Sudan - Gezira study mission (Vol. 4 of 15) : Annex II : appendix 1 : soils of the Gezira

RESTRICTEE ' *.03 UNN39 Vol. 4 G E Z I R A S T U D Y M I S S I O N unm C,D ANNEX II - APPENDIX 1 - SOILS OF THE GEZIRA ___ X G: -A - C3 C C:*  - ii - ANNEX II APPENDIX 1 SOILS OF THE GEZIRA TABLE OF CONTENTS CHAPTER PAGE 1. INTRODUCTION ... ........................ ......... 1 2. DESCRIPTION AND PROPERTIES OF NORMAL GEZIRA SOILS 3 Description ............. .................. 3 Classification ..... .O..-*. . ...........*. 3 Gezira Main soils .. . .. . 4 Managil sois............... 4 Chemical Properties ............ .......... 4 Exchangeable bases ......................5 Soil pH.................H * * * ............ 6 Soluble sals............ s0.. 6 Total sals......l......s* . 8 Nitrogen .... . . . 10 Physical Properties.. ........ . . . o.. ...... . 11 Cracking ............... to ..... *.0.0......... 11 Bulk density ....... . ... . . ... . 12 3. SOIL WATER RELATIONSHIPS ........................ . 14 Soil Moisture ......... o.................... oo.. ...... 14 Available moisture ....... ............... 14 Permeability: ........ . ... ...... ..0 .... 15 : of loamy soils............ 16 : of heavy clay soils ....... 16 Irrigation . . .. .. .... .......... . 17 Cracking and water entry . . 17 Watering intervals ..................... 18 Effects of prolonged flooding ........... 19 Drainage ...... .. . . .. . ..... . 21 Surface drainage .. ......... * ............ 21 Internal drainage....................... 21 4. SALINE SOILS IN GEZIRA ................... o....... . 23 Natural saline and alkaline soils: . .... 24 : the sodium value test............. 24 Secondary salinization: .. .............. 25 : quality of irrigation water ....... 25 : past changes in salinity in the Gezira ................ ............. . 28  CHAPTER PAGE rate of secondary salinization in the future ... ............ 29 5. FERTILITY OF GEZIRA SOILS ......................... 31 Annual fluctuations in yied .............el d 31 Variations in yield .. 31 Maintenance of Fertility ...................... 32 Fallows ................ ......... a ... .... 33 Functions of fallows .. 35 Recommendations on fallow frequency ... 36 Fertilizers ............. , ......... 36 Nitrogen: cotton ...................... 36 : other crops ................. 38 Other fertilizers . . 38 Field results ....... . ... . 38 6. IMPROVING THE PRODUCTIVITY OF GEZIRA SOILS ........ 39 Improved moisture supply .............. 39 Improving saline and alkaline soils .............................. #..... 40 : research ..... . .. ............... 41 : alleviating salinity in the field. . ....... ..... 41 Fertilizer: ........... .. ... .............41 : legumes ........................... 42 : other fertilizers ................. 42 Alternative crops .. .. ......................... 42 7. SOIL ANDLAND USE CLASSIFICATION ................... 44 Soil surveys ............. . .... .. . 44 Land capability classification ................ 5  ANNEX II APPENDIX 1 C H A P T E R 1 SOILS OF THE GEZIRA INTRODUCTION Soils and water form the major natural asset of the Sudan and are the basis of the country's economy. Though black alkaline cracking clays are widespread in the tropics and sub-tropics, nowhere are they irrigated so extensively as in the Gezira. From the start of develop- ment, agricultural scientists in the Sudan recognized that these clays had properties which made their management different from irrigation schemes elsewhere and planned their research accordingly. Before the building of the Sennar Dam, pilot pump schemes showed that the land could grow good cotton. Fertilizer trials were under- taken long before the use of fertilizers became a commercial propo- sition; experiments were carried out to select a rotation which could maintain productivity of the cotton crop and yet include the traditional crops of the area; many new crops were introduced at the Gezira Research Station so that their behavior and growth in the local environment and soils could be studied. In other tropical territories much work has been done on individual crops, and some on rotations, but the Gezira is outstanding in the amount of research devoted to evolving and perfecting an agricultural system, centering round cotton, in an area which had virtually no previous tradition of settled agriculture. Although the Scheme is fortunate in having so much information there are some serious omissions. Irrigation was not available all the year and thus there have been few investigations of the effect of continuous irrigation throughout the year, or of other agricultural systems (e.g. livestock) where cotton is not the main source of cash. The lack of precise knowledge of the possible effects of continuous irrigation is a serious weakness in present investigations. Since cotton is the most important crop, the greater part of the research work has been done on it. In the discussion of Gezira soils in this report more attention has inevitably been given to their influence on cotton than on other crops. Chapters 2 to 5 of this Appendix describe such evidence as could be obtained from published and unpublished reports, from  ANNEX II - 2 - APPENDIX 1 discussions with soil scientists at present or formerly in the Sudan, from personal observation or from records elsewhere. The management of the Gezira soils is discussed in Chapter 6. For many subjects sufficient experimental evidence does not exist to make recommendations for immediate application in the Scheme. Thus some of the recommendations will have to be tried out on a pilot scale, others will need considerable research and the mission can do no more than indicate the important problems requiring further definition before research programs can be started.  ANNEX II APPENDIX 1 C H A P T E R 2 DESCRIPTION AND PROPERTIES OF NORMAL GEZIRA SOILS DESCRIPTION The central Sudan consists mostly of a vast plain which slopes gently downwards from the south to the north and west. The 420 m. contour runs through Sennar and the 385 m. through the south and west of the Scheme. The Managil ridge varies from 430 to 460 meters ASL. (see map in Report.) Although the origin of the soils has been widely discussed it is generally agreed that the area consists largely of sediments of the Blue Nile and its tributaries, mostly derived from the basic rocks of the Ethiopian Highlands. The absence of any form of bedding in the deposits is explained by the swelling and shrinking of the clays during wetting and drying. Some clay soils formed by the weathering in situ of tertiary lavas, basement complex and Nubian series occur in parts of the Gezira and Kanana. Classification The Gezira soils generally fit the description of Grumusols, which includes the black cotton soils, the 'regur' soils of India and the black cracking soils of Australia and the southern United States. In the most recent classification, the Seventh Approximation, they have been termed Vertisols, i.e. soils that crack widely, have a high content of montmorillonitic clay and have a high base exchange capacity. PMost Gezira soils fit the sub-order usterts but some with a darker colour are classified in the sub-order aquert. The most important agricultural properties of both these sub-orders are their stickiness when wet, their hardness when dry and the rather narrow range of moisture content over which they can be worked. The physical properties of Gezira soils, especially their self- mulching character, suggest that they should be included in the great soil group, grumustert. However, their chemical properties, particularly the rather high exchangeable sodium, are closer to those of the great soil group mazustert.  ANNEX II - 4 - APPENDIX 1 Gezira Main Soils The most marked features of the soil when dry are the wide cracks and the loose granular surface. The intensity and width of cracks varies from north to south of Gezira Main, the cracking being generally less pronounced in the north than in the south. Cracking may extend throughout the depth of the dark brown or dark grey browqn surface soil (i.e. to about 60 cm) which contains calcium carbonate nodules coated with manganese; soft white streaks of calcium carbonate and very tiny crystals of gypsum may also occur. Between the surface soil and the very dark grey clay (the so-called grey layer) is a transitional layer where deep tongues of the surface brown clay penetrate into the underlying grey clay. The grey layer has some calcium and nests of gypsum crystals; when dry, it is very hard and difficult to penetrate with an auger and,when it is near the surface, it hinders root and water penetration. Below the grey layer a very firm brown clay with gypsum extends to considerable depth,but sometimes a sandy layer (usually 75-95 cm thick) is found at 2 - 2.5 m. M4anagil Soils These are of the same origin as Gezira Main soils and are,in general,similar. Few detailed descriptions are available but it would appear from reports that they are more saline and alkaline than soils in Gezira Main. Saline and alkaline patches are common, accumulation of salts being especially noticeable where there have been leakages in the banks of the canals. No information is available on the real extent of these patches though a tour through the area suggests that they are numerous but small. The surface soils of the tops of ridges in Managil often have a very high pH. In some areas of Managil where there is a hard pan at 30 to 60 cm, growth of cotton is uniformly poor; the tap roots of the plants bend at right-angles because of the impenetrability of the hard pan. CHEMICAL PROPERTIES The soil throughout the Scheme is a heavy, alkaline clay. Clay percentages vary between 50 and 70 percent, silt between 15 and 30 percent and fine sand between 10 and 25 percent; coarse sand is almost entirely absent. The light portion of the sand fraction is composed mostly of quartz with some potassium feldspars. The heavy fractions have a high proportion of hornblende, with considerable quantities of augite and epidote. These readily weatherable minerals supply magnesium and calcium. X-ray analyses indicate that the clays may be illitic or montmorillonitic in character though the physical and chemical behavior suggests that they are mostly the latter.  ANNEX II - 5 - APPENDIX 1 The soluble salts, and especially the exchangeable sodium, have received much attention from research workers because they play such an important part in soil fertility and influence the physical properties of the soil. Exchangeable Bases The cation exchange capacity is approximately one milli- equivalent per gram (m.e/g) of clay. Calcium is the dominant cation on the exchange complex throughout the profile, although in a typical soil its relative percentage decreases (in favour of sodium) at increasing depths (Table 1). Exchangeable sodium invariably has been found in all of the thousand of samples of topsoil from Gezira Main and Managil which have been examined. In the nearby Roseires Project, the consultants found that over 70 percent of samples of the top 45 cm of soil had exchangeable sodium percentages (e.s.p.) of 10-15; at greater depths (45-90 cm layer) the e.s.p. exceeded 15 percent in over 80 percent of samples._/ Table 1: EXCHANGEABLE BASES IN A GEZIRA MAIN SOIL Cation a/ Depth, cm - Calcium . Sodium -.Magnesium Potassium 0 - 15 78 6 14 2 15 - 30 73 10 15 2 30 - 45 66 16 16 2 45 - 60 62 20 17 1 a/ as percentage of total bases exchangeable in ammonium chloride. Source: Ministry of Agriculture, Sudan. Bull. 12. The relative proportions of different cations in the surface layer (0 - 30.5 cm) is not uniform throughout the Scheme. Little information is available from Managil, but the proportion of sodium is much higher in the north than in the south of Gezira Main. Typical Gezira soils have a much higher e.s.p. than the typical more freely permeable soil of the Gash delta (Table 2). 1/ Report on Rahad Project 4a.  ANNEX II - 6 - APPENDIX 1 Table 2: COMPARISON OF EXCHANGEABLE BASES IN GEZIRA MAIN AND GASH SOILS Cation a/ Gezira Main Typical Typical Turabi Wad Medani Hag Abdullah Gezira Soil Gash Soil Calcium 70 80 82 78 83 Sodium 20 9 7 8 2 Magnesium 9 9 9 12 14 Potassium 1 2 1 2 2 a/ as percentage of total bases (ammonium chloride exchangeable) in top 30.5 cm. Sources: Gezira Main samples. Ministry of Agriculture,Sudan. Bull. 12. Typical Gezira and Gash samples. J. Agric. Sci. 29, 1. Soil pH The pH of 1:5 water suspensions of Gezira soils is about pH 8.5 to pH 9.5 although more alkaline patches are found. The pH is slightly higher in the upper horizons, where sodium occurs as the carbonate, than at greater depths where the sulfate predominates ./ The pH of soils with high e.s.p. is often above pH 8.5 but there is no close correlation between the pH and exchangeable sodium per- centage. The pH is influenced by a number of factors other than sodium, soluble salts markedly depressing the value.i/ Soluble Salts Soluble salts also increase with depth and there is usually a sharp increase at the downward limit of penetration of irrigation and rain water, which usually coincides with a compact layer in the profile. Table 3 shows values for soils at the G.A.R.S., where the increase was found to occur at 75 cm irrespective of the rotation practised. 1/ J. Agric. Sci. 15, 407. 2/ Soil Sci. 71, 273.  ANNEX II - 7 - APPENDIX 1 Table 3: SOLUBLE SALTS IN A G.A.R.S. SOIL Depth, cm Percentage Soluble Salts a/ Rotation Rotation Rotation Rotation ccCC b/ CFCF CFFC CFFF 0-15 0.06 0.09 0.08 0.09 15-30 0.11 0.10 0.09 0.09 30-45 0.11 0.09 0.09 0.09 45-60 0.13 0.11 0.09 0.09 60-75 0.16 0.12 0.14 0.13 75-90 0.36 0.26 0.147 0.34 90-105 0.68 0.47 o.56 0.64 105-120 0.59 0o-45 0-49 0.52 120-135 0.43 0..44 0.41 0.45 135-150 0.50 0.49 0.50 0.36 150-165 0.77 0.51 0.56 0.35 165-180 0.82 o.56 0.63 0.38 a/ Salts percentage 1:5 soil to water extract. b/ C = cotton. F = fallow. Source: Ann. Rep. Gez. Res. Stat. 1952-3 The soluble salt content is also indicated by the conductivity 1/ of an extract of the soil. Table 4 gives such values for soils from Managil; sample 1 was from an area giving 'good' cotton yields, samples 2 and 3 from areas with a hard pan giving 'poor' yields, although the yield criteria were not stated. The notable feature is the increase in conductivities,and thus in salts, at 30 cm in the 'poor' samples (cf. the increase at 75 cm in the G.A.R.S. soil, Table 3); there was an increase in sodium content at the same depth (Table 5). 1/ High conductivities indicate high soluble salt content and vice versa.  ANNEX II - 8 - APPENDIX 1 Table 4: CONDUCTIVITY OF MANAGI,. SOIL Conductivity a/ Depth, cm 'Good' Yield 'Poor' Yield 'Poor' Yield 0-30 1.0 1.2 o.6 30-60 1.2 2.9 6.2 60-90 1.6 13.h 16.6 a/ Conductivity of 1:1 extract (mmhos/cm) Source: Ann. Rep. Gez. Res. Stat. 1962-3. Table 5: SODIUM CONTENT OF MANAGIL SOIL a/ Sodium Content b/ Depth, cm 'Good' Yield 'Poor' Yield 'Poor' Yield 0-30 1.0 1.0 1.6 30-60 1.1 3.0 13.2 60-90 1.6 5.9 15.2 a/ The same sample of soil as Table 4. b/ Sodium - m.e. Na/100 g in 1:1 extract. Source: Ann. Rep. Gez. Res. Stat. 1962-3 It has not been proved conclusively that level of total salts or of exchangeable sodium in normal Gezira soils depress yields, but it is very probably that the maximum feasible yields are not being obtained. For example, in an experiment at the G.A.R.S., cotton grown in imported soil from the Gash delta (very low in salts and exchangeable sodium, Table 2), gave higher yields than that grown in Gezira soil. It was found, however, that crops grown in the Gash soil followed the same pattern of annual fluctuations in yield as those grown in the native soil in neighboring plots. Total Salts. Some total analyses have been done on Gezira soils, one set of values showing 6.25 percent Ca, 0.59 percent Mg,  ANNEX II - 9 - APPENDIX 1 0.26 percent K and 0.05 percent P. 1 The presence of gypsum is usually indicated in soil descriptions but the quantity is not usually determined. Table 6 shows a set of values which have been published. The values for total salt are similar to those given in Table 4 for samples from the Gezira Research Station and show the same sharp increase at a depth of about 75 cm. Table 6 suggests that gypsum first begins to appear at this 75 cm layer and increases downwards, but the method of determination (sieving out on a 1 mm sieve) does not detect the very fine particles of gypsum that can be seen on examination of the upper part (0-75 cm) of the profile. Calcium carbonate vaties from about 1-2 percent in the upper 30 cm to 4-5 percent at 2 m. Table 6: SALTS AND GYPSUM IN A GEZIRA MAIN SOIL Depth, cm Salts a/ Gypsum b/ 1-15 0.07 15-30 0.08 30-45 0.09 g 45-60 0.11 60-75 0.16 75-90 0.43 0.3 90-105 0.49 0.4 120-135 0.43 0.2 135-150 0.54 0.9 150-165 0.65 1.5 165-180 0.65 1.0 180-205 0.61 1.1 a/ Salts expressed as percentage in 1:5 water extract. b/ Gypsum estimated by sieving on 1 mm mesh (see text). Source: Modified from J. Agric. Sci. 18, 518 in which depth given in 6-inch steps. Gezira soils are very low in organic matter, the general level being under one percent. 1/ Min. Agric. Sudan Bull. 12.  ANNEX II - 10 - APPENDIX 1 Nitrogen They are also very low in total nitrogen, the general level being 0.02 to 0.03 percent 1/, but levels have not changed demon- strably on plots which have been cultivated for 20 years at G.A.R.S. Nitrate nitrogen, on the other hand, is not particularly low and considerable quantities of nitrate are found in soils kept under artificial conditions favoring nitrification. In view of the consider- able need of cotton for nitrogen, much research has been done on soil nitrates, especially on the relation of nitrate level in the soil to the amount of pre-sowing rainfall, which is correlated with cotton yield (see Appendix 2). As would be expected, nitrate levels in the soil were higher in years when the pre-sowing rains were light. Heavy weed growth can reduce nitrate levels by 80 percent V and the thorough weeding of fallows by hand or with tractors was standard practice in the higher rainfall areas of southern Gezira Main until it was rendered unnecessary by the adoption of nitrogenous fertilizers. There are indications that nitrate levels have fallen in plots cultivated for 20 years without nitrogenous fertilizer. The role of nitrates in relation to 'decline in fertility' is discussed later. Nitrates show variation with depth, season and rotation in the soil profile. There is a fairly sudden increase in nitrate (as there is with other soluble salts) near the top of the grey layer. There is a rapid increase after the beginning of the rains, values being considerably larger after lubia than after dura. 3/ Some nitrogen is fixed by free living organisms, but the amount would only be equivalent to a few kg of the standard fertilizer, urea, per feddan. Much larger quantities are fixed by lubia and other leguminous p'ants. Mineralization of some organic residues in Gezira soils is slow. This applies particularly to dura roots and straw and to cotton roots, which were found to use up nitrogen from nitrogenous fertilizers for as long as four months after the beginning of incubation. Lubia and clitoria leaves, on the other hand, mineralized rapidly and released nitrogen. Nitrification of ammonium fertilizers proceeds rapidly in the laboratory but is slower in the field in the dry season; this is attributed to the very high temperatures which may exceed 600C in the top 2.5 cm of soil during April and May. Some nitrification 1/ J. Soil Sci. 8, 211. 2/ Ann. Rep. Gez. Res. Stat. 1962-3. 3/ The roots of members of the Genera Andropogonae (including dura) are said to produce an exudate which inhibits nitrification. A fallow year after dura may allow the micro-biological breakdown of this inhibitor and so improve nitrification but sufficient nitrogen fertilizer could be given to overcome nitrification inhibiting effects of the dura roots.  ANNEX II - 11 - APPENDIX 1 appears to take place between two and four weeks after application of the fertilizer to the soil. PHYSICAL PROPERTIES As far as irrigation is concerned, physical properties of the soils are even more important than chemical ones since they control much of the behavior of the soil under cropping; it is also more difficult to modify the physical than the chemical properties. Cracking A very important characteristic of the Gezira soils is their behavior under wetting and drying when they exhibit strong shrinking and swelling properties. The shrinking and swelling of these soils was measured in the field by sinking steel plates, to which rods had been attached, to different depths in the clay soil and then noting the movement of the rods when the soils were wetted. The upward movement on wetting is shown in Table 7. The expansion at the surface on wetting is very considerable and has important practical applications for it leads to 'gilgai' formation i.e. small mounds and hollows, which interfere with water control and may need leveling perhaps every four or five years. Table 7: UPWARD MOVEMENT OF GEZIRA M4AIN SOIL ON WETTING Depth, cm Upward movement, cm Surface 12.50 30 7.25 60 3.75 90 1.75 120 1.00 150 0.25 180 0.00 Source: J. Agric. Sci. 24, 42. Detailed observations have been made on the cracking patterns.!/ The width of cracks varies widely from year to year. In the 1960 season, when total rainfall was low, very wide (approx. 20 cm) cracks 1/ Ann. Rep. Gez. Res. Stat. 1962-3.  ANNEX II - 12 - APPENDIX 1 developed during the succeeding dry season. As the cracks began to fill in an unusually uneven surface was formed with an intensive pattern of 15 to 20 cm deep gullies. In the following seasons fine cracks developed across the puffs of the network of gilgai and grew wider,but some cracks also developed alongside the 1960-61 old cracking pattern. As a result the surface tended to be flattened gradually during the following season. The 1960 cracking pattern was discernable until the 1963 rainy season but such persistent patterns develop only during a particularly dry season; in normal seasons the pattern is unique and is not discernable nor repeated in the following year. Some observations have been made of the cracking of regularly irrigated soils. It was found that the same cracks opened and closed during the irrigation cycles. Moreover it was noticed that the cracks started closing gradually only after they were completely filled with water. Bulk Density The bulk density increases, and hence the pore space decreases, down the profile. Table 8 gives the values for 'badobe', a normal cracking clay, and for 'lugud', a heavier, stickier clay normally associated with poor cotton yields; both samples show a denser layer at 75-14 cm. Table 8: DENSITY AND PORE SPACE OF GEZIRA SOILS 'Badobe' 'Lugud' Depth, cm Bulk density Pore space Bulk density Pore space g/cc percent g/cc percent 0-30 1.21 54 1.45 45 30-45 1.28 52 1.45 45 45-75 1.36 49 1.52 43 75-105 1.47 44 1.61 39 1o5-14o 1.47 44 1.65 38 140-170 1.37 48 1.51 41 170-190 1.40 47 1.51 43 Source: Ann. Rep. Gez. Res. Stat. 1961/2.  ANNEX II - 13 - APPENDIX 1 The influence of these dense layers on root development has been investigated by several workers in the Sudan. By digging the soil to 1 m (3 ft.) depth and then replacing it layer by layer in its original order, root penetration to the depth of digging was very much more vigorous and it was noted that it was the compact soil which prevented deep rooting of cotton rather than the increase in salts which usually occurs at just above the compacted layer. Examination showed that roots of Upland cotton penetrated the grey compact layer more readily than those of Barbadense cotton but that the layer was a poor medium for either 1/; this simple observation is not sufficient to justify any change in cropping of 'lugud' soils. 1/ J. S. E. Agric. Coll. Wye No. 36 135.  ANNEX II APPENDIX 1 C H A P T E R 3 SOIL/WATER RELATIONSHIPS From the beginning of cropping in the Gezira it was realized that the behavior of the soils as a result of changes in moisture was the major soil factor in crop growth. In the earlier years of the Scheme much research was devoted to studies of moisture regimes of the soils under irrigation. Although these studies produced very interesting findings,much of the work stopped some years ago and none of the findings were applied on a field scale. SOIL MOISTURE Available Moisture The moisture extraction curve for a typical Gezira soil shows a very high moisture content at field capacity (39 percent) and also a very high value at wilting point (22 percent). 1/ Although this indicates 17 percent available water, i.e. by weight about 230 mm per 10 cm depth (2.7 inches per foot), plants cannot grow, but only survive, at moisture levels near the wilting point. Table 9 shows that the available water range for cotton for good growth is 30 to 40 percent, i.e. about 150 mm of water per 100 mm of soil.2/ Furthermore, it has been shown that 24 hours after irrigation only the top 15 cm of soil contains any appreciable amount of available water (Fig. 1). IWater movement into the lower layers is slow,and in fact the depth of penetration of irrigation water more or less corresponds to the depth of rain water penetration under natural conditions. 1/ M.Sc. Thesis, Osman Ahmed Ali, University of Khartoum. 2/ 1.8 ins. per foot. These calculations are based on a bulk density (dry soil) of 1.3 g/cc. Min. Agric. Sudan Bull. 12.  ANNEX II - 15 - APPENDIX 1 Table 9: EFFECT OF SOIL MOISTURE ON COTTON Soil moisture content as percent by weight Maximum for healthy growth 40 Optimum for growth 36-40 Cotton begins to need water at 30 Cotton suffers a severe check, crop ruined 23-24 Wilting point for cotton seedlings in pots 18-20 Source: Agriculture in the Sudan. (Greene) ed. Tothill p. 445 Permeability The most important characteristic of Gezira soils under irrigation is their low permeability. Direct measurement of permeability in the field is very difficult because of soil cracking. However, relative permeability may be indicated by the capillary rise; soils having the lowest capillary rise have the lowest permeability. Capillary rise varies down the profile (see Table 10). Table 10: CAPILLARY RISE IN GEZIRA MAIN AND MANAGIL SOILS Depth, cm Gezira a/ Managil b/ 0-15 86 15-30 38 30-45 3323 45-60 17 60-75 25 75-90 108 3h 90-105 93 35 105-120 4h 120-135 67 8 135-150 87 2 150-165 72 165-190 97 26 Source: a/ same sample as-Tab1le 3, Ann. Rep. Gez. Res. Stat. 1952-3. b/ Ann. Rep. Gez. Res. Stat. 1957-8.  ANNEX II - 16 - APPENDIX 1 The permeability is very much affected by the amount of soluble salts and by the levels of exchangeable sodium. The influence of exchangeable sodium on permeability is shown by the following values for capillary rise in samples of a Gezira clay:- Capillary rise cm Original soil 33 Saturated with: sodium 6 magnesium 90 calcium 152 (Min. Agric. Suda.n Bu-ll. 12.) Permeability of Loamy Soils. Much research has been done in different parts of the world on the influence of sodium on the permeability of loamy and sandy soils. The actual permeability depends on the electrolyte concentration in the irrigation water but, with water of low electrolyte concentration, the permeability is little affected by e.s.p. values below 4. Permeability decreases by about one-third when the e.s.p. value reaches 10 and by one-half when it reaches 20; between 20 and 30 permeability falls to very low values.l/ Permeability of Heavy Clay Soils. Much less work has been done on the influence of sodium on the permeability of montmorillonitic clay soils of the type found in Gezira. It has been reported that, in such soils, very strong intra-crystalline swelling of the clay lattice takes place on wetting when the e.s.p. exceeds 30.2/ At levels of about e.s.p. 30-35 clay soils tend to behave as gels when wet and to be extremely hard when dry. About e.s.p. 30 would be regarded as the upper limit for workability of these soils but the influence of smaller quantities of sodium has not been reported. From practical obser- vations it would appear that even quite small quantities of exchangeable sodium decrease the permeability to such an extent that, for all practical purposes, water entry depends on cracking. Infiltration rate has been measured. The figure given for Rahad is 10 cm/day. 3/ Lower figures for Gezira are given by Coleman and 1/ Third Cong. Irrig. Drain. Int. Comm. Irrig. Drain. 1957 3,8,115. 2/ Quirk Ph.D. Thesis London Univ. 3/ Hunting-MacDonald Rept., Rahad Scheme.  APPENDIX I FIGURE I SUDAN: GEZIRA MOISTURE-CONTENT OF A COTTON PLOT Soil moisture content per cent. 20 30 40 50 .C3 aWW 432 5 - . Soil Moisture c-o-n-en -i Before irrigation l -_- 24 howrs after 6 ______ A}_____ ,______ an irrigotion SOURCE Agriculture in the Sudan. Ed., J. D. Tothill, Oxford University Press. London, 1952. Figure 180, p.447. IBRD -3196  ANNEX II - 17 - APPENDIX 1 Spoor I/, who obtained a steady infiltration rate of 2 mm per hour, i.e. about 4.8 cm per day, with furrows 100 m long. The difference between the Rahad and Gezira values are more likely due to the methods used than to any marked difference in the soil; the smaller value is probably more realistic. IRRIGATION Such low permeability values would make the Gezira soils virtually unmanageable under irrigated agriculture if it were not for the fact that the soil cracks; in practice, nearly all of the water enters the soil via the cracks. This was, of course, recognized in setting out the system of irrigation. Although the land is ridged, the irrigation is not by the conventional furrow system in which the furrows are used to transport the water without any ponding effect. The Gezira system is, in fact, a ponding (or basin) system in which the water is allowed to stand in the furrows in order to soak into the soil. Experience on a large scale in the Gezira confirmed that ponding is necessary for the effective irrigation of heavy clays. 2/ Cracking and water entry Since most of the irrigation water enters the soil before the cracks close it is of interest to calculate the volume of water which the cracks could hold. Some measurements on these soils show a shrinkage from 77 to 57 cc per 100 g dry soil when the moisture content falls from 40 to 20 per cent, i.e. from optimum growth to wilting. 3/ Assuming that shrinkage takes place equally in the vertical and horizontal planes, the volume of cracks in the top 61 cm (2 ft.) would be about 490 m3 per feddan, i.e. about 115 mm of rain or rather more than one standard irrigation. However, the volume of water taken in will depend on the depth to which the soil has dried out; if the cracks extend only to 30.5 cm (1 ft.) depth,then they will hold only about 240 m3 per feddan, or about half of a standard irrigation. These calculations are not very accurate, as there are a number of parameters influencing the shrinkage for which values are not available. However they do show that the soil has to be dried out very thoroughly and to considerable depth before large quantities of water can be held in the cracks. 1/ Quoted in M.Sc. Thesis, Osman Ahmed Ali, Univ. Khartoum. 2/ Memo. H. S. Morrice, Irrig. Dept. Sudan 1958. In a furrow irrigation system, the water does not pond; the water supply is cut off as soon as the flow reaches the farther end of the furrow. If this were done on the Gezira clay, insufficient water would enter the soil. 3/ Road Res. Tech. Paper No. 58. Road Res. Lab.  ANNEX II - 18 - APPENDIX 1 Shrinkage is not confined to the agriculturally-useful moisture range but continues as drying proceeds beyond wilting point; in fact there is as much shrinkage between pF4 and 6 as there is between pF2 and 4. 1/ These calculations suggest that large quantities of water can enter the soil at the end of the dry season; however it is imposBible to dry them out to this extent during cropping so that not even all of the standard irrigation will normally enter the soil. This agrees writh general observations, i.e. initial rain storm water enters easily but subsequent rainstorms and irrigations leave considerable quantities of water on the surface for several days. Watering Intervals In recent years a number of experiments have been carried out on the frequency and quantity of irrigation watering. The results of one such experiment are given in Table 11-A. (Table 11-B shows the corresponding irrigation rates in practical terms.) Table 11: FREQUENCY AND QUANTITY OF IRRIGATION OF COTTON A. Cotton Yields in k.p.f. Watering Intervals Water Duty 7 days 14 days 21 days 28 days mm/day 2.5 6.87 6.63 6.66 5.92 5.0 6.07 6.15 5.74 6.47 7.5 5.90 5.92 6.41 6.58 10.0 5.94 6.02 6.37 6.08 1/ Greene-Kelly, Rothamsted, priv. comm: pF is a measure of the intensity with which water is held by the soil and is therefore a measure of its availability to the plant.  ANNEX II - 19 - APPENDIX 1 B. Practical Applications Corresponding to A. Experimental Rates Wiatering Intervals Water Duty, mm/day, 7 days 14 days 21 days 28 days Approximate Application on Plot ,____________ M3 per Feddan 2.5 I 75 150 220 300 5.0 ' 150 300 440 590 7.5 ' 220 4h0 a/ 660 880 10.0 ' 300 590 - 880 1180 a/ This application is nearest the normal routine practice of 400 m3/ feddan at a 14-day watering cycle. Source: Ann. Rept. Gez. Res. Stat. 1962-3 In this 1962/3 experiment the heaviest water duty at the shortest interval and the lightest water duty at the longest interval both depressed yield. This result differed from that in 1961/2 (when the crop benefited from heavier irrigations and shorter intervals), but was in line with much of the earlier work which revealed a marked degree of tolerance by Egyptian cotton to a wide range of irrigation regimes. Effects of Prolonged Flooding Experiments at Abu Ushar (on plots which normally gave good yields) and Ganib (where yields were poor) showed that water penetrated much more deeply in the good soil during 14 days of flooding (Table 12).  ANNEX II - 20 - APPENDIX 1 Table 12: SOIL MOISTURE CONTENT BEFORE AND AFTER FLOODING Before Flooding After 14 days Flooding Depth, cm Good Soil Bad Soil Good Soil Bad Soil 0-15 12.7 4.9 33.6 29.0 15-30 13.3 8.5 37.7 33.1 30-45 12.5 8.1 39.5 18.6 a/ 45-60 14.9 12.2 39.4 13.7 60-75 16.0 13.7 39.0 14.2 75-90 16.9 13.5 37.9 13.9 90-105 16.7 13.3 34.8 13.8 105-120 16.4 13.1 33.5 13.6 120-135 15.4 13.1 32.5 13.5 135-150 15.3 12.7 30.1 13.7 150-165 14.7 13.4 26.8 a/ 13.0 165-180 14.6 13.3 18.0 12.2 a/ limits of appreciable penetration. Source: J. Agric. Sci. 18, 531. In a subsequent replicated experiment at Ganib gypsum was applied at the rates of approximately 0, 1, 4 and 10 tons per feddan and the plots watered for 14 days. Fig. 2 shows the moisture contents down the profile at different dates; the October measurements were made after drying out, the December measurements after a further 10 days flooding. An interesting aspect of this experiment is that a relatively small dressing of gypsum gave a marked increase in water penetration. An example of how large quantities of water can lead to deeper wetting is seen in plot 4 at the Gezira Research Station, which has grown irrigated cotton every year since 1918, and the adjacent plot which has had cotton only one year in four during the same period. In the continuous cotton plot the soil is moist down to at least 150 cm whilst the rotation plot is very dry and hard at 90 cm. The maximum depth of water penetration is easily detected by the resistance to a core sampler pushed into the soil. It is easy to auger to 150 cm in the continuous plot but very difficult to penetrate to this depth in the rotation plot.  APPENDIX I FIGURE 2 SUDAN: GEZIRA EFFECT OF GYPSUM ON THE PERMEABILITY APRIL 1927 OCTOBER 1927 DCEMBER 1927 X S r>t f Z ) Z 1 81/2 X 3 C, X10 20 30 10 20 30 1 20 3 Soil moisture content per cent. SOURCE: Aigriculture in the Sudan.< Ed., J. D. Tothill, Oxford University Press, London, 1952. Figure 182, p.451 IBRD -3191  - 21 - ANNEX II APPENDIX 1 DRAINAGE Surface Drainage Excess water from over-irrigation or from storms which does not enter the soil has to be removed by surface flow or by evaporation. Under natural conditions Gezira soils must be waterlogged at certain times for the calcium carbonate concretions are coated with manganese, which is only mobilized at the prevailing high pH, in reducing conditions. Many waterlogged fields with yellowed and stunted cotton plants were seen towards the end of the rainy season. No measurements could be found of actual losses in yield due to waterlogging but a number of events suggest that yields are adversely affected. For example, in 1962/3, when serious flooding was reported from Managil 1/, several blocks yielded little more than half of the Managil average of 3.4 k.p.f. In the previous season, these same blocks had been close to the average of 5.4 k.p.f. On similar heavy cracking clays in Australia the following yields of seed-cotton were obtained: 2/ Not liable to flooding 7.1 k.p.f. (2232 lb/acre) Liable to flooding 4.5 k.p.f. (1420 lb/acre) It is very probable that losses of the same order could be demonstrabed in the Gezira. Nevertheless, the need for providing surface drains, and for proper maintenance of those that do exist, is often ignored. The diligent cultivator can minimize, to some extent, the effect of poor drainage by strict attention to watering, by using properly made ridges and by breaking down the robats after heavy rain, but where field work is done by hired labor without adequate supervision, the full ill-effects of poor surface drainage are felt. Internal Drainage From the beginning of the Gezira Scheme, the lack of drainage for removal of salts was regarded by some as a basic weakness, likely to lead to eventual destruction of the soil. Consequently early 1/ Memo SGB/EGR/2-3-1 XR/123-5 2/ Emp. Cotton Grow. Rev. 39, 81.  ANNEX II - 22 - APPENDIX 1 research included practical trials of drains and measurement of rates of water movement through the soil. It was found that there was no water movement to drains over distances greater than about 1 m, although addition of gypsum allowed movement over 2 m and also a greater daily flow into drains. 1/ If these soils are to be provided with internal drains then tile or similar drains would be necessary because of the close spacing. Spacing depends on permeability and on depth to an impermeable horizon. The grey horizon, though not completely impermeable, has to be treated as such in the calculations. The Donnan formula is used to calculate drain spacing:- S = [4P(b2-a2)I Qd where S = spacing between drains (in feet) P = coefficient of permeability in inches per hour b = distance from draw-down curve to barrier stratum at mid-point between drains (in feet) a = distance from average tile depth to barrier stratum (in feet) Qd = quantity of water to be drained (ins/hr.) Using the Rahad figure of 10 cm/day (0.166 in/hr) for permeability (although it is probably too high, see above), b = 1 ft., a = 3 in and Qd = 0.0033 (about 30 in/annum), then S (the required drain spacing) is equal to about 3.5 ft or just over 1 m. Such close spacing is clearly impracticable. To make drainage an economic proposition,both permeability and depth of placement would have to be greatly increased; even if large applications of gypsum (which might cost f 50 to fI00 per feddan) could increase the spacing to about 10 m (30-h0 ft), the cost of making the drains would be f50 to f60 per feddan using present- day methods. 1/ Tothill, Agriculture in the Sudan. p. 453.  ANNEX II APPENDIX 1 C H A P T E R 4 SALINE SOILS IN GEZIRA Salts affect plant growth directly by increasing the osmotic pressure of the soil solution and by causing accumulations of toxic ions in the plant. Plant species vary greatly in their tolerance of total salts, most of those potentially useful in Gezira being tolerant or moderately tolerant; exceptions are beans, stone fruits and some clovers. Crops also vary in their tolerance of sodium but, again, the useful ones show at least moderate tolerance, only citrus being affected by exchangeable sodium percentages less than 20 percent. 1/ On alkaline soils, it is the indirect affect of salts - poor soil structure, with extremely low water acceptance - which causes poor growth; even plants tolerant of high salt levels may yield poorly on such soils. In pot experiments where the influence on soil structure was eliminated, cotton was not affected until the e.s.p. exceeded 30. 2/ It is widely recognized that the effect of exchangeable sodium varies greatly from soil to soil. The intensity of shrinking and swelling depends on the type and amount of clay, the surface area of the clay and exchangeable cations on the clay complex. When a calcium- saturateg montmorillonite is wetted,the lattice of the clay expands from 9.6A to 20R i.e. the volume doubles, but no further swelling takes place. Thus if a soil has 60 percent clay, all calcium mont- morillonite, it would swell on wetting by a little over half its dry volume with a corresponding shrinkage on drying. When a sodium- saturated montmorillonitic clay is wetted, the lattice also expands to 19.5k but expansion does not stop there for the clay can go on swelling, up to 20-fold, due to structural disarrangement. 3/ Thus even soils with similar texture and similar exchangeable sodium levels may show wide differences in permeability and the e.s.p. value can only be regarded as a guide to behavior under irrigation. 1/ Adv. Agron. 16, 162 2/ Proc. Amer. Soc. Soil Sci. 19, 29 3/ Quirk, J.P. Thesis, Univ. London  ANNEX II - 24 - APPENDIX 1 Dispersion under irrigation is an important part of the behavior of alkaline soils for when they are dispersed the surface is sealed, the soil runs together and the pores are blocked. Sodium clays are very easily dispersed and need a high concentration of salts to flocculate them. Calcium clays are difficult to disperse and need only a low concentration of salts to flocculate them. A solution of 1 x 10-3 M NaCl, will flocculate calcium montmorillonite but it requires 3 x 10-2 M NaCl, i.e. 30 times the concentration, to flocculate sodium montmorillonite. 1/ Natural Saline and Alkaline Soils Saline and alkaline soils occur naturally, generally forming in arid climates on parent materials rich in salts or, as in the Sudan, as the product of weathering of felspars in basic rocks. With soil surveys, any substantial areas of such naturally saline or alkaline soils can be excluded from irrigation schemes during the planning stage. The soils of, Gezira Main were examined before development and saline patches were excluded. These patches were found mainly in the north and west (see map, Fig. 3) and recent surveys by consultants have confirmed that, on the basis of salinity, soils are poorer in the more northerly, drier, parts of the Gezira. No detailed soil surveys were available when the Managil extension was planned, and naturally saline areas were developed for irrigation along with the rest of the land. The Sodium Value Test. Early research at Dueim Pump Scheme did demonstrate a relationship between salt ccntent and crop yields. Sampling of good and poor areas showed that poor yields were associated with a high total salt content in the top 120 cm of soil; further, the concentration increased more rapidly down the profile in poor areas. This work also showed a relationship between sodium levels in the soil and yields which led to the development of a simple test:- "The sodium test, in its present form, measures the exchangeable plus soluble sodium per unit of clay in a composite of the first three feet of the soil profile. The result is expressed as a whole number, in milligram equivalents of sodium per 100 g of oven dry clay, and is called the sodium value." 1/ J. Phys. Chem. 39, 593.  APPENDIX I FIGURE 3 SUDAN: GEZIRA PERCENTAGE OF EASILY SOLUBLE SALTS IN THE TOP FOUR FEET OF SOIL 32'20 32" 40 5300 3 320 33 40 op .showino the petekto e of _________ _________ _________ seijts mn the first four feet Mbar CoumwC Less thdn .25% \~~~~~~1~ R 5Z - _:5o >0/ N 35 % - 50 '/ AfMore Mhen .561 4. Ulido u.a4 *D -; Ifi JIei va WA edrni 0SO A t t eSu. J '.'DTthl,''- Oxfor s P 1952.o Figure *8 8. :BD-3:2.=--Q AM a34eod 4 dd SORE Agicltr in th Suan Ed..DTtil xodUnvriyPes odn 1952. Figure 181,p.448. IBRD-32111  APPENDIX I FIGURE 3 SUDAN: GEZIRA PERCENTAGE OF EASILY SOLUBLE SALTS IN THE TOP FOUR FEET OF SOIL 3220 32 40 3300 3S20 39 40 *K4C~ ~ ~ ~ ~ ~ ~~ ~M svt inthe first four feet, K"I. w tZ Less thdn *25% A~~~~~~I s 35 -, Sao \ _ w~~~~ore Man l o It *~~ IUJ t i 1502 I C4 .== '200d .? 20 _ _ _ _ _ _ _ _d < J.~~ 4r Ulid. ' *\ : SOURCE: Agriculture in the Sudan, Ed J. D Tothill, Oxford University Press, London, 1952. Figure 181,p.448. IBRD-3211  ANNEX II - 25 - APPENDIX 1 When the yields from a large number of Gezira blocks were cal- culated from the regression equation on sodium values derived from the Dueim yields, the best agreement was found for blocks in the extreme north and the extrene south but there was no significant over-all correlation. Within small areas, correlations between yield and sodium values were significant though the correlation coefficients were quite low. 1/ In more recent investigations in the Managil extension, the yields of numbers within blocks have been compared with the sodium values of their soils but no correlations were found. 2/ Secondary Salinization Many catastrophes in irrigated agriculture have arisen from secondary salinization of the soil due to the rise of saline ground water or to the use of irrigation water with a high sodium content. The rise of ground water levels is the more common cause. Irrigation has been practised in Egypt and China for thousands of years without serious soil deterioration although, even in these countries, trouble has arisen when perennial crops, with frequent irrigations, have been grown. In other places, however, rising water tables have caused very serious deterioration and necessitated expensive drainage or under- ground water pumping schemes. In the Gezira the ground water table is at 10-20 m and, since there is an almost impermeable layer between it and the surface, there is little likelihood of it being supplemented by irrigation water and so rising dangerously close to the surface. Even a system of agri- culture using very large quantities of irrigation, e.g. rice growing, would be unlikely to raise the water table to a dangerous level. Therefore, the only likely source of secondary salinization is the irrigation water. Quality of Irrigation Water. Fortunately the Blue Nile water is of an exceptionally high quality with low sodium and a favorable calcium-sodium ratio. The low bicarbonate content is particularly important since calcium (amd magnesium) are precipitated as the insoluble carbonate, thus lowering the calcium-sodium ratio; even at Cairo, the residual sodium carbonate in the Nile waters is low (0.23 m.e/l). Analyses of water from the Blue Nile and from the Barakat III canal are given in Table 13. Over the year, the increase in 1/ Min. Agric. Sudan Bull. 12. 2/ Ann. Rep. Gez. Res. Stat. 1964/5.  ANNEX II - 26 - APPENDIX 1 concentration of salts due to evaporation of water in the canal system is fairly small but the calcium-sodium ratio (due to precipitation of carbonates in the canals) is lowered from 2.88 to 2.15. However, fairly major differences in both total salts and calcium-sodium ratio occur in months when the flow through the irrigation network is small and evaporation high; this may give rise to a flush of relatively sodium-rich water when the system is brought into full operation in August-September.  Table 13: COMPOSITION OF BLUE NILE AND 3ARAKAT III CANAL WATER 1936/37 a/ Blue Nile Water Bbrakat III Canal Month 1 2 3 4 5 6 7 8 9 10 11 12 Ca Na Ca/Na Ng S01 Alk Ca Ma Ca/Na Mg S0L Alk Mar '36 31.0 11.6 2.67 7.2 16.4 2.29 30.0 16.2 1.85 7.6 17.6 2.33 Jvne '36 24.3 9.8 2.49 6.o 13.3 1.73 30.8 16.0 1.92 7.0 18.3 2.35 July '36 21.5 5.5 3.91 4.8 8.8 1.38 28.3 13.0 2.17 5.5 13.0 2.01 AUg. 136 25.4 6.4 3.97 5.0 10.0 14I5 2i.4 10.6 2.30 5.0 12.2 1.1 Sept. 136 22.0 6.0 3.58 5.0 6.8 1.341 18.3 23.5 0.78 3.8 8.8 1.79 Oct. '36 19.6 8.4 2.33 4.6 5.6 1.411 22.4 10.0 2.24 4.8 6.0 1.57 Nov. 136 20.5 8.5 2.4.1 4.8 5.5 1.19 22.3 9.8 2.28 5.3 5.8 1.60 Dec. '36 21.0 8.0 2.63 5.0 6.3 1.51 22.3 8.5 2.62 5.0 6.3 1.59 Jan. '37 22.8 9.0 2.53 5.0 6.6 1.75 23.4 9.0 2.60 4.8 8.8 1.69 Feb. '37 24.5 8.8 2.80 4.8 6.8 1.79 24.5 9.8 2.51 4.8 7.3 1.85 Mar. '37 25.0 10.6 2.36 6.0 7.6 1.95 25.2 11.0 2.36 6.0 7.6 1.95 overal3/ 23.4 8.4 2.88 5.3 8.5 1.64 21-.7 3.2.5 2.15 5.4 10.2 L86 a/ Ca. Mg. Na. SOl in p.p.m.; Alkalinity = 0.1 N HC1 per 100 rnl of water Source: Min. Agric. Sudan Bull. 12.  ANNEX II -28 - APPENDIX 1 Past Changes in Salinity in the Gezira. Tlhile it may seem possible to determine the effect on the soil of irrigation since the Scheme started, there are practical problems. The first is that of getting the appropriate soil samples. Although there are areas within the Gezira which have not been irrigated since the start of the Scheme they have been left out for a reason, e.g. too high, too salty or too uneven, and samples from them would not be typical of the original, natural Gezira soils generally. The location of sampling points in the original survey is known and it would be possible to repeat sampling at the same points but the original analyses was not done by methods which are used today and strict comparison is impossible. In any event, it will be difficult to detect the small changes likely to be involved. Most of the land has been under irrigation for a maximum of 40 years and during this period only 10 or 11 crops of cotton have been grown. In addition, the equivalent of perhaps 5 crops of dura and 2.5 crops of lubia have been grown. Thus the total irrigation water per feddan has been less than 80,000 m3. 1/ Each 1,000 m3 of irrigation water contains, on the average, about 8.3 kg of sodium so that only some 650 kg sodium per feddan has been added over the 40 years. Evenly distributed over 2 ft (61 cm) of soil, this would raise the exchangeable sodium by only about one milli- equivalent. Since, in fact, it is most unlikely that all of this sodium has entered the exchange complex 2/, the increases to be measured are probably so small that the results of analyses would be meaningless. Conductivity measurements, giving an indication of salt content, have been made at the G.A.R.S. to compare soils (from a rotation trial) whose histories were known. Some sets of samples had received 31 seasons of irrigation, others only eight seasons. No major differences were observed, although there was some indication of leaching and redistribution of salts within the profile; in general there were marked accumulations at about 90 cm in soils which had received few irrigations but none in intensely irrigated soils. 1/ cotton 10 x 6000 m3 = 60,000 m3 dura 5 x 1600 m3 = 8,0oo m3 lubia 2.5 x 4000 m3 - 10,000 m3 Total 78,000 m3 2/ Sodium must be twice the amount of other cations in the soil solution before the sodium will enter the exchange complex readily. A complicating factor, however, is that the calcium and magnesium are precipitated as carbonate. Thus, since there is no leaching by internal drainage, the relative amount of sodium in the soil solution gradually increases and may eventually exceed the value of 2.  ANNEX II - 29 - APPENDIX 1 Sodium values were slightly higher in plots that had received 31 seasons irrigation, than in those which had had only eight; however, there was no such increase in soil from arnother area which had received 44 seasons irrigation. It was noted that calcium carbonate had increased in soils which had received many irrigations; the increase occurred in the coarser fractions, indicating that the carbonate had been precipitated around nodules present in the original soil. 1/ Exchangeable sodium was measured in another study. 2/ The exchangeable sodium in the top 15 cm was much lower where plots had received irrigation for 31 seasons, than in those irrigated for eight only. Rate of Secondary Salinzation in the Future. Nevertheless, some sodium is being added each year; if there is more intensively irrigated agriculture in the future average annual additions will increase. Most Gezira soils are very productive at present but if none of the sodium added in the irrigation water is removed, then there is bound to be a build-up of exchangeable sodium which will lead eventually to soil deterioration. Increasing sodium levels will be more damaging on soils already high in sodium, for example, northern Gezira Main and in Managil. The bad effect of additional salts on such soils is agg- ravated by two factors, the shallow depth of wetting of the soil (and thus small depth of soil in which the salts are concentrated), and the effect of salts (by raising the osmotic pressure of the soil solution) on the already often marginal moisture supply. Whilst the evidence is slender, it seems probable that the Gezira soils will be manageable under the present system of irrigation until e.s.p. values exceed 20. On this basis, the 'life' of the soils would be about 80 irrigation seasons. Theoretical 'considerations suggest that the soils may be manageable at somewhat higher e.s.p., perhaps even up to e.s.p. 30. In this case the 'life' would be 160 irrigation seasons. These 'lives' refer to soils which are not naturally saline, having an e.s.p. of 10 in the virgin state. Obviously, soils which started with a higher sodium content will have a correspondingly shorter useful 'life'. Further, the calculation refers to the time when the soils become unmanageable; though no quantitative data is 1/ Memo S. Sc. S/17. D.2. 2/ Osman Ahmed Ali. M.Sc. Thesis, Univ. Khartoum.  ANNEX II - 30 - APPENDIX 1 available, there can be little doubt that yields will be adversely affected before this critical stage is reached. Under the agricultural system proposed by the mission, 'lives' of 80-160 irrigation seasons correspond to a time span of about 110-220 years. There is reason to hope that the problem of alleviating high sodium levels will be solved during the next century; practical soil management methods which can be adopted now to alleviate the problems on naturally saline areas are discussed in a later chapter.  ANNEX II APPENDIX 1 C H A P T E R 5 FERTILITY OF GEZIRA SOILS Annual Fluctuations in Yield Wide yield fluctuations are one of the most spectacular and puzzling aspects of cotton growing in the Gezira. Such fluctuations are encountered elsewhere but can usually be attributed to rainfall variations and are not common in irrigation schemes. Annual fluctuations are discussed in Appendix 2, and one factor, the correlation between pre-sowing rainfall and cotton yields, has already been mentioned in relation to the nitrogen status of the Gezira soils. However, the importance of soil nitrification, or leaching of nitrates, as a cause of fluctuations is much diminished by the routine use of nitrogenous fertilizers. The very high yields that are obtained in favorable years confirm that the Gezira soils can be a very good medium for growing cotton. Moreover, the fact that cotton grown in Gash delta soil imported to the G.A.R.S., exhibits the same annual fluctuations in yield as crops grown on the native soil, strongly suggests that the Gezira soils play little if any part in this vitally important problem. Variations in Yield From the start of soil investigations in the Gezira attempts were made to set up standards which could be used to classify the soil with respect to its value for cotton growing. Mention has already been made of the sodium value determination, which is useful for separating the best from the worst soils but does not place inter- mediate soils in order of yield. More recently a high correlation has been found between clay content and yield, the soils with higher clay content giving better yield. On small areas, i.e. 90 feddan numbers, this correlation between clay and yield only holds good in certain areas (notably in the north) indicating that, like the sodium value, it is only one of a number of factors influencing yield. In a recent investigation of soil factors influencing yield in Managil, it was found that, within Blocks, there was a good correlation between the yield of the same number, i.e. high yielding numbers always gave more than the Block average whilst low yielding ones always  ANNEX II - 32 - APPENDIX 1 gave less. Both the Block averages and the yields of each number varied greatly from season to season. Clearly soils play a part in the variations of yield from place to place. However, the level of crop husbandry has such a strong influence that correlations between any particular plot and yield have to be treated with caution. Irrigation and drainage problems may also distort the overall picture, especially since the highest salt contents are often found in low-lying areas where flooding may depress yields. MAINTENANCE OF FERTILITY From time to time statements are made that the fertility of the Gezira soils is declining, by which is implied that their productivity for cotton is decreasing. The suggestion that nitrogen response is increasing in the Gezira observation plots has also been used as a supporting argument. Because of the violent annual fluctuations in cotton yields it is difficult to adduce anything from yield records, but there is no evidence of any decline in yields due to soil deter- ioration. There are a number of ways in which soil deterioration could have taken place. These include increasing salinity, deterioration of soil structure, accumulation of toxic salts (e.g. boron), loss of organic matter and lower supplies of plant nutrients. As has been shown in the previous chapter, there is no evidence of salt accumulation, nor any proof of structural deterioration. Few investigations on boron have been done but toxic levels (300 p.p.m.) have been reported in citrus leaves, after 20 years' irrigation. 1/ In cotton, the levels in mid-stem leaves ranged from 56 to 68 p.p.m. at the Gezira Research Station and from 94 to 262 p.p.m. at Managil. 2/ Organic matter contributes to the well-being of soils by its influence on structure, as a medium for micro-organisms and as a supply of nitrogen by mineralization, but Gezira soils, in common with other cracking clays in sem-arid areas, have a very low organic matter content (about 0.7 percent) even in their virgin state. It is doubtful whether change in such a small amount of organic matter could have any marked influence on the structure of Gezira soils, although the influence of organic matter may be more than its quantity suggests, for it is very well dispersed through the clay colloids. Nitrogen comes from mineralization of organic matter. This occurs in flushes, a large flush taking place after the first rains. Nitrate nitrogen levels as high as 50 p.p.m., equivalent to perhaps 100 kg nitrogen in the top 30 cm of soil per 1/ Ann. Rept. Gez. Res. Stat. 1960/1. 2/ Ann. Rept. Gez. Res. Stat. 1961/2.  ANNEX II -33 - APPENDIX 1 feddan, have been reported but the average values are of the order of 7 to 12 p.p.m. 1/ In the absence of fertilizers this is the only source of nitrogen for crops and, unless nitrogen is being fixed by legumes or free living organisms or supplied by adding organic matter, cropping would reduce the supply. But, even if soil supplies of nitrogen are decreasing, the recommended dressing of 2N/feddan fertilizer, properly applied, should be ample for the crop. Because of their basic volcanic origin the soils have good reserves of magnesium and calcium and fair reserves of potassium. These reserves are depleted very slowly for the cotton crop removes, in the lint and seed, only about 4 kg of potassium per feddan. Of course, the potassium supply would be depleted much more rapidly if forage crops or legumes containing large amounts of potassium were regularly grown and removed from the field. Gezira clays have better reserves of phosphate than cracking clays in other parts of the world but cotton seed and lint remove small quantities of phosphorus. Cotton has shown some responses to phosphate in fertilizer trials where very heavy dressings of nitrogen were applied. 2/ As with potassium, a change in the agricultural pattern might show up a need for phosphorous fertilizers for the reserves in the soil are not large. The only reported trace elemert deficiency is that of iron in groundnuts, but no increase in yield occurs when it is cured. Soils are of course only one of the factors determining the productivity of an area. Environmental, biological and crop husbandry factors are often more important. There have been very great changes in crop husbandry over the past 15 years, the most important being the introduction of nitrogen fertilizers and insecticidal sprays in the late 1940's. In spite of these there has been no marked increase in yield and the question must be asked whether the new techniques are being applied effectively or whether other crop husbandry factors, e.g. cul- tivation and weeding, have deteriorated to such an extent that they have nullified the potential benefits from fertilizers and insecticides. The overall picture is that while productivity in the Gezira may be static in spite of fertilizer and insecticide use, there is no evidence of a decline in soil fertility. FAILOWS The Gezira is probably unique among large scale irrigated schemes in that the land is uncropped for long periods of time. 1/ Emp. J. Expt. Agric. 11, 1. 2/ Ann. Rept. Gez. Res. Stat. 1957/8.  ANNEX II - 34 - APPENDIX 1 In Egypt the land rarely remains uncropped for more than a few months in any year, and in some parts of the world, where temperatures are suitable and water available, irrigated areas may be cropped virtually continuously. In other areas an intensive crop, e.g. rice, is inter- spersed with a few years of an extensive crop, such as grass. In the Sudan however the practice is to have one period of "bor" i.e. agri- cultural land that remains uncropped and uncultivated for at least one whole year before each cotton crop. In addition to the fallow years, there is invariably a period of at least three months after each crop when the land is not in use. The implications of these periods are discussed in Appendix 2. More intensive use of land in an existing scheme is better economically than bringing new areas under irrigation and increasing attention has been given to intensification and diversification of cropping in the Gezira. Such intensification could mean the cropping of land during the fallow year and/or during the uncropped period. Even today however there is apprehension, in the Board and elsewhere, that the point has been reached where intensification may become excessive and lead to a deterioration of cotton yields. Before the start of the gravity irrigation scheme many rotations were tried at Tayiba. Yields were often low because of weeds and deterioration of soil structure from too much water. Good soil structure was preserved and good.cotton.crops obtained only after a fallow year had.been interposed before the cotton crop. 1/ The value of the fallow year in increasing yields of cotton has been demonstrated in many experiments. In most cases, these have received no fertilizer and there has been a very high incidence of disease in the continuous cotton plots, making it difficult to identify the factors that lead to a depression of yield in the absence of fallows. In one experiment at G.A.R.S. the long term means were:- Continuous cotton 1.50 k.p.f. Fallow - Cotton 2.62 k.p.f. Fallow - Fallow - Cotton 3.48 k.p.f. Fallow - Fallow - Fallow - Cotton 4.08 k.p.f. There was a very high response when nitrogenous fertilizer was given to some of the continuous cotton plots, but the yield dropped 1/ S.G.B. M1/150  ANNEX II - 35 - APPENDIX 1 immediately to the previous low level when fertilizing was dis- continued. Functions of Fallows With moderm agricultural techniques it is possible to replace many of the functions of fallows. Nitrogenous fertilizer can replace nitrification during the fallow,and herbicides and pesticides can control most weeds and pests. It is unfortunate that, until recently, there have been no long-term intensive rotations, which could give information on whether fallow periods were needed in the Gezira under modern agricultural systems of field management and, if so, how long these periods should be. In the absence of such information one can only examine the evidence for the effects of fallows on soils and discuss what might happen should fallows be abolished. Whilst the chemical and some biological effects of fallows can be replaced, there is no economic way of replacing their effect on soil structure. When dry Gezira soil is first rewetted, it breaks down into an attractive loamy texture, which gives place to an increas- ingly clay-like consistency with continual watering. After a period of fallow the soil returns apparently to its original condition. 1/ Many of the field management problems of Gezira soils result from the presence of exchangeable sodium on the clay complex. Sodium clays disperse easily but in the presence of salts they flocculate; in doing so they become impervious to water. The flocculent has to be dried to a granular state before it regains its good structure. During irrigation, soluble salts are washed out of the top of the soil, the solium clays disperse and are then flocculated during drying out. This goes on throughout the season with probably more and more dis- persion as the season goes on, particularly below the surface. At the end of the season the soils dry out, regaining their granular, loamy texture. lWhen irrigation stops the soil is dried out by evapo-trans- piration and shrinkage takes place. However, drying curves show that water has to be extracted to well below wilting point before air enters the pores of the soil; 2/ intense drying over some months would probably be necessary before significant aeration occurred. As a corollary to this, of course water would require a long time to enter these pores so that entrapped air, a source of oxygen for plant roots, probably remains throughout much of the growing season when the surface is sealed over by each irrigation. 1/ Trans. 3rd. Int. Cong. Soil Sci. 1, 21. 2/ Tech. Paper No. 58. Road Res. Lab.  ANNEX TI - 36 - APPENDIX 1 In addition to structural deterioration, salt accumulation as surface crusts may take place on the ridges in some areas. Some of these crusts contain 30 to 40 percent salts, sufficient to kill groundnuts, though apparently not harming cotton. 1/ But when the soils are wetted by the first rains the swelling dislodges pieces of the crust and some fall down cracks; in this way there is a very thorough mixing of the soil which amongst other things, probably greatly increases germination of weed seeds. Mixing is most intensive in the surface layers but profile examination shows that there is some mixing to 60 cm depth. Recor,mendations on Fallow Frequency. From the point of view of the soils, there is virtually no experimental evidence on which to base a recommendation on fallow frequency, but practical observations and theoretical considerations alike show that a period of desiccation is necessary to restore the structure of Gezira clay after it has deteriorated under irrigation. The most intense drying takes place during March, April and May and the system of making the first ridges before or during these months probably assists intensive drying and the entry of air into the pore spaces. On present evidence, this drying period must be considered an essential part of farming system. alternative systems in which the soil is irrigated during this period should be restricted to large-scale experiments or pilot schemes until they are proven. There is no evidence to show that the fallow year benefits the soil structure to a greater extent than the annual drying during M4arch- May. Fallows are, of course, essential for agricultural reasons other than soil structure, which are discussed in the Annex II. Careful measurements of soil properties will be needed in the experiments of experimental rotations with less frequent fallows recommended by the mission. FERTILIZERS Nitrogen Cotton. Under well-controlled experimental conditions the use of nitrogen fertilizers has given large increase in cotton yields. Table 14 gives the results of a typical trial,comparing times of application of the standard dressing of fertilizer. 1/ Ann. Rept. Gez. Res. Stat. 1959-60.  ANNEX II - 37 - APPENDIX 1 Table 14: TIME AND LEVEL OF NITROGEN APPLICATION Cotton yield k.p.f. Treatment 1954/5 1955/6 1956/7 1962/3 No fertilizer 5.67 6.oo 6.66 3.03 2N a/ at sowing 6.47 7.62 9.45 3.80 1N at sowing, 1N on 1st Oct. 6.23 7.92 10.33 3.65 2N at 1st Oct. 6.83 8.59 10.27 3.64 1N at sowing, 1N at 1st Nov. 6.99 8.14 10.80 4.20 2N at 1st Nov. 7.52 8.50 10.04 2.27 1N at sowing, 1N at 1st Dec. 6.17 7.66 9.62 3.93 2N at 1st Dec. 6.05 6.73 8.74 2.80 a/ 1N = 45 lb per feddan. 2N = 90 lb per feddan. Source: Ann. Reps. Gez. Res. Stat. 1954-7 and 1962-3. Split dressings or application of all the N after sowing gives increased yields provided the fertilizer is not applied too late. However, in a year of very poor yield (e.g. 1962/3) the application of all the fertilizer at sowing time is almost as good as the more labor-consuming split applications. The value of nitrogen fertilizers for cotton was known from the early days of the Syndicate, and the Board introduced its use as soon as it was economic to do so. Urea is used because it has a high nitrogen content, reducing freight and transport cost per unit of nitrogen. About 16 percent of the area received 1N in 1948 and the area increased each year until by 1953/4 virtually all of the Scheme received 1N. Much of the area was receiving 2N by 1956/7, but it was not until 1964/5 that the whole area received 2N. 1/ 1/ Agron. and Plant Phy. Sect. Gez. Res. Stat. AP/GB/126E.7. Newly-opened blocks in Southern Gezira received only 1N for the first cotton crop.  ANNEX II - 38 - APPENDIX 1 Other Crops. Nitrogen fertilizer is available for, but is seldom used on, dura and is generally used on wheat. Nitrogen dressings greatly increase the yield of dura, both of grain and of straw; in some experiments, 80 lb N per feddan gave a three-fold increase grain yield. Nitrogen dressings also decreased the attacks of buda. 1/ Increases of the same order, i.e. from about 272 to 658 kg (600 to 1450 lb) grain per feddan, have been obtained with 60 lb N per feddan on wheat. 2/ In experiments with groundnuts only small increases were obtained. No records could be found of any experiments on fodders or pastures but it was observed that pastures at Huka were showing obvious signs of nitrogen deficiency. Other Fertilizers. The complete fertilizer experiments on the G.A.R.S. have shown quite good responses to phosphate on cotton, especially in years when yields were large; trials in the Scheme have not, as yet, given any response, so phosphate fertilizers are not used. However with new phosphate-demanding crops and heavier nitrogen applications, a need is likely to arise fairly soon; a close watch should be kept on soil and plant phosphate levels. There are no indications that other fertilizers are likely to be necessary on these soils, at least for some very considerable time. The only trace element on which any information is available is iron, the deficiency of which causes yellowinrg in groundnuts which can be cured by iron sprays. The yellowing also disappeared after each irrigation so that the appearance of deficiency symptoms may be connected with the moisture stress. 3/ Field Results. The discrepancies between the responses that have been obtained in experimental plots and on a large scale, and the short- comings of the present methods of application of fertilizer are dis- cussed in Appendix 2. 1/ Ann. Rept. Gez. Res. Stat. 1959-60. Buda (Striga hermonthica). A sem-parasite weed of dura. While buda produces leaves and has a rudimentary root system it draws most of its food supply from the host plant which, in several cases, is stunted. 2/ Ann. Rept. Gez. Res. Stat. 1961-62. 3/ Ann. Rept. Gez. Res. Stat. 1959-60.  ANNEX II APPENDIX 1 C H A P T E R 6 IMPROVIMG THE PRODUCTIVITY OF GEZIRA SOILS The management of Gezira soils was planned more than 40 years ago after many investigations, starting in 1911, at Tayiba. The basic methods evolved have remained unchanged in spite of the mechanization of cultivations, the changes in rotations to deal with pests and diseases of cotton and the introduction of two new crops, wheat and groundnuts. That they have stood the test of time is a tribute to the agriculturists who developed them; nevertheless it is essential to examine them in the light of technological advances - ample mobile power, fertilizers, chemicals for pest and disease control, increased water supplies, a better understanding of soil techno- logy, etc. The physical factors necessary for good growth include ample nutrients, soil moisture, and soil aeration; these combined with optimum biological conditions (absence of pests and disease, suitable varieties) and with proper management ensure a good crop. The aim of management must be to identify the main limiting factors and alleviate them. Moreover, a superabundance of one factor cannot compensate for deficiencies in another: a point is reached when improvements to one factor alone cease to give economic - or indeed any - response because some new factor has become limiting. Thus, for example, extra fertilizer cannot compensate for lack, of moisture, costly sprays cannot undo the harm done by flooding, etc. Improving M4oisture Supply While there is little experimental evidenco to support the view that moisture supply is limiting yields, thure is plenty of evidance that soil moisture conditions are often far from optimum and that this may make the difference betweon average and really good yields. Improved moisture relationships in these soils consist essentially of getting more water into the soils in a shorter period. Excess water lying on the surface of the furrows is not only a waste of a valuable resource: it is also detrimental to the plant if the soil is left in a waterlogged condition for several days each month. Possible ways of obtaining better moisture conditions include wetting the soil to greater depths and maintaining the ridges in a moist but not waterlogged condition for as much of the period during growth as possible.  ANNEX II - 40 - APPENDIX 1 It has been shown (Table 12) that at least some Gezira soils can be wetted to 150 cm by allowing water to stand on the surface for a period of 14 days. A suitable site -- say half a rotation unit -- should be selected 1/ and leveled and bunded so that, at the end of the dry season, it could be kept flooded for 2-3 weeks. On drying, the germinating weeds would be cultivated and the soil kept bare to conserve moisture until the cotton crop was planted. Large ridges, exposing the greatest possible volume of soil to the air and good structure forming conditions are always desirable. Anothor method of obtaining better moisture conditions is worth testing on a field scale. Experiments at Abu Naama show that it is possible to keep the tops of ridges well aerated and moist even when the bottoms of the furrows are waterlogged. High, and thus large ridges, would be needed and the level of the irrigation water would need to be carefully controlled, necessitating levelled sites. Overhead sprinkler irrigation is unlikely to have any commercial application for cotton, because splashing may aid the spread of blackarin. However, the operating costs of such systems have been markedly reduced over the past decade; costs of 8 shils per acre inch or about PT.4O per 100 m3 have been recorded on large, diesel-engined units in Africa. Overhead sprinkler systems might be worth testing for perennial crops -- sugar cane, pastures, etc. -- especially as a way of utilizing "resting" land which is too uneven to be economically levelled for flood irrigation. Improving Saline and Alkaline Soils The rate of salinization brought about by irrigation water has been discussed and it has been shown that there is no immediate danger of increasing the salinity and alkalinity of Gezira soils. Though these calculations show that many years of irrigation would be required before the soils became unmanageable, there is virtually no information on the influence on yields of gradually increasing salinity and alkalinity levels. No doubt they will have an adverse effect but it is quite impossible to say how significant this will be. Thought should be given now to ways of preventing the accumulation of sodium. If the irrigation water can be got down to 1,5 m then salts will be spread over a greater depth of the profile. Better still, of course, would be the complete removal of the sodium salts from the plant rooting depth of the soil. This would require intensive drainage which, as has been seen, would be extremely expensive. 1/ Not all sites would be suitable for such an experiment: test borings in the bottom of dry abu ishreens would indicate soils where deep penetration of water could be expected.  ANNEX II - 41 - APPENDIX 1 Research. Much more research is needed on the rate of accumulation of salts and the increase of exchangeable sodium under irrigation. Soils from the continuous cotton plots form useful material for this. At the same time laboratory and pot investigations on the behavior of Gezira soils with varying levels of salts and exchangeable sodium are needed. Even more urgent is an accurate estimate of the influence of salts and exchangeable sodium :n yields on these soils. More work is needed, too, on the influence of gypsum on the soils as well as on crop growth. There is at present no answer to the question as to what can be done with these soils if and when they have been allowed to reach high e.s.p. levels. Practical trials of field drainage are needed, both to obtain more precise estimates of cost and to measure the effect of drainage on soil conditions and crop yields. Alleviating Salinity in the Field. If the sodium, added in the irrigation water only, could be removed then this would prevent deteriora- tion and allow perhaps an improvement by leaching to greater depths. One possible method for limited areas is growing saltbrush (Atriplex), which would be removed from the fields and burnt. One trial showed that 250 kg sodium per feddan (the equivalent of that added in irrigating 4 cotton crops) was removed by one saltbush crop: the 45 kg potassium and 90 kg nitrogen removed at the same time could be replaced by fertilizers but it is difficult to suggest a safe method of disposal of the sodium-rich ash. Elsewhere saline soils have been improved by growing deep rooting crops, e.g. berseem, which dry out the lower horizons, causing deeper cracking and thus deeper water penetration. Under present conditions the roots would probably not penetrate the dark grey subsoil horizon any better than cotton roots. Mechanical treatment - subsoiling - wDuld be prohibitively expensive but repeated heavy irrigations might allow enough moisture to enter the grey layer to stimulate root penetration. Under irrigation, salt accumulates at the top of the ridges whilst the furrows are nearly salt free: thus the more salt-sensitive plants, like groundnuts, are liable to suffer during germination. Planting on the side of the ridge, though requiring more careful water management, may be worthwhile in very salty areas. Fertilizer There is ample evidence from both ground and air observation that nitrogen fertilizers on cotton and other crops are very poorly distributed and that this is the cause of much uneveness of growth. The first requirement is therefore uniform distribution. Furthermore there is no evidence that economic returns are being obtained from nitrogen under the present cultural conditions of the cotton crop. Therefore large- scale pilot trials of 40 lb N/acre and 80 lb N/acre (properly applied)  AU7NEX II - 2 - APPENDIX 1 should be started within the Scheme immediately. It is quite probable that when other aspects of the husbandry improve, even higher dressings of nitrogen will give economic returns. Meanwhile there is need for work on the placement of nitrogen fertilizer in the cotton rows, on methods of applying it near planting time and on split dressings, the advantages of which have been shown by experiments at the G.A.R.S. Split dressings could easily be applied by hand, but are difficult with machines (see Appendix 3). There is convincing evidence that dura yields can be greatly increased by nitrogen fertilizers; in terms of water use it is totally uneconomic to grow an irrigated crop without fertilizers. Similar con- siderations apply to wheat. Legumes. The introduction of legumes solely for nitrogen supply is not likely to be economic but, in view of the high cost of nitrogen fertilizers, advantage should be taken of every opportunity to introduce useful legumes and some account should be taken of the nitrogen fixed when planning fertilizer dressings for succeeding crops. The need for microbiological research on nitrogen-fixing organisms is stressed in Appendix 4. It is unlikely that the present crops of lubia add much nitrogen but "berseem" crops at Kuku were very dark green: cn the other hand, pastures nearby showed obvious signs of nitrogen starvation emphaeizing the need for research on the use of legumes in pastures. Other fertilizers. More intensive cropping, the removal of more cash crops and utilization of crop residues and the future possibility of larger nitrogen applications may necessitate the use of phosphate fertilizers, probably fairly soon in some areas. There are no indications that other fertilizers will be necessary. Deficiency symptoms can be treated, e.g. iron deficiency in groundnuts by iron chelates, but there is no evidence that this increases yield. Alternative Crops Alternative crops are considered in detail in Appendix 2, Agronomy. Experience at the G.A.R.S. and from other parts of the world with similar soils (notably Kafue in Rhodesia and Kimberley in N.W. Australia) shows that a wide range of crops can be grown on clay soils of the type found in Gezira provided that the soil structure is restored by periodic drying. Grazing of irrigated pastures can do serious damage through trampling and puddling of the soil. Rice is an interesting crop because of its special effect on the soil. The very heavy irrigations (perhaps 6000 m3 per feddan) needed increase the danger of secondary salinization: however, there is no  ANNEX II - 43 - APPENDIX 1 danger of rising the water table in Gezira and the heavy applications might in practice wash salts down to greater depths and build up a reserve of water. In the absence of any experience with rice in Gezira (although work has started at Abu Naama) it would be worth starting trials at some convenient site, which would need to be fairly large to ensure proper water control. Forestry is often the best use for poor soils: Eucalyptus microtl.eca can be grown on saline or poorly drained soils and will withstand long periods of drought.  ANNEX II APPENDIX 1 C H A P T E R 7 SOIL AND LAND USE CLASSIFICATION Soil Surveys During the initial development of Gezira Main, the soils were examined and sampled on a grid basis but no soil maps were made. Certain areas, especially in the north and west, were omitted because analysis showed high salinity. More recent surveys by consultants confirm that the soils of the more northerly, i.e. the drier,parts of the Gezira are of poorer quality when salinity is the main criteria. A project ("Strengthening of the Soil Survey Division") has been: set up under the Development Plan and is to operate for five years from 196h. The United Nations Special Fund has allocated $694,900 and the Government $1,866,456 to this project. Amongst its original commitments were the following:- a. Survey of selected areas, amounting to 500,000 feddans, in the Gezira-Managil for purposes of intensification and improvement of existing irrigation. b. Provision of a basis for soil classification and land classification. c. Reconnaissance soil survey of 2 million feddans in the El Hawata-Dinder region to select areas for more detailed surveys: such areas would be irrigated from Roseires. d. Survey of 500,000 feddans in the above area for planning the overall irrigation layout. e. Reconnaissance survey of 1 million feddans in the Kenana Scheme for determination of suitability for irrigation by a left bank canal from Roseires dam. More recently it has been agreed that reconnaissance and semi-detailed surveys on 600,000 feddans in the El Hawata-Dinder area should be carried out by consultants. The Soil Survey Division is using the survey of the Gezira- Managil area to confirm the present criteria, or develop new ones,  AN EX II - 45 - APPENDIX 1 by which soils can be classified as to their suitability for growing irrigated cotton. Some soil survey work has already been done and data on cotton yields of certain selected areas have been collected. These will be used to establish correlations between cotton yields and soil characteristics and, if necessary, to set up a new land capability classification. Some information on soil capability for other crops will be collected also, but since yields are not known accurately, it will be impossible to establish correlations. Land Capability Classification In their project areas, the consultants are adopting the usual basis for classifying alkaline soils, i.e. exchangeable sodium per- centage. Their criteria for class 2 and 3 lands are:- a. Exchangeable sodium percentage: 0-45 cm - less than 15 45-90 cm - less than 25 b. Conductivity o-45 cm - less than 5.3 millimhos per cm 45-90 cm - less than 8.0 millimhos per cm c. Clay content 0-45 cm - 35 to 50 percent 45-90 cm - more than 50 percenit There is little experimental evidence for selecting these levels for Gezira soils; the value of e.s:.p. 15 is suggested by the United States Bureau of Reclamation for soils usually quite different from those found in Gezira. Whilst the influence of salinity and exchangeable sodium on Gezira soils is recognized, there is ample evidence that the levels applicable elsewhere do not apply in the Gezira and in fact there is little evidence that they are the dominating factors in the productivity of the soil. Although classification within an international system, e.g. the Seventh Approximation, is desirable for relating soils found in the Sudan to those located elsewhere, some of the properties used .or separations of the higher categories appear to have little agronomic significance in the Sudan. There is a need therefore for a local classification to take into account what appear to be the most sig- nificant features in agronomic terms. Since water enters the soil via the cracks and since the depth to which roots can penetrate is very important, depth and extent of cracking and depth to a hard layer should have high priority in the soil classification system. Another factor in land use is the evenness of the soil. Since non-uniform soils will need expensive  ANNEX II - 46 - APPENDIX 1 levelling, degree of unevenness shou'ld be a factor in the land use classification. The nature of the surface soil, whether self-mulching or forming hard or brittle crusts should be included in the local soil classification. Laboratory methods of soil examination should include both physical and chemical determinations. Determinations of physical properties, e.g. shrinking and swelling measurements, flocculation values, would be useful. Measurements of capillary rise should be useful too in arranging the soils in some order between the very good soils, i.e. those with little or no sodium, and the very poor ones, i.e. those with very large quantities. Chemical measurements most likely to be useful are exchangeable sodium, soluble salts and gypsum. Different criteria from those normally used will have to be evolved for the exchangeable sodium. Full use cannot be made of the various physical and chemical criteria until they are correlated with crop yields. At first sight this might appear relatively straightforward since cotton yields are accurately recorded over the Gezira and field management standards are theoretically on the same level. However closer examination soon shows that this is not correct for the skill and diligence of the individual cultivator has an enormous effect on yields. In some of the blocks e.g. Feragin, the layout of the irrigation system influences yields and in others e.g. Managil, yields are greatly influenced by flooding. Whilst it is possible to get a broad picture of the average yield distribution of different areas,such a picture does not give any information on the reasons for the yjield variation. Values of certain factors e.g. clay percentage, salts, exchangeable sodium,can be correl'ated with yields but this does not necessarily mean that they are the cause. More reliable information can only be obtained by a well-designed research program to examine in detail certain specific areas where, under uniform field management conditions, the various soil factors (salt, exchangeable sodium, cracking, bulk volume, moisture penetration etc.) can be measured and correlated with yield. Such detailed work will take several years in a well organized research program and is certainly most desirable. Meanwhile it is urgent to get information for use on the potential new irrigable areas which would be irrigated from the Roseires dam. The consultants appear to have made full use of any information available and it seems unlikely that any further information leading to a drastic alteration of their land use classification will arise from the current soil survey programs.  ANNEX II - 47 - APPENDIX 1 Their arrangements of Class 2 and Class 3 lands are probably of the right order but some of the Class 4 and even Class 5 lands might have higher potentials than have been ascribed to them. Generally the better areas seem rather similar to the Gezira and have about the same land use potential.