 to 5 produce a smaller yield than the shorter, and water-stressed plants where the RI is less than 3. The mechanism of this latter feature is discussed in Section 3.1.5 but it is particularly important in the context of improvement of drainage and the irrigation schedule. The form of the relationship, idealised from knowledge of the physiology, is shown by the broken line. A more rigorous form is shown by the linear relationship above a RI of 3. The linear regression is highly significant with an of 91Z. It is recognised that the relationship is non-linear below RI ■ 3, as shown, but below RI - 2 the formal curve is shown a9 horizontal at a yield of 55 q/ha. The reason for this is that, improvement in drainage may be expected to reduce the RI to a limit of about 4, and increase yield accordingly. Beyond that point further yield increase could only be achieved by additional improvement to the irrigation schedule and in water-management. Below a RI of about 2, there is a maximum yield potential of more than 60 q/ha which is unlikely to be achieved. For this reason the maximum potential yield for use in 'with project* estimates is taken to be 55 q/ha of seed cotton. 3.4.2 Banana The methodology successfully employed to establish the influence of water logging on coteon yield was less successful with bananas. The reasons are firstly that yield in bananas cannot be estimated with accuracy in a 9hort study. Secondly, decline in bananas at AIP is due to several factors besides waterlogging and they cannot be independently quantified. Thirdly, variation in the degree of soil compaction between the sample plots is likely to have exerted as great an influence on the growth of the bananas as waterlogging as measured by the RI. The data from the five plots, in which the post-irrigation moisture profile wa9 monitored, are given in Table 3 (plots 3 to 7) and in Figure 2. The best growth was in plot 3 where the soil was clearly more granular and less compact silty clay than in plots 4, 5 and 6. Before irrigation the free watertable was deep (125cm), with saturated soil at 90cm. The profile was not completely saturated by the irrigation and there was some evidence that the temporary rise in watertable was caused partly by lateral flow from outside the plot area. Surface water was less persistent than in other plots and the watertable receded to below 50cm more rapidly than plot 7.-19- The poorest growth was in plot 5 on the opposite side of the same field as plot 3. Yield is zero, plant population sparse and vigour of surviving plants very low. The plot no longer is being irrigated but as Figure 2 shows, the already high watertable rose steadily due to lateral flow once irrigation in the rest of the field began. The WI was exceeded only by that of plot 7 and would have been close to the maximum of 25.0 had the area been irrigated directly. As the area appeared to be something of a drainage basin it is not surprising to observe that the profile is seriously salinized. It is not possible to attribute definitively the decline in vigour to one or other factor but it seems most likely that the rise in the watertable was the origin of both declining vigour in the crop and the salinity in the profile. Plots 7, 4 and 6 represented a sequence in vigour of crop between the best in plot 3 and poorest in plot 5, and the five plots represented an almost linear section across the field and down the slope. The moisture profiles did not correspond to this pattern. On the basis of yield score (Table 3) plot 3 was markedly better than the other four, despite the steady grada tion in average height of pseudostem. It seems most likely that plot 7 was poor on account of the moisture profile, plot 4 on account of the salinity profile and plot 6 on account of some other factor since it had the lowest WT. 3.5 Effect of Salinity on Yield The field work to study this effect is described in Sections 2.1 and 2.3 above, where the Salinity Index (SI) is defined. The data are given in Tables 2 and 3 and in Figures 1 and 2. 3.5.1 Cotton The effect of salinity on the physiology of the crop is explained in Section 3.1. The expected features of salinity effect were observed: darker green foliage, thicker leaves which curl upwards, shorter plants and smaller bolls. Lower plant population, caused by salinity affecting germination, generally produced more vigorous growth in the surviving plants and more obvious 9ympodial branching. Soil salinity generally is associated with some degree of waterlogging which, as shown above, has a marked effect upon yield. In order to dis entangle the independent effects on yield the waterlogging relationship was established using data from plots in which the soil was non-saline. Having done so, it is then possible to make an adjustment to the measured yield of a plot to estimate the likely yield had the plot not been affected by waterlogging; this is shown in Table 7. The amount of the adjustment, for those plots in which the WI had been measured, was calculated as 1.94 x WI. This factor is the 9lope (b) in the linear regression shown in Figure 13. As it was not possible to measure WI in some plots (as they were not irri gated as expected), an alternative and less accurate adjustment using the Colour Index was used. Adjustments in Table 7 in brackets were calculated as 2.35 x CT, where CI is the Colour Index of the leaves and the factor is the slope of the linear regression shown in Figure 13.-20- A number of plots were located in areas bare of crop, where it can be assumed that the surface salinity at planting was too high for satisfactory germination. The effect of surface salinity on germination is shown in Figure 14. Due to variable degrees of leaching during the irrigated period prior to the fieldwork, the salinity of these areas was very variable; the Salinity Index ranged from 1.3 to 6.0 mS/cra. Only a single plot was found (plot 50) where .salinity had increased during the season from a level low enough for the seeds to germinate, to a level almost but not quite high enough to kill the mature plants. Figure 15 shows the effect of Salinity Index on the Wl-adjusted yield. The majority of sample plots were non-saline and the scatter of points around a mean of about 55 q/ha reflects a combination of error in the estimation and adjustment of yields, and the effect of other factors which influence yield such as planting date, soil fertility and previous water management. Plot 50, with an ST of 9 and tiny yield, provides an extreme point. Between those points it is convenient to assume the relationship to be basically linear, with the regression as shown in Figure 15 and an R of 88Z. The points measured in areas bare of crop form a scatter along the X-axis with a mean SI around 4.5. In roost cases a mid-season SI of this value is 2 likely to reflect an early-season surface salinity too high for germination (an EC of about 20 mS/cm), so the more likely form of the relationship is e curvilinear. In Figure 15 Chis is shown as a broken line and represents the threshold beyond which production of cotton effectively ceases. Thus yield-loss from salinity may be expressed approximately, as follows: Z Loss ST 5 1.1 10 1.8 20 2.5 30 3.3 100 4.5 3.5.2. Bananas Plots numbered 11 to 15 in Table 3 were a sequence selected to illustrate the effect of salinity on the growth of newly planted bananas at a site where the watertable was found to be below 2m. As no flowers had appeared, even 12 months after planting in the apparently healthy bananas of plot 12, it wa9 not possible to estimate yield. The effect of salinity therefore is shown in Figure 15 in terms of the average height of the pseudostems (R) (to the petioles of newly emergent leaves) and the score of vigour (V). The two agreed closely, and form a positively asymptotic curve in contrast to that for cotton in Figure 15. Plot 13 in the sequence was markedly sandy, freely drained and non-saline, and undoubtedly was suffering nutri tional deficiency. The other plots are shown in the figure and all fall below the real salinity envelope on account of vigour being suppressed by other factors, principally waterlogging. Thus bananas are confirmed to be highly sensitive to salinity, with a Salinity Index greater than 0.25 seemingly having a catastrophic effect on growth. Section 2.3 described the weighting factors required to show a good relationship between Salinity Index (weighted mean ECl:2) ancj vigour.-21- Root activity surprisingly seems greater between 50 and 100cm depth than at the surface as expected. The reason for this is likely to be that on account of moisture deficit prior to irrigation during about 8 months of the year, the roots are penetrating deeper to retrieve moisture (see Annex C). 3.6 Results of the Sample Survey of Cotton The data are given in Table 4, yield having been estimated from the nomogram in Figure 6. The yield-nomogram was derived from the separate effect of crop height and Colour Index on yield, determined from the Yield-Sample Plots and illus trated in Figures 17 and 18. The relationship of yield with crop height is a airly good one. The explanation for an expected maximum yield at an average crop height of 180cm is given in Section 3.1. The scatter either side of the fitted curve in Figure 17 is caused mainly by salinity and waterlogging, which have opposite effect on the colour of the crop. For this reason the Colour Index is used to adjust the estimate of yield from crop height, the nomogram giving more than twice more weight to the influence of crop height than Colour Index. r The overall potential yield is estimated to be 55 q/ha of seed cotton from the results of the Yield-Sample Plots, in which the greatest unadjusted yield was found to be 66 q/ha. As explained in Section 2.6, each site was coded on the likely hazard which may be responsible for the failure of the crop to achieve this potential yield. A hazard code does not indicate severity of the condition, nor is it likely that in general only a single condition exists at a site. In most cases of salinity, the level of water logging is serious. Chemical infertility is only suspected since neither soil nor foliar analyses were possible. It is associated with predominantly coarse-textured soils, where in a few cases temporary waterlogging due to poor external drainage was observed. An analysis of variance was made to isolate the variance associated with predominant hazard code, field and samples-vithin-fields (Table 8). The distribution of yields by hazard code are presented in Table 9. As a note on the stratification of the sampling procedure used, it is clear that the coefficient of variation of samples within fields was somewhat smaller than that between fields. Considerable variation in crop growth was observed within fields which is the reason for taking 10 samples from within each of only 20 intended fields. Time permitted the sampling of only 15 fields which contributed to the larger CV of the between-fields effect. From these observations it appears that about 6 samples per field and more fields might have distributed the variance more evenly. Using source 2, in Table 8 as the error term, there was a significant difference between hazards in their impact on yield. Salinity and water logging were much the same in both mean yield and in yield distribution, both producing a median value between 35 and 40 q/ha of seed cotton or a yield decline of some 15 to 20 q/ha. Chemical infertility was less serious with a median value above 45 q/ha and a decline of only 10 q/ha, which confirms the conclusions from fertiliser trials done previously (Ref 4).-22 In view of the observed relationship between both salinity and infertility with waterlogging, and because 72Z of the area was regarded as suffering predominantly from waterlogging, then the overall summary is the most mean- ingful from the survey. The estimate of overall yield was 39 q/ha of seed cotton. If the present net area under cotton is 9,000 ha then the distribution of seed cotton production may be sunnarised as follows: Yield Classes (q'ha of seed cotton) 0-9 10-19 20-24 25-29 30-34 35-39 ao-45 45-49 50-54 Z of area Area (ha) Mean Yield (q/ha) Production (c) J/ Z of production Yield loss (q/ha) Production Ion (t) Z of lOBl 1 1 3 11 14 22 22 12 15 90 90 270 990 1260 1980 1980 1080 1350 4.5 14.5 22 27 32 37 42 47 52 40 120 530 2410 3630 6590 7480 4570 6320 0.1 0.4 1.7 7.6 11 21 24 14 20 Total Exoected Production 31.690 t seed cotton 50.5 40.5 33 28 23 18 13 B 3 410 330 800 2490 2610 3210 2320 780 360 3.1 2.5 6.0 19 20 24 17 5.9 2.7 Total Estimated Los: Production 13.31C t seed cotton l 1/ Assuming that only 90Z of crop yield reaches market mainly due to harvesting losses Thus 80Z of the loss of potential production is in fields producing from 25 to 45 q/ha or 60Z of the area. Only 5Z of the area is more seriously affected than this and the reason is mainly salinity. Less than 9Z of production is lost from the 27Z of the area with yields above 45 q/ha. Land suspected of suffering chemical infertility, and where a response to fertiliser might be expected, falls mainly in this latter category. The conclusion to be drawn from this survey must be, therefore, that at present neither salinity nor lack of fertiliser is responsible for much lost production. The loss of more than 10,000 t of potential seed cotton production, or a third of the expected marketable crop this season, is the consequence of moderately serious surface waterlogging of the soil. The marginal economic value of this lost production is considerable and may be assumed co be sufficient to pay for substantial remedial action. The extent to which this action would be successul in improving yield is Che subject of Section 4.-23- 3.7 Other Causes of Yield Loss 3.7.1 Cotton If the expected overall yield of 35 q/ha (allowing for harvesting losses) is reached in 1984/85, then by worldwide standards it is a good achievement. Drainage and water management aside, a potential yield of 55 q/ha reflects a high standard of management. It is understood that the cotton seed generally planted is acid-delinted Acala 1517/70 originating from Tendaho. This cultivar is known to be resistant to bacterial blight (and very little was observed during field work), but in trials in 1974/75 at the LAR (Ref 4, p.23) the average boll weight at 6.839 was significantly less than most other cultivars on trial. In 1982/83 the estate average was said to be 6.3 q (Ref 3). Table 5.1 shows that the average in 1983/84 is likely to be only about 5.50. This apparent decline and the observation in the growing crops of massive heterogeneity on the basis of crop height, suggests there may have been a serious decline in genetic quality of the seed stocks. The effect on yield potential is difficult to predict but it seems most likely to be worth up to 10 q/ha in average yield. For reasons discussed previously, any tendency to rankness of growth creates a number of management problems and may in itself lead to a loss of yield. Work at the IAR consistently has demonstrated that the period which is optimal for planting is relatively short; it begins in mid-May, and crops planted after mid-June are liable to yield substantially less. The planting period in previous seasons has been very protracted on account of machinery problems. In 1984 the situation is said to have improved with planting beginning in mid-May and the majority completed by the end of June. This is undoubtedly one factor which has contributed to the expected improvement in yield in 1984. There is potential for perhaps a further improvement of 5 q/ha by narrowing the ’window’ even more. The analysis of previous data showed that pre-plant irrigation signifi cantly improved yield by up to 8 q/ha (Section 3.3.4). In 1984 the majority of fields were pre-irrigated compared with around 20X only in 1983. This difference is likely to have contributed some 5 q/ha to the expected yield improvement in 1984/85 but leaves little scope for further improvement on this account. Achieved plant populations varied from 16 to 79 thousand plants/ha in Yield-Sample Plots (Table 1). There was little evidence that plant population affected yield except in very saline and waterlogged plots where the population was less than about 25 thousand plants/ha. This range exactly corresponds with that tested by the IAR (Ref 4, p.28) and found to have no effect on yield. It appeared that seed rates were higher in fields where germination problems due to salinity were expected. A plant popu lation above about 50 thousand/ha equivalent to an in-row spacing of 22cm in 90 cm rows) is likely to produce excessively rank growth under good growing conditions, with consequent management problems.-24- Pest and weed control was at a reasonable level in most fields though considerable variation between fields was apparent. Weed competition is most serious in areas where the vigour of the crop is less, on account particularly of waterlogging. This increased competition will contribute further yield loss in the cotton crop, overall perhaps by 2 or 3 q/ha (but by up to 10 q/ha in fields most seriously affected). Cyperus spp. is the dominant weed in waterlogged areas and Portulaca spp. in saline areas, both of which are notoriously difficult to control in cotton. Where Che cotton forme a dense canopy only an occasional Hibiscus and Abutilon spp. escapes hand weeding, although two vines (a cucurbit and an Ipomoea sp.) are commonly troublesome to pickers. During Che period of the study, major pests were adequately controlled. There was some evidence that pink bollworm earlier had been problematic in certain fields producing serious boll-shedding. Aphids and Red Cotton Mite were becoming prominent pests in some fields in August and serious by November. Spraying to control aphids was not strikingly successful and that to control the mite (reputedly the most serious outbreak recorded at Amibara) was too late. Some yield will be lost on account of these pests, but since it will be in the very late season crop, only a proportion of which is likely to be harvested (and which mainly contributes to the expected 'harvest loss’), the overall loss will be small, perhaps 2 q/ha. 3.7.2 Bananas By the international yardstick of intensive, irrigated bananas the yield at Amibara is very low. By no means the only reasons for this are high water table and salinity. It seems likely that improvement in land drainage will have only modest impact on yield unless there i9 an overall improvement in husbandry. The three components of yield in bananas are the rate of bunch formation, the number of bunches per unit area and average bunch size. Yield is low on account of all three factors being below expectation. Field 3A2 was replanted in November 1983 but even apparently healthy plants had not flowered 12 months later. In older plantations the appearance of fruiting pseudostems suggests their age to be perhaps 18 months at harvest. The fruiting-interval would be expected to be about 13 months initially, falling to about 10 months thereafter. The reason for the apparent delay seems to be lack of vigour, particularly evident in the ratoons. Maiden suckers are planted at a spacing of 2.2m square, fairly standard for cultivar Dwarf Cavendish. Matt distribution in older plantations is extremely irregular, particularly as fields are said to be replanted every four or five years, and density varies widely both above and below the optimum. Some 2000 pseudo9tems/ha should be fruiting at any time unless there is marked seasonality in growth, vigour is diminished or the matt population is beLow optimum. During this study the value was judged to be about half of expectation even in the better fields, perhaps for all three reasons.-25- I HI R A R Y Bunch size is small with both number of hand and size of hands below expectation. Average b kg but this must apply to selected bunches, perhaps inte for 5) to be export. By observation, average bunch weight in October/November was assessed to about 6-8 kg only. The reason may be seen in the general lack of vigour < the fruiting pseudostem and its age. Typically it bears only three effective leaves, the remainder being seriously yellow or dry. The effective leaves characteristically are inclined upwards, as if unable to fully expand and extend out to the horizontal. Once flowering has begun insufficient assimilates are entering the peduncle to sustain the develop ment of more than 5 or 6 rows of female flowers and the remainder abort. Those that set develop slowly and mature at inferior size. These are the symptoms of poor general vigour and probably a number of factors are responsible, of which waterstress, nutrient deficiency, pest and weed control and matt-management are prominent. The present irrigation schedule produces an apparent overall water deficit during eight months of the year (see Annex C). The real deficit is likely to be much smaller, partly on account of roots tapping the high watertable in much of the area. However, it is impossible to schedule frequent but small applications by surface irrigation of small basins. A deficit will develop during a 14 day interval before the next irrigation in the drier months. Bananas have a high nutrient requirement particularly of nitrogen and potassium. Urea reputedly is being applied at the rate of 6 q/ha twice per year (equivalent to 550 kg N/ha). From the appearance of the crop only the newly planted and most productive areas are indeed receiving nitrogen and it seems unlikely that the rate is this high. This policy probably is justified since an economic response is unlikely in most of the area. The only other fertiliser in store is triple superphosphate but apparently this is not being used. Soil analysis shows that the level of exchangeable potassium (K) in absolute terms is sufficient for crop growth. Bananas though are one of the most K-consumptive crops. Furthermore, an absolute criterion of In sufficiency is not adequate when the exchange complex is dominated by divalent cations and cations compete during absorption. Soil pH at Amibara generally is well above neutral, at a level where orthophosphate (P) is transformed into an unavailable complex. For these reasons it is extremely likely that banana growth and development are being seriously limited by deficiencies of these two nutrients. Minor element deficiencies also are possible although no clearly identifiable foliar symptoms were seen. Serious levels of root/stem weevils and nematodes have been identified in the past. The root-burrowing nematode (Radopholus similis) was the subject of control measures by the IAR (Ref 4) where a marked response was recorded. A soil-injectable nematocide/insecticide was used regularly in the banana plantation when it was under private management but no control measures are 12 be of-26- currently employed. Furthermore, the normal field sanitation of removing or chopping and dispersing old pseudostems to discourage weevils is not being practiced. The few roots and stems examined in this study were not obviously affected by either pe9t but there are many reports from previous studies and from field workers which suggest that a significant part of the poor vigour could be attributed to this cause. Weeds are controlled by manual slashing but only after a serious level of competition has developed, and hardly at all in the more open canopies. In newly replanted fields, weed competition almost certainly is contributing to a slow rate of development. In declining plantations, as the LAI falls with decreasing vigour and stem population, so more light reaches the ground and the weed competition becomes more serious, which in turn further reduces the vigour of the matts. Therefore weeds are contributing to the unusually rapid decline of the plantations. Matt-management is considered an important feature of preserving the long evity of a plantation. The replanting cycle in comparable plantations normally is at least 10 years, more than double that at Amibara. Plantations older than 100 years are known. The objective is to maintain the ideal number and distribution of fruiting pseudostems and to avoid matts 'wandering into clusters or declining. No matt-management is apparent in the plantation at AIP. 1 Bananas are very sensitive to soil texture and microstructure in that growth is poor on soils with either excessively good or bad internal drainage. The texture characteristically with the poorest internal drainage is silty clay, since it often forms a weak microstructure compared with a base-saturated clay, and is often compact and of very hard consistence. Such 9oils are normally avoided in Banana plantations. .Areas □f silty-clay and silty-clay loam topsoil are estimated to cover about 45Z of the banana plantation. Neither fungus nor virus disease was apparent at a serious level on the crops inspected.-27- 4. SURFACE WATERLOGGING AND SALINITY In recent years more cotton production has been lost at Amibara as a direct result of surface waterlogging and salinity, than from a rising watertable and its related salinity or for that matter, from any other single technical cause. In 1984 it is estimated to total more than 10,000 t of seed cotton (Section 3.6). This source of loss has almost stabilised, whereas that from rising groundwater will increase progressively, to lead ultimately to the probable collapse of the enterprise, unless measures are taken to improve sub-surface drainage. The Rooting Index was used to measure the degree of waterlogging and its effect on yield, and the Random Sample Survey revealed the magnitude of the problem of waterlogging. The Rooting Index makes no distinction between the possible causes of saturated 9oil in the rootzone, which are: low soil porocity, perched watertables, excessive water application and a high level of groundwater. The latter cause, poor external drainage, is the main subject of this study and is considered in detail in other Annexes. Low porocity and perched watertables arc not necessarily related but together account for the state of Internal Drainage. 4.1 Soil Porosity The pore-size distribution in soil is important in ensuring the ideal combination of moisture storage and gaseous exchange with the roots. In a compact soil of low hydraulic conductivity, the pore-size distribution is skewed towards the finest pores, which are non-draining, and air only slowly displaces water as it is drawn cowards the roots and surface by mass flow. After thorough wetting such soils remain saturated for so long as to have a detrimental effect on crop growth. This is an inherent feature of certain soils, particularly silty clay and silty clay loam. It is influenced only slightly by tillage and the organic matter content of the soil, factors under control of management. The greatest concentrations of compact, fine-textured soils are in the vicinity of Dinile and Che western half of Algeta, but smaller pockets are scattered throughout. Considerable areas of somewhat less-seriou9ly compact soils are present and which with careful management are satisfactory for cotton. Compaction by machinery and over-irrigation interact with the nature of the soil to produce extensive areas at present in which, particularly after planting, soil saturation persists to Che detriment of the crop. Bananas are more susceptible to thia condition chan cotton, and such soils are normally avoided for banana plantations. Tracts of these soils are present in the MSAE and prolonged saturation was observed during field work (Fig. 2). Yield potential is limited 3nd will not be improved by deep drainage. This has been taken into account in yield projections of bananas and of cotton at Algeta.-28- 4.2 Perched Watertables Most soils at Amibara show some degree of stratification and compact horizons of silty clay are common, even in predominantly coarse-textured profiles. Although often narrow, they are of such low hydraulic conduc tivity that saturated horizons above them can persist for days after irrigation. Most of the observed profiles, which are illustrated in Figures 1 and 2, exhibit some evidence of temporary perched vatertables. Due to the uneven distribution of root activity in cotton, revealed by the need to use very different weighting factors in surface and deep soil horizons, the impact of a perched watertable above is many times greater than below 80-90cm depth. This is the approximate maximum depth of mechanical subsoiling (ripping), designed to break the impermeable horizons. Soils in which stratification is least apparent are those mapped in Annex A as very fine-textured vertisols and ’undrainable' soils, and those as coarse-textured soils. The former group are those described as having seriously low porocity. The coarse soils occur scattered throughout the estates as small pockets too small to map except for somewhat larger consolidated tracts in the banana estate and near Algeta. Tn the Random Sample Survey (Table 9) about 9Z of the area was identified as predomi nantly coarse textured and apparently so well-drained as to produce nutrient deficiency in the crops. The remainder, some 70 to 80Z of the estate land, has soils which show some degree of stratification. Tn perhaps half of this area, transitory though significant watertables are perched on relatively impervious horizons in the surface soil. Tt is impossible to be precise about either the amount or location of these soils due to their small-scale heterogeneity and the amount of field-work necessary to identify them. 4.3 Microtopography Land level was measured along a few selected sections passing through (or close to) sample plots, as shown in Figure 4. This revealed serious variability in microtopography within the designed slopes of the scheme. Doubtlessly, contracted land-levelling initially achieved more-or-less the slopes intended to optimise water distribution, even though this proved more difficult than had been envisaged by contractors. Equally likely is that small-scale variations remained, which have since been aggravated by subsidence of in-fill and irregular annual tillage and smoothing. The result is land characterised by high spots, steep slopes, levels, depres sions and, of course, areas of water sufficiency. High spots and steep slopes are apparent in Figure 4.1 (both sections), Figure 4.2 (Section AB) and Figure 4.5 (field 3/2/24). Plot 11 on Section AB of Figure 4.1 on a relatively steep slope, produced the post-irrigation moisture profile illustrated in Figure 1.3. Plot 16 on Section AB of Figure 4.2 was on the steep side of a high spot and received no surface irrigation water. The moisture profile of plot 16 in Figure 1.5 showed that during irrigation water moved laterally into the profile; the soil surface remained quite dry meanwhile. Although the section through plot 4 was not measured (because passers-bv had removed the plot posts), the moisture profile (Fig. 1.2) showed it to be located on the summit of a high spot.-29- Tt wa9 clear during field-work that the crop on high spots and on steep slopes was showing symptoms of water-stress: short stature, green but sparse and, in late season, necrotic leaves, and precocious boll-splitting. Where salinity was not serious, the yield was good despite the growth form, making such plants the ideal for the estate. Level areas and depressions were often associated. Plot 20 on the section in Field 1/2B/56, illustrated in Figure 4.3, was located on a wide level area of vertisol all of which showed symptoms of excessive standing water (even though the moisture profile, after an abnormally light irrigation, did not confirm this). Plot 51 was located in a level area along the section through field 3/2/25 (Fig. 4.5). The crop was submerged for several days as shown in Fig. 1.9, but the section revealed the reason to be not only the lack of slope, but the ponding of water by a high spot further down the furrow. The moisture profile in this case was excep tional, in Chat surface water persisted and the soil remained saturated to 50cm. This was due in part to a leaking canal, but the soil below 50cm, a very compact silty clay, was of particularly low hydraulic conductivity. Depressions were not invariably characterised by compact soil. Plots 8 and 9 (featured on sections CD and AB, Fig. 4.1, and moisture profiles Fig. 1.3) were of medium textured yet very stratified soil. By contrast plot 17, in a marked depression (Fig. 4.2), was almost unstratified, coarse sand with groundwater at about 75cra (Fig. 1.5). This plot was inundated by run-off from the surrounding basin, yet on the third day after irrigation stopped, surface water rapidly disappeared revealing this to be a site referred to as a ’sink* to the watertable. Areas of water-sufficiency tend to fall between areas of excess and deficiency. They are characterised by exceptional growth for reasons explained in Section 3.1 and regions of such plants are evident in most of the sections illustrated in Fig. 4. 4.4 Irrigation and Leaching The present irrigation schedule is designed to ensure an adequacy of water in three years out of four in the parts of a field most difficult to irrigate. This is achieved by using the 75Z probable rainfall and a field efficiency of 0.7 in the calculation. This implies that regions of the field, and particularly in the early part of the season, receive water far in excess of the consumptive use by the crop. Added to this is the clearly evident problem of ’eaking field canals. At its upper end the field canal may be irrigating for several hours then leaking into the field for a further six or more days, while the remainder of the field is irrigated. Some canals were observed to be left running for much longer without being U9ed to irrigate. There is a clear interaction between water application and the features described above, in producing excessive soil saturation, but also in avoiding or encouraging salinity in the surface horizons.-30- The salinity profiles of plots 4, 16 and 33 (Figs, 1.2, 1.5 and 1.8) demonstrate the consequence of inadequate irrigation. Tn plot 4, leaching is adequate to 40cm but insufficient below that to avoid a significant accumulation of salts in the root-zone. In plot 16 leaching is moderate below the depth of intrusion of water by lateral flow but above that depth, salts are accumulating. Plot 32 is close to the main canal near Melka Sadi village in the area deeply cut to achieve the design slope. Levelling is not satisfactory and irrigation water breaks across the ridges to flow parallel to the field canal, leaving a deficit in the area. The soil is known to have been saline-alkaline before development and the evidence of inadequate leaching is apparent in the salinity profile. The section through field 3/2/24 (Fig. 4.5) is an example of leaching being influenced by slope. From the canal down the furrow, the crop gradually deteriorates but flourishes between 20 and 28 m. Plot 42, at that point, showed the highest recorded yield of 66 q/ha and a non-saline profile (Table 2). The next 80 m are bare of crop and the soil very saline (EC in e the ridge was 35 mScm“^). The section reveals a sharp ridge, parallel to the canal, at 27m down the furrow, from which the land falls steeply at an average slope of 0.34X. The ridge clearly is ponding water and ensuring adequate leaching in the vicinity of plot 42. Water which tops the ridge, flows so rapidly down the slope that insufficient infiltrates to leach the profile. The break of slope at 110 m down the furrow ponds water, leache9 the profile and allows crop establishment. Section EF (Fig. 4.4) demonstrates a similar feature, where plot 29 is on bare and saline land on a short but steep slope, whereas plot 28 is located about 20 m further down the furrow in growing cotton. Figure 1.8 shows the moisture and salinity profiles of these plots. Textural stratification is apparent below about 75cm which in both profiles prevents the leaching of the subsoil. Above that depth, sufficient water infiltrates in plot 28 to leach the soil and permit reasonable crop growth, but on account of slope, there is almost no leaching in plot 29. Plots 12 to 15 are a sequence up a furrow in which slope and micro topography cannot explain the yield variation from nil to 38 q/ha. Their moisture and salinity profiles are shown in Figure 1.4. The soil ia markedly stratified but the depth of the perched watertable seems to be wholly responsible for the variation in salinity and yield. Plots 24 and 25 (Figs 1.7 and 4.3) similarly are highly stratified and illustrate the effect of impervious horizons in the upper profile in minimising leachability to the extent of producing an unacceptably saline profile. A similar effect is seen in plot 52 (Fig 1.9), a non-stratified soil, where the whole profile is so impervious that even an excess of irrigation water fails to leach the salts below 50cm depth. It is important to emphasise that in the majority of the plots studied in the field-work, and in those used to illustrate surface waterlogging and salinity, the groundwater level at the time of study was close to or below 2 m deep. Tn some cases it is likely to be contributing to secondary salinisation of the surface during the fallow period, but the phenomena discussed as influencing yield are not the direct consequence of the groundwater level or salinity. A system of drains at 1.8 m deep would have minimal effect in overcoming this source of yield loss, other than to prevent its aggravation by the watertable continuing to rise and perhaps by improving the lateral drainage of perched watertables.-31- 5. OPPORTUNITIES FOR IMPROVEMENTS The main objective of the Project is to prevent the gradual decline of production which is expected to occur as the watertable rises. The means of doing so is the installation of a drainage system, the design of which requires to be optimised in order to maximise the benefit. The benefit merely is to maintain the present level of production, with the return to production on the land which has been abandoned already. Since the Random Sample Survey has revealed the extent of the essentially separate problem of surface waterlogging, it is clear that there is considerable scope to expand the benefit by simultaneously making improvements to water management and internal and external drainage. The problem of surface waterlogging is manifest as much in areas where the watertable is deep, as in the priority areas. Equally, the installation of the drainage system will make only slight improvement to surface waterlogging. Yet improvement in water management, its application and distribution, and in permeability are instrumental in achieving both objectives, so there is considerable benefit to be gained from linking them in a unified project. The benefit will then be derived from an increasing yield rather than a static one. The improvements required to achieve this combined benefit are in the optimisation of drain spacing, the irrigation schedule, improved tillage and in land smoothing. The field-work in this study has demonstrated the present potential cotton yield to be about 55q/ha of seed cotton. Plots unaffected by internal or external drainage problems, or salinity, consistently yielded around Chis value. Lower yield, in most cases, was clearly attributable to either waterlogging or salinity or both. Overcoming these problems must lead to achievement rising from the 39 q/ha gross yield at present to something approaching the potential yield. Tn the broadest figures, the proportional contribution to this increase is likely to be some SOX from improved land smoothing, 30X from improved infiltration and internal drainage and 20X from an improved irrigation schedule and application of water. Section 3.7 discussed the background to the scope for improvement in general crop husbandry. In cotton the effect of implementation would be to raise the potential somewhat above 55q/ha, but such improvements are not assumed in this Project. In bananas, the situation is more complex in that it is believed the present level of production is determined as much by shortcomings in general husbandry as by problems of drainage and salinity. Therefore, in order to justify investment in an improved drainage scheme, every effort should be made to effect a complementary improvement in general husbandry. Such improvements are assumed in the with-project yield projections for Bananas in Section 6. Restructuring of the management of the RRC Settlement Scheme is taking place with the objective of increasing the settlement of families and improving their productivity. It is assumed that the plans and objectives of the scheme will be implemented immediately and that the benefits to a drainage project will improve with time. These more general opportunities are discussed in Section 5.4.-32- 5.1 Optimisation of Drain Spacing Deep field drains, spaced closely, clearly would control the watertable at a level below the rooting depth. Widely spaced and shallow drains would be cheaper but the watertable may rise into the root zone during and for a period after irrigation, producing loss of potential yield. Between these extremes there is an economic optimum at which the benefit, of output minus cost, is maximum. 5.1.1 Cotton For different soil types, drain depths and spacings, the modified Glover- Dunn equation was used to generate sets of watertable recession curves following a simultaneous rise (h ) during irrigation (Annex B). The ten- Q day recession curve was used in the procedure described in Section 2.4 to calculate the consequent Rooting Index (RI) designs. From Figure 13 the expected yield a gross margin analysis was completed. The shown in Appendix 1 of this Annex. The gross margin for each ofthe drainage for each RI was estimated, and sequence ofcalculations is is the ’profit’ associated with each drainage design, so the specification with the largest value is optimal. Using data for an average soil type, both drain depth and spacing were varied initially. To test sensitivity to assumptions about the amount of harvesting loss in cotton (which has been high in recent years) and the loan repayment schedule, the variation in gross margin for different drain spacing is shown in Figure 19. It should be emphasised that the cost of in-field drains only were included in the calculation, since the estimation of main and collector drain costs was not possible for different drain depths. The maximum marginal return (at the optimal spacing) is plotted in Figure 20 for different drain depths. Drains as shallow as 1.0m are ineffective in controlling the fluctuation of the watertable in the rootzone and yield and gross margin are seriously reduced. As depth of drains increases, the gross margin increases to a maximum with drains at a depth of about 1.6m. Were the pro rata cost of the main and collector drainage system to be included, then the marginal return curve would be expected to reach a maximum and then decline, as deeper drains require increasing investment in the whole system. This is shown by the hypothetical curve C in Figure 20. Assumptions about yield loss and repayment schedule have almost no effect on this conclusion, within the ranges tested. Two additional factors were then taken into account: the possibility of cotton being replaced by a more sensitive crop, and the impact of secondary salinisation during the fallow season, the amount of which is dependent on the depth of the watertable. Many possible crops for the area are more sensitive to salinity, if not to waterlogging, than cotton. In a coarse-textured soil, the secondary salinitv at the soil surface from a watertable below 1.8m is about nil. In a fine, compact soil, however, the secondary salinity from a watertable at 1.6m is considerable (Annex R) and a depth of 2.0m is required to minimise the hazard. The proportion of such soils at AIP is small so a compromise depth of 1.8m for the drains was selected.-33- Assuming field drains at a depth of I. Am, the recession curves in soils of different drainability classes were generated. Typical curves are illustrated in Figure 21, together with the zone of soil which is saturated as the irrigation water passes through the profile. It was observed during field-work that the surface horizon of fine-textured soils remained saturated for several days, partly on account of persistent surface water unable to infiltrate, the formation of an incipient plough-pan and partly because of reduced porocity due to sorting and orientation of soil particles at the surface by the irrigation flow. This feature accounts for a much higher RI value in fine soils and a lower maximum yield. The model described in Appendix I was adjusted accordingly, and ho was set to occur one day later in fine than coarse-textured soil. The gross margin was estimated for several selected drain-spacings, in fine and coarse-textured soil, each of four drainability classes (defined in Annex AL). A complete example and the sumnary estimates are reproduced in Appendix 1, but the estimated gross margins are given in Table 10. A quadratic equation was fitted to each data set in order to estimate the optimum spacing. The estimated maximum gross margin at the optimal drain spacing is given in the cable. It is clear that this model tended to under-estimate optimal spacing for coarse-textured soils and over-estimate it for fine-textured soils (see Appendix 1). Since these textures represent extremes, the fitted optima are more representative of the soil textural classes as defined in the Project Area. However, the drainability class ’fine* is a 1ocally-relative term and includes medium-textured soils. As such it corresponds to a soil approximately mid-way between those modelled as ’fine* and ’coarse extremes for drain optimisation. The optima! drain spacing is as follows (in m): 1 Texture Drainability Class A B D C 1 (Coarse) 2 (Medium/Fine) 41 63 85 108 28 43 58 74 Drainability classes do not correspond conveniently with field boundaries but the system is designed as far as possible to correspond to these values and maximise the benefits. 5.1.2 Bananas The irrigation schedule for bananas is very different to that for cotton. Sections 3,4 and 3.5 confirmed bananas to be particularly sensitive to waterlogging and salinity but that the critical rooting depth, estimated to be 1.0m at MSAE, is greater than it is in more humid climates. Using the revised irrigation schedule, the recession curves were generated for the drain configuration estimated to be optimum for cotton. The conclusion is that the watertable is not expected to rise above the critical depth at any point in the year, assuming median rainfall. Drainage configuration suitable for cotton, therefore is recormnended for the banana plantation. Should a future decision be to replace bananas by cotton, then no modification to the drainage system would be required.-34- 5.1.3 Other Crops The area is climatically suitable for a wide range of alternative crops (Ref 4). Those planned for the RRC Settlement scheme, under irrigation, are: pasture, maize, vegetables, forestry and cotton. Pasture and forestry are perennial and would benefit from year-round irrigation. Vegetables for home-consumption are in constant demand, and the return on and demand for maize is such that land should be double-cropped. All these crops there fore have a more protracted irrigation schedule and shorter fallow than cotton. With the exception of some potential tree species, they are all tolerant of a somewhat higher watertable than cotton. The additional consideration is that if legumes are successfully established in the pastures, then periodic rotation of maize, cotton and vegetables on old pastures would be of considerable benefit to them. The drainage system must be designed therefore for the most critical crop. The recommendation, based on the logic used for drainage of MSAE, is that the system designed for cotton would be the most suitable for the Settlement Scheme. 5.2 Water Management Improved water management is one of the factors most likely to lead to substantial improvement in yields. It embraces both the refinement of the irrigation schedules, in the light of experience, and improvements to the organisation of the application and achievement of the proposed schedules. 5.2.1 Revised Irrigation Schedule for Cotton As discussed in Section 3, it is clear that the present irrigation schedule for cotton is producing periodic waterlogging and excessively luxuriant growth associated with boll-shedding. In the early season soil conditions fluctuate between water deficit for short-rooted seedlings and excessive waterlogging and discharge to the groundwater. The parameters used in the calculation of irrigation requirements are shown diagramatically in Figure 22. They are as used previously (Ref 2) except that the moisture depletion factor (1-p) is modified to take account of the observed problems and in the hope of producing a higher yield from a more manageable crop. The available moisture curve for maximum yield (1-p.) Sa.D is shown in Figure 23, together with water requirement (Wn). Irrigation is required at the points of intersection. The intervals have been slightly rationalised as shown to avoid overloading the canal system. The figure also illustrates how irrigation relates to the growth stage of the crop. The net requirement in early irrigation is smaller than the minimum which can be practically applied by a surface irrigation system. This presents no problem, as discussed in Annex B, because there is a need for leaching of salts accumulated during the late season and fallow. Fine and medium- textured soils overlying saline groundwater will require a heavier leaching regime than the majority of soils on the estates. Due to constraints imposed by the capacity of the canal system and the proposed drains, this can be achieved only by an extra preplanting irrigation of the amount presently applied once. It is proposed to slightly increase the early post-planting irrigations compared with present practice, in addition to changing the intervals fairly drastically. The proposed new schedules are shown in Table 11.-35- 1 The most obvious change in schedule is the lengthened interval between irrigations 3 and 4, between days 40 and 80 from planting, or more importantly during the period of maximum flowering. This is designed to stress the crop and discourage shedding of flowers and young bolls, in turn avoiding a stimulus to vegetative growth. It coincides with the peak in the average rainfall distribution curve, so variation in rainfall could have a marked impact on the irrigation requirement in this period. The rainfall used in the calculation is the median so that on average, two years in five or one year in three mav be so dry as to merit shortening of this interval (in those years) to avoid excessive moisture stress. The number of post-p anting irrigations has been reduced from eight to six. This is partly on account of the growing season being reduced by the precocity of the crop. Tn order to ascertain the importance of late-9eason irrigation, a study was made of boll-splitting in relation to the size and form of the crop. The result is described in Figure 24 where it can be seen that the yield loss is dependent on the age of the crop at the Finn1 irrigation and it9 growth pattern as indicated by the Waterlogging Index (Section 2.2). Three ’type' crops are indicated. Crop A is a shortish crop, whose height has been limited by careful control of water, as is proposed in the above schedule. Crop B is the tall and luxuriant crop which results from an adequacy of water during flowering. Crop C is seriously waterlogged, short and precocious. (The background to these contrasting crop characteristics is given in Section 3). If the above schedule successfully achieves crop type A then the sixth and final irrigation at 142 days (shown by the dashed line) would enable the whole of the potential crop to mature. Crop C exemplifies a large proportion of the estates and the presently achieved crop which is seriously waterlogged. Yield loss from this type is small if irrigated at 142 days, corresponding to late October for a crop planted in early June. This observation justifies present practice given that waterlogging is such a widespread problem. By contrast, crop B is the type occasionally encountered on the estate, mainly in small scattered patches within fields. No further irrigation after 142 days would lead co a potential yield loss of 10 q/ha. which is not taking into account the likely further loss due to the protracted crop and its destruction before boll split is complete, more than nine months after planting. It is as important to avoid type B crops as Che seriously waterlogged type C crops. Considerable benefit should result from strict water management according to the schedule proposed above. 5.2.2 Revised Irrigation Schedule for Bananas Analysis of water application in the past three years reveals a marked moisture deficit in eight months of the year, if the irrigation parameters are assumed co be as shown in Table 12 (see Annex C). Tn practice, the crop coefficients shown in that table for the mature plantation, are too high for some of the fields still in production and being irrigated. Many fields are irregular in growth with poor areas of reduced Kc. Furthermore, if growth within a basin is poor in an otherwise good field, then that basin is left unirrigated.-36- Banana Is a crop particularly sensitive to water deficit and in the past, soil moisture depletion factors have been so high as to seriously limit growth potential. On account of other problems it is not clear if this is the overriding limitation on production. Indeed there would be little to be gained by improving water management if there are other limiting factors. A rationalised irrigation schedule for mature bananas is given in Table 12. The maximum permitted depletion of available soil moisture is about BOZ. The result is a schedule of application of mainly 90 mm at intervals ranging from 10 to 15 days, and lOOmn at 8-day intervals during the desiccating months of May and June, ie there should be two or three irrigations monthly. The crop coefficients used in the calculation of the schedule are those of a well-grown mature plantation. Also shown are the coefficients for the crop as it develops from planting in (say) July. Maiden Suckers are planted and take three or four weeks to produce the first leaves, so the Kc value is that of the irrigated bare soil. As the canopy develops the Kc increases, particularly so during the months of high advective energy. From observation, the first harvest follows some 18 months later, after which the canopy should be thinned and the Kc value falls. With time and irregular growth of replacement pseudostems, the harvest period becomes more protracted and the Kc value responds more to season then harvesting. In the early stages of development, the total available moisture is less than the 220mm to the mature crop so that the irrigation interval should be reduced to irrigate four times monthly. By so doing, the time to first harvest could be reduced to 12 to 14 months. The gross application of water is close to the minimum which could be applied by this type of surface irrigation system. For high yield, bananas are intolerant of any degree of moisture deficit and a very short irrigation interval of small gross application is required. The desired schedule can only be applied by sprinkler or drip systems, and it is recommended that a study be made of the economic feasibility of converting at MSAE to one of these. Careful cons’deration must be given to the possibility that frequent applications of water to the typical silty clay soils of the plantation might maintain soil saturation to the detriment of the crop. This is the reason for soils of this type being avoided by high- technology producers. 5.2.3 Other Crops Maize at the RRC Settlement Scheme was observed co be seriously water- stressed. The irrigation system apparently was well-prepared but the problem was said to be shortage of labour. The crop stand was less than half that required and some 50Z of potential yield already had been lost by four weeks due to inadequacy of irrigation. To justify the investment in irrigation and drains for pasture it must be well managed in order to maximise production. A schedule is proposed in Table 13 which consists of 15 applications of 143mm (gross) each time, with-37- intervals varying from 30 days in July to 14 days in December and 12 days in May. It is assumed chat the main canal would be closed for maintenance in the months of January to March. Were the period to be shorter, then extra irrigations would be justified in order to increase the total net application (1500™n) to the net requirement (1800mm). The irrigation of the proposed forestry area would require a schedule much the same as the irrigated pasture. tn the year after planting seedling trees out in their permanent stations the irrigation should be more frequent to compensate for a small rooting depth. The Kc value of mature trees is likely to exceed 1.0 but by that time the roots should be tapping groundwater for part of their requirement. They are likely to lower the level of the groundwater substantially. The installation of subsurface drains would be a measure to permit the establishment of faster-growing species which would be sensitive to saline groundwater close to the surface. 5.2.4 Management of Irrigation Recommendations have been made previously (Ref 8) for changes in the management structure to improve the supervision of the irrigation teams and the application of water. These changes seem not to have been made yet. The problem is threefold: schedules, amount of water applied and its distribution (or irrigation efficiency). Schedules in cotton have been fairly closely observed but there is clear room for improvement. In the other crops there is serious departure from the proposed schedules. Some changes to the schedules are proposed above and it is hoped that it will be possible to incorporate these into the operational plans without delay. Analysis of water application is presented in Annex C where the conclusion is that there is broad agreement between the total water distributed by the supply canals and the previously estimated total requirement of the crops. There are serious doubts, however, about the distribution of this water both in time and space. The problem of irregular distribution within fields, as observed in crop characteristics, has been discussed in previous sections. In part, its solution is dependent on improved land smoothing and infiltration rate. However, it is important to ensure the correct use of syphons, inadequate control of which at present is contributing significantly to the overall problem of surface waterlogging. 5.3 Tillage Included under tillage are both land smoothing and improvement of infiltration rate. Grading of the field slopes was part of the land development, and with the exception of a few fields is considered to have been completed satisfactorily. At present annual tillage in preparation for planting is a five-stage operation: ploughing, harrowing, smoothing, ridging and finally reforming the ridges after preplant irrigation in order to control weeds.-38- 5.3.1 Avoiding Variation in Microtopography Ploughing is presently done by non-reversible disc ploughs pul’ed by small and medium sized, wheeled tractors. Ploughing of the headland away from the field-canal has aggravated the problem of the ’borrow-strip’ formed during construction of the canal, which leads to considerable loss of the precision possible with the use of svphons direct into furrows. Starting furrows and finishing-ridges, the consequence of the use of rigid plough bodies, are a major contribution to the problem of maldistribution of water. Harrowing is by light disc-harrow, quite inadequate to level ridges created by ploughing. There are two possible solutions to this problem: the reversible-plough and the heavy disc-harrow. The most satisfactory reversible plough, now that stones and tree roots have been removed, is the mould-board type. However, its use requires a higher level of skill from the operator. For this reason and on account of higher work-rate and effectively a lower operating cost, the heavy duty disc is the preferred alternative. This unit, such as the trailed ’Rome' harrow, cultivates to 2O-3Ocm, and in a single operation, chops and incorporates residues and prepares the seedbed. Tt is a heavy implement requiring a heavy-duty tractor, which can be either a track-laying tractor or a large wheeled tractor. Recent experience favours the latter alternative, on account of work-rate and operating cost. Operating first along the furrows and then parallel to the field canal, the work output even with the double pass could be as high as 2 ha per hour on lighter-textured soils. This provides an additional benefit in perhaps removing a constraint to the rate at which fields can be planted, the optimal period for which, between mid-May and mid-June, is rather short. It is recommended that when present machinery becomes due for replacement, consideration should be given to a change to a heavy-duty harrow system. Meanwhile operators should be encouraged to use the available reversible disc ploughs, if at aH possible, as was previously advised. Land-smoothing is an annual operation carried out by long base land-plane, and demanding of considerable skill. There are significant but small undulations of a few cm, caused by ploughing and subsidence of ’fill* during the earlier grading of slopes. On a shallow design-slope a minor ridge across the field, over which the furrows pass, creates a deficit beyond it (see Section 4.3). Several cases were observed where impounded water had broken through the ridges and flooded across the slope. Tn some cases the irrigation teams deliberately break the ridge9 in order to drain such areas. Tn the light of experience the conclusion is that this vital operation is not being carried out to an adequate standard to avoid serious problems. Rased on the experience of similar large-scale surface irrigation schemes in arid regions, the recommendation is to replace the present land-planes by more expensive, laser-controlled equipment, rather than to trust in a significant improvement in operator-skill. This is a high-technology solution, but paradoxical1y it demands less from the unit operator since the control is automatic and has a smaller margin of error.-39- An optimal operational unit, most suited to working in fields of approximately 1000 x 250m, consists of a single laser-transmitter on a tripod, and four 80-120 bhp tractors, each with a simple drag-scraper, laser-receiver and control-box. The transmitter is erected about 250m down the field-canal and the inclination of the laser-sweep is set to the design slope of the field. This operation Cakes about 30 minutes. If the field i9 known to be reasonably uniform Chen the four tractor units directly may begin work, driving round in a spiral, in an off-set tandem pattern. Control of the blade, through the controller and the tractor hydraulic system, is automatic. The operating width of the blade is 3 to 3.5q and the speed of the tractor may be fairly high. Output clearly depends on the amount of 9oil being moved but in a well-graded field the workrate of the four-tractor unit may be about 6 ha per hour, or a total elapsed time of 5 hours for Che average field. The accurate radius of operation from the tripod is 300m so it commands half the area of the average field. Where the field is less well graded then a survey of it i9 required before smoothing can begin. A laser-sensitive, semi-automatic survey staff is available, used in conjunction with the transmitter on its tripod, which is impressive in improving survey rate. When the pattern of high and low- spots is known then the suitably trained surveyor can economise on the tractor work involved by directing the pattern of work. Some of the equipment on the market is robust and well proven but the range is small. For this reason tender-documentation must be prepared with great care. Spare units would be necessary and local repair and maintenance facilities essential. Machinery requirements to effect these improvements in the Stage 1 area have been estimated at two tractors and subsoiling tine sets, and one set of laser levelling equipment for the Melka Sadi area, and one complete set (tractor, tine set and laser levelling equipment) for the Amibara area. The cost of this equipment has been included in the project investment cost estimates. An assessment has also been made of the machinery requirements to effect similar improvements in the AIP areas which lie outside the designated Stage I area. Costs for the required machinery, which amount to Birr 1.92 million, have not been included in the investment cost estimates. The corresponding numbers are four tractor/tine sets and three laser levelling sets for Melka Sadi, and five tractor/tine sets together with three laser levelling sets for Amibara. 5.3.2 Reducing Compaction and Improving Infiltration As described in Section 4, 9oil porosity is a largely immutable property of soil. It can be slightly improved by good management of organic matter, but that is impracticable at AIP. Machinery passing over and tilling moist soil breaks down structure and reduces porosity. The merits of the preplant irrigation lie in the opportunity it presents to leach accumulated salts from the surface, co recharge Che moisture profile and to stimulate Che germination of weeds, permitting their mechanical-40- cultivation before planting. The danger lies in the real possibility of the cultivating and planting tractors passing over soil when it is too moist to avoid compaction and destruction of structure. It is recommended that there is greater care to ensure sufficient interval between preplant irrigation and cultivation, for the soil to dry out. Secondly, it may be possible to combine ridge-reforming, cultivation and planting by mounting the implements in tandem on the same toolbar, or by front-mounting the ridger. Operating costs and risk of compaction would be reduced. The widespread existence of compact horizons close to the 901I surface is responsible for considerable yield loss. These horizons cause transitory perched watertables and serious reduction in leaching efficiency. The existence of these horizons was recognised in previous studies and annual subsoiling was recommended. In practice land has been sub9oiled once only, before the first crop. It was recorded in 1982 that the operation was not being carried out satisfactorily (Ref 3). Subsoiling or ripping requires a powerful tractor to pull a tine through the soil to a depth of 80 to 100cm laterally shattering the profile for some distance. It is done before ploughing. The spacing between tines or passes should be about lm and the most efficient method is a double tine unit. In soil of hard consistence the drawbar-pull required is consider able and a track-laying tractor may be necessary. The work-race is slow and it is an expensive operation. Furthermore, the benefit is shortlived since the cracks quickly reseal. It is impossible to predict how much land is affected by compact horizons, sufficiently serious in effect to produce an economic crop-response to subsoiling. A very crude estimate is 4000 ha. A detailed soil survey would be required to identify Che fields which should be 9ubsoiled annually. Meanwhile it is recommended that provision should be made for subsoiling all fields under cotton. The practice should be reviewed from time-to-time, particularly if the results of a survey become available. The number of subsoiling tine sets required to alleviate compaction problems and poor infiltration has been outlined in the preceding section. 5.3.3 Minimum Tillage In recent years the International Centre for Research in the Semi Arid Tropics (ICRISAT) in India has been demonstrating substantial benefits from minimum tillage of vertisols. Crops are grown on permanent raised-beds which in section appear as in Figure 25. A similar system is becoming popular in Europe too. Adapted for AIP Che average interrow spacing would be 90cm as at present. The furrows are established permanently at 180cm intervals but in being 50cm across are somewhat larger than at present in order to take the greater volume of water required by there being only one furrow per two crop-rows. The furrows are permanent tractor wheel-tracks which become compacted. At crop eradication a heavy tractor with two subsoiling tines set at 130cm would pass. The tines on the line of the crop rows would uproot the old plants and fracture the soil to 80-I00cm deep. Evaporation from the open-41- cracks would reduce mass-flow to the soil surface reducing secondary salinisation there. The pre-plant irrigation would be applied as norma’, and before planting the cultivator'bed-former would pass to reshape the bed and eliminate the germinated weeds. The crop is planted on the original lines so that the roots would develop down the soil-fracture. The potential benefits from such a system are likely to be considerable in both improved yield and reduced land preparation costs. The foreseen problems are likely to be from the effectiveness of irrigation and the possibility that leaching of salts from the centre of the bed may be difficult. For these reasons it is strongly recommended that some experi mental plots are established around the estate on different soil tvpes. It may be necessary to vary the slope on the furrows in order to improve the effectiveness of irrigation. An improvised bed-former could be constructed in the local workshop. Bed dimensions may need to be adjusted in order to fit present tractor wheel-spacing. 5.4 Other Possibilities for Improvements This project is concerned with surface and subsurface drainage. In the assessment of benefits the yield of cotton, already fairly high, is presumed to respond only to the improvement in drainage. The present yields of other crops are low for an irrigated scheme and a modest improvement is anticipated. Indeed a yield increase, resulting from more than improvement in soil drainage, might be considered essential in order to justify the total capital investment and the availability of good land and water, scarce resources in an arid land. Section 3.7 discusses the observed causes of loss of potential yield. 5.4.1 Cotton The scope for improvements in cotton was identified as: improved cultivars and 9eed quality; - planting all fields between mid-May and mid-June; fertiliser for particularly sandy areas. Uniformity of seed stocks must be preserved once the mo9t suitable cultivars have been identified. Rogueing of selected seed-stock fields is necessary followed by careful processing and storage of the seed from them. One hectare of seed cotton will produce seed for planting about 50 ha in the following season so that 90me 200 ha of designated seed area is required each year at AIP. This seed should be acid-del inted and dusted with mixed inaecticide/fungicide . The scope for completing all planting in the optimal period is limited by constraints on land preparation and the heavy demand for water on the canal system. There wa9 a big improvement in 1984- which should be maintained. There is scope for further improvement however.-42- Where large uniform areas of sandy and freely drained soil exist a compound fertiliser could be drilled with the seed at planting using a combination seed drill. Further nitrogen could be topdressed after three weeks. The rates required should be determined by soil and foliar analysis and by field trials. Meanwhile an application of SON: 20P: SOK in kg nutrient^ha should be considered, with most of the N topdressed. Small areas could be treated by hand using scoops. The problem of poor weed control in some areas will be relieved by improvement in drainage and crop vigour. Pest control generally is effective. Pink bollworm seems to be a recurring problem in the early season and aphids in the late season. Given the other improvements then the bollworm is the more serious and every effort should be made to control this pest on the early bolls. If mites show signs of reappearance in the numbers of 1984 then control measures at an earlier stage would he of benefit. 5.4.2 Bananas The scope for improvements in bananas is in encouraging a substantial increase in general vigour. Tn addition to improvements in drainage and water management this will require: - improved matt management: - more fertiliser of the appropriate type; better hygiene and pest control. Standard matt management aims to achieve a uniform distribution of the optimal population of fruiting pseudostems. To do this suckers are pruned regularly to ensure developing pseudostems on the side of the matt which faces the greater opening in the canopy. If a matt dies then it should be replaced immediately after disinfecting the area with a fungicide/ insecticide drench, by a healthy maiden or bull sucker. Each matt should bear, in addition to the fruiting pseudostem, one other about half grown and a third at no more than sword-leaf stage. Any supernumerary suckers should be removed regularly. If the leaf canopv is too dense then lower leaves should be removed, as should those which show significant yellowing. Old stems should be cut down immediately after cutting of the bunch. The stem should be cut with care at the level of the corm, and then chopped and scattered. Old leaves and stems form a useful mulch and organic layer and, with attention to pest control, should not be heaped as at present. Poor quality bananas and chopped pseudostems can be fed to stalled pigs (with supplementary feed to balance the diet) and the manure returned to the plantation. This is a useful method of control of banana weevil, which lives and breeds in old stems, unless they are quickly desiccated. Each field could have its own pig-sty to minimise the cartage involved.-43- Some foliar analysis is recommended. Meanwhile four quarterly applications of 60N:IOP:40K in kg nutrient/ha are recommended. The most convenient method is the use of a compound fertiliser of this ratio. Failing that then these nutrients may be supplied bv 130 kg/ha of urea, 55 kg/ha of triple superphosphate, 80 kg/ha of muriate of potash. If mixed, urea and TSP react chemically and the mixture becomes very wet and evolves amonia gas. Muriate of potash is a much finer crystal than the prills of urea or granules of TSP and tends to separate out during transport and handling if mixed. They are better applied separately using scoops. Assuming 1600 matts per hectare, then the scoops may be cut From plastic or metal to hold (when filled level with top) 81 g of urea, 34 g TSP or 50 g muriate of potash. It is not necessary to bury the fertiliser in the soil as 1*9 the present practice. Surface mulch should be moved 2m from the matt, the fertiliser scattered and the mulch replaced. Irrigation should follow without delay. The recommended solution to the reported problem of banana weevil and nematodes is the use of the chemical Carbofuran which is a systemic insecticide/acaricide/nematocide. It is manufactured by the FMC Corporation of USA and Bayer A.G. of Germany and sold under the brand-name of Furadan and Curaterr respectively. It is very toxic if eaten but is extremely safe to handle (LD50 is 8-20 orally and over 2500 dermal ly in mg a.i.'kg). It decomposes in the soil in about 30-40 day9 in hot areas and the decomposition products are non-toxic. It i9 not very soluble in water and is unlikely to leach in drainage water at a concentration sufficiently high to pose a hazard to either fish or people. Applied quarterly by scattering granules at the same time as fertiliser, it provides good pest control within the roots and stems for about four weeks. Granules are available containing 5 or 102 a.i. and the rate of application should be 50 to 20 kg/ha respectively. The corresponding scoop size would be 25 or 12.5 g respectively. If bananas are to be exported and compete with New World production then much greater attention to quality is required, during growing, harvesting and packing. Regular inspection of developing bunches i9 necessary to remove flower bracts and leaves which touch the fruit and cause marking. For harvest, the maturity of the bunch must be carefully judged. Dehanding of bunches should be done in the field and hands packed in stackable plastic crates to be transported to the packing shed. Grading, washing, fungicide dipping and drying are necessary before boxing. 5.4.3 Other Crops The range of alternative crops which could be produced at AIP is considerable but a survey of markets would be necessary before any could be recommended for volume production. None are outstandingly better than the two main volume crops and for this reason no changes to the present cropping patterns are recommended. Nor is an intensification of the land use recoomended for the present cotton growing area. On this area the essential land preparation which at present requires a five stage operation, must be completed during the fallow to ensure that planting dates are met for the following season. The period available extendsapproximately from December to June. Land preparation can be delayed or interrupted through late harvesting or unfavourable weather conditions (short rains during period January-March are unreliable) such that there is little scope for the production of a second crop following cotton. Those crops proposed for expanded production on the RRC Settlement Scheme represent the most favourable for providing a satisfactory mixture of fuel, food subsistence and dietary requirement, cash income and social acceptability. Maize: The maize observed was a poor crop and showed considerable scope for improvement, mainly in water management. Yield projections, with project, assume a steady increase in yield. This will require the achievement of a uniform plant population of 35 to 40 thousand plants per ha, grown without water-stress. Fertiliser should be applied, with about 20N:20P in kg/ha drilled at planting and a further 40 to 60 kg/ha of N topdressed after 3 weeks. This is equivalent to about 45 kg/ha of urea and 100 kg/ha of TSP at planting and about 100 kg/ha of urea topdressed (see comments on fertiliser for bananas). Stalk-borers and corn earworm are likely to be a constant problem and armyworm a periodic possibility. A contingency insecticide provision is recommended and a sprayer to apply it. Granular insecticide placed in the funnel is the best solution to stalk borers but the other pests would require the use of either a granular systemic or a sprayable product. For the purpose of costing the insecticide, a contingency of a single application of Carbofuran has been al lowed. Vegetables: Mixed vegetables have been proposed and the yields averaged. The following data are assumed: Type Z of area Yield t/ha Yield contribution (t) Spinach 30 20 8.00 Tomatoes 10 20 2.00 Green maize 20 20 4.00 Dry beans 10 0.3 0.03 Onions 10 20 2.00 Red pepper 20 8 1.60 (capsicum) Average 17.63 t/ha The fertiliser required co sustain Chis level of production is 50N:20P:25K in kg/ha although the amount and ratio will vary with the type of vegetable, with most nitrogen on spinach and less on the others. Contingency 9prays of both insecticide and fungicide are allowed, for convenience in terms of Carbofuran and Benomyl. The management are advised to maintain a stock of a range of products and knapsack sprayers for loan or hire. Pasture: It is proposed to plant pure stands of Rhodesgrass (Chloris gayana) and mixed leys with alfalfa (Medicago aaciva). Alfalfa has been tried in the area on a commercial 9cale (Ref 8) vithout much success and the maintenance of mixed stands is particularly difficult when the pasture-45- is grazed (see for example Ref 4, pp 137 and 153). As an alternative it is suggested that a well-fertilised pure stand of Stargrass (Cynodon plecto- stachyum) would prove more economically productive and easier to manage. The closely related but less vigorous species Cynodon dactylon outyielded all other grasses and legumes in a trial at Che IAR (Ref 4, pj7 156-158). The grass is best planted as stolons, of which an abundant supply is available locally along canal banks. It is perhaps debateable that the species may rather be C. aethiopicus, but some authorities believe them to be local races of the same species. It clearly is well adapted locally and prolific in growth if well-irrigated. This grass is of particularly good persistence throughout the year, either cut or grazed, in both yield and feed-value. Its fertiliser requirement is considerable and quarterly applications each of 50N, equivalent to 100 kg/ha of urea are recommended. An annual dressing of 20P:25K/ in kg/ha as 100 kg/ha TSP and 50 kg/ha muriate of potash are advised, broadcast before the first irrigation after reopening of the canal. A cheap mounted reciprocating-arm or spinner is recommended for broadcasting fertiliser. The pasture will benefit from the enforced dry period during canal closure <19 long as it is not overgrazed during this (or any other) period. Grass was observed to be too mature when cut for hay at MSAE, so attention to this detail, to ensure high feed value, is advised. Armyworm may appear periodically and a contingency spray of insecticide for its control is included in the budget. Forestry: The objective of irrigated forestry is a high level of production of firewood and building poles for subsistence. Deaffore9ration in the vicinity is now so serious that this is a wise precaution, even though the valuation of project benefits is difficult. It is assumed chat the laud least-affected by high watertable and salinity will be reserved for crops of greater value and sensitivity to these conditions. The choice of tree species then is vital and the suitability list is limited, as follows: a) Tolerant of salinity and high watertable: Acacia auriculiformis Conocarpus lanctfolius Eucalyptus occidentalis Eucalyptus sideroxylon Melaleuca leucqdendron Tolerant of high watertable but not salinity: Eucalyptus robu9ta c) Tolerant of salinity but not impeded drainage: Caauarina equisetifolia Eucalyptus camaldulensis (Septentrionales ( northern race) Proaopis juliflora 3-46- As this crop is long-term it is important that the nursery is established immediately. Tf the area chosen for the plantation is as advised then the choice of species must be from group (a). Even if the land is drained subsequently, planting should be well advanced, so that the choice of species is not much different for the ’with* and 'without' project scenarios. For growth rate, the Eucalyptus are preferred. Of the two in group (a) E. occidentalis is the more tolerant of salinity but in other respects E. sideroxylon is the more versatile and likely to have a faster growth rate. For plantations after the improvement of drainage other Eucalyptus and other species may be considered. Meanwhile it is recommended that a trial of several species is established by the IAR. The nursery area required is about 25m^ of beds for each hectare of plantation. It is advised that not more than 36 ha of plantation is established in each year. A total net nursery area of 0.1 ha of beds, surrounded by pathways, is required. A fence, windbreak, shade over the beds and a good and reliable source of non-saline water for irrigation are essential. About 500 g of Eucalyptus seed is required to produce the 1600 plants required per hectare of plantation. The soil for the raised-beds, about Im wide, should be carefully mixed and fumigated for which plastic sheets and canisters of methyl bromide gas are required. This chemical is very poisonous and should be used with care. Some fertiliser should be mixed with the soil, and contingency sprays of fungicide and insecticide are allowed for in the budget. The area for the plantation should be cleared and marked out with planting stations at 2.5m square. Furrows to facilitate irrigation should be drawn at 2.5m intervals from the field canal, and bunds constructed. Planting holes 25-30cm deep must be prepared in advance, and ideally a quantity of anima! manure and some triple superphosphate placed in each before planting. Seedlings should be in the nursery seedbed for a year before being carefully lifted to retain as much soil on the roots as possible. Transport to the planting area should be rapid, and the seedlings planted without delay. Small batches are advised, to avoid desiccation of exposed roots, and irrigation should follow immediately after planting. The irri- *at ion schedule is discussed above. Regular attention to weeding and pruning will be necessary if a high rate of growth is to be achieved. Animals must not be allowed into Che plan tation until the trees are mature, by which time their canopy should ensure a minimum of weed growth. For this reason, and to discourage casual extra ction of firewood and poles it is recommended that the plantation is fenced, and local people advised of the importance of a well managed plantation. Should subsurface drainage be installed in areas under forestry there is a hazard Chat the tree roots would locate and penetrate the field laterals, special precautions would have to be taken in such circumstances, such as Che introduction of a root inhibiting solution into the laterals or special deepening of the laterals in affected areas. It is however unlikely that Che investment costs of subsurface drainage itself, or of such inhibitors, would be warranted for forestry.-hi- fi. FUTURE LARDOSE AND CROP YIELDS IN STAGE T AREA The groundwater varies in salinity and is projected to rise to an equilibrium depth close to the soil surface (Annex B). The areas where this rise ha9 already occurred, or is expected to occur in the next five to six years, has been designated Stage I area. Watertable characteristics are a feature of the area and the detail of deterioration of soil conditions wilt not be quite the same in parts of the estate outside the Stage T area. Tn principle, the effect on yield will be common to the whole area. The estimation of average yield without a drainage scheme is calculated from parcels of land all at different stages of degradation, which weight the overall average in proportion to their contribution the land area involved. The year in which a drainage project is implemented also exerts an influence on the projected yields since the calculation must be made in real-time rather than relative time. For these reasons, the procedure for estimating overall average yield is based on the real-time calendar and requires the landuse pattern to be established first. 6.1 Present and Future Landuse The Stage I area is divided into its Melka Sadi and Melka Warer components. These are further sub-divided into MSAE Banana plantation and the state Farm cotton in the former area, and Melka Warer and Algeta State Farms, under cotton, and the RRC settlement area, in the latter. Part of the Institute of Agricultural Research (lAR’l is included in the latter area but the benefits of draining the land for experiments have not been evaluated. The reason is that it is not possible to place an economic rate of return on field trials without a full study of the whole Agricultural Research Sector. Further sub-division is on the basis of the calendar year in which it is proposed that the sub-surface drainage works should be implemented. An implementation year is assumed to begin in July and end in June, excepting the final period which is restricted to July-December in 1990. The first implementation year is 1987/88. Three alternative development scenarios are considered for the RRC Settlement Scheme. 6.1.1 Groundwater Characteristics The methodology for predicting the rate of rise in the watertable is described in Annex B. The maps show the boundaries between zones of different groundwater salinity, the thresholds of which are <10, 10-20 and 20-30 These data are derived from observation well records. The model of secondary salinisation in Annex B and the marked skewing of the groundwater salinity distribution indicated the need to have data in smaller classes at the lower end of the range. The transformation wa9 made by plotting the cumulative distribution curve for areas <10. <20 and <30 mScorU The land areas in classes with 4 mScm”l intervals, to 20 mS were obtained by interpolation, after first increasing the conduc tivity values by 60Z. The results are shown in Table 14.-48- The need to adjust groundwater salinity by 60Z was on account of the water in the observation wells being consistently and markedly less saline than the watertable sampled by augering nearby. Tt is believed that the water in the well is mixed during sampling whereas water in the soil matrix becomes stratified, the surface layer being more concentrated than the groundwater at depth. The model of salt balance demonstrates the reason for the accumulation of salts in the surface layer of the watertable. The second reason is that using the original values, the model of progressive salinisation suggested only 60 ha would have been abandoned at present whereas the real area is around 125 ha. The third reason is that ground water sampled by augering had a conductivity as high as 54 mScm“l. and several samples were greater than 40 mScm“i. Mo observation wells were more saline than 30 mScm~l. 6.1,2 Abandonment of Land The average rate of salt accumulation from the irrigation water supplied to cotton is calculated to be equivalent to a fallow-season increase (A EC«) in the plough-layer of 3.15 mScm-1 each year. By the end of the cotton irrigation season most of these salts have been leached down the profile. Movement of salts is dependent upon the depth of the watertable and the leaching efficiency, the latter being a function of soil porosity, the presence of compact horizons (on which watertables are perched) and the efficiency of water management. When the watertable is more than 2m deep the capillary wetting front during the fallow season does not reach the soil surface. As the groundwater rises above that level then secondary salinisation becomes increasingly more significant, as described in Annex B. Annual accumulation from irrigation water and secondary salinisation are separate phenomena. They both contribute to the salinity of the plough layer however, which each year becomes greater. Leaching studies in the Pilot Drainage Scheme, situated on a soil which is representative of the whole area, have shown that 20Z of the salts remain after the first and preplant irrigation. Using the procedure described in Annex B it was possible to calculate the number of years which would elapse, from the time the watertable of a specified salinity reached a specified depth, before the salinity in the plough-laver reached a critical level. The results are shown in Table 15. The critical plough-layer ECe for germination of cotton and maize is believed to be 10 and 2 mS cm“l respectively and the equivalent value for survival of pasture, 20 mS cm-1. The year in which the critical level is reached is the 'schedule year'. Tt is the year when productivity, though having declined gradually over previous years, finally collapses due to failure of germination or a marked fall in yield of a perennial crop. Tlie overriding effect on the data of Table 15 is the rapid rate of rise of the watertable, estimated to be 9 months between 2m and 1.5m and 12 months between 1.5m and 1.0m. Even with saline groundwater, most crops would survive for many years were the water table to remain below 1.5m. There is an annual fluctuation in the depth of-4Q- the watertable due to irrigation exceeding evaporation during the cropping season. Once the annual average depth is above 1.0m then irrigation would be expected to raise the watertable to or just below the soil surface during the cropping season. The level would fall rapidly after the final irrigation and is assumed to be at a depth of 75cm at the start of the fallow. From this depth, secondary salinisation is serious and the schedule year is rapidly reached. There is considerable variation over very short distances, even within fields, due it seems to variation in soil characteristics, and more importantly, in the microtopography. Since this variation at present is manifest mainly in surface waterlogging, the random-sample survey provides the evidence for the form of the area distribution. Values have been adjusted siightly to calibrate the model Co fi t t he area already abandoned, and they are as follows: Year relative to 'schedule -year ’ -3 -2 -1 s + 1 *2 >3 X of area affected - cotton/pasture 5 15 30 70 85 95 100 - maize 0 5 25 75 95 100 100 S - ’Schedule Year’ The first sign of serious trouble in an 'average* cotton field will be in about five percent of its area, which will be so saline at planting that the crop will fail to germinate. As the watertable rises these areas rapidly expand, due to the added secondary salinity, to cover 15Z of the 'average' field in the next season. By the schedule-year, 70Z is affected and at that point it is assumed that in the case of cotton the field would be abandoned. It is therefore irrelevant that 15Z of the area would still support a reasonable crop in the following year were one to be planted. Barren areas, from a few square metres upwards, are visible in all parts of the estate, but particularly near Melka Sadi. They are characterised by soil which is darker brown, which is due to the effect of high pH on 9oil organic matter, giving a wet appearance, which often it is. Around the areas the cotton is often short and sturdy, due to low plant population, and with dark green foliage. Generally these 9aline areas have formed, even though the watertable is well below 2m deep, on account of local variation in leaching efficiency. They are almost stable in extent but form the nuclei for advancing decline as the effect of the rising watertable is reinforced by inadequate leaching. By assigning values to the loss of yield which is expected to occur in the years before the schedule-year is reached, and by relating the yield estimates in relative-time to the areas affected in real-time, shown in Table 14, the rate of abandonment of cropped area is projected. These estimates are shown in Table 16 and 16a. With the RRC Scenario I the area expected to have been abandoned in the 1985/86 cropping season is 1086 ha, most of it old banana plantation and all of it near to Melka Sadi. About 40Z of the net area within the Melka Sadi Stage I boundary already has been or is about to be abandoned. Without a drainage project this figure is expected to increase steadily over twentv years, to 83Z of the area.-50- Tn the Melka Warer/Amibara region of Stage T there need not be abandoned land at present. Part of the RRC settlement area has been abandoned for arable cropping but the area is most suited to the forestry plantation which is planned in Scenario 1. The establishment of tree species tolerant of salinity and waterlogging should be possible. Assuming that landuse is deliberately optimised at the RRC then the proposed cropping pattern should not be affected by rising groundwater until 1989 when the first loss of planted pasture may be anticipated. By 1992 subsistence cropping within the Stage T rationalised boundary will no longer be possible and the carrying capacity of the planted pastures will have fallen to about a quarter of their potential. After sixteen years only rough grazing and the forest area will be available for the settlers, and overgrazing is likely to minimise even that. About 90Z of the land presently under RRC admini stration falls within the Stage T boundary and the deterioration of the remainder will follow rapidly on from that shown in the table. With development Scenario 3 and the plan to produce a relatively large area of the salinity sensitive crop maize, the rate of abandonment of land will be somewhat more rapid as schown in Table 16A. The land of the state farms, Melka Warer and Algeta. which falls within the Stage I boundary will not be seriously effected immediately. Once the watertabte reaches the upper horizon in the fine-textured soils of Algeta the land will be rapidly abandoned. The effect on the Melka Warer portion will be delaved by a few years, but on both units within the rationalised boundary, cotton production will cease, as at Melka Sadi, after about 18 or 19 years. It is possible that on account of an increasingly unfavourable ratio between overhead costa and total gross margin from the residual cultivable land, that a decision to cease production might be taken somewhat earlier. 6.1.3 Future Landuse With a drainage project the future landuse more-or-less will be the converse of land abandonment. Even with relatively low yields bananas at present are the most profitable crop grown in the area. Without a study of the national banana sector it is not possible to advise on whether or not they would be better produced at another site. On account of the relative proportion of uncropped to cropped land in the banana estate, it is believed that an equilibrium exists exists in the depth and salinity of the groundwater. Provided this equili brium is not upset by either a marked change in the irrigation regime or a massive increase in plantation on the abandoned fields then there would seem to be no reason to move out of banana production. Without a drainage project it is assumed therefore that the present productive plantation, and the area soon to be replanted, totalling about 450 ha, will continue indefinitely. There is scope within a drainage project for a considerable increase in production, both bv planting-up the abandoned area, which was developed originally for bananas, and by a substantial increase in yield (see Section 5). Tt is considered that as. the ideal planting period is limited to the months of July and August only, then 200 ha is the maximum area which could be planted each vear. At present the replanting cvcle is short, some 4 to 5 vears, so that when increasing the size of the plantation, some resources-51- must be diverted to replanting old fields. A recommended replanting schedule is shown in Table 17, but in practice it would need to be tailored to suit the size and state of individual fields. By 1991 the whole banana estate could be replanted provided that the schedule is adopted in 1986 in anticipation of the installation of drains in the following two years. For the RRC Settlement Scheme, a suggested schedule is shown in Table 18A for the development Scenario 1 proposed by the RRC Management. It is similar in principle to that for bananas but more complex in practice, on account of crop diversity and the more variable groundwater features. As it is intended to accelerate the settlement of families, the infrastructure and cropping pattern is to be reorganised, partly to accommodate Che extra families, and partly to make the scheme self-sufficient in food and fuel. Potential land use in the RRC scheme, under the alternative development scenarios 2 and 3, are given Tables 18B and C. The procedure used to establish Tables 18A and C was to allocate gardens and maize to the land in which Che watertable either has not yet reached lm depth or is in Che lowest salinity class. When land becomes too saline for these crops it is assumed that it may be replanted Co cotton or pasture, thereby minimising the projected rate of land abandonment. As already explained Che rate of abandonment of land is such that the plans of the RRC will have a short life unless the area is drained. With drains installed in 1989 and 1990 the landuse plans could be achieved easily with minimal disruption Co the schedule. Installation of drains on Che land currently in cotton on the state farms and RRC Scenario 2, will ensure their continued production and permit the reclamation of Che relatively small area of abandoned land at Melka Sadi. The with and without project projections of productive cropping area for the various administrative sub-units are shown in Table 19. 6.2 Present and Future Crop Yields Earlier sections of Chis annex have sought to draw attention to the existence of two essentially independent drainage problems. Surface waterlogging and the associated salinity are caused by poor internal drainage. It is a problem which seems likely to have nearly stabilised in its effect on crops, having been responsible for a major part in the overall yield decline in recent years. Poor external drainage is the cause of the rising watertable and its associated salinity. This problem may be said only to have contributed to the abandonment of banana and cotton fields near Melka Sadi and the RRC Scheme. However, projections indicate that its contribution to future decline in productivity will be paramount. The installation of a subsurface drainage system should satisfactorily solve the problem of external drainage so that if indeed the effect of surface waterlogging has more-or-less stabilised, then the benefit should be the prevention of yield decline. Present yields have already declined from those of the crops on newly planted land, and fieldwork has shown that the annual loss in seed cotton production, because of surface waterlogging, may be as much as 13,000 tonne. If internal drainage can be improved then the benefit is to be measured in yield increase.-52- The remedial works to overcome these two problems are very different. To improve surface drainage would contribute almost nothing to the solution of the rising watertable. To eliminate the subsurface drainage problem alone would be to forego the considerable potential benefit of restoring yields to their initial level and perhaps beyond. Furthermore the effectiveness of either measure on its own is likely to be improved by their concurrent implementation. For these reasons it is to be assumed that the Project will encompass both measures. A programme designed to improve surface drainage is recommended to begin as soon as it can be financed and organised, thought to be from 1986 onwards. The implementation of the subsurface drainage system has been scheduled tentatively to begin in 1987 and to last three-and-a-half years. The 'without project' scenario is without either measure. The projection of future yields in their decline to nil, as the land is abandoned, is somewhat more involved than for the ’with project’ scenario, in which a steady increase to a maximuni is predicted. 6.2.1 ’Without Project Yields The greatest influence on yield is the effect of reaching the critical surface salinity in the schedule-year (Section 6.1.2) at which the yield on the affected land is nil. Four other factors influence the average yield: 1 land is not uniform, and different areas are at different stages of decline; some yield is lost before the schedule-year due to profile salinity: yield already is affected by surface waterlogging: some yield is lost by subsurface waterlogging when the watertable is shallow. The effect on yield of these factors is illustrated in Tables 20 and 21. Once the vatertable annual average depth reaches 1.0m, then its level is assumed to fluctuate due to irrigation as shown in Figure 26. The annual contribution to the watertable is 226mm divided between 8 applications. The porocity is assumed to be 0.08Z, so the rise (ho) would be 350™. The fluctuation in vatertable is assumed to be either side of a cropping-season mean of 50cm depth, that is between 32 and 67cm, The surface soil is assumed to be saturated temporarily by irrigation as illustrated. Using the Rooting Index (RI1 concept (Section 3.4), which is calculated to be 7.5 in Chis case, the yield reduction in cotton is estimated to be 28Z. With higher levels of groundwater salinity it is to be expected from Figure 15 that profile salinity would cause further yield loss. These two sources of loss are combined in Table 20. . The non-uniformity of land is discussed in earlier sections (see 6.1.2). The random-sample survey indicated the area distribution of degrees of severity of surface waterlogging. It was observed that in general the surface salinity was inversely related to the severity of surface-53- waterloggxng. Surface waterlogging and salinity therefore can he related in time, relative to the schedule-year, as shown, with the corresponding yields, in Table 21. The area distribution data have been modified slightly during the calibration of the model to match the cotton area already abandoned. It is assumed that better-than-average land would be chosen for maize production by the RRC so the range is smaller. Tn time relative to the schedule-year the average yields are 22. Each yield estimate is a weighted average calculated as shown in Table follows: (yi.*l•li• where y is yield (Table 21), a is (X of area affected t 100) and I is (1 - (X yield toss ? 100)) (Table 20), in year i. (Table 21), Example: Cotton, 3 years before schedule-year, ECwt lOmScnrl; (50x0.05x0) + (45x0.1x0.67) + (40x0.15x0.72) ♦ (36x0.4x0.72) * .... ....(30x0.15x0.72) * (25x0.1x0.72) ♦ (15x0.05x0.72) » 23.3q/ha The value of 1 for year -3 is zero because cotton on that 5X of the land area has already failed to germinate. Example: Maize, 1 year after schedule-year, EC 2mS cm"!; (26x0.05x0) * (24x0.2x0) ♦ (20x0.5x0) ♦ (17x0.2x0.8) + .... wt .... (10x0.05x0.85) - 3.1 q/ha The yields in relative-time are then placed in real-time by the use of Tables 14 and 15. The gross area affected by a high watertable of specified salinity in a particular calendar year in Table 14, is multiplied by 0.87 to convert it to net area within fields. The number of years to the schedule-year is taken from Table 15 to which the average yield in Table 22 under 'S’ applies. Tn the years before and after that year the corresponding average yields shown in Table 22 apply. As an example of the procedure, Table 23 sets out the schedule for the units of land designated for cotton in that part of the RRC which is recommended for drainage in 1989/90, as shown in Table 18. In this table, some 40 ha of land will be affected by the groundwater of EC t 4-8 mScm-l w rising above a depth of 1.0m in 1987. Table 15 shows that 10 years crops 1 may be expected before crop failure. Inclusive of the crop in 1987, the schedule-year is therefore 1996. In Table 22 the expected yield in the schedule-year in a non-uniform field with groundwater of that salinity is 5.6 q/ha. This table shows the yields for most of the state farm cotton area. Algeta on account of very poorly drained, compact soil and the RRC cotton on account of poorer husbandry, are expected to yield slightly less. For this reason the yield in year 1996 is entered in Table 23 as 5.5 q/ha. There are 43 ha of land with the same salinity of groundwater which are affected the following year, the expected yield of which is 15.5 q/ha in 1996. A similar schedule is developed for the land of lower salinity. The production is calculated for the land units of similar salinity in each year, and these are added to estimate total production in that year. The productive area is the total land under cotton less any units which have reached the critical year and been abandoned. The weighted average yield on the still productive land may then be calculated for each year, until production finally ceases in the year 2000.-54- A similar procedure has been used for most crops and land units within the Stage T boundaries. The results are shown in Table 24. The exceptions are bananas at MSAE and forestry at RRC. There is an active period of banana replanting at present which is scheduled to continue. As a result,yields are recovering from previous damage to rise to an average level of production, believed to be about 10 t/ha. The schedule for Scenario I forestry at the RRC is calculated in a similar manner to the example for cotton above. The difference is that land units are the sequence of annual plantings of 30 ha beginning in 1986, this being the amount which is believed possible. The second difference is that during the life of a plantation the yield in of wood of different qualities generally increases as shown in Table 25. Even without subsurface drainage a reasonable level of productivity is expected from carefully selected species. As the trees develop they are expected to maintain the watertable at a lower level but they are unlikely to prevent the gradual salinisation of the soil and a slow decline in production is forecast. 6.2.2 ’With Project’ Yields With the implementation of a surface drainage project annually from 1986 (see Section 5) yields are expected to rise for a few years. With the implementation of the proposed subsurface drainage scheme from 1987 to 1990 the decline in yields due to the groundwater will be prevented and yields will then rise, to stabilise after a few years at levels somewhat greater than have been achieved in the past. Contributing to the rise in banana yields and those of crops at the RRC Settlement Scheme, is an increased standard of husbandry and higher level of inputs (Section 5). The projections of expected yields 'with project*, to include both surface and subsurface measures, are shown in Table 24. Allowance has been made for damage to the crop during the installation of the subsurface drains. Most of these works will take place during the fallow season between annual crops and before the replanting of the more permanent crops. A loss of 5 X of yield in cotton and 5Z in bananas during the implementation year is incorporated within yield projects, which in most cases are increasing as a result of the improvements in tillage and water management. 6.3 Timing of Drainage Installation Tn financial terms drainage installation would be justified when the water table and/or salinity conditions are such as to induce a level of income loss due to a yield decrement which equates to the annuity equivalent of the capital and operating costs of drainage. The average cost of drainage installation in the Stage T area is estimated at some 7000 Birr/ha (Annex D). An annuity equivalent of this sum, neglecting operating costs, would be in the order of 700-800 Birr/ha, based on 25 year life at, say, 10Z rate of interest. gross margin on a 45 q/ha cotton yield is estimated at approx. 3080 Birr/ha (Annex H. Table Hl), such that a yield decrement of some 10-12 quintals would equate to the annuity equivalent of the drainage installation. This corresponds to an approximate loss in yield of some 25Z.With present average cotton yields in the AIP area of some 35q/ha (Main Report, Table3), a reduction in yield to some 2^q/ha due to high water tables and'or salinity effects could justify drainage investment. As many other factors are involved in yield reduction it is impractical to define a specific yield reduction as a means of justifying the timing of drainage. Such reduction would however serve as a useful indicator. As discussed in Section 6.2 and illustrated in Table 20, an appreciable yield loss is sustained when the annual average depth of the groundwater becomes less than 1.0m. At this level of water table a minimum yield reduction of 28Z is estimated, increasing with higher levels of groundwater salinity. On this basis alone the investment in drainage is justified at or prior to the onset of a water table within lm of ground level. Reference to Table 15 indicates that for low salinity water tables, the schedule year for crop failure can occur many years after such a high water table is attained. The recommended rate of drainage installation should be therefore in advance of the projected rate of land abandonment - the latter only occurring when the schedule year is reached. As a fundamental part of the study, projections have been made of the rate of rise of groundwater and attendant rate of land abandonment. Tn preparing the implementation progranvne, the concept employed has been to ensure that the risk of land abandonment is minimised. By this expedient, continued productivity is ensured and the cost of land reclamation avoided. Apart from the relatively small area of land currently abandoned, the proposed implementation programme is designed to ensure that further areas are not abandoned, based on projections of groundwater rise and secondary salinisation. Thus, for example, in the Melka Sadi Stage T area the imple mentation of 2660 ha of drainage is programmed to be completed by 1989, although the projected abandoned area by that time is estimated co be 1326 ha (Table 16)I I I I I I I IREFERENCES 1. AIP Master Drainage Plan, Discussion Document, Sir William Halcrow and Partners, August 1984. 2. AIP Water Management Manual, Sir William Halcrow and Partners, 1983. 3. AIP An Interim Evaluation, Hal crow- ULG Limited 1982 4. Institute of Agricultural Research, Melka Warer, Progress Report, September 1976. 5. Melka Sadi/Amibara Proposed Irrigation Project. Feasibility Study Italconsult, Rome 1969. 6. Angelele/Bolhamo Feasibility Report, Sir William Halcrow and Partners 1.9.1975. 7. Operations Manual. Irrigation and Drainage Works, Sir William Halcrow and Partners, 1983. Personal communication. Local manager, Meat Corporation. 1984.I I I I I I I I I [ II 0 [I G 0 B II U U u un V* 1! -u It TJ i—• -r- K 1 * -., 2 2 • *r 1- *H J2 g □ »- <J M ~c 1“ ■-• —1 a c i-> a a —• O Du P-. £ ■a >V <X 4 * s> ba V«»O* M *♦ *1 a *K nw H : da : *i . —• a r, w« ** »- ^1 £S 2* •-•1 «■ o* ■a- -a. ♦Q h h> cu V« M e. * a. a a aa a. a. •u a «j ». a. M qa ■■ io s ii Cl - fk. a. - ! n/v 4/1/5* 4/191 20/9 76 4 2.2 3 1 J 1 H 4iU r-. .J --■ o • ki UM X a. lai a* -Lr ” 5 •• u qa z a. 4 h u LJ C O a. k* o at - !r XJ s; ,X r? *1 l-M 3* k, 5 2 rB ■■+ ■-. eac a fc o Mn p UJ k- o r-r u A’ *■ "S- •u ■ • rv Sj: Cl '-■ *1 -a i-4 W «4 4 41 H Sa ♦H tr «—• re. /— «W < bbl — - hrn■ 3 n/» 4/1/4* 4/ 14 ID 20/9 M3 1 1.4 1 1 i IL2 4*1 1916 17, L Mb ■ i 4. n 10.6 p 15 57 45 B i.b 1/3 47J 4.45 d.6 m 9 * 2b b J 23/9 4/1/5* 4/ 14 IE 10/9 121 1 1.9 3 t 4 Jfc/9 4/1/31 5/22 3 «/9 115 > 1.1 3 7 4 164 7* 2OM 12.1 455 531 4.93 0.4* t 144 64 1747 13.3 404 its 4.56 i .0 m > 7 17 7 5 n/s 4/1/31 7/24OW 21/9 42 1 1. 7 4 II 1 93 41 9ll& 9,5 414 421 4. 19 u .i »n 5 a 18 2 4 24/9 4/1/51 J/n» J1 /* 112 1 2.4 1 7 1 111 W 1546 13.6 111 507 4. 16 1.1 5 9 JI .4 a 0.4 43 5 •1 O T3 u n P- J 24/9 4/1/5B J/i»3 >1/9 175 J 1.1 2 1 1 1 J J tu IM 13,1 US Mi 4.73 Sir • 2U/9 4/4 9 26/9 4/4 1/7 U 24/fl 111 3 1,1 4 5 2 137 64 1125 19,1 a>27 440 4.41 0.9 Fit I/25H 24/B 93 4 L.B 4 7 2 143 65 910 6,9 164 422 4.03 1 IN 10 25/9 4/4 l/Bl M/a 156 11 2I/-9 4/4 1/41 24/6 122 3 1,1 I 1 1 194 71 1359 a, r 354 540 5.57 2.1 IM 5 26 W. M 11 29/9 3/ V U 2/lou 1/9 14/ 1 2.2 7 4 1 Bi 19 l«4F 21 U 450 462 4.44 0.4 FH Itl J 4| ,b u 29/9 3/VH 7/1341 1/9 nil 4 2.0 1 1 1 104 4b 15 >0 14./ 424 4J2 4,29 5. ? F> 5 14 30.7 14 79/9 3/ 5/31 7/95 1/9 /□ 5 1.1 J ] 2 79 15 /JO 9.1 426 412 4,29 4.2 fR 5 40 14.4 u 2»/9 1/2/24 9/122 J 3/9 Mil 4 1.7 1 2 1 135 41 1140 9,9 414 516 4.76 0.9 F 5 6 MJ.3 ii 39/9 V i/24 9/ZJ5 13/9 43 5 7. t 4 11 1 1/4 19 233 5.7 UiO 314 1.17 0.3 r 5 t LA.4 I 1 .0 1 I 1 1 H tiJ H)i 26,3 354 540 5.57 l.l TH w 14 59. fi IQ 39 40.a IQ lb 19 4 IS 10 W.4 IB w/v 1/21*34 9/135K 11/9 205 1 L» i r io 1/7*/56 2/JO Ji/9 171 1 1.9 2 4 1 93 41 Jrua 11.a 514 549 5.40 ID 0 46 4t|. 4 20 1/10 1/ 2V 54 2/50 26/9 H5 3 1.7 3 10 3 61 27 I6H 25.9 4iU 521 5.M 7 0 26 >4.7 21 J/ 10 3/1/9 2/7K 4/ 10 B? 5 2.6 4 10 1 91 40 1404 16.1 UD 551 5. lb 5.1 4 JU 33,3 1.4 1 2 1 1 JU 59 2/70 11.5 M3 M3 1. W 6J F -13 41 44.6 22 3/ID 3/1/9 2/25 4/ II) 117 1 2.2 1 4 4 /4 73 7/>U 3/1/9 1/10 7/10 05 4 2.4 1 J I 91 40 1319 14.2 512 543 5.2B 1.5 F 1 9 II .fl II IB 19 74,9 vu 551 1. 16 1U.3 F 10 43 46,1 24 1/ID 3/1/9 »/4J M/JO 91 4 2.1 3 B 1 •9 40 1 iiw 14.6 - • 5.2B 7. L 36 JU.a 22 J/io 1/1/9 9/57 fcl/JCj 112 1 1.5 1 1 1 14 7 6? 1491 11.5 N3 Nd 5 50 r 10.2 r n 28. 53.5 26 1/10 3/1/9 VJ> LO/IO 12/ 3 1.5 1 2 i 97 44 1147 12.0 - - 5.36 10.a F 5 IB 12. 1 11 4/ ID VI/1 1/5U 4/10 90 5 2.5 2 5 J Bi 19 946 II.1 w4 3 570 4.Bl 0.2 32 4/10 3/1/4 11 4/10 3/1/4 i/a 9/10 45 3 2.5 3 « I 176 7* 44 ti 2.5 - - 4,34 0.5 r a 3 20.1 5 21 9.1 I//W 9/10 50 5 1.1 1 1 1 HI H 325 J.B - 4.40 l.fl r 5 76 7. 1 54 ■/ 10 1/1/13 4/ 211 1 / 10 III 3 l.l 1 7 1 LUO 44 1453 14.1 5Zfl 5BI 5.35 3.1 r LU 27 i9.4 lb 11 J ID J/l/4 1/124 io/y IH 3 2.5 2 5 I /7 13 16 35 21.1 4B4 604 5.45 5.1 r 17 26 li 11/10 1/1/7 4/12 19/9 150 2 2.4 1 1 2 111 54 lH4b 15.6 517 57B 5.4B 3.9 0 II. 1 *4.3 40 11/ 10 1/2/M 10/ 112 17/10 11* 3 3.0 1 3 1 119 55 1694 15.9 Ns MS 5.55 5.4 r 5 35 34,7 <J 11/10 1/2/M 10/ w 5/10 132 2 1.1 1 1 3 111 51 J 71* 23.6 4t IB/LO 1/2/M 10/JID 25/9 75 a 2.9 1 b 1 al 27 215 11.7 - - 5.46 5.6 0 H bi . * - 1. W 3.6 0 31 LI,J 30 39/LQ 5/7/24 10/200 25/9 10 5 1.2 2 2 2 - 50 - .. - - - - U.L 11 12/10 5/2/23 2/'JO ib/V b« 5 5,0 5 13 1 JL 14 262 LB M5 MU 4.20 fl.5 0 b 5.3 11 32/ 10 3/1/15 2/55 24/9 105 4 1.5 5 IB 1 /3 31 9W U.3 Id MS 4, ID 0.6 0 5 LB,4 >7 27/10 3/7/25 7/45 24/9 120 3 1.7 5 11 1 lUf 4i U64 15.4 H-l NB 4.50 7.4 a H 31.7 54 33/10 3/7/15 2/20 26/9 H4 1 2.1 5 4 1 99 44 2263 n.i N9 N9 4.60 i2,o r i St 51.1 First Waive IB tBKtL* af di at* <14* along t Laid canaL lr<* tertiary ca/ul and / ••JimiJ vbIm* Ib Jiataikca ut plot duwa furrow Nm field canal in bi. V Ltlilaplan calendar. f\ Visual Biart with I * bast, m>at grttro, I wait necrotic him! 5 » wmaC, Barat y» 11 dw/imc r**414. 4/ Sec cast tor eBplanation (I • nP rv4n«BJt 5 - cuutrlatal y rad). 5/ Product fil Kedncsa Iikdea. arid idl-.vacaa Kcuie . J&/ F * aCill ILauerinCk I • bom regroutli. 2/ r-rcuntD^. aacra NqIIb aBpectad |w Mturv r^-a rud with the initial oUmIw. ®/ Taking *C4cWttt of plot width, ui*«b«r at bolls per plat, oarage bat I-utei girt unit erffuiHtMiic tat tor. 9/ ‘Mlaafng* plot liuatbcrM era of pinta wikcra th* crop hail dle-d Ere* Bdllfkiry.Table 2 Soil Data from Cotton Yield-Saaple Plats PT*TX CpT /T iop«ii r* •) "»p i u — fuElld L :MiA ( ‘t‘bj II1HS1 r — >*n (’>**] »P“1 i^acT zirJi »i«y«ji3»n J“ as («>] jet ?, ms (■3) =»< /“ ;o w3tf*a JB4W.S 3®U<1^ - S- - 09’t - Ml - - - - I f 'FUWU ( TOI ■ JI”1 UJ1oji( - W -(> 06 - 01-0 - OM I £ 1) M 9 (ft c nri i»r irf ! t -'j io a- m i— >< £ Plot: NtMber Field fluwber * •xj 3 IM — — •- *-J M" 4fr« & J- W Q Q o o &O Q IM *v. Slight of QQ M1 LJ Pieu do u tetis («a) Score for Vigour £1-10) Score for Tellowwe <l-5) Score for PIibe Populitiop (1-S} ATwr4|i Huaber fleade per ftunch Score fur Field (1-10) Depth of War er table fcni) EE of Groundwater | >j 5/cw) Salinity index (■£ / to.) Water Logging Index footing Index Doh i event Soil TeKture in TopiciI JB ! H sqou ajdnres Buuurg iso3? jtcg”pue doj^ F F~^1gfxTable 4 Data from Random Sample Survey in Cotton ■9 s n 5 V a> s X *u * Coo rd-1/ inat«« ab 0 • *1 a •■ ■ • • — 5| Mu 4 V — — ST W V —• V - U ~ ?5 2 4’ “ J T ££ ■» H :I su I -S 1/3/39 J 3 4 5 6 7 1 9 10 11 XS 3/4/50 12 1 120 1 155 2 115 3 125 140 11 1 L 6 20 50 N 50 125 7.7 1 7L 6 115 140 M/H - 130 7.1 2 L 5 125 140 9.4 3 0.49 5 4 237 110 10 4 L 5 3 219 140 9.1 5 VL 5 6 140 150 12 6 L 4 6 166 190 3.5 6 L 4 1 96 no 9.9 8 L 4 9 242 no 7.8 9 L 3 so L 44 V - VL 42 W 50 M 42 u - VL 45 w - VL 35 V 5 VL 45 w 55 X 45 w y 1 13 I* ; 15 16 17 is 19 20 21 XS l/ZB/74 22 23 I 12 223 4 36 5 26 6 U2 ? 91 1 at 9 211 • 224 ISO 7.3 I L 13 :«o 6.5 2 L 13 170 5.4 4 VL 13 iso 5.3 $ L 11 no 11.2 6 L 11 LIO 10.3 7 L IS too 6.5 a L IS 85 10.2 9 L 17 120 12.8 9 T 17 105 25 85 65 50 135 50 - 125 - ISO 10 L 54 10 267 no 10.4 10 L 17 I Bl 125 10.8 9 T 18 I 177 120 7.5 f b T, 19 t 24 15 26 21 1 :« 29 30 31 XS 3/4/56 32 33 34 15 16 37 38 39 40 xw 6/4 42 45 44 45 46 2 20) 3 214 4 74 5 122 > 16 10 SI 10 177 10 247 1 10 170 5.3 J 0?51 19 130 9.5 3 0.13 20 130 10.8 4 T, 20 no 12 5 L 20 100 8.3 9 0.35 20 120 5.3 13 L 20 ns 10.4 10 71 20 too 6.3 10 L 20 no 3.8 I L 18 X/R 55 X/R - 9- X7R 145 X7R 45 VI 140 X5 !! 65 M/R 50 M 50 M/L 50 L 50 ▼ 50 1 40 W 1 IOC ▼ is 130 X .X 1 1 X X X L t. 125 80 X- L- so 1 50 I 114 140 8.6 1 18 2 31 no 10.8 2 L 18 2 154 4 22 7 190 9 147 9 192 10 192 10 233 I 210 130 3.0 2 L 18 100 10.6 4 L 19 90 6.6 7 0.37 19 115 10.6 9 L 19 100 11.2 9 L 19 ns 3.6 10 L 19 us 7.1 10 L 19 90 13.0 1 L 25 - L 36 V 30 X 36 •J 100 S -8 w/r 100 < 46 VT 100 I 53 T 100 9 49 r 100 R 34 IfT 90 N 35 V/J 90 H 34 J 50 X 26 35 X 16 7 - VL 15 7 - VL 19 7 - VL 40 7 - VL 53 (S) - VL 42 7 - VL 41 d i - VL 34 7 IS X 33 7 70 X 41 7 - 7t 36 7 - VL 34 7 100 8 40 F .00 H 45 t '.OO R 37 7/y 90 N 45 a 50 X 31 7 | - VL 30 7 | w R 36 7/T 55 X 31 w 75 R 40 r 3 188 110 10.5 3 1 25 .10 2.9 4 1 25 190 3.9 4 1 25 120 9.8 4 L 26 110 2.3 5 L 26 120 2.6 5 1 26 170 1.6 6 1.17 26 165 4.0 9 L 14 140 120 - - - - - - - - 200 - - 145 , 47 46 i 49 50 4 122 4 207 4 234 5 43 5 122 6 40 9 220 10 :oo ns 4.8 10 1 14 - - ’.90 - 125 X 50 X 50 L 70 .1 :oo L• X 70 M 75 L 75 X/A 100 W 140 X/R IS L5 1 10 X3 1 13 L 35 X/X L X) 1 SO 50 X 42 7 10 L 25 W 70 I 35 W/F , 150 VI 54 F 73 I 54 F 145 VI 38 F 50 X 40 » 115 X 43 F ♦ VL 52 S 20 X 52 T -0 X S3 F 1fl Z9 TA - - li a 06 TA os 09 B 59 51 T € 9*6 CT I 06 51 T z 6 OCX 8 01 6/fl B TA 5/fl 9V TA A vc TA fl Ot "la E 51 10Z z 6*9 oct 66 8 fa 51 15*0 X 6*9 OVT 51 5 601 8 t/r 8C Ta r. OC XI H V u u 6 ta X 6*9 06 CI TA X 6*9 001 s TA X B’9 ozt i OCX I 0*5 OZ 1 fl n Ta fl ZC TA fl 8C TA a 61 TA a 6Z TA kA 55 S TA B 8 TA B B II z 6*9 oci z 001 001 oc; C xoz C 59t Z C9 z 901 I 96 X fl fl fl s: h/T 09 06 HA sz TA * CT lz z B‘H CZI CT TA z 9*1 0Z1 CT V. 1 Z*8 06 CT *1a z 9*6 06 • B 06 09 CT TA z C’6 OE a VC T 01 001 k zz K/T OB OOT M OP 5Z TA z 0*8 OX CT CT TA z 9*9 OCT XCT 6 XIZ • zc 8 981 X 06 X 6X 4 ZCZ 5 001 66 86 X6 96 56 96 u Z6 61/C/6 «* 16 06 68 68 X6 98 56 •rS fl IX T M CH H a 9Z K 59 5X k fl BC TA fl X9 17. a ZV TA a ZC TA • )r 9C TA fl 6C TA fl 19 TA a ll TA a zz TA 7fl Ll TA K K k K K K k k k k k - CT TA z 0*f 06 XZZ : Oil 5Z TA z 0*51 06 zo: 1 CT TA X 8*6 0Z1 9Z : OC TA 1 5*5 0Z1 orz 0] OC "* z 9*1 OCT ozz 6 OC TA z 6*6 OTT 9C E oc 69 0 1 8*9 OTT oc T z 1*9 Oil oc T z 0*9 0Z1 oc T z 0*9 06 oc 1 z 5’C 09 oc Z60 z 9*8 06 061 X 89 9 59 5 001 9 C61 C Ot Z fl ll TA oc T z 9*01 06 CT 1 a C£ TA a BC TA B OP OX 5 TA z L ’L 081 Z9T 6 k 5 TA z 9*5 ox: ♦9 6 Cl :e 80Z/9 * Xi Of 62 BX Ll 9X 5X 9X CT ZX tCZ/85 IK XX 01 ♦9 ct T OS oct HA SZI 05 5 TA z 0*G1 06 1ZZ 6 86 fl Z* TA fl VC TA fl 6C *1A /a 9Z K S8 59 A zc U. fl w TA fl B TA )fl B T E 59 5 T z x*z oc: 991 1 H/k on 511 5 T w € 5*9 oo: C61 A K • 5 TA X 6*5 5X1 nz 9 k 091 5 T X Z'l 06 6CT 5 k k K - 5 T z 5*6 06 OC 9 R 5 T X 0*5 OK czz I - 6 T 1 0*9 on Z5 t H/K OOT 09 6 K/T 1 9*Z 5XX 99Z 01 A 6C T/K oz 55 T - 6 TA 1 C*9 CT t 06 01 fl 9* T/K 59 59 M/T 001 6 TA 1 z*s on cc 01 fl 59 T/k cr 09 fl ZC T/K 5C CT fl 99 T OS oc T T T - 01 TA X 5*1 09’. 81Z 6 - 01 T X 9*9 0X1 96 6 fl 97 T/k CT 5C T A 9f T/K oc 55 T - 01 1 z 0*5 5C1 • 01 T X C*9 oc: - 01 T 1 9*9 081 tzz X X6 X 991 9 fl ZC TA T • 01 1 1 TV 091 6Z1 9 fl 99 T/K oz 56 H/K 55 t T 1 0’S 0C1 OZ 1 r. = 9w •< AT. ?S> *R o RB IB B X ■ n ab R A — c. r • R 9 P- f R R R X r a© T1 R "C R M r* f S’ r s’ ©=9V C H IT R- R R Dr • R Ry 9 ©J R X5 99 59 99 C9 Z9 5/Z WV 19 05 65 85 £5 95 55 95 C5 Z/C OKV 15 •w ni i * C3 9« M *-«. ■ n O °3 • *C c I*- rv 33 CR R ST B n r» i» O R- • 9> B «A R R- R- 9 R *e B — IA 2. K e Z r^«o R• ■ fc FT B k £► at o? • e ■ * r. — ? M •< • 1 n w> » & ©© 9a ■ X ©R B r Hs I* R n s (panuiauoD) aiq*lTable 4 (continued) • X 5 cn 2 x w X Coord-2/ inatea 1 w * V X u a a. *1 a i h 3 3 "1 • u e V VI a ■ o u w ■ X *1 M B fl C >. C a «n • vn e— -M on « M •» ►. w Wv a -« a 9 M £ X OV MW • JS W w3 au ue a v* 'g w • 3a -• W x: w WV Xw 93 a3 Ji •fl 3 X • J s* fl SI •n x> o -c ■ X" w— a. - a 1 in *• ae u sS 4.2 £5 *1 • • e i 9 Wt ^1 a w <v a M ■ JJ • ■o — « ab *• a• WT S3 101 XS 3/2/24 102 103 104 105 106 107 108 109 110 111 XW 58/3 112 in iu L15 116 in 118 118 120 2 133 3 190 5 54 5 189 7 186 9 216 10 102 10 no 10 196 70 5.9 2 L 17 23 95 VH 150 6.6 I L 17 35 125 8 120 9.3 2 0.89 17 30 - 8 110 10.1 2 0.68 17 55 130 n 90 10.8 2 0.63 18 8C • M 115 6.2 I 1.04 18 110 135 8 120 6.2 2 0.74 20 45 60 TH 7L 21 w - 71 48 w - 71 37 w/s - 71 34 ’!($) - 71 26 WCS) - 71 37 W/s - 71 39 d(S) 120 9.2 2 0.49 20 40 150 H 71 38 W(S) 10 206 s 88 W 6.5 1 2.62 20 90 8 w 9.5 1 5.14 20 55 8 120 9.9 2 1 19 • M/H 140 7.8 2 1 to 90 8 no 9.5 2 71 19 ;ia 8 - 71 10 s - 71 6 5 - 71 39 w - 71 45 w - 71 35 w 100 9.4 2 71 19 - 90 3.6 2 71 19 - no 9.3 2 71 19 55 x/x 71 33 •8 X/R 71 29 7R 100 50 1/X 35 w 90 9.0 2 71 19 145 W 71 29 w 100 9.6 ■» 71 19 40 150 7R 71 32 w 80 6.0 2 71 19 - X/R 71 27 w 30 5.9 -X 19 TO 100 VS 1 til 122 123 124 125 126 12? 121 Ilf 130 58/3 5 IB 5 27 6 33 8 196 T • 49 7 216 7 210 9 53 10 2 2 174 108 6 109 7 131 190 8.7 I n TO - .x/h 160 3.0 e • a 20 — x/s 180 5.5 2 rt 20 - X/H - 7t 27 w - TL 51 w 71 50 w - 71 54 w 130 8.7 2 L 20 95 145 H 35 10 1 42 V i •40 5 233 160 6.3 2 n. 20 - 135 7.7 ■» 71 20 - x/x • n 51 W I x/w 71 49 w 10 23 130 9.4 e t 20 - 10 US 120 1.3 ■* 71 20 35 9 192 130 9.9 2 •X 20 70 1 71 52 w x/x 71 41 w 7W 40 10 r 39 7 1)1 1U 10/3/1 1 132 10 189 130 9.6 2 •T 20 30 8 140 :o 71 41 w .00 3.4 1 1.-8 16 - 98 123 6.5 1.30 16 125 103 134 135 136 137 138 139 140 14! XS 1/3/16 142 14) 144 145 4 56 140 5.1 J 1 16 - 4 113 150 5.7 1 1 16 115 4 171 ♦0 12.4 2 0.61 16 - 6 136 no 1.0 I 1 16 125 7 38 110 8.1 I 1 16 - 9 106 130 7.7 1 1 16 - 3 140 10 1 42 w 1 90 60 X 36 J X 95 25 X 41 W(s) X 145 5 47 w X 50 •0 X 47 V TH 140 10 8 26 w(s) X 45 55 X/H 42 w X 115 35 X 36 w 2 30 160 4.3 I 1 14 - 190 3.0 2 71 14 - 175 3.5 I L 14 - X 2 232 4 221 6 20 7L 51 6 54 6 IIS 7 201 170 6.0 2 vt 12 - 140 6.2 I VL 12 > 180 3.5 2 vt 12 - 130 5.3 2 1 10 X 145 5 L 53 W I X 71 54 w X/H 71 52 •J X/H 75 65 X 46 •4 X 71 54 X • ••x 44 7 IJ Coordinate (a) n tenth* along the field canal froe inlet and (bJ if w 4ovn furrow. 2/ Fee 70s is score )1 ■ none, 5 ■ severe) indicates naistura aerate, ta a Srv area. T/ Xarard :c>4e: indicates aoet likely cause of yield lot* where V - waterlogging (of surface hue high wetertable in a few placet), V ■ nutrient infertility, S ■ Salinity. V Tense like Salinity Indea <ST), Waterlogging Todas fWT), Colour Index - q.v. VI ■ very |0Vi L X - atdiue, X • high. .’H • vary high-Table 5 : Average Cotton Yields for Farms in Project Area Farm Year Irrigated Yield Area (ha) (q/ha) Melka Warer & Melka Sadi Amibara/Angelele Amibara Settlement Farm 1980 1981 1982 1983 1977 1978 1979 1980 1981 1982 1983 1977 1978 1979 1980 1981 1982 1983 220 41.77 1800 35.82 3346 28.46 5252 26.32 2263 22.84 2653 21.12 2672 21.29 2653 17.81 2627 30.74 2652 27.81 2437 21.02 Not Available 34.06 Not Available 31.40 Not Available 27.18 Not Available 36.55 Not Available 38.70 Not Available 27.83 Not Available 19.66TAI* I i o.<• • ! FIrH OhihIwm, Tl«H ■*< TliU Cb—— f 1905, ft kl«<U< Va<ra »t Fl*14v <1901 !• Ml IJ Taar F laid *<■■•« r Fair l*r C«Mpalian« IHI I7J/H) 1107 <14/751 imj <m/ioi 4/1/11 4/1/IM 4/1/1 4/J/IJA 90/14 54/1< 4/MA M/1U 1•1215•4 10/4 1/7 1/9 ll/l 4/9 !/• 11/0 V/1 19/1 1/9 11/9 0/1 M/9 PR 4/1/11 •/5/IM 4/1/1 4/1/llA 10/14 10/14 4/MO 10/154 111t1544 4/1/11 1 4/1/MA 1 4/1/7 1 4/1/114 V 50/14 1 M/14 1 4/104 4 50/JM 4 Data# (Klkirplra Calamlar) riiMgKlR* Rar rpwi«( U««lll*« Fre-Uvital l*« Planting Irrlgail** IPPl 1 1 1 4 1 • 1 Cappia* TWImIm ■ m4 Wr*4lM 1 1 1 4 1 Narhlne Cultivating I 1 - 17/7 ii/r 5/9 IV/IO M/V 14/10 14/M 11/10 NR 10/5 M/9 IV/10 m/9 17/10 11/10 10/10 10/0 4/9 H/9 m/io 14/9 10/11 n/ie • M/V 11/9 4/M 70/10 11/10 11/11 ii/ii VII 11/V 15/9 IV/IO ll/ll 9/11 M/1! 10/11 Mill 10/10 14/M |4/|| IV/11 15/11 ll/l 0/1 M 10/11 •/II n/n 11/1 1/1 1/1 ll/l Ml 17/11 M/1 14/0 1/0 15/0 M/0 15/0 11/0 M/ll 11/0 M/V 11/V 5/10 15/10 11/10 1/M V/10 io/v 1/M 0/10 50/10 15/10 11/10 1/10 11/10 11/V 11/10 11/10 10/10 11/10 IV/10 10/10 >5/10 1/0 15/V 11/0 NR 4/0 4/4 15/0 >!/• M/0 M/1 >5/9 Ml M/V V/V M/10 V/IO 11/0 V/IO 10/M 00 M/14 M/V 17/10 11/M I5/I0 ll/to M/IO 11/10 1/10 71/10 I//M - --- - 50/0 1/V - - - - - 14/V 11/10 11/10 70/10 H/IO IV/IO 10/10 75/10 15/V 11/10 11/10 l/ll 5/11 0/11 10/10 15/11 v/v 19/IC 14/10 I/ll M/10 0/10 15/10 M/10 4/10 M/iO IV/IO V/H 4/11 14/10 l/ll D/10 11/10 ll/ll ll/ll ll/l 11/11 M/11 19/11 4/11 M/IO M/ll ll/ll l/l ll/ll 11/11 V/01 ll/ll 19/11 >/i V/1 19/V 10/1 10/1 M l/l •/II 0/1 10/1 ll/l 5/1 ll/l M/1 ll/l 17/1! 10/1 V/1 4/1 15/1 11/1 M/1 1/1 11/11 M/11 17/11 10/1 19/1 10/11 11/1 14/1 14/11 77/1 17/1 >0/1 »/l ll/l M/1 17/1 ll/ll n/i ll/l H/l ll/l 15/9 IM - 1/11 1/5 14/5 1/1 V/1 V/5 10/4 M/1 70/1 M/1 5/1 15/1 4/1 - --- 15/1 -• - V! 77/1 • M/1 11/1 ll/l - - IV/4 - 17/1 19/9 - 17/4 14/10 >/ii ll/ll I/ll II/II 11/10 ll/ll ll/ll 1/10 •/II ll/ll 10/11 ll/ll - ----• WO 14/11 - M/ll - 0/1! 5/11 4/11 M/1 10/10 M/ll M/ll 1/11 11/11 - l/l 11/1 ll/ll - M/ll V/IO r/li 14/11 M/IO 15/IV •/II •/Il 14/10 10/V 14/11 l/ll 11/11 17/11 17/11 ll/l! 14/11 0/11 >/i> ll/l! M/ll 1/11 10/1 5/1 11/11 14/10 15/11 0/11 M/11 10/11 9/l| 0/11 7/15 •/II M/17 5/11 5/1 5/11 M/11 4/1 0/1 l/ll 1/11 17/11 5/11 IV/11 10/1 M/ll 21/11 1/19 1/15 - 17/11 11/5 9/1 1/1 - ll/l 4/1 M/1 14/1 M/1 M/1 M/1 >1/1 0/1 15/1 1/4 - ll/l V/11 ll/ll M/1 M/|l 71/11 ll/l M/1 11/1 l/ll 10/11 - --• -•- -- 4/15 - - - ll/l • - M/1 IV/1 - 10/9 M/V M/ll 10/11 •/Il I/ll IM 11/10 VII 11/11 - 14/11 - 14/11 lip - 4/10 ll/ll 4/11 M/ll ---- 70/J -- |4/|] M/11 4/11 • l/l 74/V l/ll 1/11 IV/11 M/ll l/ll M/ll M/ll M/11 0/1 * 17/10 14/11 ll/ll >1/11 4/11 ll/ll V/11 10/11 - - 4/11 - - - 5/4 9/9 10/> 11/4 11/4 H/4 1 TlaU (</!»•> Flail ebangt ( in l/Ka to MOI ( Io claaae* 1/4 11/4 m/5 M/1 ll/< II/* f J/4 J4/4 M/5 M/5 0/4 11/5 M/4 l/ll ll/l M/4 - ll/l - • - - - iv.o 44.7 M.l •l.l u.J 11.1 10.4 •1.1 •4.4 -7V.F -4.7 -l/.t l.l -II. • -7.1 -11.• • -1 • - 1 0 - • -1 - 5 51.0 79.1 11.5 14.4 »O.I 14.7 11.0 11.1 • 5.1 -14.1 -0.1 -1.1 -5.0 -10.V -4.V • 1.0 0 - 1 -0 0 1 1 -I o 54.1 15.0 54.4 M.l M.l 11.• 11.V 15 1 >-) it I*e*r4*4Table 7 Cotton Yield Adjusted for Effect of Waterlogging ^1 <*) 1 XJ •u <D C —'I JC u •a cr 1 « x 0 ex nj xj a g e Q <—• Q) C O- cr XJ o -lx M cr W -1 c u o r7| co J B i o □ e •H <V v xx -o '*■" u 5?: <3 XJ JX •MU-J e < •r* >• £ B < •H XJ 3 U - B ■g 3 z C c* CT u u B3 o C 4J •r-» J- 01 u Oi •w O CL <D CL u- w < M- >- CO H *U' O o C XJ X u < >- Um 3 •H U 0J o CL W’ 1 26.8 7.8 14.6 E 2 45.8 10.3 56.1 E 3 37.7 12.2 49.9 E 4 45.5 (17) 63 DE 5 18.7 (26) 45 E 6 31.6 (17) 49 E 7 40.8 ( 5) 46 E 8 19.8 15.1 34.9 re 9 30.4 9.7 40.1 FE 10 59.8 1.9 61,7 E 11 39.8 0.6 40.4 DE 12 41.6 8.5 50.1 13 30.7 16.9 47.6 14 14.4 25.6 40.0 16 30.5 8.9 39.4 D 17 16.4 14.9 31.3 F 18 44.6 6.6 51.2 19 48.6 3.5 52.1 E 20 36.7 3.5 40.2 E 21 35.3 21.3 56.6 22 46.1 6.6 52.7 23 31.8 5.8 37.6 26 30.8 16.7 47.5 27 53.5 3.7 57.2 28 32.3 25.2 57.5 31 20.9 (12) 33 F 32 9.1 1.2 10.3 F 33 7.1 1.0 8.1 D 34 39.4 ( 5) 44 D 36 38.3 (12) 50 38 46.5 ( 7) 54 40 50.7 ( 7) 58 42 66.4 ( 7) 73 49 13.7 (14) 28 50 0.1 0 0.1 51 5.3 45.6 50.9 52 18.4 39.6 58.0 53 33.7 23.9 57.6 C 54 51.1 26.0 77.1 C V Final yield estimate from Table 1 2/ Amount by which yield is expected to have been greater than measured had there been no waterlogging problem at the site. See text for calculation. Brackets indicate adjustment by CI. Adjusted yield assuming no waterlogging effect. 4/ Explanation’ as in note 3J Table 2.Table 8 Overall Analysis of Variance of Yield Data from Random Sample Survey . GT = 5653 N = 145 GM « 39.0 CFM - 220389 Source of Variance df SS MS F CV Sign’f. 1. Between hazards (H) 2. Between fields (?) within H 3. Between samples within F+H Total : 2 1105 552 4.45 60 ★ 42 5199 124 2.21 29 *** 100 5598 56 19 144 11905 Table 9 : Summary of Yield Paca from Random Sample Survey Yield Classes (q/ha seed cotton) 0-9 10-19 20-2<» 25-29 30-34 35-39 40-45 45-49 50-54 Waterlogging -Number of samples -X of area Infertility -Number of samples -X of area Salinity -Number of samples -X of area All samples -Number of samples -X of area 0 0 4 13 19 24 26 13 15 0 0 3 9 13 17 IB 9 10 Mean yield 38.8 q/ha, 72X of area • 000001435 000001323 Mean .yield 46.6 q/ha. 91 of area 1I0317212 110215111 Mean yield 34.6 q/ha, 12Z of area 1 1 4 16 20 32 32 17 22 1 1 3 11 14 22 22 12 15 Mean yield 39.0 q/ha seed cotton (SE ± 7.5)Table 10 : Gross Margin at Different Drain Spacings in Soils of Different Drainability- Class and Estimated Optimum Spacing. Soil Drain ability Class 2/ Model u 2/ Drain Spacing (run number) 12345 Fitted Optimum 4/ 1A IB 1C ID 2A c c c F c F 2B 2C 2D c F C F C F Spacing GM Spacing GM Spacing GM Spacing GM Spacing GM Spacing GM Spacing GM Spacing GM Spacing GM Spacing GM Spacing GM Spacing 81 65 41 25 20 1430 3189 4214 4175 4144 107 89 63 44 36 1437 3200 4237 4219 4206 136 117 85 61 51 1443 3207 4249 4236 4227 168 146 108 77 64 1447 3211 4255 4245 4238 68 54 34 21 16 1781 3296 4201 4154 4117 54 34 21 16 2795 3292 3366 3362 74 52 36 30 3310 4229 4207 4191 74 52 36 30 2808 3319 3419 3436 97 70 50 41 3318 4242 4227 4216 97 70 50 41 2817 3332 3439 3461 139 120 88 62 51 1801 3323 4249 4237 4228 120 88 62 51 2822 3340 3449 3473 34 4329 53 4400 72 4446 91 4473 28 4290 21 3372 42 4294 34 3435 57 4320 47 3459 74 4433 59 3472 J_/ Soil Drainability Classes: 1 - Coarse-textured soil, 2 ■ Medium/Fine-textured soil. See Annex B for description. 2/ Models developed for coarse (C) and very fine-textured soil (F), being more extreme than Drainability classes. Drainability Class 2 corresponds approximately to mean between models C and F. 2/ Drain spacing in m and Gross Margin (GM) in Br/ha. 2/ Fitted by y • a + bx + ex and optimum when dy/dx ■ 0, but not a perfect fit to data near true optimum. 2Table 11 ; Proposed Leaching/Irrigation Schedules for Cotton (values in mm) Preplant Irrigation Postplant Irrigation Irrigation Number 12 123456 Total Interval (Days) > 20 20 20 20 40 30 30 Fine-Textured Soils and Medium-Textured Soils with Saline Groundwater Consumptive use Excess for leaching U Net Application Gross Application - 60 30 60 80 185 130 125 227 167 80 35 15 5 6 11 220 220 105 91 91 182 130 130 286 286 150 130 130 260 185 185 Medium-Textured Soils (except with Saline Groundwater) and Coarse -Textured Soi Is Consumptive use Excess for leaching J/ Net Application | Gross Application -- 30 60 80 185 130 125 - 186 93 47 27 28 22 27 - 220 105 91 91 182 130 130 - 286 150 130 130 260 185 185 670 546 1169 1617 610 430 949 1326 1/ See Annex B for explanation of calculation.Table 12 : Irrigation Parameters for Bananas and Estimated Schedule July Aug. Sept Oct. Nov. Month Dec. Jan. Feb. Mar. Apr. M-y June Kc Crop Coefficient Year 1 0.40 0.55 0.60 0.65 0.70 0.75 0.80 0.85 0.90 1.00 1.15 1.20 Plant Year 2 1.00 1.00 1.00 0.95 0.90 0.85 0.80 0.85 1.00 1.15 1.25 1.25 First Harvest Year 3* 1.15 1.05 1.00 0.95 0.90 0.85 0.85 0.90 1.05 1.20 1.25 1.25 Barve st more spre ad the reafter For Mature Crop • ETo 219 198 195 208 178 176 171 173 205 205 227 245 ETC 252 208 195 198 160 150 145 156 215 246 284 306 ?e(50) 101 94 34 12 0 0 0 33 52 48 21 22 Wn 151 114 161 186 160 150 145 123 163 198 263 284 216 163 230 266 229 214 207 176 233 233 376 406 Interval 12 12 12 12 12 12 15 15 12 10 8 8 Application 90 90 90 90 90 90 90 90 90 90 100 100 Annual total: grots requirement 3000am scheduled application 2955 net requirement 2100Table 13 : Irrigation Parameter! for Irrigated pasture and Estimated Schedule July Aug. Month Sept Oct. Nov. Dec. Jan. Feb. Mar. Apr. M.y June ETc 1/ 219 198 195 203 178 176 (17111/ (173) (205) 205 227 245 Pe(50) 101 94 34 12 0 0 0 33 52 48 21 22 Wn 113 104 161 196 178 176 (171) (140) (153) 157 206 223 Interval 30 29 13 19 21 14 canal closed 19 12 16 Humber ^/ 1 1 2 2 1 2 2 2 2 Annual total.: gross requirement 2200mm approx. 21 scheduled application 2145 -ft/ net requirement 1300 approx 21 _1/ Assumed canal is closed for three months. Without irrigation Kc would fall so ETc and Wn would be reduced from values in brackets. 2] Number of applications at the stated interval in the month. J/ Allowing for reduced evapotranspiration during January-March. 4/ (l-p). Sa.D assumed to be lOOcnm (with depletion factor (50Z), gross application constant at 143mm each irrigation. 5/ Kc assumed - 1.0 henceTabic Sroci Land Areaa in Nationalitad Crovndvater Salinity Claacaa Iha) Fam I?' Znplth Year eI7 Tear Groundwater Salinity Claas mSem“l 2? 0-4 4-8 8-12 12-16 16-20 20-32 32-48 Total VT >ln MS 1987 Z88 c 0 1 2 J 4 5 182 103 65 47 35 54 40 526 228 29 28 23 17 It 15 22 145 83 17 17 10 4 0 0 0 48 35 10 10 6 2 0 0 0 28 7 J211000 70 0000000 00 Total 754 1 10 67 78 65 59 51 65 0 385 0 .Total 185 1988/19 c 0 1 2 3 a 5 6 10 MW 1989/90 c 0 1 2 3 4 5 5 3 2 1 1 3 0 15 768 25 26 22 18 14 20 20 145 623 22 23 19 17 14 26 22 163 480 38 37 26 21 15 33 35 205 275 50 47 33 24 17 16 0 187 83 15 17 13 11 10 14 0 80 8 42I1000 80 Total 783 740 277 73 29 7 4 0 1130 Total 1130 0 0 0 0 0 0 0 0 128 0 0 0 0 0 0 0 0 128 0 0 0 0 0 0 0 0 128 10 6 3 1 0 0 0 20 101 34 19 10 3 0 0 0 $6 42 | 23 • a m 6 I 0 0 0 42 0 Total 128 | AU 1990 c 0 I 2 0000000 477 3 4 5 31 20 11 3 0 0 0 45 412 - 50 29 16 3 c 0 0 98 314 1 62 33 1? 4 0 0 0 116 198 28 27 21 14 10 12 0 112 86 1 22 20 15 10 7 7 0 81 5 6 3110000 50 Total 477 1 uc 1989/90 0 1 2 3 4 140 157 13 46 20 28 0 482 913 85 51 21 7 0 0 0 164 741 90 51 22 7 0 0 0 170 578 91 52 22 7 0 0 0 172 406 243 55 22 6 0 0 0 326 BO 5 25 18 10 3 0 0 0 56 24 6 10 8 5 1 0 0 0 24 0 Total 139a 1990 0 1 2 3 A 5 Malka Sadi Total 0 1 2 3 4 5 6 Total Malka Varar Totall^ 0 I 2 3 4 5 6 0 0 0 0 0 0 0 0 135 0 0 0 0 0 0 0 0 135 0 0 0 0 0 0 0 0 135 4 3 2 1 0 0 0 10 125 60 39 16 6 0 0 0 121 4 2 1 1 0 0 0 •0 4 o Total 135 994 461 205 136 94 126 40 2056 996 54 54 45 35 25 35 42 290 706 39 40 29 21 14 26 22 191 515 48 47 32 23 15 33 35 233 282 33 49 34 25 17 16 0 194 88 15 17 13 11 10 14 0 30 8 4211000 80 1207 670 359 252 175 250 139 3052 140 157 83 46 28 28 0 482 1652 116 71 32 10 0 0 0 229 1423 140 80 38 10 0 0 0 268 1155 167 94 44 13 0 0 0 318 837 363 140 69 29 10 12 0 625 212 72 31 32 15 7 7 0 183 29 13 9 6 1 0 0 0 29 0 1 Total loth Areas Total 1013 602 304 123 45 47 0 :i34 2220 1272 663 375 220 297 139 5186 Motet; _1_Z Par: of the LAA land falls in tha Malka Varar unit of the Stat* X Project. The water table is estimated to be lass than 20 mScm’l in conductivity, and that io 121 ha and 143 ha in paars 4 and 5 respectively, eha vatartabla will ba at lass chan In below thr surface. Thais areas ara not included in the above table. 2/ Year 0 is 1984/85. Year 1 ia 1985/86. ate. T/ MS • Melka Sadi, MW • Malka Varar, ALC • Al|eta State Tams. A*C - Settlement Scheme. 4/ Crop C - Cotton. 3 - banana (others are mixed cropa).Table 15 Estimated Number of Years Before the Schedule-Year Crop Failure Occurs, from Time at which Groundwater of Specified Salinity Reaches an Annual Average Specified Depth. Watertable Salinity (ECwt in mScm"^) Crop Annual Average Depth of Watertable (m) 0-4 4-8 8-12 12-16 16-20 20-32 32-48 Cotton Maize Pasture 2.0 1.5 1.0 2.0 1.5 1.0 2.0 1.5 1.0 16 12 9 5 2 2 2 15 11 8 4 1 1 0 14 10 7 3 0 0 0 321 0 0 0 0 2 0 0 0 0 0 0 2 0 0 0 0 00 3 3 2 2 I 1 0 2 2 2 1 1 0 0 Bananas 2L 2 1 0 0 0 : present, almost in equilibrium - s ee text 00 ______________________Table 16 Fan. Iln it J/ Crofil/ 1985 J986 J?f»7 1988 1989 1990 1991 1997 I 2 1 4 5-6 7 fl Tear 1993 1994 1995 1996 1991 9 10 || I? |) 1998 1999 2000 2001 2002 14 15 16 17 18 200J 2004 19 20 Total Area MS MW Ai.c RRf C (ha) n c c c M G F pV All All 204 258 1)0 395 446 464 484 565 614 660 759 837 881 1097 1175 1224 1266 1313 1333 -------► 1333 880 ■ Ip i .. 2. .. 1 Ma 0000001 5 5 8 17 22 26 43 53 62 91 111 —------ 111 0 0 0 20 31 38 41 53 82 86 101 136 174 201 253 314 368 192 411 —-------► 411 0 0 0 0 0 0 0 0 0 0 0 40 113 124 149 186 186 0 0 0 0 0 16 166 186 0 0 0 0 0 10 60 66 ---------- m ■ "• I Fir• 1 Aiiu / _ I _ _ ->— A_ -1 _ —— 1 - - 3 ml. I _ 0 0 0 0 113 262 374 442 465 508 5)1 w 186 66 1 Bo 705 0 0 0 0 111 28fl 600 694 717 760 785 825 89ft 909 934 971 - ■* 1129 lAft 0 0 0 5 29 50 79 111 in 111 166 229 ------------ -> 229 MS All (ha) 1084 1138 1210 1275 1326 1344 1367, 1445 1494 1520 16)9 1717 1763 1977 2055 2104 2146 2193 2213 —► 2661 (X) 41 43 45 48 50 51 51 54 56 57 62 65 66 74 77 79 81 82 83 -------- —► 100 MW All (ha) 0 0 0 25 173 376 721 R64 917 964 1060 1190 1323 1367 1459 1567 1630 1683 1722 -------- ---- ► 2080 (X) 0 0 0 1 8 IB 35 42 44 66 51 57 64 66 70 75 78 81 83 -------- ---- * 100 All Stage 1 (ha) 1084 1138 1210 1300 1499 1720 2085 7JO9 2411 2484 7699 290 7 1086 1344 3514 36 71 1776 38 76 3935 ---------—*■ 4741 £/ 1/ MS * Melba Sadi, MW - Nelka Murer, A1.G • Algeta State Fa rata, IIRC • Settlement Scheme, JAR « Institute o( Agricultural Research. 2j Crops! c • Cotton, 8 • Bananas, M = Haile, C ■ Carden (mi and vcgrtnbles), F - Forestry, P • Planted Pasture. _3/ Some land abandoned, some never planted as scheduled because nf anlintty. 4/ All Stage I men of 4741 ha la programmed for drainage installation by 1990 when projected abandoned total area ia 1720 ha; the remaining areas would be adversely influenced by higher watertablea/salinity although not yet abandoned.Tabic IGA : Project Rate of Abandonment of Land Within RRC: Alternative Scenarios 2 and 3 (Net cropping area in ha) 1985 1986 1 2 1987 3 1988 1989 1990 1991 1992 1993 1994 Year 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Scenario 2 : All cotton 68 48 48 88 94 100 108 189 210 230 388 465 519 693 849 944 1033 1296 1320 1330 Scenario 3 : 50Z cotton i 25Z maize 25Z pasture 49 89 117 141 244 305 364 420 442 664 779 828 859 987 1101 1162 1240 1317 1326 1330Table 17 : With Project lanaaa MeplaBtini Schedule end Yields, J n 6 n ? H Rn 9n 10 r: 50 50 50 50 50 50 50 "0 9 18 16 14 12 8 ZOO 200 ZOO 200 200 200 200 70 ”T 12 18 17 15 14 12 16 200 200 200 200 202 200 200 140 70 0 13 20 19 17 15 13 11 10 200 200 ZOO 200 200 200 200 140 B0 II 22 20 11 17 15 12 10 20C 200 200 200 200 200 200 200 150 0 20 24 22 20 18 16 13 12 141 141 141 14] 141 |4j 141 141 T a 25 23 20 18 16 13 10 n 12 12 : i 59 59 59 59 59 59 59 59 T 22 24 23 20 IB 16 13 141 141 141 141 141 141 141 | 0 22 25 23 20 11 16 : 137 137 137 137 137 137 0 22 25 23 20 IB ii LJ n 14 15 16 17 ii ii ii XT 50 50 50 50 50 50 T 22 25 23 20 11 130 130 130 130 130 0 22 25 23 20 130 130 130 130 22 25 23 130 130 130 S 22 25 130 130 0 22 130 Total ll' 1 9 9 9 209 337 337 337 337 337 337 337 337 337 337 337 337 337 Average U 6.0 6.0 2.0 2.0 1.0 a.o 13.2 16.2 14.8 12.8 9.1 7.1 13.6 23.4 23.3 20.9 18.5 16.3 Total II u 329 479 629 729 700 700 850 991 991 99] 991 991 991 991 991 991 991 991 Average £j 1.0 7.0 6.1 8.1 9.4 1.8 6.4 8.1 10.8 13.5 13.9 18.7 17.1 15.1 15.2 15.5 15.9 16.4 1/ Net Area (ha). Arae ia underlined in replanting year. 1/ Yield (t/ha). X • area iwpleneated io 1967/16. II - area iwplmented lb 1980/89. Motet TieIda vary eonually for agronomic reasons related to banana husbandry techniques. Thia requirea individual parcels, of land to be replanted successively. Yields increase co a naeinun following replanting, before fra du. Uy dact«aaL fluctuate a on an annual baa is. Q | prior co cbo oast plancinf. The total yield for the whole slntatioo thereforeTable 18 : Potantia* Uodun in R1C Settle—nt Sehe— in Pelaeioe eo Groundveter Featoraa (Wet area* in ha) Salinity of Uacertable (nSce“0 Year Inplcn Y..r 1! 0-4 4-8 8-12 12-16 16-20 20-32 32-48 Crop Sy ary 6 (19901 : Total P9P7P4P1 974 I 000 P 21 T 21 5 (1989) I Total C6 P 16 P 16 P 9 P 3 22 16 9 3 0 0 0 G6 P 44 T 50 II Total P2P1PI 1 21•0000 P4 T4 - (1988) • Total c 50 M 120 C 20 C 43 P 21 P 5 P 19 P 5 211 48 19 5 0 0 0 G 50 N 120 C 63 P 50 T 2S3 II Total M 30 C 17 c 30 P 5 p 4 P 14 P 5 M 30 C 4? P 28 52 34 14 5 0 0 0 T 105 3• (1987) Total 11 Total c 10 M 36 25 V 40 P 6 p 5 P 19 ? 6 r* i 79 45 19 6 0 G 10 M 36 C 65 P 38 * a 0 149 2 P3p3P P1 P9 33210c0•9 a ♦I (1986) local C 11 P 67 p 44 P 19 P 6 c 11 ? 136 78 44 19 6 0 0 0 7 14 7 1• P 74 p 44 P 18 P 6 (1985) Total 74 44 18 6 0 o 0 p 142 I 142 0 (1984) I Total P 113 p 110 P 12 F 10 F 28 F 60 F 40 F 24 F 24 123 134 72 40 24 24 0 p 235 F 116 T 421 All Tears I II Total 594 342 160 67 24 24 0 X 1211 57 38 17 6 0 0 0 II 118 651 38C 177 73 24 24 0 T 1329 Total Crop Area* C Cotton 186 ha (net) 142 [ M Maiae 186 14 G Cardan 66 5 F Foraatry 186 14 P Feature 705 53 T Total 1329 10C 1/ Source: WC Lasduae Plan. {j Inpleaentation I • 1989/90, II - lace 1990,Table 18A Potential Land Use in RRC Settlement Scheme in Relation to Groundwater Features Scenario 2 (all Cotton). Net Area in ha Salinity of Watertable (mScm“’) Year Impltn Year U 0-4 4-8 8-12 12-16 16-20 20-32 Crop WT >lm Summary 0I II 1I II 2I 11 3I II 4I • II 5I II 61 II 122 137 72 40 196 181 90 46 274 225 109 52 353 270 128 58 3321 564 318 147 63 55 37 16 6 586 334 156 66 57 38 17 6 595 341 160 67 57 38 17 6 793 1164 118 118 651 1164 118 118 503 1164 118 118 353 1164 109 118 70 1124 4 118 21 1118 118 1112 118 \J Implementation I - 1989/90, II = late 1990Table IBB Potential Und Uee in MLC httlmwt Schema in delation to Groundwater Feature*? Scenario 5 (50? Cotton, Z5t Mtitc, 25* Pastura Met Area in ha 1 Tetr Xnpltn Tear 1/ Crop Salinity of Watertable (reSce’-l 0-4 4-8 8-12 12-16 16-20 20-32 WT > la Crop Sunniry 0 (1984) I M C 100 no 50 332 332 287 547 p 22 26 22 40 24 24 174 308 . T 122 136 72 40 24 24 793 1187 II C 118 Ilf 1 (BBS) I M c 150 145 52 332 332 200 547 p 46 35 38 46 24 24 119 284 T 196 180 90 46 24 24 651 1163 It C 118 118 2 (1986) I M C 210 175 62 332 332 100 547 P 64 49 47 52 24 24 72 244 T 274 224 ’.09 52 24 24 504 1123 | II c 118 118 3I (1987) M 332 332 C 270 200 62 15 547 P 83 69 66 58 24 24 5 216 T 353 269 128 58 24 24 355 1095 II C3321 109 118 4I (1988) M 250 C 240 230 62 50 300 15 547 P 7i 8’ 85 63 24 24 5 244 7 564 317 147 63 24 24 70 1071 II C 55 37 16 6 4 118 5 <1909) I M 195 20 215 C 317 230 1 547 P 74 103 156 66 24 24 206 T 586 333 156 66 24 24 21 968 n C 57 38 17 6 118 6 (1990) i M 20 C 321 226 20 547 P 254 115 160 67 24 24 340 T 595 341 160 67 24 24 907 ii C 57 38 17 6 118 Total C rop Areas C Cotton 065 ha 501 M ? Mai re Planted Pasture 332 332 25 25 • T Total 1329 100 i------------- 1/ Implementation I • 1989/90, XI • lere l»90.Tahir IV J rejected LandlUaa With and Without Praiaifce rioj«ct* (*«•< craprcit ■ •*» »»*> Fatal Unit Crop |«ap|n Tern 1985 1 1966 198/ 2 ) 1988 4 1989 1990 1991 1997 199) 1995 5 6 7 a 9 IO 1995 11 199* 17 1997 11 NS HA//AA C WM 502 502 502 502 502 502 461 656 446 44) 441 Ml MB 154 259 7)5 220 AW9 u 9 V 9 209 J17 “ NS I’nn/m c w 62/ 573 501 4// 441 426 408 IAS 356 119 IB 761 7 »0 u 62/ 5>) 501 B W 4)9 u 4/9 62 9 7/9 WO JOO 850 991 KU IW/W C w Hl 111 III 111 111 ill 110 107 106 106 |0) 94 09 w HI tear l?e 199V 1000 2001 2CHJ2 2001 2004 2«W5 2004 21107 ?*Xia 2009 14 IS IV H 18 I* 20 21 22 23 24 25 52 7/ 12 3 0 -------- ► Air iw c w 411 411 411 J9I MIO 37) )70 u 4JI 411 411 )9I >80 Jr • 158 All )29 )25 110 275 711 me 1909/90 c w 1)9 1)9 1)9 119 1)9 1)9 1)9 119 119 1 19 119 99 56 in 91 64 20 f.fl 58 49 20 0 ► %A 92 43 19 0--------- 45 20 0 u 1)9 >1 V 154 156 156 156 154 140 70 0 c w Gt 66 66 66 66 56 6 0 ------ a- w 66 u - 30 60 90 120 150 186 r w WO 400 405 502 )90 248 150 86 69 w 300 400 60 5 502 190 SUU 600 1990 c W 42 47 47 47 47 47 4? 47 4 7 4 J 47 4/ 17 u 47 H w 30 W XI >0 W )0 n —— a Fu )0 60 90 120 150 If If 0 u rM 30 - 9 40 )9 )2 IB 14 a 7 0 ------ w u — 9 -.0 )9 )2 40 Both R w 657 527 403 125 408 679 029 971 946 MB 9 1014 1061 11 16 u 657 527 40) )25 ‘.08 275 1 )l 6ft I 1122 1209 -------- ► Hot.: a * »o-.th «r.ri«» tr.t.l -< .... DJ* ha. W . U.tL.l Project a, - - With hr.i-n. Project.. C . to,,.... (n/ha) « - ho............................................... .. (t/M». •< • H»« Oh. . e. - Garden U...4 *••**«- pr«d«ti« land. pl— « -• h '<^ -*« »»•<• ‘ ‘^ Lld’in t/h»> ’ « ’ .7.^7 ’ r - Fn.e.tr, rjrrit.c.4 - ...ra«a field firewood. pol„. <H.ildin tlmher I.. -Vhi», f - In.rated Pa...................................................................... . <r/ha 0 dry matte.>, '-P'" »«■ deep d..I<Mge project (.urf.c. 3rai»a«c prnj.ct impl.-ent-d O- IWI NS ■ H-U« Sadi, HU • HelM Worn.. MX e *l,et. Star. Farm., MC Setrl—.t Scheme. Area within IAR narlude.1 fcoa abnva a» not contributing to aaatjanm^ut nf prnjn.cl Imiwl II •». Bruble m • Projected Land U*e With and Without Drainage Project* for Alternative Cropping Pattern* on 88C (net eropped area in ha) Input Year Crop 1985 1 19B6 2 1987 3 1988 1989 1990 1991 Year 1993 1994 1995 1996 199? 1998 1999 2000 2001 2002 200 3 2004 4 5 6 ? 1992 8 9 10 11 12 11 14 15 16 17 18 19 20 Scenar io 2s all cot ton 1*89/90 C V 1164 1164 1164 1124 1118 1117 1105 1029 1008 988 812 769 716 544 421 329 243 32 9 0 V 1164 1164 1164 1124 1118 1212 199<1 C w in 11? 117 117 II? 117 116 III 111 III 109 95 94 91 57 56 51 2 0 V 117 Scenario 3: 50Z cotton, 251 seise, , 25Z pi 1989/90 c IF 547 547 547 547 y 54 ) $4? 54? 54? 54 7 542 547 441 406 176 251 171 111 36 11 4 0 H if 332 n? 332 100 215 20 w 332 332 332 JOO 215 112 0 ■ P w 284 244 216 224 206 140 107 251 729 7 0 w 284 244 216 224 706 332 1990 C V 117 117 117 117 117 117 116 111 111 III 109 95 94 91 57 56 53 2 0 w 117 ------- (Definition of symbol* ■ * on Table 19).Table 20 : Percentage Loss of Crop in the Years before Schedule-Year Once Annual Average Depth of Groundwater <e Less than 1 m. EC Groundwater Crop ’---------------------------- Years Before Schedule- Year u 0-4 4-8 8-12 12-16 Cotton Maize Pasture 1 2 3 1 2 3 4 1 2 3 4 5 28 28 33 38 28 28 28 33 28 28 28 28 20 - - - 15 - - - 10 - - - 5- -- 20 25 30 35 15 20 25 30 10 15 20 25 5 10 15 20 0 5 10 15 1/ See text. Table 21 : Average Crop Yield (q/ha), in Years Before and After Schedule-Year as Affected by Waterlogging. Poly. Crop Year Before and After Schedule-Year (S) U Weighted -3 -2 -1 s ♦ 1 ♦2 *3 Average U Cotton yield Z of area Maize yield Z of area Pasture yield Z of area 50 50 45 40 36 30 25 15 35.2 0 5 10 15 40 15 10 5 26 26 26 24 20 17 10 0 20.0 0 0 5 20 50 20 5 0 100 100 95 90 80 70 65 60 80.0 0 5 10 15 40- 15 10 5 _1/ See text. 2/ Kvetagt yield weighted by Z of area. Rate of onset of salinity is assumed to be almost the converse of waterlogging. Area distribution from Random-Sample Survey (slightly modified).Table 22 Decline in Crop Yields due to Rising Watertable and Salinitv Salinity of Groundwater (mScm~l) Years Before and After Schedule-Year (S) X! -5 -4 -3 -2 -1 s ♦ 1 +2 Cotton - Average of State ? arras (in q/ha) 0- 4 4- 8 8-12 16-26 16-20 29-32 32-48 25.3 25.3 25.3 20.3 16.0 5.6 2.4 0.6 25.3 25.3 25.3 20.3 16.0 5.6 2.4 0.6 25.3 25.2 23.3 20.1 15.4 5.4 2.2 0.5 25.2 25.0 23.0 19.3 14.8 5.2 2.2 0.5 25.0 24.5 22.1 18.4 14.0 4.9 2.0 0.4 24.0 23.8 20.2 15.8 12.2 4.4 - — 23.0 22.3 18.0 13.0 9.4 3.7 Maize - RRC Settlement Scheme (in q/ha) 0- 4 19.6 18.8 17.9 15.9 11.3 3.1 0.4 — Pasture - RRC Settlement and MSAE (in q/ha of dry matter) 0- 4 78 75 67 57 43 17 8 2 4- 8 75 71 64 53 41 16 7 2 1/ See text Section 6.1Table 23 anple of Calculation of Yield Declint of Cotton Enterprise at XAC without Subsurface Drainage Salinity of Groundwater (aSccT^) Year 0-4 4-8 Tota 1 Prodn (,) Productive Av. Yield Area on Productive (ha) Area (q/ha) Areal/ TicUi/ Prodni' Area Yield Prodn 1-8 56 24.6 1378 83 24.6 2042 | (1985/92) 34 20 139 24.6 9 — 0 19.7 0 (1993) Total 1 40 24.6 984 43 24.6 1058 56 24.6 1378 83 24.6 2042 3420 139 24.6 10 0 15.5 0 40 19.7 788 43 24.6 1058 Total 56 24.6 1378 83 22.2 1846 3224 139 23.2 11 Total 0 (5.5) 0 40 15.5 620 43 19.7 847 56 24.6 1378 83 17.7 1467 2845 139 20.5 12 Total 11 19.7 217 0 (2.3) 0 25 24.6 615 40 (5.5) 0 20 24.6 492 43 15.5 667 56 23.6 1324 83 8.0 667 1991 99 20.1 13 ( Total 11 15.5 171 0 <0.6) 0 25 19.7 493 40 (2.3) 0 20 24.6 492 43 (5.5) 0 56 20.6 1156 83 0 0 1156 56 20.6 14 Total 11 (5.5) 0 0 0 0 25 15.5 388 40 (0.3) 0 20 19.7 394 43 (0.4) 0 56 14,0 782 83 0 0 782 45 17.4 15 Total 11 (2.3) 0 0 0 0 25 (5.5) 0 40 0 0 20 15.5 310 43 (0.3) 0 56 5.5 310 83 0 0 310 20 15.5 16 11 (0.3) 0 25 (2.3) 0 20 (5.5) 0 Total 56 0 0 83 0 0 000 17 Total 11 0 0 25 (0.3) 0 20 (2.3) 0 56 0 0 83 0 0 0 00 J \J krv* in hectare. . 2/ Yield in q/ha. Values in brackets indicate that the field would have been abandoned before that year, 3/ Production in q.’•M, >4 PMferled Oyerag^Ctaj Viable Willi and Wither Orlinage rrajecti. !S'opr | A.ra IM It • -r i<i <*•« Crop 1104 IORA IMF IMO imp •wo ion 1117 1111 IMS • Ml im till im IMP 7ODC1 jom 7001 2001 >U.H 2005 2OQ6 200I 70»U» loop 1 > 1 4 5 4 i R 1 IO II 17 13 14 15 16 11 IA IP ?n 71 W 71 14 n feat IM 7/M c w n.t 13.1 24.4 w 13.1 13.1 H.O a w 4.0 1.0 0 V 4.0 1.0 7.0 I1HR/RP r w 20 6 16.1 U.l w 70 A 71.1 11.0 R w 10 1.1 l.l w 1.0 4.1 0.1 25.0 15.0 24.6 n.o >4.) >1.0 ».4 B o 40 0 65.41 ■——»- 1.0 5.0 1.0 2.0 0 no 5 « 1.0 1.0 11.1 K.t 14.R 11.0 3.1 74.1 70.7 12.1 jo n ia.i M.l 16.1 1.0 1 0 0 no 5.0 1.0 7.0 >.l II 6 11.4 n.i 20.0 10.5 16.1 0 0.0 1.0 1 0 1.0 0 a.n 16.0 13.4 0 —“ 16.0 IM I>I1 /'Ml C w w hi B1 Hl HI 11.1 ta.o 10.1 11.1 13.1 27.6 >5.1 40.n 75.0 43.0 >3.1 >5.0 >4.1 >4.1 >4 6 >1.3 77.1 71.3 11.6 11,1 I6.| 0 ----------*• 1110 c w 14. a 14. a 24.6 23.1 >4.6 >4.4 >4. r 24.1 >4.1 >1.3 ». 1 Ml n.i 11.1 11.7 It.l 16.4 11.0 40.0 Al a 11.4 10.6 10.0--------- *• 1.4 0.0 4.4 R 1 IO.a 11.3 14.1 10.1 U 1 ii.i 13.7 13.3 15.1 16.4 11.0 M.O 10.0 41.3 41.n 74.1 24.5 n*< w is.a ia. a 20.0 IMI/30 c w 2* 6 14.6 >4.6 w 74.6 76.6 >1.0 N w 1.00 3.11 1.4a w 1.08 1.11 4.0 14.3 14.1 74 1 >4.1 >1 n ll> >1.1 77.0 11.1 IP.a 10.8 m.o 16.6 a • 14.6 24.6 14.6 24 6 14.6 14.6 n.i 10.0 11 0 H 0 13.3 M.w 16.3 3l.o l.ao I.M 0.41 0 4 0 ------- ► 4.5 4.5 5.0 3.1 6.0 6.3 r n I’l- 5 20 1 10.4 11.6 13.5 <1 * 11.5 M.o '■. 5 IP 0 IP. 5 40.0 c w 11.6 io.a B. a U.7 10.5 6.3 4.0 0 ------ ► w 11.6 ii.a 15. a 13.1 10.0 11.6 r w - 0 o 0 0 0.IG 1.3 1.0 3.1 4.J 14.1 14.4 13.0 11.1 II.1 21-3 ia o 16.6 17.6 15.1 13.1 3.7 7.0 1.1 w - 0 □ . 0 0 0.11 1.1 7.n 4.o 4.1 14.5 13.1 in.a 14.5 13.2 31.1 71.4 11.0 >1.1 IIP ip.a 4.1 4.0 4.1 1.6 o a 10 0 1.0 4.1 p w 5 0 i.a 1. > w 3.0 1.3 1.1 c w 14.6 14.6 14.6 w 24.6 >4.6 11.0 1.5 1.1 l.i 1.6 l.l 0.6 0. 1 1.1 1.1 4.(1 6.0 1.0 1.1 0.0 14.6 24.6 14.6 14.0 14.6 24.6 15 6 10.0 11.0 31.0 11.0 11.5 16.0 V. 5 •nJf- 4 u11 _ 0.1 21,4 0.4 1A 1 0.6 ,1 i. 0.a 1itvt. r 10.1 P.O IlS>.M1 l.l n —- 1.4 mo il n 11 4 10.0 11.0 11.3 40 O ii w 1.1? 1.76 .1.31 1.10 1.76 0.62 O -------• W 1.11 1.76 4.0 F fl - w - 4.1 4.1 4.3 5.0 3.0 3.3 Cl n 4.4 t n 14 o.n ---- ► 1.4 7.3 l.l 1.4 0 f 0 6 0 2 0 ► • 1.4 2.3 1. 1 4 O r.o 1.0 1.4 an n. > n.4 0.4 0.a 1.0 1.1 1.4 0.4 i.a m.o a a m in 100 03 101 150 104 >04 W1 7IR Ml 71> 741 >40 154 747 _____ a W HI in 100 03 101 >3 44 11 ------- ► .. Nntri t W|<»i.Mt Rr.i.age Prefect!, U - With Drainage Projecta, C - Cnttim (irrigate^ averse? yl*14 firrwood, golra, | •-— 4rnli«M* rro|*rt laplewetaA In — T--------- --------- , J ■ Cut t im *- - RR• - iwwi, ft/•<■). tl - Itair? (t/M), G - Carilen (aiaed vegetables - average glell ia t/hri), F • Forestry b'liiiling, ti*brr in « 7l«a>, f - Irrigated rr»«|.i«- (t/hv e| dry amltrr), lapl" fear • gear *»< ••1»l ?•?«! al I o* nf 4e?g drainage project lanrfece • MF). M - Nr | ladl, »•! M>|ba Vater. Air - *»r-ta <1.1. Imm, war - Irl 11aivent Ochrwe, ■ • lUwgh gracingTable 244 I Projected Avarage Crop Yields With and Without Drainage Project a fat Alternative Cropping Patterne at RRC Input Year Crop Year 1985 1986 1987 1988 1989 1990 199) 199? 1991 1996 1995 1996 199? 1998 1999 2000 2001 2002 200 ) 2004 1 2 3 4 5 6 ? 8 9 10 II 12 I) |4> B 16 11 18 19 20 SScceennarariio 2 » al o 2 I alll cotton cotton 1989/90 C W 24.6 24.5 24.1 24.4 24.3 74.1 23.8 24.2 23.6 72.8 23.5 22.8 21.6 21.7 21.5 19.0 16.2 16,2 W 24.6 24.6 27.0 30.0 31.0 35.0 35.5 16.0 16 5 37.0 17.5 38.0 38.5 1990 C w~ 24.6 20.6 24.6 24.6 24.5 24.) 24.2 24.6 24.3 71.8 23.3 24.3 22.6 39.0 39.5 21.2 24.2 19.6 15.6 15.5 w 24.6 24.6 27.0 30.0 33.0 33.0 35.0 Scenario 3 : 501 cotton. 25X me ire l pasture 1989/90 C W 24.6 24.6 24.6 24.6 24.6 24.6 74.6 35.5 34 0 36.5 37.0 37.5 38.0 38.5 39.0 39.5 40.0 24.6 73.7 22.5 23.5 22.4 20.5 20.4 19.3 17.4 17.3 17.1 15.5 0 V 24.6 24.6 27.0 30.0 31.0 15.0 35.5 16 0 16.5 37.0 37.5 38.0 38.5 39.0 39.5 40.0 H V 4.0 4.0 4.0 3.6 3.3 3.3 0 w 4.0 4.0 4.0 4.5 4.5 5.0 5.5 P w 5.2 4.7 3.0 2.4 1.8 2.9 1.7 w 5.2 4.2 3.0 2.4 1.9 4.0 6.0 6.0 6.5 7.0 7.5 8.0 * 0.9 7.0 0.1 8.0 1990 C w 24.6 24.6 24.6 24.6 24.5 24.3 24.7 74.6 w 24.6 24.6 27.0 10.0 33.0 33.0 35.0 35.5 0.3 7.5 24.3 36.0 0 8.2 8.4 8.6 8.8 9.0 9.2 9.4 9.6 9.8 10.0 23.8 23.3 24.3 22.6 16.5 37.0 37.5 38.0 21.2 24.2 19.6 15.6 15.5 38.5 39.0 39.5 40.0 0 -------- - (Prfinicion of eymbnle aa on Table 24)Table 25 : Expected Yield of Firewood and Poles in RRC Irrigated Forest Areas (m^/ha) Year from Planting Firewood Small Poles Building Poles Total 5 6 7 8 9 10 11 12 13 14 15 1 1 1 2 2 3 2 3 4 4 5 10 10 40 20 100 1 11 1 12 2 63 2 3 4 4 105 16 (coppice) 4 4 8 17 18 19 20 21 22 23 24 25 4 4 3.5 3.5 3.0 3.0 2.5 2.5 2.0 3.5 3.0 3.0 2.5 2.5 2.0 2.0 1.5 1.5 7.5 7 6.5 6 5.5 5 4.5 4 3.5 Note: Without drainage it is expected that this level of production would not be sustained due to gradual 9alinisation.APPENDIX 1 PROCEDURE FOR OPTIMISATION OF DRAIN SPACINGANNEX E Appendix 1 PROCEDURE FOR OPTIMISATION OF DRAIN SPACING The object of this procedure is to calculate the drain spacing which produces the greatest marginal return. To do this for cotton the following data are required: - the post-irrigation recession curve of the watertable; - the relationship between the watertable and total yield; an estimate of harvestable yield; the proportion of the harvested yield on the first and second pLcks; - the value of cotton; the linear cost of drain, installed in the field; (ideally, the equivalent per ha cost of the main drain and collector system); any other yield related costs (eg harvesting cost); the annual rate of loan repayment and interest. Recession curves were generated by the modified Glover-Dumm equation for different drain spacings and depths and for soils of different drainability classes (Annex B). The effect of the watertable on the yield of cotton was estimated from field work. A rooting/waterlogging index was calculated (Section 2.4) and this is related to yield in Figure 13 (Section 3.4). The loss of potential yield, and the harvestable yield which is picked in the first and second harvests are shown as percentages in Figure Ap I (in this Appendix). Both are functions of the growth form of the crop, which like yield itself are related to the rooting/waterlogging index, for reasons which are discussed in Sections 3.1 and 5.2.1 The remaining financial data are given in Table Ap 1. The inclusion of an average value of other direct costs makes the marginal return the actual gross margin. The recession curves were generated for different spacings of drains at a fixed depth in each of the drainability classes of soil. The calculation of the gross margin at each of six selected drain spacings, with the drains at a depth of 1.8m in a coarse-textured soil of drainability class 1A, as an example, are given in Table Ap 2. For each spacing, the program calculates the Waterlogging Index (WI) each day after irrigation and the simple mean of 10 days. The total yield is estimated and the percentage harvesting loss. The net yield is subdivided so that the harvesting cost may be estimated. From the linear run of drain, the total in-field drainage cost is calculated and this is reduced to an annual repayment cost. The total cost subtracted from the gross output of the crop produces the gross margin.The program then summarises these estimates, as shown in Table Ap 3. A quadratic regression of the form G - a + bd + cd^ is fitted, where G is gross margin and d is drain spacing. The analysis of variance shows that overall, the regression is an excellent fit to the data. A square-root quadratic G « a * b\Td' ♦ cd was found to be a slightly poorer fit. During fieldwork, coarse and fine-textured soils were observed to have different moisture profiles following irrigation, in addition to their features used differently in the Glover-Duram equation. This is illustrated in main-text Figure 21 where the fine-textured soil is shown to remain saturated throughout for significantly longer. Its surface also tends to remain saturated for several days. Two separate models, representing extremes of soil type, therefore were developed in the program. Soils of the Stage I area have been classified on drainability, and grouped into two predominant textural classes 1 and 2, the letters A to D referring to the depth of the impervious layer. Class I is the group of coarse- textured soils which corresponds to the profile labelled ’coarse’ in Figure 21. Drainability Class 2 includes profiles of the 'fine' type in Figure 21 but also includes medium-textured soils which are intermediate in their moisture profile characteristics. The optimum spacing for Drainability Class 2 therefore falls between the modelled 'coarse' and 'fine' textures. The summary tables for the different drainability classes are presented in Table Ap 4. For the reason given above, recession curves for Class 2 have been optimised using both 'fine' and 'coarse' soil texture models, the optimum lying between the two. The final adjustment is consequent upon the fitted regression not being a perfect fit to the calculated gross margin in the region of the optimum even though highly significant (P ■ 0.01) . This is illustrated in Figure Ap 2 for a fine and coarse-textured soil. In the case of the fine texture the calculated optimum is at a spacing somewhat wider than where it should really be. Conversely, in the case of the coarse-texture a sharp inflexion occurs close to the real optimum, whereas it is estimated by the model as at a somewhat closer spacing. Thus it is suggested that for the Drain ability Class I (coarse) soils, the inflexion indicates the real optimum. For drainability Class 2 (fine and medium-textured soils) the mean between the 'coarse* and 'fine' model optima would appear to be a reasonable opt imum.Table api Basic Financial Data Used in Drain spacing Optimisat i on Drain depth = 180 cm Cost per linear metre of drain » Br 4.15 First harvesting cost = Br 7 /q Second harvesting cost* Br 10 /q Price of cotton ex field * Br 115.8 /q Repayment rate • 6.34 % p.a. Other fixed costs: Br 1032 /ha»f-Qv\^s xavuin at birtttconiJDraLn apaclnuai tXdin^l’C. JteAfck r.^v Ocsivsc V^iuvvd & acvU Lvaiilina LbL'vaLLiinay CbllaaLa 1ALiy .CLaaa_lA 4 IW. <V»K^S£ Trutia itti 7»Vt * T KEl<«TkAeQ\E PLAINS* K T DEPTH DAILY W/l ITS <> x 24 1V 3 23.5 3 ITS A 23 4, 1T4 6 5 o*•o«■ IT* S 21 5 * 1S9 11 19.5 ItsS 14 IS S IT 16.5 ««B >* i-Sd 20 15 1ST 1SS wsTer-loses;sg Sudex * 19.5 Ti-eld of sse«i ootxon «• 15.9 q ha yield loss - 5 23 13.5 ve‘ yield^q - 15.2 Tfe£d lit first piofccq. ha> - 12.1 Ttel<± m secca: picfc - 3 r rtreer mu of draiQ<as/tta> « 116 Zraj_x cgsC< dejc!I. rnstaE’n Br.ha> = AS3 □tuss outputxBr ta> IT58 Ctests < Er Ua^z Srsest -rosiz pncfc I - 35 i=amest cost plcK 2 ^sp^nnpnr » 3 £ orize-r fixed cgsts ■» 1032 Total costs - LL78 Sro^s ■nar^Lt. = 580 5021 typer >3^xSE IzraJT spsdsa® * 83.2 in 30 5^-3 7* T 2E5^A3CME D&W3; Z69 168 167 365 162 169 157 15^ 151 147 W/T DEPTH 11 12 13 15 18 20 23 26 29 33 DAILY W/r 19.5 19 18.5 17.5 16 15 13.5 12.2 11.3 10. 1 iA<l8% ’ 15,26 71 >14 6ft &r*rf CGltM - 24,3 q/ha p^ceTri^ae P/ss - 5,11labia AuH-CunLl/iusd Nat • Z'i>2 Yield in first pjcfcf'j/tw' - iff 2 Yield In second pick - 4,9 Unatif run of drains*/faff • iW Drain costOncl, Insrsl'n-ftr/?^/ " 'J3 dross output' fir/tin > - 2f>ffi Cottr.H (Br/haff Hnrveat cost pick l * 128 Harvest cost pick 2 • 49 Repayment - 22 Other fixed costs • 1022 Total costs • 1241 Gross margin • 1440 sol l type: COARSE Drain spacing ■ 64.9 m DAY W/T HEIGHT(ABOVE DRAINS? 1 2 3 4 5 6 7 8 9 10 139 138 136 134 131 128 125 122 118 115 Water-logging index = 5.42 Yield of seed cotton - 43.4 g , ba Percentage yield loss = S.61 Net yieldlq/ha) « 39.7 Yield in first pick<q’W = 21.S Yield in second pick * IS Linear run of drain<a.ha> « 154 ■4 T DEPT 4t 42 44 46 49 52 55 58 62 55 Drain cost<incl. instal*n-Br ba' * S3* Gross output(BrZha> • 4602 Costs <Br/ha>: Harvest cost pick 1 * 152 Harvest cost pick 2 • 180 Repayment • 41 Other fixed costs • 1032 Total costs • 1405 Gross margin • 3197 soil typo: coarse Drain spacing = 40.3 m DAY W/T HEIGHT!ABOVE BRAINS* x T bETTV. nuw x 3 I 99 81 2 97 83 J 9<\ 85 s X *. A 90 90 X w •» •/6hTable ap2 Continued 5 6 7 8 9 10 86 82 78 74 71 67 94 98 102 106 109 113 1.55 1.35 1 . 19 1.07 . 98 .86 Water-1ogging Yield of seed index - cotton 1.5 55 q/ha Percentage yield loss = 10 Net yield<q/ha) = 49.5 Yield in first pick(q/ha> a 24.8 Yield in second pick = 24.8 Linear run of drain(m/ha) = 247 Drain cost(incl. insta1’n-Br/ha) Gross output<Br/ha) = 5732 Costs (Br/ha): - 1025 Harvest cost pick 1 Harvest cost = 173 pick 2 = 248 Repayment = 65 Other fixed costs = 1032 Total costs = 1518 Gross margin « 4214 Soil type: COARSE Drain spacing = 25.2 m DAY W/T HEIGHT(ABOVE DRAINS) 1 2 3 4 5 6 7 8 9 10 78 74 69 63 58 53 48 44 40 36 W/T DEPTH 102 106 . Ill 117 122 127 132 136 140 144 DAILY W/I 1 . 19 1.07 • 92 . 74 . 59 .46 . 36 . 28 .2 . 12 Water-logging index » .593 Yield of seed cotton - 55 q/ha Percentage yield loss 10 Net yield(q/ha) = 49.5 Yield in first pick(q/ha) = 24.8 Yield in second pick ■ 24.8 Linear run of drain(m/ha) ■ 397 Drain cost(incl. instal’n Br/ha> = 1647 Gross output(Br/ha> = 5732 Costs (Br/ha): Harvest cost pick 1 - 173 Harvest cost pick 2 • 248 Repayment » 104 Other fixed costs « 1032 Total costs = 1557 Gross margin = 4175Table ap2 Continued Soil type: COARSE Drain spacing • 19.5 m DAY 1 W/T HE1GHT(ABOVE DRAINS) 72 66 58 52 45 40 35 31 27 24 water-logging index • .349 Yield of seed cotton - 55 q/ha Percentage yield loss = 10 Net yield<q/ha) = 49.5 Yield in first pick(q/ha> = 24.8 Yield in second pick « 24.8 Linear run of drain(m/ha) • 513 Drain cost(incl. instal’n-Br/ha) = 2128 Gross output<Br/ha> - 5732 Costs (Br/ha>: Harvest cost pick l = 173 Harvest cost pick 2 • 248 Repayment = 135 Other fixed costs = 1032 Total costs = 1588 Gross margin « 4144 W/T DEPTH 108 1 14 122 128 135 140 145 149 153 156 DAILY W/l 1.01 .83 . 59 .44 .3 .2 .1 2E-2 0 0"able Ap3 Drain Optimisation. Analysis of Variance and Summary. ixample: Model for Coarse-Textured Soil. Drainability Class 1A *NOVA SS df MS F Regressi on 1.242938E7 2 6.21469E6 433.49 Error 57410.98 4 14352.75 Total 1.248679E7 6 Actual drain spacing Actual gross margin Estimated gross margin 86 580 694 81.2 1440 1349 64.9 3197 3094 40.5 4214 4321 25.2 4175 4244 19.5 4144 4048 Drain spacing that maximises gross margin^ 34.7 m Gross margin at optimum spacing^ Br 4369 /haTable ApA Estimation of OPtimum Drain Spacing for Soil Drainabilty Classes Model for Coarse-Textured Soils Soil Drainability Class 1A Actual drain spacing 86 81.2 Actual gross margin 580 1440 64.9 3197 40.5 4214 25.2 4175 19.5 4144 Estimated gross margin 694 1349 3094 4321 4244 4048 Drain spacing that maximises gross margin^ 34.7 m Gross margin at optimum spacings Br 4369 /ha Soil Drainability Cl ass .IB Actual drain spacing 106.5 Actual gross margin 1437 89.4 3200 62.6 4237 43.9 4219 36.1 4206 Estimated gross margin 1507 3060 4306 4316 4111 Drain spacing that maximises gross margin= 53 m Gross margin at optimum spacing= Br 4400 /ha Soil Drainability Class 1C Actual drain spacing Actual gross margin 136.4 1443 117. 1 3207 85.3 4249 61.2 4236 50.6 4227 Estimated gross margin 1538 3024 4325 4360 4 115 Drain spacing that maximises gross margin- 72.2 m Gross margin at optimum spacing® Br 4446 /ha Soil Drainability Class ID Actual drain spacing Actual gross margin 168.4 1447 146 3211 107.5 4255 77. 1 4245 63.5 4238 Estimated gross margin 1561 2999 4338 4383 4115 Drain spacing that maximises gross margin- 90.7 m Gross margin at optimum spacing^ Br 4473 /haTable Ap4 Continued Mode]for Coarse-Textured Soils Soil Drainability Cl ass 2A Actual drain spacing 67.6 54 33.7 21 16.2 Actual gross margin 1781 3296 4201 4 154 41 17 Estimated gross margin 1813 3227 424 1 4207 4061 Drain spacing that maximises gross margin** 28.2 m Gross margin at optimum spacing- Br 4290 /ha Soil Drainability Class 2B Actual drain spacing 74.2 51.9 36.3 29.8 Actual gross margin 3310 4229 4207 4191 Estimated gross margin 3315 4200 4264 4156 Drain spacing that maximises gross margin- 41.9 m Gross margin at optimum spacing- Br 4294 /ha Soil Drainability Class 2C Actual drain spacing 96.8 70.2 50.1 41.4 Actual gross margin 3318 4242 4227 4216 Estimated gross margin 3326 4207 4293 4177 Drain spacing that maximises gross margin- 56.6 m Gross margin at optimum spacing-* Br 4320 /ha Soil Drainability Class Actual drain spacing 138.5 119.8 87.7 62.4 51.1 2D Actual gross margin 1801 3323 4249 4237 4228 Estimated gross margin 1895 3149 4315 4356 4124 Drain spacing that maximises gross margin* 73.7 m Gross margin at optimum spacing- Br 4433 /haTable Ap4 Continued Model for Fine-Textured Sulls Soil Drainability Class 2A Actual drain spacing 54 33.7 21 16.2 Actual gross margin 2795 3292 3366 3362 Estimated gross margin 2796 3289 3372 3358 Drain spacing that maximises gross margin- 21.3 m Gross margin at optimum spacing= Br 3372 /ha Soil Drainability Class 28 Actual drain spacing 74.2 51.9 36.3 29.8 Actual gross margin 2808 3319 3419 3436 Estimated gross margin 2810 3312 3433 3428 Drain spacing that maximises gross margin^ 34.1 m Gross margin at optimum spacing= Br 3435 /ha soi1 Drainabi1ity Class 2C Actual drain spacing 96.8 70.2 50. 1 41.4 Actual gross margin 2817 3332 3439 3461 Estimated gross margin 2819 3322 3457 3450 Drain spacing that maximises gross margin= 47.3 m Gross margin at optimum spacing= Br 3459 /ha Soil Drainability Class 2D Actual drain spacing 119.8 87.7 62.4 51.1 Actual gross margin 2822 3340 3449 3473 Estimated gross margin 2825 3328 3470 3461 Drain spacing that maximises gross margin^ 59.1 m Gross margin at optimum spacing^ Br 3472 /haFIGURESFIGURE l.l MASTER DRAINAGE PLAN FOR MELKA SADI ANO AM I SARA AREAS FEATURES AND DESCRIPTION OF MOISTURE ANO SALINITY PROFILES IN COTTON AND 8ANANA FIELD SAMPLE PLOTS USING COTTON PLOT 6 AS AN EXAMPLE. See Figure Nos. I 2 to 1.9 and Figure Nos 2.1 and 2.2 DEPTH (cm) moisture profile SALINITY PROFILE YIELD PLOT 6 ■ (J) Days offer irrigation stops EQ|:2 Suspension- mS/cm Yield (K) q/ha FEATURES A Surface water persisted for several days in some plots; e.g. Plot 17 and this is indicated in the Figure. B Soil which is saturated (above Field Capacity) following Irrigation. C Soil which has drained by gravity to Field Capacity. Soil from which (mainly) roots have evaporated moisture to a level somewhere below Field Capacity. E Soil dried normally, well below Field Capacity prior to irrigation. r Soil behind the wetting front wetted (but not saturated! by capillarity. Example of a temporary perched watertable on a thin layer of compact, impervious horizon, associated with a peak H in the salinity profile. Dries between irrigations. Peak in salinity profile due to an impervious soil horizon. Note that salinity measured as Field Capacity of a 1:2 (vol) suspension. Watertable which persists during irrigated period but which is perched where deep augering has established dry soil below. J Plot number - see Table 5.1, 5.2 and 5.3 for other details. K field of seed cotton measured in yield-sample plot by boll-count. In corresponding figures for bananas (Fig. 5.2) the yield is replaced by height (cm) of mature pseudostem. SIR WILLIAM HALCROW S PARTNERS JULY 1905I I I I IDepth below So*I Surface - cm FIGURE 1.2 MASTER DRAINAGE PLAN FOR MELKA SADI ANO AMIBARA AREAS MOISTURE 9 SALINITY PROFILES a ESTIMATED SEED COTTON YIELDS IN SAMPLE PLOTS (For description see Figure II) PLOT 4 Doys offer irrigation stops E C | 2 Suspension - mS/cm SIR WILLIAM HALCROW a PARTNERS JULY 1905Depth belo* Soil Surfoce - cm FIGURE 1.3 MASTER DRAINAGE PLAN FOR MELKA SADI AND AMIBARA AREAS MOISTURE 8 SALINITY PROFILES a ESTIMATED SEED COTTON YIELDS IN SAMPLE PLOTS c- - +1- ' -t'- .1 L .'.4™ j lx’*' ■ r-rl-.'- v - 1" ’L r —- — —------ IB fit $3 \ l haxl___ L r- *71 37 Ba/ho I PLOT II Si % L r/4 0 10 20 30 40 50 60 Doys otter irrwjofion itops Yield E G i 2 Suspension - mS/cm SIR WILLIAM HALCROW 8 PARTNERS JULY 1985I I I I I I I I I I I I (I a a n ■Depth below Soil Surface - cm FIGURE I 4 MASTER DRAINAGE PLAN FOR MELKA SADI AND AMIBARA AREAS MOISTURE a SALINITY PROFILES a ESTIMATED SEED COTTON YIELDS IN SAMPLE PLOTS 37 Bq/ho PLOT 12 I PLOT 13 PLOT 15 II. All YlELO NIL Days of far «rrtqofion stops 0 IO 20 30 40 50 60 Yield EC 12 Suspension - m /cm SIR WILLIAM HAL CROW 8 PARTNERS JULY 1985». I I I I 1Depth below So«l Surface FIGURE 1.5 MASTER ORAINAGE PLAN FOR MELKA SADI AND AMIBARA AREAS MOISTURE 8 SALINITY PROFILES a ESTIMATED SEED COTTON YIELDS IN SAMPLE PLOTS Ooys afttr irrigation stoos EC|;Z Suspense - mS/cm SIR WILLIAM HALCROW a PARTNERS JULY 1985Depth below Soil Surface - cm FIGURE I 6 MASTER DRAINAGE PLAN FOR MELKA SADI AND AMI0ARA AREAS MOISTURE 8 SALINITY PROFILES ft ESTIMATED SEED COTTON YIELDS IN SAMPLE PLOTS ■ i A IT PLOT 20 7 - 30 3<j/ha P! PLOT 23 Days offer irrrqofion stops 10 20 30 40 50 60 E.Q|:2 Suspension - mS/cm Yield SIR WILLIAM HALCROW a PARTNERS JULY 1985Depth below Soil Surfoce - cm FIGURE I 7 MASTER DRAINAGE PLAN FOR MELKA SAOI AND AMIBARA APEAS MOISTURE 8 SALINITY PROFILES a ESTIMATED SEED COTTON YIELDS IN SAMPLE PLOTS — 30 0<yno n PLOT 26 ■ 46 5q/ha PLOT 27 Doys offer irrigation slops --- . -------------------- 4 0 50 6 0 Yield E.C.I 2 Suspension - mS/cm SIR WILLIAM HALCROW ft PARTNERS __ _____________________________ July 1995Depth below Soil Surfoce - FIGURE MASTER DRAINAGE PLAN FOR MELKA SADI AND AMIBARA AREAS MOISTURE a SALINITY PROFILES a ESTIMATED SEED COTTON YIELDS IN SAMPLE PLOTS PLOT 30 Bq/ho 28 PLO 33 Days after irrigation slops SIR WILLIAM HALCROW 8 PARTNEDeplh below Soil Surfoce - cm FIGURE MASTER DRAINAGE PLAN FOR MELKA SADI AND AMI0ARA AREAS MOISTURE 9 SALINITY PROFILES a ESTIMATED SEED COTTON YIELDS IN SAMPLE PLOTS 48-7q/ho PLOT 54 Doys offer irrigation stops O i‘o 2 0 30 «0 E.Q|:g Suspension - 60 Yield mS/cm SIR WILLIAM HALCROW 6 PARTNERSii I i i I. [ [ r I DE FIGURE 2 I MASTER DRAINAGE PLAN FOR MELKA SADI AND AMIBARA AREAS MOISTURE AND SALINITY PROFILES AND AVERAGE MAXIMUM HEIGHT OF PSEUDOSTEMS IN BANANA PLOTS 250 PLOT 3 Granular g l«ti comp- ZC i 8 1U J § 5 $ 5 1 ] ] ] J J ] J ] a PLOT 6 Compact ZCL SIR WILLIAM HALCROW 8 PARTNERSDeplh below Soil Surface - cm FIGURE 2.2 MASTER DRAINAGE PLAN FOR MELKA SADI AND AMI0ARA AREAS MOISTURE AND SALINITY PROFILES AND AVERAGE MAXIMUM HEIGHT OF PSEUDOSTEMS IN BANANA PLOTS SIR WILLIAM HALCROW 8t PARTNERSFIGURE MASTER DRAINAGE PLAN FOR MELKA SADI AND AMI8ARA AREAS VARIATION OF CONVERSION FACTOR OF EC|:2 TO EC BETWEEN a CONTRASTING SOIL TYPES AND WITH SOIL MOISTURE CONTENT NOTES PWPV FCV SPv Permonent Wilting Point Field Capocity Saturation Point (All % by volume or mm/mm %) SIR WLLIAM HALCROW ft PARTNEf JULY 191Lond Elevotioa (cm) Crop Height (cm) Land Elevation (cm) Crop Height (cm) FIGURE MASTER DRAINAGE PLAN FOR MELKA SADI ANO AMI9ARA AREAS GROUND ELEVATION AND CROP HEIGHT ALONG SELECTED SECTIONS FIELD 6/4 - Example of Wafer Oeficit to Waterlogging Distance from Canol (m) PLAN OF PLOTS B SECTIONS tB SIR WILLIAM HALCHOW 8 RftRThLoad Elevotion (cm) Crop Height (cm) Lond E levo I ion (cm) Crop Height (cm) FIGURE 4.2 MASTER DRAINAGE PLAN FOR MELKA SADI AND AMIBARA AREAS GROUND ELEVATION ANO CROP HEIGHT ALONG SELECTED SECTIONS FIELD 3/2/6 PLAN OF PLOTS B SECTIONS tb Distonce along Section A0 (m) SIR WILLIAM HALCROW 3 PARTNERSFIGURE 4 3 MASTER orainage plan for melka saoi and amibara areas GROUND ELEVATION AND CROP HEIGHT ALONG SELECTED SECTIONS FIELD 3/1/9 PLAN OF PLOTS a SECTIONS fF N/ ^PUJT2I RjQT 240 A O IO 20 30m I____ Z___ _i_____ i Apprai Scote FIELD 1/28/56 (Deep vertical) SIR WILLIAM HALCROW 3 PARTNE JULY ISLond Elevation (cm) Crop Height (cm) Land Elevalion (cm) Crop Height (cm) FIGURE 4 MASTER DRAINAGE PLAN FOR MELKA SADI AND AMIBARA AREAS GROUND ELEVATION AND CROP HEIGHT ALONG SELECTED SECTIONS FIELD 3/1/9 SECTION AB (Seepion) PLAN OF PLOTS 3 SECTIONS N I PLOT 26 PLOT 20 PLOT 29 SECTION EF (See pion) EXAMPLE OF SALINITY SIR WILLIAM HALCROW 8 PARTNER JULY 196FIGURE 4 5 MASTER DRAINAGE PLAN FOR MELKA SADI AND AMI9ARA AREAS GROUND ELEVATION AND CROP HEIGHT ALONG SELECTED SECTIONS Drsfonc# from Carol (m) FIELD 3/2/24 - Example of Salinity SIR WILLIAM HALCROW 8 PARTNE JULY ISmaster orainage plan for melka sad AND AMI BARA AREAS LOCATION OF FIELDS SAMPLED IN RANDOM SAMPLE SURVEY Jul. IQMFIGURE 6 MASTER DRAINAGE PLAN FOR MELKA SAOl ANO AMIBARA AREAS nomogram to estimate seed cotton yield from AVERAGE CROP HEIGHT AND COLOUR INDEX HEIGHT tI80 YIELD ■170 160 150 140 130 ISO -t-55 (Max. Potential Yield) COLOUR INDEX I ■2 50 ■110 ® u 100 - e 90 jj 80 c a. e % □ 70 ® 60 50 40 4 6 8 ■10 12 14 ■16 ■10 ■■20 ■22 1-24 30 20 O USE OF NOMOGRAM: Pioce straight-edge from the measured height (left) across to measured colour index (right). Read yield estimate off centre scoie. Note that the nomogram gives more thon twice the weight to crop height os to colour index. SIR WILLIAM HALCROW 6 PARTNERS JULY 1905FIGURE 7 MASTER DRAINAGE PLAN FOR MELKA SADI AND AMI0ARA AREAS STYLISED GROWTH PATTERN OF COTTON PLANTED ON I JUNE AT AIP Month | MAY | JUNE I JULY | AUG I SEPT I OCT | NOV 1 DEC 1 Doys Stage PI P O 30 60 90 1^0 + 150 + 18 210 FH FI SH........ . ......... _E.- N.B The time-socle is not much affected by planting date but is morfcedly affected by moisture regime and salinity. NOTES PI - Pre* irrigation P - Planting FH - FI - First Harvest Finol Horvest SH - Second Harvest E - Eradication (Width indicates rote from nil to maximum) SIR WILLIAM HALCROW 3 PARTNERS JULY 1905R qans*-*FIGURE 9figure io master drainage plan for MELKA SADI AND AMIBARA AREAS YIELD DECUNE OF COTTON BETWEEN 1981 ANO 1983 SEASONSI I I I I I I I I 1 I I 1 I 1 I I I I I IFIGURE II MASTER DRAINAGE PLAN FOR MELKA SADI AND AMIBARA AREAS RELATIONSHIP BETWEEN ROOTING INDEX AND AVERAGE HEIGHT OF COTTON PLANTS NOTE Saline ond chemically infertile plots ond plots in which irrigation was exceptional are omitted. Curve fitted by eye. SIR WILLIAM HALCROW ft PARTNERS JULY 1985FIGURE 12 MASTER DRAINAGE PLAN FOR MELKA SADI AND AMIBARA AREAS RELATIONSHIP BETWEEN ROOTING INDEX ANO COLOUR INDEX IN COTTON AT NON-SALINE BUT CHEMICALLY FERTILE SITES NOTE Curve fitted by eye. Form is the most likely on o physioloqtcol basis Solinity suppresses yellowness and hence CI. SIR WILLIAM HALCROW 3 PARTNEF JULY 196II B B ■ I I I I I I I I I IFinol adjusted seed cotton yield (q/ha) FIGURE i: MASTER DRAINAGE PLAN FOR MELKA SADI AND AMIBARA AREAS THE EFFECT OF ROOTING INDEX ON THE ESTIMATED YIELO OF SEED COTTON NOTE: Rooting. Index is o measure of the quolity of root environment It takes account of permissible rooting depth ond the extent of temporary saturation of the root zone following irrigation. sir wiluam halcrow a partners JULY 1985■ ■ ■ I I I I I I I I I I I 1FIGURE MASTER DRAINAGE PLAN FOR MELKA SADI AND AMlBARA AREAS EFFECT OF SOIL SOLUTION CONCENTRATION ON GERMINATION OF COTTON SEED Soil Solution Concentration (mmhos/cm) EFFECT OF SOIL SOLUTION CONCENTRATION ON THE HEIGHT OF COTTON SEEDLINGS ONE WEEK AFTER GERMINATION SIR WILLIAM HAUCF ft PAR TNI JULY IIFinal adjusted seed cotton yield (q/ho) FIGURE 15 MASTER DRAINAGE PLAN FOR MELKA SADI AND AMIBARA AREAS THE EFFECT OF SALINITY INDEX ON THE ESTIMATED YIELD OF SEED COTTON NOTE Doshed line is likely form of curve since generally an SI>4 5 of mid-season reflects a surfoce ECe>20mS/cm at plontmg which completely prevents germinotion. SIR WILLIAM HALCROW 8 PARTNERS JULY 1985Average Height of Pseudostems (cm) Score for vigour (best -►-) FIGURE l€ MASTER DRAINAGE PLAN FOR MELKA SADI AND AMIBARA AREAS THE EFFECT OF SALINITY ON HEIGHT AND VIGOUR OF BANANAS 2OQ- nl------------------------------------ --------- --------- - lT ■ I Im 0 05To 1-5 Salinity Index 202 5 KEY —-Or—Score for vigour (V) ~ * Height of Pseudostems F Sandy, infertile plot o Sites scored for vigour but where other problems are opparent. SIR WILLIAM HALCROW 8 PARTNE JULY 19FIGURE 17 MASTER DRAINAGE PLAN FORMELKA SADI AND AWIBARA AREAS RELATIONSHIP BETWEEN AVERAGE HEIGHT OF CROP AT MATURITY AND ESTIMATED YIELD OF SEED COTTON NQIE Curve fitted by eye. Yield is expected to fall oto plant height>190cm due to on excessively high LAI and consequent low NAR. SIR WILLIAM HALCROW 8 PARTNER JULY 196FIGURE 18 MASTER DRAINAGE PLAN RDR MELKA SADI AND AMI0ARA AREAS RELATIONSHIP BETWEEN THE COLOUR INOEX OF THE LEAVES AT ABOUT 145 DAYS AND THE ESTIMATED SEED COTTON YIELD AT NON - SALINE SITES. COLOUR INDEX IS A MEASURE SIR WILLIAM HALCROW 0 PARTNER! JULY 198!Marginal return 0/ha Marginal return B/ha FIGURE 19 MASTER DRAINAGE PLAN FOR MELKA SADI AND AMIBARA AREAS THE EFFECT OF INTERDRAIN SPACING AT DIFFERENT DRAIN DEPTH ON MARGINAL RETURN «— Repayment 10 yrs. •— Repayment 25yrs. x Maximum marginal return Drain depth shown on curves HIGH LATE-SEASON LOSSES (L) Interdrain spocmg (m) LOW LATE-SEASON LOSSES SIR WILLIAM HALCROW ft PARTNERS JULY 1985figures 20 a MASTER DRAINAGE PLAN FOR MELKA SADI AND AMIBARA AREAS FIGURE 20: THE EFFECT OF IN-FIELD DRAIN DEPTH ON MAXIMUM MARGINAL RETURN TO DRAINAGE 3700 3600 L25 LIO Assuming late season loss of potential crop remainsat 10-20% Assuming reduced loss of potential crop to I - 5 % Note Main drain cost is not included. It moy be expected to increose with greater dram depth thereby accentuating the maximum marginal return at about 16m os illustrated by hypothetical Curve C H25 2900 2400 2300 FIGURE2I- Repayment over 10 years □t 5% interest = 12 9% pa Repayment over 25 years ot 6% interest = 7 8% 2 0 Depth of drains (m) ILLUSTRATION OF THE MOISTURE PROFILE FOLLOWING IRRIGATION IN SOILS OF CONTRASTING PROFILE 5A7 ** , ,9 JQ Days after irrigation stops FINE TEXTURED SOIL COARSE TEXTURED £ SOU- Q I S- Soil saturated by irrigation water draining through profile FC = Soil drained to field capacity WT- Watertable R= Watertable recession curve I - Irrigation < (See Fig. 1.1 for general description of moisture profile) SIR WILLIAM HALCROW ft PARTNERS JULY 1985FIGURE 22 MASTER ORAINAGE PLAN FOR MELKA SADI AND AMI0ARA AREAS PARAMETERS FOR CALCULATION OF COTTON IRRIGATION SCHEOULE ETo Penman Reference Crop Evaporation ETc Crop Evaporation (ETo Kc) Wn Net Water Requirement Pe(50) Median Effective Precipitation Kc Crop Coefficient (I-p) Moisture Depletion Foctor Sa Available moisture (Av. 22%) 0 Rooting Depth 90 Aug. 100 120 140 Sept I Oct 160 Nov. Days from plonting (P) and approx, equivalent calendar month —— SIR WILLIAM HALCROW 9 PARTNERS JULY 1985FIGURE 23 MASTER DRAINAGE PLAN FOR M ELK A SADI AND AMI8ARA AREAS OPTIMISATION OF COTTON IRRIGATION SCHEDULE Deys from ptonting (P) ond Approx. Equivalent Calendar Month (l-p)So 0 Wn PR 11,12 etc Jy 2/ Total Available Moisture for maximum yield (mm) Water Requirement (= consumptive use) (mm) Preplant Irrigation Number of postplant irrigation Irrigation Interval (days) Amount (net) of irrigation required (mm) GROWTH STAGE Maximum rate is indicated by width of arrow and location of letter V Vegetative growth (stems, roots, leaves) F Flowering SW Boll swelling (requirement of assimilates) SP Boll splitting SIR WILLIAM HALCROW 3 PARTNERS JULY 1985FIGURE 24 MASTER DRAINAGE PLAN FOR MELKA SADI AND AMI8ARA AREAS THE EFFECT OF AGE OF CROP AT FINAL IRRIGATION ON THE LOSS OF POTENTIAL COTTON LINT YIELD FOR THE DIFFERENT TYPES OF CROP CREATED BY VARYING DEGREES OF OVER AND UNDER IRRIGATION Waterlogging Index (WI) and position of Coses A, 0 and C —160----------- 200 approximate maximum plant height (cm)-------------------------------—75 —2 0------------3-5 approximate LAI -10 SIR WILLJAM HALCROW 0 PARTNERS JULY 1985FIGURE 25 MASTER DRAINAGE PLAN FOR MELKA SADI ANO AMIBARA AREAS INDIAN RAISED - BED SYSTEM FOR MINIMUM TILLAGE SIR WILLIAM HALCROW a PARTNERS JULY 1985FIGURE 2 MASTER DRAINAGE PLAN FOR MELKA SADI AND AMIBARA AREAS MOISTURE PROFILE OF SOIL AFTER IRRIGATION WHEN AVERAGE ANNUAL DEPTH OF WATERTA8LE IS LESS THAN I Om. Days after irrigation stope 01 23456789 O Yield reduction 28% NOTE Total overage contribution to groundwater 226 mm Assume 8 irrigations i.e. 28mm each Porosity 0 08% i.e. 353mm rise each irrigation. SIR WILLIAM HALCROW a PARTNER’ JULY 198!% of yield in 1st. harvest Yield loss {%) FIGURE APP I MASTER DRAINAGE PLAN FOR MELKA SAOI AND AMI9ARA AREAS EFFECT OF WATERLOGGING ON HARVESTABLE YIELD AND ITS DISTRIBUTION BETWEEN FIRSTAND SECOND HARVESTS OF COTTON SIR WILLIAM HALCROW ft PARTNERS JULY 1985FIGURE APP 2 MASTER DRAINAGE PLAN FOR MELKA SADI AND AMIBARA AREAS COMPARISON OF CURVES FITTED TO GROSS MARGINS AT DIFFERENT DRAIN SPACINGS, FITTED BY EYE AND BY QUADRATIC REGRESSION SIR WILLIAM HALCROW & PARTNERS JULY 1985 'Coorw' Model Gro»i Margin (G in Br/ha)
🔊 Read Aloud
⏹ Stop
<
Summery
<
×
Summery