. I J 0 • <, \, - . • " THE EQUiATORIAL NILE PROJ:E,CT • • AND ITS EFFECTS IN TH.e ANGLO-EGYPTIAN SUDAN f • • Being the Report -.. • of the longlei Investigation Team, • , - . VOLUME I • .; . S. 2. EXPERIMENTAL FISHERIES: SOBAT (1953). 3. A TYPICAL CATCH OF DISTICHODUS, HETEROTIS, TILAPIA AND HYDROCYON. VII I. SOUNDING IN THE MAIN CHANNEL OF THE BAHR EL JEBEL (NEAR TERAKEKA). 2. TYPICAL POOL OF OPEN WATER (ALIAB VALLEY). 3. LEVELLING ON A TYPICAL CROSS-SECTION (ALIAB VALLEY). VIII 1. CHAINING THROUGH TYPICAL INUNDATED TOICH (NEAR GEMMEIZA). 2. PAPYRUS-LOCKED POOL ON THE LINE OF A CROSS-SECTION (NEAR GEMMEIZA). 3. CONDITIONS ON ONE OF THE MAIN CROSS-SECTIONS (ALIAB V ALLEY). • " ,, CHAPTER L HYDROLOGY 1. TOPOGRAPHY The area with which we af(~ concerned extends from the southern border of the Sudan to Kosti on the White Nile, from latitude 3° 30' N. to latitude 13° N. The tribal m:eas in which the effects of the Equatorial Nile Project will be felt directly are those on the banks of the main rivers, the Bahr el Jebel, the Bahr el Zeraf, the White Nile, and parts of the Bahr el Ghazal and the Sobat. They lie roughly between longitudes 29° and 34° East. The description of this area from the point of view of hydrology will hav~ to include oth~r areas whose rivers run into and affect the Jonglei Area itself. For convenience, the topographical account which follows is divided into four parts, the first dealing with the main rivers, and the remainder with each of three natural hydrological divisions which are as fonows: (I) East of the Bah"l"el Jebel and south of the White Nile and the Sobat-the South-Eastern Area. (2) West of the BahT el Jebel and south of the BahT el Ghazal- the South-Western Area. (3) North of the White Nile to Kosti-the Northern Area. SOUTHERN, CENTRAL AND NORTHERN ZONES Later in this report it will be necessary to refer frequently to thr~e Zones which have been known throughout the investigation as the' Southern, Central, and Northern Zones ' . The division of the Jonglei Area into Zon~s is based on the hydrological effects ofthe Equatorial Nile Project; in each Zone one set of hydrological conditions will exist. The exact reasons for the divisions will be explained later when the effects of the Project are described; here we wish merely to define the Zones, and describe them geographically. The Southern Zone extends from the Sudan frontier at Nimule to the beginning of the Canal at Jonglei. The Central Zone begins at Jonglei and is bounded on the north by a line drawn roughly through Yoynyang, Buffalo Cape, Fangak, aNd Abwong. The Northern Zone covers all the area north of the Central Zone to Kosti. It will be seen at once that the , Northern Area' of the topographical divisions already described very nearly coincides with the Northern Zone, but that the' South-Eastern' and' South-Western Areas' cover eastern and western parts of both Central and Southern Zones. THE BAHR EL JEBEL AND THE WHITE NILE The principal physical characteristics of the Bahr el Jebel and the White Nile are sum- marized in Table 1 (p. 49). The remarkable longitudinal profile of the river is shown in the column of average slopes, which begin at almost 100 cm/ km. between Nimule and Rejaf and decrease progressively from south to north to less than 1 cm/ km. between Jebelein and Khar- toum. Table 2 (p. 49) is a list of selected cross-sections of the Bahr el Jebel and the White Nile which show the level of tht9 river in relation to the ground alongside it. Between Juba and Bor (Figs. A 1 to A 12) the main features of the conformation are: (i) A main river channel (or channels) which has a deep section bounded by high alluvial banks. Cii) A broad flood-plain Cor plains) the level of which is lower than the water-level in the river except when the latter is at its lowest. (iii) The whole river with its flood-plain is contained by definite banks of hard ground, above highest river levels and roughly 10 km. apart on the average. (iv) The banks of the river which merge into the flood-plains are of alluvial formation and are up to one kilometre wide. Butcher(l) shows that this type of formation of river valley continues on the east bank to a point about 15 km. north of Bor, 612 km. from Lake No, whereas on the west bank high ground can be traced as far north as Lake Jur (or Nuong), 356 km. from Lake No.(2) The cross-sectio.n s at Malek (Fig. A 13) and at Shambe (Fig. A 14) illustrate this high ground east and west of the river. From the two points mentioned above and nearly to the White Nile at Lake No the same formation of river and flood-plain is to be found, as shown on the section of the Bahr el Jebel at Peake's Track (Fig. A 15). But the difference between this portion of the river and the southern section already described is that there is no high ground outside the flood-plain to set a limit to its extent. The river, with its alluvial banks, is at the highest point of the cross-section, and the ground on both sides lies at a level below that of the water surface in the river. The cross-section from the Bahr el Jebel at km. 60 to Bentiu is an example of this feature (Fig. A 18). Westwards from the Bahr el Jebel at this place the groun({ falls below the river level by 1·0 m. in the first 8·5 km., is flat for the next 20 km., and at 30 km. from the river it is still half a metre below river level. On the cross-section from Buffalo Cape to Fangak (Fig. A 19) it can be seen that the ground is below normal high river level for 14 km. east of the river, and at only one point in the 65 km. between the two points does the ground rise to 0·75 m. above the river. The remaining sections cover the White Nile, which will be seen to be similar to the Bahr el Jebel between Juba and Bor in that there are definite high banks which limit the width of the flood-plain. The White Nile has a deep water channel (or channels) with small alluvial banks, and a flood-plain (or plains) bounded by high banks which are 2 km. apart on the average. Referring to the cross-section north of Fanyikang (Fig. A 22), we see that the river bank is a ridge with lower ground on its inland side. At 1·5 km. from the river there is a marked rise of about 2 m., after which the ground falls slightly to km. 4.8, where the slope changes to a gradual rise continuing as far as the section was taken, 10 km. from the river. Exactly the same formation is seen on the section west of Wau (Fig. A 22). In this case the ridge of high ground which limits the width of the flood-plain is more outstanding; the depression next to it is less than one kilometre wide, and the rise beyond that is nearly 3 m. high. The plateau inland of the rise is saucer-shaped as at Fanyikang, though its slopes are negligible. On the east bank, the Khor Dungaru section (Fig. A 23) indicates that the formation is the same as on the west bank. (Due allowance must be made for irregularities which appear where the survey line approaches the khor, as at km. 7.) On the section west of Detwok (Fig. A 24) the ridge can be seen to be about 4 m. high and not more than 100 m. wide. Inland of the ridge is a broad depression, 10 km. wide and on the average not more than 0·5 m. above normal high river level. From the end of the depression to the end of the section, 29 km. from the river, there is a slight steady rise. Along Khor Adar (Fig. K 24) the ground formation is the same as at Detwok. The ridge is 1·5 m. above normal high river level; the depression behind it is about 15 km. wide, and the steady slight rise continues to the end of the section nearly 80 km. from the river. On all the White Nile sections a characteristic feature is the high bank, or ridge, which marks the edge of the river flood-plain. The Shilluk have their villages and some of their cultivation on this ridge, which is therefore suitably called the ' Shilluk ridge'. The other cross-sections taken by the Egyptian Irrigation Department between Malakal and Jebelein indicate, where they have been continued inland (particularly just north of Renk on both west and east banks), that the same type of formation is found in that part of the river; that is to say a high plateau, either very flat or slightly dished, outside a depression near the river, which is bounded by definite banks. At Jebelein and northwards the river appears to lie in a valley with steady rises inland on both sides. This concludes the description of the topography of the Bahr el Jebel and the White Nile. Little remains to be said here about other main rivers. The Bahr el Zeraf is very similar, as would be expected, to tha.t part of the Bahr el Jebel which it parallels, as can also be seen in Table 1. It will be observed that the Bahr el Ghazal is extremely flat~in the whole of its 250 km. from Meshra er Req, and the slope on the River Sobat in its last 117 km. is not much more than that on the Bahr el Ghazal. THE SOUTH-EASTERN AREA (See Figs. A 26 and D 27) We have defined the South-Eastern Area as that part of the Jonglei Area which lies east of the Bahr el Jebel and south of the White Nile and the Sobat. In general this area can be regarded as a single hydrological unit ; as one catchment whose watercourses drain from south to north, and, with a few obvious exceptions south of Juba, make no contribution to the Bahr el Jebel. A further convenient sub-division can be made on the line of the River Veveno, where it runs from south-east to north-west, extended to meet the Bahr el Jebel near Gemmeiza, since there is only one watercourse whick crosses this line between the Bahr el Jebel and the Veveno. c Beginning in the far south, the outstanding topographical features of the South-Eastern Area are the mountains on the southern border of the Sudan. In a previous report these mountains were given the general name of the Imatong Mountains,(S) though, as was pointed out, this name applies strictly to only one of the mountain groups. On a small-scale map they appear as a range which roughly follows the Sudan frontier. In fact they are not a range at all, although they do divide the streams into those flowing northwards to the plains, those flowing westwards to the Bahr el Jebel, and those which flow southwards to join the Assua. 2 They are a g{oup of isolated massifs, mainly outcrops of the basement complex, rising abruptly from the plains and separated by broad flat valleys largely formed by the degradation of the mountains. Referring to the map of the south-eastern mountains (Fig. A 26), we see the first of these groups in a line stretching roughly south-eastwards from Juba for a distance of 100 km. It consists of a series of low, rocky hills, thinly covered with forest. The most conspicuous of these is Jebel Lokiri (1,691 m.), about 30 km. south-west of Torit, which is visible from many different and distant view-points. The next group to the south-east is the Imatong-Acholi group. These hills cover an area of some 2,000 sq. km., and are topped by Mt. Kinyeti (3,187 m.); they are tlle largest and highest of the south-eastern mountains. The rock is basement complex, and much df it is precipitous. The highest part of the group is heavily forested plateau with few rock outcrops, but, to the east, the mountains end abruptly above the valley of the Koss with an escarpment. The Imatong and Acholi Mountains form a giant horseshoe open to the north, between whose arms lies the Kinyeti. To the north-east lie the Dongotona and Lafit hills. The former are separated from the Imatongs by the 15 km. wide valley of the Koss, and cover an area of about 400 sq. km. Their highest peak is Jebel Emogodung (2,623 m.). Their rock is basement complex, their slopes are the most precipitous of all the groups, and they are only thinly covered with trees. Around the base is a ' beach' of talus up to 4 km. wide. The Lafit Hills are an outlier of the Dongotonas, about 15 km. to the north-west. They ext~nd some 60 km. in a north-west direction almost to the 5th paraJllel, and their highest point is Jebel Lodio (1,930 m.) at the southern end. They are precipitous and have many bare rock faces, but where the steepness allows there is a thin cover of forest, and in some sheltered kloofs heavy forest. The eastern side is steeper than the western. Farther east again are the Didinga Hills on the eastern side of the Kidepo valley. These have an important geological difference from those already described. They are formed of tertiary lavas overlaid oy sedimentary rocks, resulting in rounder outlines and more gentle slopes. There is considerable forest cover, especially on the central plateau, but the north-east corner, the upper catchment of the Singaita River, is desiccated and subject to erosion as a result of over-grazing. The total area of the Didingas is some 2,000 sq. km., and the highest point is Jebel Lotuke (2,795 m.) in the south-east corner. Like the Dongotonas, the Didinga Hills have a conspicuous outlier to the north-west, the Boya Hills, which extend to latitude 4° 45' N. and whose peaks, though not higher than 1,400 m., are jagged and extremely precipitous. The Nangeya Hills lie outside the Sudan border, but they are important to the Sudan because their run-off is either into the Kidepo or into the Pager, a tributary of the Assua. They extend for 50 km. in a line from south-east to north-west, with peaks at the northern and southern ends; Lonyili (2,290 m.) and Rom (2,372 m.). They are fairly heavily forested. Finally, mention must be made of one more group which lies in Uganda to the south-east of the Didinga Hills. They have no generic name, but are topped by Mt. Moruongole (2,750 m.). They are the farthest sources of the Kidepo and the Pager, and form a watershed between the Nile basin and.an internal drainage system to the north-east ending in the Lotagipi swamps. Tables 3 and 4 (pp. 50-1) su~arize some of the data concerning slopes of the torrents and peak heights of the southern hills from which these torrents emerge. The geological map of superficial deposits(4) shows the south-eastern mountains to be covered with thin superficial deposits, whereas there are ferruginous deposits round their bases. To the east of the Didinga Hills surface clays extend to the Sudan border. These geological factors have an important bearing on run-off, the ferruginous deposits being permeable and the clays impermeable. Before leaving the mountains, mention must be made of the area lying between the Acholi Mountains and the Bahr el Jebel. In the south the country consists of a plateau extending from the foot of these mountains to the gorge of the Bahr el Jebel. The Ateppi, the largest of the loc~ torrents on the east bank, cuts a gash in the plateau and joins the Assua a few kilometres upstream of its junction with the Bahr el Jebel. Farther north the country is more rolling, and the sides of the valleys of the Uma and the Kit have sufficient slope to be obvious to the eye from the air. Dense forests of mahogany are found at the foot of the Acholi Mountains, but the farther one goes from the hills, the sparser the vegetation becomes. There are numerous smaller torrents flanking the Bahr el Jebel, and a few khors which join it in the neighbourhood of Juba, but no other channels of any significance in the eastern area run into the Bahr el Jebel. 3 { ( From the 5th parallel and northwards we find nothing but clay plains, so flllt that, as will be seen later, the torrents quickly fan out and lose their channels altogether. In Table 1 we see that the slope of the Bahr el Jebel between Mongalla and Jonglei varies from 18 to 10 cm/km. It is very probable that the slope of the plain to the north of the south-eastern mountains is the same. From surveys we know that the difference in level between the ground at Jebel Lafon, where the Kinyeti has a dtdined channel, and Pengko is 70 m., and from the map the distance is 142 km., the average slope on this portion of the plain being therefore 50 cm/km. We now turn to the plain north of the Pibor-Gemmeiza line (Fig. A 27), and, from surveys carried out by the Egyptian Irrigation Department, we find that the average slope in any direction between north-west and north-east from Pengko exceeds 10 em/km. Directly towards Pibor the slope is not more than 2·5 cm/ km. for the first 40 km., after which it steepens to 11 ·5 cm/km. The Duk Ridge traverses the plain north of Pengko from south to north. It lies along one of the main drainage lines (Figs. A 28 and A 29), mostly to the east of the , Direct' Line of the Canal, though it crosses the latter in the north where the khor is called Chieth, the upper part of Khor Atar. The' duks ' themselves are large sandy dunes on the banks of the drainage-channel, and were probably formed many years ago when, possibly, the whole plain was part of a swamp and the khor a bigger waterway. Clumps of fine old trees and groves of dam palms characterize the' duks " and form prominent features in a region of seemingly endless grass plains. Though there is a definite slope northwards over the Pengko plain, on any cross-section along a parallel from the Bahr el Jebel we would expect to find little change in level. We hav~ already pointed out that near the river in this region the water-level is higher than the ground bordering it, and the cross-sections from km. 60 to the west and from Buffalo Cape to the east confirm the assumption of flatness east and west of the river. In fact so featureless is this vast area that there is little else to describe. THE SOUTH-WESTERN AREA (See Figs. A 30 and D 53) We can consider the area west of the Bahr el Jebel and south of the Bahr el Ghazal also as a single hydrological unit. We need not include the Bahr el Arab, the Jur, and the Bahr el Ghazal upstream of Y oynyang, because the Equatorial Nile Project will affect only the the lower 90 km. of the latter, between its mouth at Lake No and Yoynyang, and the first two are tributaries of the Bahr el Ghazal which join it upstream of Y oynyang. Hence the boundaries, for investigation purposes, are the watershed between the Tonj and the Jur on the west, and the Nile-Congo Divide to the south, the boundary between the Sudan and the Belgian Congo. It should be mentioned that, strictly speaking, remedial measures for losses on the Bahr el Ghazal might involve control works on the Bahr el Arab, the Jur, the Sue, the Pongo, the Bahr el Ghazal, and other rivers. But since only a negligible portion of the Bahr el Ghazal wi11 be affected, such a wide survey could neither be justified nor carried out with the staff and the time available, especially in view of the greater importance of the losses on the Bahr el Jebel and the need to make surveys of other rivers which affect conditions within the Jonglei Area more directly. In fact we were able to carry out some work on the Sue and the Jur, and our observations are included in this report. 0 We have seen that the country east of the Bahr el Jebel is essentially a sedimentary clay plain out of which, in the extreme south, rise mountainf>us peaks of the inselberg type. West of the river the clay plain ris limited in extent, being found only in the north between the Bahr el Ghazal and the Bahr el Jebel, and as a thin strip bordering the Bahr el Jebel as far south as Juba. The remainder of the South-Western Ar~a consists of two peneplains, covered with ironstone deposits, which incline northwards and disappear below the clay at the 400 m. contour, some distance north of the road which connects Wau with Tonj, Rumbek, and Yirol. In the south the plateau rises to over 1,000 m. in a steep eastern escarpment flanking the Bahr el Jebel between Nimule and Juba. As a result of this geological difference the rivers on the west bank, in contrast to those on the east which lose their form south of the 5th parallel soon after debouching from the mountains, have defined channels and valleys as far north as the 7th and 8th parallels, where they encounter the clay plains. L Hilly peaks are not as high nor as frequent as on the east bank. The eastern escarpment and the country westwards to the Yei River is hilly with steep-sided inselbergs, but elsewhere the country is rounded and rolling with few rocky outcrops. In the far south we find Jebel Gumbiri (1 ,708 m.) and a point on the escarpment west of Nimule (1 ,594 m.) the highest peaks. Next to them are the Bala Hills, with Korobi (1,501 m.) and other summits averaging 1,350 m. The topographical differeNces between the country east and the country west of the Bahr el Jebel are brought out in a comparison of peak heights and river slopes (Tables 3 and 4). In 4 the west the\e are no mouNtains to compare with Mt. Kinyeti; the rivers there are longer and their slopes in the hills are similar to the slopes of the rivers in the east outside the hilly region. The two torrents, the Kaia (with its tributary, the Kijo) and the Luri, are exceptioNs, but they discharge direct into the Bahr el Jebel 64 km. south and 6 kID. north of Juba respective- ly. They, together with the Tapari (Pap Gel), drain the eastern and higher plateau. The last named also joins the Bahr el Jebel some 20 km. north of Bor. The western and lower plateau is drained by the Yei, Na'am, Gel (Rumbek), Ibba (Tonj), and Sue, which rise on the Divide and traverse considerable distances in defined valleys in ironstone country before reaching the plains. The Sue, one of the farthest sources of the Jur, has several upper branches which cover about 220 km. of the frontier. Here the plateau is so flat that the streams are slow flowing and are overgrown with papyrus and grasses. Farther downstream they pass over a series of rocky bars and are swifter Funning with sandy beds. From somewhere south of the Wau-Yirol road, these rivers begin to fan 01;1t on to flood- plains, deltaic in formation and increasing in width, until they reac.h approximately the 400 1iI!l. contour. There they spread out into a mass of interconnected swamps and lose their indi- viduality completely. North of Shambe we find the typically small slopes of the clay plains. From Adok there is little difference between average slopes to the north and to the west of the Bahr el Jebel (Fig. A 27), and the Bahr el Ghazal takes a circular course from Mesilla er Req to Lake No witla the extremely smatl slope of less than 1/100,000. Over most of the ironstone country there is forest cover aRd grasses. The latter are burnt off annually for grazing, with the result that trees never reach any great height. In the swamps are grasses which remain wet for most of the dry season. THE NORTHERN AREA (See Fig. D 56) We will consider the Northern Area to b(l roughly from the line of the White Nile where it flows from Lake No eastwards to the mouth of the Sobat, to Kosti in the north (Fig. D 56). In longitude the investigation was confined to the fiat valley bounded on the east by the Ethiopian and Ingessana Hills, and by the Nub>a Mountains on the west. We have already mentioned the small slope of the plain from south to north, where the river runs, and the lack of appreciable differences in level (last and west of the Bahr el Jebel, as shown on sections ftom krn. 60 and Buffalo Cape. North of the Sobat and east of Malakal the major slope from Khor Machar head through the marshes to the mouth of the Adar at MeIut averages 7·85 cm/ km., or a difference of about 22 m. in 280 krn. North of Melut the country is more undulating, especially inland from Renk, and there must be a bluff of land which causes the Khors Tombak, Ahmar, and Yabus to take a southerly course after leaving the hills, into the Machar Marshes. The' Shilluk ridge ' and the country flanking the river have been described together with the topography of the White Nile. West of the river we assume, from the lack of marked watercourses, that an almost uniform plain reaches to the foot of the Nuba Mountains, rising steadily with a gentle grade all the way. At the latitude of Renk and northwards large dunes are found , as much as a kilometre in length and 100 m. wide, on both sides of the river. These qazes, as the dunes are called locally, are not entirely composed of sand, and the run-off from them has been most successfully collected into hafirs sited at their bases. It is hoped that the t~pographical information given in this part of the report is sufficient to enable the reader to form a correct appreciation of the lie of the land ; alii. essential preliminary to a discussion of its hydrology and j)f its potentialities for development for alternative liveli- hood and remedies where alternatives cannot be provided. 2. CLIMATE According to the generally accepted classification of world climates, the Jonglei Area lies within two major climatic zones : zone of steppe climate and zone of grassland-savannah climate. Since climate is the resultant of a number of interacting factors there are no sharp boundaries, but wide transition belts, between major climatic zones. However we must specify some sort of boundary between the two zones within the J onglei Area, and a study of vegetation and climari.c data suggests that the 640 mm. (25 inches) isohyet is the boundary between them (Map 3); steppe climate being to the north, and grassland-savannah climate to the south of it. Generally speaking temperature and rainfall have more effect on climate than any other factors. In the Jonglei Area the temperature range is small and, on the whole, biologically favourable. On the other hand rainfall is characterized by an irregular and seasonal distribu- tion, and also varies according to latitude, that in the northern part of the area being often critical in quantity. 5 Rainfall is therefore the principal limiting climatic factor. It holds this dominant position in forming environment not solely because it is the chief source of the supply o(moisture but also because of its influence on other factors of importance to plant life, such as humidity, temperature, insolation, and local air movements, which are responsible for evaporation and transpiration. We therefore devote most attention to rainfall, but at the same time, so that they are not overlooked, we include the available data concerning the other climatic factors. RAINFALL The South Atlantic Ocean is the main source region of the moist air-masses which bring rain to the Southern Sudan when the south-west winds blow. Precipitation is due to three causes; convection, orography, and the meeting of the cold, dry northerly with the warm, moist southerly winds. Frequent afternoon thunderstorms are evidence of convective activity, and the aFlnual movement of the sun within the tropics is responsible for the variation in .the length of the rainy season from 2 or 3 months at Kosti (lat. 13° N.) to 7 or 8 months at Juba (lat. 5° N .). The map of isohyets for the Southern Sudan (Map 3) might well be taken as a text book example of the influence of topography on rainfall. It is higher over mountain, hill and plateau, and lower over the flat plains. The highest rainfall, which averages 1,408 mm. per year, occurs over the highest mountains to the south-east; the next highest, 1,224 ffiffi . per year, over the south-west plateau. On the plains at the same latitude the rainfall is only 865 and 900 mm. per year on the average, east and west of the river respectively. In the Northern Area the lowest rainfall occurs at Kosti, 400 mm. per year ; and from Malakal northwards rainfall on the lowest central part of the valley, on the White Nile, is lower than on the Kordofan and Ethiopian hills on its flanks. . The proximity ofthe swamps appears to have some effect on the rainfall at Tonga, Malakal, and possibly at Fangak, but it is obviously very localized. The reader should note that over the larger part of the Jonglei Area, particularly where grazing will be affected in the Central Zone, the rainfall averages between 800 and 900 mm. annually. Between Malakal and Kosti the average rainfall decreases from 800 to 400 mm., and in between lies the interesting inter- regional boundary, the -640 mm. isohyet. This line crosses the White Nile at Melut and turns northwards, to swing round the Ingessana Hills in the 0ast and the Nuba Mountains in the west. The change from one climatic region to the other has one immediately important effect on the problem of grazing within the Jonglei Area. Though there is a transitional belt, the grasses to the north of the 640 mm. isohyet are, on the whole, short annuals and are palatable and nutritive in the dry state (sweet veldt), whereas those which grow to the south of it are mainly perennials which have the opposite characteristics (sour veldt). Another aspect of rainfall which has an important effect on agriculture is its equatorial double maximum intensity. This is noticeable in the figures for average annual rainfall as far north as Jonglei, but in individual years it can often be clearly observed as far north as Malakal. STATISTICS Until recently rainfall recording stations could be set up only where literate and reliable recorders could, and would, live permanently, with the inevitable result that over most of the Jonglei Area there have been only a few stations at centres of habitation. Over the vast Eastern Plain where data would be most useful there are no recor~ ; so we have to assume the average annual rainfall from gauging stations on the circumference. In the last few years the number of recording stations has been increased, and monthly recording gauges have been in use, but until the Eastern Plain is more accessible the rainfall there will not be measured. We have analysed the records of 16 stations in the Jonglei Area, which represent rainfall on mountain, hill, plateau, and plain (Fig. Bland Tables 6-21 , pp. 52-9). In addition the Research Statistician of the Ministry of Agriculture has been kind enough to analyse the records of monthly rainfall for five stations according to a method, developed by H . L. Manning and modified by the Statistician, which presents the data on variability in a form more suitable for agriculturists (see Tables 23- 27, pp. 61-3). The main feature of the analysis is the high variation of rainfall all over the area, particularly in individual months. H cannot be over-emphasized that in this area average annual rainfall figures mean very little. The great variations annually, and the even greater variations monthly, are definitely the most important single environmental factor affecting the present and future economy of the Jonglei Area. The variability as measured by the standard deviation tends to decrease where the rainfall is highest, over the hills in the south, and increase where it is lowest, over the plains in the north. 6 , The re~der must also realize that there are long dry periods between the rainy seasons. From the southern border at Niruule to about the 7th parallel rainfall is negligible for 5 months from November to March; at Malakal there is no rain for 6 months from November to April ; at Renk for 8 months from October to May; and at Kosti for 9 months from mid-September to mid-June. On the average it is wet for half the year, and dry for the other half. (The rainy season is defined here as the period during which the rainfall in anyone month exceeds 50 ffiffi. on the average.) This is stressed so that it is clearly understood that if green grazing is needed from the river late in the dry season it must be irrigated, or obtained from particular places where the combination of soil and topography results in the retention of sufficient moisture in the ground to produce green regrowth after the grasses have been burnt. The five tables giving the estimated percentage points of the monthly rainfall are self- explanatory (Tables 23-27). The information given includes first the mean monthly rainfall and secondly the median monthly rainfall. This latter is the rainfall figure which is reached in 50% of the years. Since the distribution annually is skew, the median is in all cases less than the mean; that is to say the rainfall is more frequently less than the mean. The tables also give the higher and lower percentage points. The figures of monthly rainfall presented against 'lower percentage points' are estimates of the rainfall which is not reached in the corresponding percentage of the number of years, whereas the figures given against the' higher percentage points' are estimates of rainfall which is exceeded in the corresponding percentage of the number of years. The estimates are not exactly correct, both because they have been calculated from a finite number of years and because the mathematical assumptions are not strictly true. Of special interest are the 25% high and low percentage points, because the chances are fifty-fifty that the rainfall in each month will be between the two figures shown against them. EVAPORATION Whereas rainfall data represent the credit side of the soil moisture account, evaporation and transpiration represent the -debit side. There are no known simple relationships between evaporation from free water surface (as measured by the Piche method) and evaporation from soil surface, except when the latter is saturated. Piche evaporation is usually the only measure- ment available which indicates the general trends on the debit of the soil moisture account. We have converted the Piche readings for the few available stations by the generally accepted conversion factor of 0·5, and have tabulated the evaporation in mi1limetres per month (Table 22, p. 60). We have underlined the figures for those months in which rainfall usually exceeds evaporation. The average annual evaporation varies from just over one metre at Nagishot to over three metres at Kosti. It is lowest in July, August, and September; since rainfall is highest in July and August we find conditions of maximum soil moisture occurring in September. RAINFALL AND SOIL MOISTURE We draw the reader's attention to the reason for the existence of critical conditions during the rains on the impermeable plains of the Jonglei Area. The structural and textural character- istics of the soil types cO.ffiffion in this area (which are discussed in detail in Chapter 2) result in the useful range of soil moisture between wilting point and saturation being small. When the soil moisture is below wilting point, no water is available for vegetation. Above saturation point, since both surface and sub-suface drainage is poor, no more moisture can be absorbed and additional rain causes immediate widespread flooding, from which arises the phenomenon which we caII ' creeping flow'. When one compares the wide range of variability of rainfall mentioned above with the narrow range of useful soil moisture, it at once becomes evident that conditions frequently occur which are outside the critical limits, with the result that drought and flooding can follow each other in quick succession. Crops planted in the early rains often suffer from drought in the middle months owing to the equatorial distribution of rain, and in September the soil is as often saturated, so that moisture conditions are normally unfavourable to the majority of crops. The difficulty with pasture, on the other hand, is that the grasses are unpalatable during th.e dry season unless the ground is subjected to sufficient flooding to ensure fresh green regrowth after burning, or unless drainage is good enough to dry the soil quickly and produce' sweet veldt'. RAINFALL DURING THE PERlOD OF THE INVESTIGATION Any observations made during our field investigations, whether hydrological, pastoral, agricultural, or veterinary, must be related to average rainfall conditions, and the effects of 7 extra high or low rainfall noted. So that this can be done, we have tabulated the aCFual rainfall figures for the years of investigation (Table 5, p. 51), for comparison with the average, separated into different topographical units. Average conditions occurred in at least one of the four years in each unit and, with a ratio of high to low figures 11 to 9, it cannot be said that our observations are stressed too highly in either one direction or another. TEMPERATURE Temperature data for three typical stations, Juba, Malakal and Renk, are given in dia- grammatic form (Fig. B 2). A study of the diagrams reveals the influence of three factors, geographical position, seasonal occurrence of rainfall, and seasonal air movements, as will be explained below. The geographical position causes temperatures and temperature ranges to fall from north to south, as one would expect. However the variations of these ranges are comparatively small and rarely greater than 5° Centigrade, so that the Jonglei Area, as far as temperature is concerned, can be considered a climatic entity. The general shape of the curves for the three stations shows the predominant influence of rainfall, and the double minima at Malakal and Renk together with the single minimum at Juba show clearly how the effect of the cold north wind decreases with latitude. To complete the picture of temperature as an ecological factor, the highest maximum and lowest minimum temperatures recorded at these three stations are worth noting: STATION HrGH1!ST0 cM. AxIMuM I LoWEST M!NJMuM °C. Kosti (Rabak) 44·1 (April) 10·1 (February) MalakaI . .. 43 ·6 (March-April) 11 ·3 (January) Juba 43·7 (May) 13·2 (November) HUMIDITY As one expects, the seasonal occurrence of rainfall and air movements are the main causes of changes in relative humidity (Fig. B 3). The influence of rainfall is evident throughout the area, but the effect of the north wind disappears somewhere between Renk and Malakal. This can be deduced from the absence of a secondary minimum relative humidity at Malakal, which can be found both at Renk and Kosti. The wide range of relative humidity throughout the year is noteworthy, parti~ularly in connection with evaporation and transpiration, which in turn reflect on the soil moisture status upon which the behaviour of crops and pasture depends so much. INSOLATION The quality and quantity of insolation are important in affecting plant growth directly through photosynthesis, and indirectly through their effect on soil moisture. Unfortunately suitable data are lacking in the Jonglei Area, with the exception of twice-daily recordings of cloud density. These, though useful in giving a general idea of seasonal changes in the amount of solar energy available, are less suitable for exact calculations of its physical and biological effects than direct measurements of light intensity and hours of sunshine. The cloud density data are given graphical form (Fig. B 4). The general increase in cloudiness and decrease in seasonal variations from north to south should be noted. Apart from light, the length of day is an important ecological factor which affects the behaviour of plants and of some animals. The variations in the length of day are given here for three typical stations: LENGTH OF DAY Station Latitude Variation June December Kosti 13° IS' 12 hr. 54 min. II hr. 22 min. I hr. 3~ min. Malakal 9° 33' 12 hr. 39 min. II hr. 37 min. I hr. 0" min. Juba 4° 51 ' 12 hr. 20 min . II hr. 56 min. o hr. 24 min. AIR MOVEMENT The movement of the main air masses, apart from being a major factor in forming climate, has a considerable physiological effect and also influences the distribution and configuration of 8 plants. Th¥ force of the wind is important as well as its direction. Wind velocities for three typical stations are given in graphical form (Fig. B 5). The velocities increase from south to north, and the effect of the equatorial rain distribution can be seen on the curve for Juba. 3. HYDROLOGY MAIN RIVERS The survey of the hydrology of the main rivers can be conveniently recorded under the following main divisions: The Bahr el Jebel from Nimule to Mongalla. The Bahr el Jebel from Mongalla to the Tail of the Swamps, including the Bahr el Zeraf. The White Nile from the Tail of the Swamps to Kosti. The Bahr el Ghazal and the Sobat. This brief survey is based on the records of the Nile made by the Egyptian Irrigation Department in the Sudan, which were begun in the south as early as 1905, and on other publica- tions (')f the Egyptian Ministry of Public Works. We begin by describing the characteristics of the rivers, and then explain their importance in relation to grazing and the physica:l processes which result in green grazing on the river in the dry season, the most important aspect as far as the effects of the Project are concerned. THE BAHR EL JEBEL FROM NIMULE TO MONGALLA Strictly speaking the information in this section should apply only as far as Rejaf or perhaps Juba, but as Mongalla is the key point it makes a more convenient dividing line. The important features of the Bahr el Jebel in this reach are: first, there is a steep average slope of 1 mjkm.; secondly, as a result, the river is swift-flowing and confined to a definite narrow channel for most of its length, and swampy flood-plains are not found except north of Juba; thirdly, numerous torrents join the river between Nimule and Mongalla on both banks, the largest of which are the Assua, the Kaia, the Kit, the Vma and the Luri. The catchments of these torrents lie within the limits of latitude of the river itself, with the exception of the Assua catchment in V ganda. The Assua is joined by a local torrent, the Ateppi, a few kilometres from its junction with the Bahr el Jebel. The intensity of rainfall in this latitude has the equatorial double maximum, and consequently the torrents have two maxima (Table 28, p. 63, and Fig. C 1), in May and August. The flow reaching Mongalla is therefore made up of two components, one from Lake Albert and one from the torrents. The former, in a normal year, varies only from 68 to 56 mid, and there is fair correlation between flows in successive years: these are features which one would expect from the stabilising effect of the two great lakes, Victoria and Albert. The latter component depends on the rapid run-off from local torrents which, in turn, depend on local rainfall." In a normal year the flow of the torrents rises from nothing to 29 mi d, and in successive years the flows bear no relation to each other. For instance in the five-year period 1915-19 the total flow at Mongalla averaged 110 mi d, while the torrents reached a monthly mean discharge of 100 mi d in September 1916. In 1918, however, the torrents never exceeded a monthly avera~e of 12 mi d. Both the highest and lowest recorded flows of the torrents occurred within three years of each other. A detailed description of the principal torrents and the investigation of dam sites on them is given later (Vol. II, pp. 359-62). Any schemes for control of the 'Nile must include control of the lakes' discharge, with adequate storage for flood protection and a reservoir (or reservoirs) sited in this reach (or on the torrents themselves) to control the torrents so"as to impound them directly as far as possible. THE BAHR EL JEBEL FROM MONGALLA TO THE TAIL OF THE SWAMPS, INCLUDING THE BARR EL ZERAF A full technical summary of the hydrology of the swamps is unnecessary. We must describe the characteristics of the river briefly, and in particular draw attention to those points which affect livelihood, particularly the vegetation on which the present pastoral economy of the inhabitants largely depends. The Sud!! region proper can be considered as beginning at Mongalla, the key gauging point. The end of the Sudd region is the line of the White Nile from Lake No to the mouth of the Sobat. The river enters this region in one channel, and leaves it in one channel. In between it spreads out like a delta, forming vast swamps of papyrus and other vegetation which is described in detail in later chapters. The White Nile from Lake No can be considered as the channel which collects the discharge from the ends of the swamps and leads it to the one and only outlet to the north-the White Nile at Malakal. 9 In a normal year 27 milliards passing Mongalla are reduced to 23·4 milliard~ at Jonglei, and to 14·3 milliards at the Tail of the Swamps (Tables 29-3"1, pp. 64--6, and Figs. C 3-C 5). This last quantity is obtained by subtracting the flow of the Sobat, measured at its mouth, from the total flow at Malakal, and is thus a figure for the Tail of the Swamps' discharge which includes the Bahr el Jebel, the Bahr el Ghazal (only a fraction of a milliard in a normal year) and the Bahr el Zeraf. The main effects of swamps are to delay the passage of water and to facilitate the evaporation of much of it. Consequently the size of a swamp can only be in- creased by a long period of high discharges and reduced by continuously low ones, though the process of drying up is the quicker. For this reason the average of several successive high and low years gives a better indication of the physical conditions in and around the swamps than one particular year. So we have tabulated the averages for the five-year periods 1915-19 for high discharges, and 1921-25 for low ones (Table 29), and have followed them through to Jonglei (Table 30) and to the Tail of the Swamps (Table 31). From Malakal northwards the variations of the river depend on the Sobat, which flows directly from the Ethiopian mountains without passing through large lakes, so that five-year averages would not illustrate any important point concerning the White Nile. If we examine the flow into and out of the swamps between 1915 and 1919 (Table 32, p. 67) we see that as the discharge at Mongalla increased, so a smaller percentage was discharged at the Tail of the Swamps, until 1918 when at Mongalla the flow was less than in 1917, while at the Tail of the Swamps it was greater. (The flow of the Bahr el Ghazal could hardly affect this since it rarely exceeds three-quarters of a milliard.) This illustrates the delaying action of swamps, and probably the intense saturation in 1917. Similarly, in 1921-22 and in 1924--25 the flow in the second year of each two-year period appears to have been affected by the flow in the previous one. The great flood of 1916-17 was, of course, disastrous, and the five-year average of 110 mid is an indication of what to avoid in the future. Even so it is probable that part of the flood-plain was exposed and some grazing was available in the dry season since the average discharge at Mongalla varied from 150 mid in October to 85 mid in March. What happens in the swamps themselves can best be understood by referring to the plan of the Sudd region (Fig. C 2) and to the several cross-sections of the Bahr el Jebel (see list of figures, Table 2, p. 49). The river goes past Mongalla in a single channel. Up to a flow of 65 mi d at Mongalla the river runs in defined channels and the water passing Mongalla reaches Bor without excessive losses. At higher discharges the river spills into a series of valleys, or flood-plains, which border it between Mongalla and Bor. Typical of such valleys is the Aliab which terminates at Lake Papiu where water spilt into the valley returns to the main river. The losses between Mongalla and Bor are not as great, even at high levels and discharges, as farther north because the extent of the swamps is limited by high ground. On the east bank high ground ends 15 km. north of Bor (612 km. from Lake No), whereas on the west bank it can be traced as far north as Lake Nuong (356 km. from Lake No). Such a difference is to be expected when one considers the differences in topography between the east and west banks of the river (see pp.2-4). Between Bor and Lake Papiu three spill-channels lead water into the swamps of the Atem- Awai system. At high discharges, beginning at 60 mi d at Mongalla, the level of the Atem is about one metre above the plain east of Jonglei, where spill occurs which never returns to the main river under normal conditions, since there is no hi~ ground to contain it. At Jonglei the main flow is carried in parallel channels in the Bahr el Jebel and in the eastern and western channels of the Atem. The total sum of the discharges at Jonglei, known as the , Jonglei latitude flow', bears a direct relation to the discharge at Mongalla at all stages, and all the flow can be accounted for except what spills on to the Jonglei plain. At Jonglei latitude the channels are not affected by the backwater from the tail of the Awai system where the flow in the Atem-Awai swamps returns to the Bahr el Jebel, so the discharges there can be related to gauges at Jongiei and at km. 522 on the Bahr el Jebel by normal gauge-discharge curves. Similar stable conditions exist at Peake's Latitude where the swamp flow is the sum of the flow in the Bahr el Jebel and in the Bahr el Zeraf downstream of the second Cut. At the downstream ends of the depressions bordering the river, where flow returns to the Bahr el Jebel, there are backwaters and lakes of open water. The river runs on a natural deltaic embankment at higher levels than in the depressions and swamps, and flow can only return from these after the level has built up to the level in the river at the point of return. This is clearly illustrated by the cross-sections and contour plans of the Aliab and Mongalla- Gemrneiza valleys (Figs. H I to H 18). The plan (Fig. C 2) also shows clearly the principal spill-channels, channels by which flow returns, and backwater lakes. 10 Between Mongalla and Bor the river spills equally over both banks (Fig. C 10). Between Jonglei and Peake's Latitude, htnd levels suggest that 25% of the spill gO€S to the east, in the first 100 km. north of Jonglei, and 75% goes to the west, north of Lake Nuong where it percolates towards the Bahr el Ghazal. North of Peake's Latitude the river spills over both banks, the larger portion probably going to the west. Buffalo Cape is really the end of the Bahr el Jebel system. The discharge then~ is practically constant whatever th€ :fj.ow at Mongalla. There is no lower limit to the discharge there, however, b€cause the swamps function only to a limited extent as a reservoir, and the flow cannot be maintained for long after the natural supply fails. The backwater effect of the Sobat reaches as far as Buffalo Cape on the Bahr el Jebel and Fangak on the Bahr el Zeraf; so river levels and losses north of those points depend almost entirely on the height of the Sobat. The total area of the swamps, based on the air survey maps of 193oding was known to be extensive in those years as well. Since the western part of the plain is subject to flood from the river, to complete the picture we give below discharges oftlhe Bahr el Jebel at Mongalla in the years under consideration: YEAR 1 1912-42 1 1948 1949 1950 1951 AVERAGE Discharge in milliards 27·0 31·92 26·68 22·35 18·98 1 Overland flooding, or ' creeping flow', on the Eastern Plain was worst when the local rainfall was heaviest. River spill, or torrents' run-off, cannot travel far unless assisted on its way by heavy rain. Turning to the northern part of the plain (Fig. D 28), we see that there are three main discharge systems east of the Bahr el Zeraf: Khor Atar, Khor Fullus and its tributaries, and Khor Geni. In their upper reaches they are wide and shallow and heavily grassed. They drain the ' creeping flow' area, and as one proceeds northwards, so the spread of the flood decreases and eventually becomes channelled. All these watercourses usually dry out in the hot weather, except for the deeper parts of the channels and several large pools such as at Thul. KHOR ATAR This is a main drain•a ge-channel running down the depression on the east side of the Duk Ridge, where it is called Khor Chieth. It has no tributaries, and can be seen from the road as far south as Duk Faiwil. The map of the Eastern Plain (Fig. D 27) compiled from air photographs shows that this drainage depression can be traced as far as Pengko. KHOR FULLUS Khor Fullus has four tributaries which converge into one channel south of Ful Turuk. They are Khor Kwanjor, the upper Khor Fullus, Khor Maadin, and Khor Thul. They begin as mere traces in the' creeping flow' area and are more defined farther north until they become depressions about one kilometre wide. Typical sections are shown in Figure D 29. Between Nyerol and its mouth the Khor Fullus flood-plain narrows down to not more than 200m. KHOR GENI Khor Geni is similar to Khors Atar and Fullus except that it runs eastwards into the River Pibor. It was thought to hold water all the year round, but in 1951-52 it dried out like the other channels. Discharges are measured by the Egyptian Irrigation Department on Khor Fullus at its mouth (0'145 milliard in a normal year), and at the mouth ofKhor Geni. The flow at the latter 21 amounts to practically nothing because in July, August, alld September the flw is in an upstream direction. It should be noted in connection with the excellent grazing grasses to be found in them, that the gauge at Pibor Post at the mouth of the Lotilla and Veveno normally registers for 4 months from July to October. THE ZERAF ISLAND Finally, inoluded in the South-Eastern Area though actually a separate hydrological unit, we come to the Zeraf Island, the triangle of country enclosed by the Bahr el Jebel, the Bahr el Zeraf, and the White Nile. A map of the inland water-systems of the Island has been sketched from the air survey maps (scale 1/100,000) prepared by the Sudan Survey Depart- ment (Fig. D 30). The southern apex of the triangle is extremely swampy and there are very few permanent habitations there now. In the chapter on the inhabitants (p. 207) the migration of the Gaweir Nuer from the neighbourhood of Adok first to the eastern side of this area and later to the east side of the Bahr e1 Zeraf is described, and from this it seems possible that the country was formerly much drier than it is now. Much of the present flooding is said by the inhabitants to date from the construction of the Zeraf Cuts, joining the two rivers together, in 1910 and 1913. Though the Cuts have had no significant effect on the discharge at the Tail of the Swamps, it is possible that they increased spill from the Bahr el Zeraf north of them, and decreased it in the Bahr el Jebel between them and Adok. If this were so, one would expect to find evidence of more swampy conditions near Adok before 1910. Sir William Garstin,(8) in his report of 1904, mentions that Captain Gage of the 7th Dragoon Guards followed the course of Khor Waard for 64 km. (one assumes in a boat) where he was stopped by sudd. This certainly seems to confirm the supposition, though when Sir William observed the channel it was 66·5 m. wide and had an average depth at low water of one metre, which is not very different from present conditions. It will be seen from the map that the country of the Zeraf Island is intersected by several large watercourses crossing it from west to east and from south to north. The cross-section ' eastwards from Buffalo Cape, which follows the road (Fig. A 19), shows how extremely flat the country is in this region. It is interesting to note that the high land where most of the villages are to be found, the so-called backbone of the Island indicated on the map by the motor road from Khor Wang Gai in the south to Khor Yirr Gwol (Yergo1) and Wath Kech in the north, in fact follows the main drainage-channel. This illustrates one of the fundamental features of life in the clay plains. The banks of the watercourses are of alluvial formation, and provide the high ground for roads and villages, and at the same time the channels ensure sufficient local drainage during the rains for habitations and crops (see cross-section Khor Nyin Yar, Fig. D 31). Between 1945 and 1948 the Egyptian Irrigation Department collected records from gauges set inland on Khors Liet, Longtam, Yirr Gwol (Yergol) at Twaibor (Jwaibor), and Khor Ban. We have plotted the reduced levels of the water surface in the khors inland, together with the reduced levels of the rivers (Bahr e1 Zeraf and White Nile) at the points where the khors join them, and these are also shown on the map. If we consider the diagram of levels for the Khor Liet gauge (Fig. D 32) and the Bahr el Zeraf gauge at R.P.29, it is difficult to draw any inferences. Neither rainfall nor the river level appears to have any consistent effect on the inla£d gauge-readings, though at times the latter seem to respond to both. This is probably because the connection between the Bahr el Zeraf and the khor at the gauge is somewhat obscure. The gauges at Khors Longtam and Ban are not far inland, and these khors are directly connected to the river at their mouths by sizeable channels. During the rains the level inland rises with rainfall (Fig. D 33), showing that these khors provide natural drainage to the river. In the dry season the inland and river water-levels almost coincide, showing that backwater from the river maintains the level inland. Rain-water flows to the river in the rains, and river water flows inland in the dry season. The first movement provides natural drainage, and the second natural irrigation of khor-bed grazing grasses. The Khor Yirr Gwol (Yergol) gauge-readings (Fig. D 34), though only for one season, confirm this two-way natural function of these inland watercourses. . THE SOUTH-WESTERN AREA (Fig. A 30) The inland water-systems of the South-Western Area comprise local torrents which join the Bahr el Jebel, such as the Kavu, Kaia, Luri, Koda, Khor Gwir, and Tapari, one river, the Yei or Lau, whose destination is doubtful, and others which are tributaries of the Bahr el 22 .1 \ Ghazal an~.the swamps betweell it and the Bahr el Jebel, such as the Na'am, Gulnam, Gel, Tonj, and Jur. TORRENTS JOINING THE BAHR EL JEBEL The high, steep-sided escarpment which flanks the Bahr el Jebel from Nimule to Jubil forms an effective barrier through which only two torrents have been able to penetrate. A few kilometres north of Nimule we find the Kavu, which hardly deserves mention, draining a small catchment of about 250 sq. km. into the Bahr el Jebel. TIIE KAIA Farther north, about 60 km. south of Juba, the Kaia, after being joined by its tributary the Kijo, cuts sharply through the ridge in a remarkable trough fault, 30 m. wide and soIlie 50 m. deep with steep, rocky sides, and plunges down over a series of rock bars to the main river. A fuller description of the Kaia is given in a later section of the report (see p. 361). A rough estimate of its run-off can be made from its catchment area, 5,000 sq. km., its average annual rainfall being 1,300 mm. Assuming run-off at 7% we get a figure of 0·455 rriilliard a year, or about one-tenth of the total average flow of all the torrents south of Mongalla. TIIE LURI AND THE KODA These two rivers have many similar features. The Luri rises near Loka, some 130 lan. from its mouth which is 6 km. north of Juba; the Koda rises some 40 or 50 km. farther north, passes to the south of Jebel Lado, and joins the Bahr el Jebel opposite to Mongalla, a similar distance north of Juba. Their upper catchments are hilly and forested, but whereas the Luri hugs the western escarpment closely for some way before turning through it towards the river, the . course of the Koda lies well clear of it. The northern parts of their valleys are in flatter and gently undulating country, except for Jebels Kunufi on the Luri and Lado on the Koda. Gfwlogically their catchments consist largely of ironstone and other permeable rocks, and losses by absorption in the river beds are high. This is clearly illustrated by sections of the Luri taken first at its mouth, 30 m. wide and 2 m. deep, where the slope is 63 cm/km., and secondly at the escarpment 70 km. upstream, 35 m. wide and 4 m. deep, where the slope must exceed 400 cm/km. The mouth of the Luri was partly surveyed in April-May 1952 as far as time and the river, which was flowing, allowed (Figs. D 35, D 36). It will be noticed that the water slope to the river, 63 em/ km., is high by Sudd region standards. The alluvial nature of the river ban'k and the wide, flat ffood-plain is clearly shown on the cross-sections. KHOR GWIR This is another small local torrent, rising some 60 km. inland from the river at the latitude of Jebel Lado, and flowing about 140 km. to join the flood-plain of the Bahr el Jebel in the middle of the Aliab Valley south of Minkaman. Since its course lies almost entirely in a clay plain, its run-off is probably a highoc percentage of precipitation than in the case of the Luri or of the Koda. THE TAPARI The Tapari is the northernmost of the local torrents to join the Bahr el Jebel, is con- siderably larger than those described above, and is therefore of greater importance. Its source is in hilly country some 80 km. east of Juba. For the first 100 km. of its length it runs in a strip of a valley and has only one sizeable tributary, the Kifu. Farther north it is joined by the Witu and Bagilogo, which run in from the west and drain the hills south of Amadi, and by many small watercourses. It is crossed by the Amadi-Terakeka road 8 km. above the Bagilogo eonfluence and by the Tali Post-Terakeka road 40 km. downstream of the confluence, where the Tali joins it. Near Tali Post its bed width is between 25 and 30 m. and it is about 3 m. deep. These dimensions do not differ appreciably from those at the 5th parallel N . near its source and at Pap near its mouth. On the other hand the ground on the sides of the river channel varies considerably. In the south the Tapari lies in a basin 95 m. in width ; near Tali Post it spills on to a toich or flood-plain, one kilometre wide; and at Pap itself, where the river winds througb 23 r, ) a flat clay plain, it does not spill, though toiches upstream and downstream of Pap are heavily flooded in the rains. The tributary, the Tali, is typical of watercourses in the plain, being wide, shallow (200 m. wide by 0·80 m. deep) and heavily grassed. The survey has brought out the importance of the Luri, Koda, Khor Gwir, and the Tapari, since they are streams which irrigate part of the Bahr el Jebel toiches. The next step is to erect gauges on th€m. This can be done conveniently on the Luri and the Koda where the Juba-Terakeka road crosses them, and on the Tapari at Pap. It will be difficult to find a suitable gauge reader for Khor Gwir, as it runs through sparsely inhabited wooded country near its outlet. The estimate of the probable run-off of these streams, 0·76 milliard in a normal year (Table 43, p. 76), based on catchment areas, rainfall, and percentage run-off, does no more than help us to compare one with the other, because it is difficult to assess exactly where th€ useful catch- ment ends and where spills take place. Percentage run-off is also a matter of judgment. As with torrents as a whole, it is probable that the flows of these four vary from one-third of the average to two and half times the average. NORTHWARD-FLOWING RIVERS We now turn our attention to those important rivers which feed the swamps between the Bahr el Jebel and the Bahr el Ghazal. Gauges and discharges are given in Tables 44 to 49 (pp. 77-82) and provisional gauge-discharge diagrams in Figs. D 43 to D 50. These four rivers have received the attention of the Team since the First Interim Report (1946). It was pointed out there that run-off from them was estimated at 6·4 milliard-a far from negligible quantity-and the suggestion was made of controlling them to produce artificial pastures on an ample scale to remedy losses on the flood-plains of the Bahr el Jebel resulting from the Equatorial Nile Project. In the Second Interim Report (1947) a thorough statistical analysis proved that the level of the Bahr el Ghazal at Meshra er Req was not affected either by spill from the Bahr el Jebel at Shambe or even by the size of the flood at Tonj. In the Third Interim Report (1948) evidence was given to show that drainage lines exist by which spill from the Bahr el Jebel can travel to the Bahr el Ghazal, reaching it either via Khor Doleib or by another channel south of Bentiu, and in the 1948-49 Progress Report this was confirmed, with further instances of water flowing from east to west, and with notes on depths of water. In this report we have described one of these rivers, the Gel, in greater detail and the others more briefly. This is followed by a more recent estimate of run-off, based on Egyptian Irrigation Department records since 1942, and a sketch map made from air photographs of the network of channels north of the 8th parallel N. (Fig. D 53). THE GEL (MARlm) (Table 47, Figs. D 46 and D 47) Topographical differences between the country east and west of the Bahr el Jebel largely account for the contrasting hydrological characteristics of the rivers. Whereas the south- eastern torrents fan out in clay plains and lose their identity in grassy swamps at about the 5th parallel N., the rivers on the west drain catchments recognizable as valleys in undulating country as far north as roughly the 7th or 8th parallels N. There they encounter clay plains and follow the pattern of the eastern torrents, though the transition from river to swamp is more gradual. This is well illustrated by the River Gel (Fig. D 37). ~he reader will note the long, strip-like character of the catchment area; the upper plateau, the descent to the plains, the head of the delta where spill begins, the widening of the flood-plains, and the river's 'eventual disappearance into swamp. The Gel is some 400 km. long from its source at the foot of Jebel Ambeh on the Nile-Congo Divide to where it merges into the swamps north of Maper, and its average slope is in the region of one metre per kilometre (Table 3). In its southern reaches it is called the Maridi where it flows through the town of that name. Here it is a small and insignificant stream consisting alternately of reaches of papyrus swamp and of rock bars (Fig. D 38). North of Maridi it passes through uninhabited undulating country, up to about the parallel of 6° IS' N. At the Tonj-Mvolo road-crossing the Gel is flanked by a flood-plain about 0·5 krn. wide (Fig. D 39) where spill occurs when the river is at its highest levels in September and October. " The beginning of the flood-plain is thought to be about 50 krn. north of the 500 m. contour, or about 60 km. upstream of the Tonj-Mvolo road-bridge. 40 km. farther north, at the Tonj- Rumbek road-crossing, the flood-plain is 4 krn. wide and is apparently flooded to a depth of 75 cm. at the height of the flood. The river channel there is 20 m. wide with banks 3 m. high. As we follow the course of the river northwards, its delta widens out until it loses its identity in grassy swamps at about the 8th parallel. 24 PLATE I l. The Bahr el Jebel below Nimule: the Fola Rapids. 2. The Bahr el Jebel at the Karpeto Junction. 3. The Bahr el Jebel ; the Makedo Rapids. 4. The Bahr el Jebel near Terakeka: cattle-camp in the background. PLATErTI ------------------~-----------------------------------_---~ I. Flood-plain of the Bahr el Jebel a t low river: near Bor. 2. Flood-plain of the Bahr el Jebel: note cattle grazing in the background. 3. Cattle grazing on typical toich grazing grasses not yet full y exposed . 4. Papyrus fire: Aliab Valley. 5. Aerial view of the White Nile Rood-plain: upstream of the Sobat mouth. PLATE III J. The Kinyeti; at about 7,000 feet; not far from its source. 2. The Assua not far from its junction with the Bahr el Jebel. 3. The River Shylor (Koss) below Toreteinia. 4. The Khor Wol near Akoke. PLATE TV ) .} 1. The steamer Dal with sur- vey party ; Bahr elJebel, 1951. 2. Survey party: Aliab Valley, 195 1. -- 3. Crossing the Ni le (in Uganda). 4. Cross-sect ion in A1i ab Valley, 1951. , . <: 5. Cross ing ~he Khor Wol : Decem ber 1949. " \ t The Egyptian Irrigation Department built a permanent gauge on the west bank of the Gel, 450 m. downstream of the northern road-crossing, in February 1942. The 10-day mean gauge-readings to date are given in Table 47. Since this gauge is inaccessible by road during the rains, discharge measurements are made at the southern road-crossing and water-levels are referred to a datum on the bridge. It is impossible to correlate exactly the gauges at the northern bridge site with the discharges at the southern one, both because tributari€s join the river between the two places and because spill occurs at the highest levels . We have plotted diagrams (Figs. D 46 and D 47) showing discharges at the new bridge corresponding to gauge- readings at the old one, and the gauge-discharge curve at the new discharge site. The first shows a wide range of discharges particularly between readings of 13 and 14 m. on the gauge. The second reveals an interesting change of slope atadischarge of 4 mi d (47 m3./sec.). Referring back to the first diagram we see a close cluster of points at the same discharge which correspond to about 14 m. on the gauge. At the level of 13·80 m. on the gauge there is a very definite muddy high-water mark, indicating a steady level for a long period which could only happen if excesses over a certain amount were spilled on to the flood-plains . In order to arrive at an estimate of the annual discharge of the Gel at the Tonj-Mvolo road, we have had to assume that the mean of the discharge measured twice monthly at the site represents the discharge during the month. The average of the years 1943 to 1950 has been taken. THE RIVER TON] (Table 46 and Fig. D 45) The Tonj is much the same kind of river as the Gel though somewhat larger. It rises on the Nile-Congo Divide 50 km. south of Ibba, by which name it is known in its southern reaches. As on the Gel, there are stretches of papyrus swamp interrupted by rocks and rapids, like those at Ibba where the river is 30 m. wide with banks 3 m. high (Fig. D 38). The Lesi, draining an area of about 1,000 sq. km., joins it on the east about 80 km. north of Ibba. From the Lesi confluence for some 65 km. to Jebel Gumbelli it follows a line of hills on its east bank, the river channel being tortuous and punctuated with rapids. North ofJ ebel Gumbelli the hills give place to undulating and well-forested country. The river is next seen at Tonj, the gauging site, where it is 70 m. wide and 3'5 m. deep in a flood-plain 3'5 km. across. From high-water marks on the gauge it seems that the river spills between readings of 13·2 and 13·5 m. The point on the river's course where spilling begins must be well south of Tonj, possibly 80 km. away at about latitude 6° 40'. At Wan Alel, 40 km. north ofTonj, an ironstone sill is exposed in the river bed; this is 60 m. wide between banks 2 m. high in a flood-plain 3 km. wide. Between Wan AIel and Fanombuil, a distance of 100 km., the flood-plain is very extensive, but it contracts again to a width of 3 km. at the latter place where the river section is 50 m. wide and 2 m. deep. The cross-sectional areas of the river at Tonj and at Fanombuil are roughly in the ratio of 2: 1, and since the slope must be smaller farther north, more than half the flow at Tonj must spill on to the flood-plains between there and Fanombuil. The Tonj eventually loses itself in swamps at approximately the latitude r 40' N., some 400 km. from its source. THE RIVER NA' AM (Table.48 and Fig. D 48) The head waters of the Na'am are its two tributaries, the Era and the Yalo, which rise on the Nile-Congo Divide 50 km. sooth-east of Maridi. For the first 125 km. it runs through hilly country thinly populated. This is followed by 50 km. of undulating country where the river has a sandy bed with occasional rocky rapids, such as those just north of the Amadi- Rumbek road-bridge at Mvolo (Fig. D 39). The Egyptian Irrigation Department gauge and discharge site is nearly 2 km. downstream of the bridge. Between the bridge and the gauge the river cascades through an astonishing collection of large boulders and massive crags of rocks. Immediately below, it flattens out, and, although it still runs between banks 3·5 m. high, the ground alongside is flat and grassy and appears to be a flood-plain on to which the river spills at the highest levels. This is probably the point at which the spill begins. At the Rumbek-Yirol road-bridge 90 km. farther north, the size oC the river has decreased from 35 m. wide by 3·5 m. deep (at Mvolo) to 23 m. wide and 2·5 m. deep, indicating considerable spill between the two places. Here the flood-plain is 5 km. wide, and is sometimes covered to a depth of over one metre. It is not known exactly how far the Na'am continues as a definite channel before dis- appearing in the swamps, but it probably starts breaking up into minor channels at about the same latitude as Lake Nyubor. In the winter of 1950-51 a new bridge was erected over the Na'am in place of the ferry on the Rumbek-Yirol road. Considerable ingenuity was shown by the District Commissioner 25 • f r ) in building, with local materials and labour, massive timber "cribs filled with stone to act as piers and abutments. As there were large borrow pits near the river from which spoil for the approach embankments was excavated, the opportunity was taken of digging test pits in them to find the level of the water-table. Levels were taken on 22nd January 1951 when the river was almost dry. The water-table rose underground to one metre above the river bed at 60 m. from it and thereafter fell at a steady slope of about 1 in 100 (Fig. D 40). At 160 m. from the river a test pit was dug to a depth of 3·5 m. and failed to find water. The soil consisted of one metre of clay overlying coarse sand. The water-table was thus clearly out of reach of the roots of existing grasses and the porous nature of the sub-soil possibly accounts for the quick drying out of vegetation on the surface. THE RIVER LAU (YEr) (Table 49 and Figs. D 43, D 44, D 50) The Lau, or Yei as it is called in its upper reaches, begins in the Bala Hills on the Nile- Congo Divide near the source of the Kaia. If Lake Nyubor is considered to be its tail, it is nearly 400 km. long. Northwards from Yei for about 100 km. there are reaches where the river is comparatively slow-flowing, interrupted at first by shallow falls and later by two waterfalls each about 6 m. high with navigable reaches in between. For the next 100 km. to Mundri the Yei passes through a succession of rapids, which continue beyond that point but with less frequency. In its upper reaches the river is 40 m. wide and 3 m. deep (Fig. D 41). At Mundri the section is 75 m. wide with banks 3·5 m. high. At Koli, 40 km. north of Mundri, the Yei reaches the plains, and some 30 km. farther on crosses from Moru to Dinka country where it is called the Lau. From Koli onwards the river runs in a flood-plain where spill occurs at high levels until Lake Nyubor is reached, where it disappears in lake and swamp. Between Mundri and Koli there is little change in size, but once the plains are reached and the slope decreases, its width increases to about 100 m., only to decrease again northwards as spill reduces the flow in the main channel. At Allel, about 17 km. northwest of Yirol, the river runs over an ironstone sill 50 ill. wide. The cross-section of the flood-plain from high ground one kilometre west to a point one and a half kilometres east of the Lau is shown in Fig. D 42. Its deltaic nature is plainly shown; as· are the minor side channels which distribute the flow over it. The flood-plain is altogether 6 km. wide. GAUGES AND DISCHARGES An accurate estimate of the run-off of the Tonj, Gel, Na'am, and Lau is of great importance in order to assess their contributions to the swamps between the Bahr el Jebel and the Bahr el Ghazal which will be deprived of Jebel spill in the future. Previous estimates based on catch- ment areas and rainfall have arrived at the figure of 6-4 milIiards (The Nile BaSin, Vol. Y, p. 173). Since 1942 these rivers have been gauged and their discharges measured by the Egyptian Irrigation Department in the Sudan, and it is to the Inspector-General, Southern Nile, Malakal, that we are indebted for the basic data of a new estimate. But we must first say something about gauging sites. It will have been noted that at Tonj, where the gauge is built, the river spills on to the flood-plain at the highest levels, and it is thought that spill begins some 80 km. farther south. On the Gel, the gauge is some 100 km. north of the point where spill begins and the discharge site is about 40 km. south of the gauge. , The Na'am is gauged and measured at Mvolo which is roughly the apex of its delta, and the Lau (or Yei) is gauged and measured at Mundri where there is no flood-plain, and gauged again at Yirol where the flood-plain is about 5 km. wide. The figure for run-off measured by the flow in the channel of anyone of these rivers must be qualified by reference to the position of the measuring site in relation to the effective catchment and to the delta of the river. The ideal site would lie at the downstream end of the useful catchment, and upstream of the point where spill begins. The hydrology of these rivers and the amount of water reaching the swamps cannot be cOIIllPletely understood and estimated until gauges are erected at many more places along their courses than at present. For instance we have seen that more than half of the flow of the Tonj passing the gauging site must be spilled on to the flood-plain before reaching Fanombuill00 km. downstream. Turning now to the data, we have estimates of the flow of the Tonj at Tonj, of the Gel at the Mvolo-Tonj (or southern) road-bridge, of the Na'am at Mvolo, and of the Lau (or Yei) at Mundri and at Yirol (Tables 46 to 49). In Table 44 we compare the new estimate with the figures given originally in Volume V of The Nile Basin (p. 173). 26.. In the fk"st place it will be no.ticed that the most recent estimate amounts to 3·25 milliards per year (taking the Lau at Yirol) as against 6-4 milliards previously, or roughly 50%. This quantity is measured approximately on the line of the Wau-Tonj-Rumbek-Yirol road. Com- paring the measured discharg@s of the Lau (Yei) at M undri with the discharges at Yirol estimated from 10-day mean gauges, it is clear that 0·61 milliard, or 37'5%, is lost between the two points, a distance of 130 km. (Table 49 and Fig. D 50). Since the gauging points are roughly a similar distance south-west of the swamps between the Bahr el Jebel and the Bahr el Ghazal, and since the deltas of the four rivers becom€ wider and spill increases progressively as the rivers proceed northward, the amount of water reaching those swamps must be a small fraction of the 3·25 milliards measured at the gauging points. Normal discharges of the south-western torrents are illustrat€d in Fig. D 52. In order to complete the picture in connection with the Bahr el Ghazal we have included the most recent estimate of the discharge at Wau of the River Jur, which is directly connected to it (Table 45 and Figs. D 49 and D 51). This amounts to 4·5 milliards in an average y€ar, as against a previous estimate of 5-4 milliards. Of this about 3-4 milliards, or 75%, is thought to be lost upstream ofWangkai (Ghabat el Arab), as stated in The Nile Basin (Vol. V, p. 181). From this survey it is perhaps a little less surprising that the flow of the BahT el Ghazal at Lake No is only 0·65 milliard in an average year. Some of this must be due to spill from the Bahr el Jebel, because at Khor Doleib some 30 km. from Lake No the mean annual discharge is only 0·132 milliard. THE WESTERN PLAIN This area has been known in previous reports as the' Eastern Ghazal Swamps '. We summarize the results of previous investigations and then bring the information up to date. Broadly speaking, this interesting area is flooded heavily by spill from the Bahr el Jebel, by the northward-flowing south-western torrents, and by spill from the Bahr el Ghazal. In years of high flood the swamps formed from these three sources must mingle together and form one vast swamp. An Egyptian Irrigation Department survey party, levelling westwards from Adok, after travelling 80 km. was still 1·70 m. below high-water level in the Bahr el Jebel. Though river levels at the time were high enough to induce spilling, the plain was dry except in khors and no flow of water was actually observed. But this fact alone proves that topographically it is at least possible for Bahr el Jebel spill to reach the Bahr el Ghazal. Statistical correlation of gauges and discharges proved the following: (i) The Meshra gauge is closely related to the total annual flow of the Jur at Wau. (ii) Since there was no significant correlation between Shambe and Meshra gauges, Jebel spill does not flow to Meshra in any appreciable quantities. (iii) Neither does the Tonj flood have any appreciable effect at Meshra. (iv) The heig,h t of the Meshra gauge does depend on its previous minimum. . The topographical information is collected and shown on the .map of the inland water systems on the west bank of the Bahr el Jebel (Fig. D 53). This map is largely constructed from the (American) air photographs north of Maper. The important hydrological points to note are first that 6·0 milliards of Bahr el Jebel spill, it is estimated, flow westwartls between Lake Nuong and Buffalo Cape, divided into 4·6 milliards between Lake Nuong and Peake's Latitude, and 1·4 milliards between Peake's Latitude and Buffalo Cape. Secondly, the rivers ~onj, Gel, Na'am, and Lau add to this quantity a fraction of 3·25 milliards in an average year, that figure being the discharge measured on the line of the Tonj-Rumbek-Yirol road. Though there is no large channel between Shambe and Khor Waard at Adok carrying water westwards, it is known from administrative records that at one point on the road between Ler and Shambe there is a 20 km. stretch of swamp. It is not easy to see exactly where the enormous volumes of spill from the Bahr e1 Jebel go to, but the losses on the Bahr el Jebel are real enough and we can only assume that transpiration and evaporation from the permanent swamp and spill westwards, which is later channelled into Khors Chier and Neang and thence to the Puragwier and Bilnyang, account for them. A cross-section of Khor Neang at Dorthir is given ill Fig. D 54. For example in February 1950, on the Bilnyang where the Ler-Wun Shwai road crosses it, a large area of Hyparrhenia rufa grass was seen still under water. This point is at least 100 km. from the nearest source of Bahr e1 Jebel spill. The total annual discharge of the Bahr e1 Jebel measured at Mongalla was 27 milliards in 1949. Rainfall at Ler that year was very heavy, though normal over the catchment of the southern torrents, and had ceased by the middle of October. Four months later and 100 km. away from the river, the ground was covered with water over a large area. No doubt the spread of river spill was assisted by the abnormally heavy rainfall, but it confirms the direction which river spill takes, 27 and we must assume that in other years, when rainfall is light or normal, evapbration and transpiration account for most of the losses on the Bahr el Jebel. As an illustration of a typical watercourse in the Western Plain we have included the cross-section of Khor Tiak, a continuation of Khor Waard, at Chotyil (Fig. D 55). KHOR MOleR The American air photographs do not cover a large area between the 7th and 8th parallels, and the area east of the 30th meridian remains a blank on the map. The only information to be added to that given in the Second Interim Report concerns Khor Moich. In the middle of March 1951, while on a reconnaissance flight, we followed the course of Khor Moich from Lake Nuong towards Lake Nyubor. The khor is clearly defined, narrow and not grassed, to within 40 km. of the latter. There it splits up into several small channels which eventually disappear, though something in the nature of a depression all the way to the lake was indicated by the deeper green of the regrowth of grass. It is very unusual to observe such a defined channel in the swampy area, and one must conclude that the slope from Lake Nyubor to Lake Nuong is steep by Sudd region standards. Therefore the possibility of levels in Lake Nyubor being affected by levels in the Bahr el Jebel at Lake Nuong can almost certainly be ruled out. Although the channel ofKhor Moich is not continuous between the two lakes, flood-water from the Lau almost certainly reaches Lake Nuong. From there northwards the direction of flow as indicated by channels and slopes is parallel to the Bahr el Jebel, swinging westwards at the 8th parallel N . The channels of the Moich are deep. In April 1949 on a trek from Shambe to GaingW, members of the Team had to swim to cross it. It was then flowing towards Lake Nuong. In December 1952, the Assistant District Commissioner, who was bridging Khor Moich near its mouth at the site of a new road trace from Gainglil to Shambe, reported that it was 57 m. wide. In 37 m. the water was knee-deep, and from 2·5 m. to 4·25 m. deep in the remaining 20 m. It was floy.'ing strongly towards the Bahr el Jebel. In March 1953 the water was stag- nant, 15 m. wide and } ·o m. deep. THE NORTHERN AREA EAST BANK OF THE WHITE NILE We now come to the area north of the line of the Bahr el Ghazal, the White Nile from Lake No to the mouth of the Sobat, and the Sobat itself. This can be sub-divided conveniently into areas east and west of the White Nile. Reference should now be made to the map of inland water-systems in the Northern Area (Fig. D 56). Beginning at the south-east corner of this area, the first and most important inland water-system to note is Khor Machar and the swamps associated with it. It forms the subject of a separate special investigation, the report on which will be found in Volume III to which the reader is referred for fuller details. THE MACHAR MARSHES (Fig. K 15) The Machar Marshes are supplied by spill from the north bank of the Baro, of which a quarter passes through the Khor Machar; by the eastern torrents (described fuIIy below, pp. 31- 2) such as the Daga and Yabus draining the Ethiopian plateau; and by direct rainfall. They may also be supplied by spill from the Sobat. The average yearly quantities supplied by Baro spiII, by the eastern torrents and by direct rainfall ar'e of the order of 3, 2 and }5 milliards respectively, with a total annual variation from the mean which may amount to over 3 milliards. On the average, between September and February the Wakau takes out of the marsh about a quarter of a milliard; but apart from this there is in most years no evidence that any appreci- able quantity of water drains from the marsh to either the White Nile or the Sobat. In 194~7, however, there was definitely a flow into the White Nile between Malakal and Paloich of something between a half and one milliard. In that year the Sobat also received a total of between three and five milliards, though it seems likely from the general slope of the ground that the greater part, if not all, of this came from the southern plains. Since there is in most years very little drainage out of the marsh, it is evident that these variations in the annual supply must cause considerable changes in the volume of water held in it and so in its depth· and extent, particularly when several years of rainfall above or below average occur consecutively. The records of rainfall from seven stations round the swamps indicate that the years 1940-45 were about 12% below and those from 1945-49 about 5% above the normal. The scanty information available from the few officials who have travelled in the swamps confirms that these abnormal series of years caused them to dry up and then to overflow in the way that we should expect. 28 The air. survey maps made from the American air photographs taken in 19~5 indicate that the water passes through tIi.~ swamps by three main 'foutes-from the eastern torrents through the Daga to the Adar; from the eastern part of the Baro through one brancll of th~ Khor Machar to the Adar ; and from the other branch of the Khor Machar and from spill west of it through a long khor called the Tierbor, running parallel with the Sobat, to the Khor Wol and so to the White Nile. This last route connects with th~ Sobat through the Khor Wakau which at first supplies and then partly drains th~ marsh; it also has conn~ctions farther west through the Nyaiyin, Mairik, Loing and Banglai, though it seems libly that the exchang~ through these are very much smaller. The existence and importanc~ of these main routes through the marsh are confirmed by the report of an Assistant District Commissioner who crossed some of the main channels and got local information about oth~rs in 1950. Th~ rate of travel of the floods along these routes appears to amount to about on~ kilometre per day. This rate is obtain~d from the time-lags between both the spill from th~ Baro al'ld the flushes on the eastern torrents, and the reversal on the Wakau each year and the heavy dis- charges into the White Nile in 1947. This rat~ of travel, which is about one-t~nth of that of fluctuations on the Sobat and ten times that of water through the swamps of the Bahr el Jebel, indicates a type of flow intermediat~ between the two. Probably the water advanGes on a wide front flanking the main channel of each route with its surface high~t in the channel, where the grass may be overtopped and also flattened by the flow. The general picture is therefore of an area of swamp which may vary in extmt considerably from one year to another because the varying annual supplies are not l'lormaliy followed by any appreciable drainage from the swamps. The supplies of water from the east and sOl!1th-east travel through the swamps by three main routes, from which they spread out and fill uhe many interconnected khors and depressions. In exceptional years these overflow into the White Nile and possibly into the lower reaches of the Sobat. Developm~nt of the aliea would require the coI1trol of the eastern torrents including the Bam, possibly some banking of the Baro, and the deepening and widening of the main channels through the swamp. Further lines of investi- gation should therefore be directed, em the engineering side, to these ends. Surveys are needed of both the eastern torrents and the Baro to determine the possibilities for storage, and means must be devised for measuring levels and gauging disGharges within the swamps themselves. THE ICHOR WOL INLAND WATER-SYSTEM The Khor Wol system of major and minor watercourses is a typical one am.d will be described rather more fully than most. The d~cription is based on air survey maps, confirmed by an expedition on foot covering the lower reaGhes in January 1950. The main sources of water in the Khor Wol system are (i) swamps north of the Sobat formed by river spill from the Baro and Sobat, (ii) local rainfall, and (iii) backwater from the White Nile. The reader should refer to the map of the Machar Marshes (Fig. K 15). The primary sources Qfflooding in the swamps north ofth~ Sobat which drain to the Khor Wol are spill from the Baro via a westerly branch of the Machar Marshes, and spill from the Sobat through Khor Wakau and possibly through other channels farther west. The swamps resolve themselves into Khors Weikirr and Urel which are the main drainage-channels leading to the lower Wol. There is also a small spill-channel nearer the Sobat mouth, Khor Banglai, which runs only at the highest Clischarges. Khor Weikirr enters Dunjol Dinka country from the Toich Malau (see Ch. 3, p. 218) at Biyunki. It describes a series of rrl'eanders, turning north-west from Biyunki to Telber and north to Temchuk where it joins the Wol which continues to the White Nile. There are two other loops on Khor Weikirr, one near Adoich south of Biyunki, which in fact carries most of the flow, and another at Riakdau farther north. Khor Weikirr was 250 m. wide and 0·8 m. deep on the average in January 1950. . Khor Urel drains westwards to join Khor Wol where it is called Khor Gowding. Between the confluence and Tongkok the country is uninhabited and the khor is 150 m. wide between well-defined banks, and holds water to a depth of 0·9 m. The maximum depth of water encountered during the expedition was 1·2 m. at the Khor Wol-Khor Weikirr junction, where the width was 300 m. Khot Banglai discharges into the Iyidit swamp where it disappears at the end of a defined channel 2 km. from the river. The Iyidit swamp is drained by Khor Gowding which meets the Urel to form the Wol. At the Urel-Gowdingjunction theformer islOO m. wide with 0·6 m. depth of water, and the latter 20 m. wide with 0·3 m. depth of water. The whole country is criss-crossed with a maze of minor depressions which are filled by local rainfall. The following minor interconnections are of interest. Khor Aywal from Mamai on Khor Weikirr runs north to join Khor Wak at Akoke. The latter has two 29 connections to the Wol, one south of Akoke, and the other north of the Malakal-J?:aloich road. Khor Yinemajak runs from Khor Weikirr near T<.- (Flood Soils) E,~" eo • f<.>- ~'" Non-Flooded Soils (J Loam Soils :'>", (of high land) I I: Sandy Soils 8 Clay Soils (For further classification see pp. 116- 18) I ~~-=-Vl';: ~ o::! c: Cl 0 Non-Cracking Sandy-Clay Soils (Gardud Soils) :; IrE Soils of the ~ '.c Semi-Arid Region ,.., 0'" ~-o.g ~x«=·-: <;it: I Eluvial Complex E~ ~ (Semi-Arid Soils) <.>'" :E.g ""'0 f-~ (For further >. o..!!a: (J 0.'" Sandy Soils -=~ e .""." .«u-=: Qoz Catena Colluvial Complex classification see Table 82) " "' 0 o~ ~~-g c.'" ~(J Illuvial Complex ~§e:-! "0", - ~ <.> .... , Permanently Waote~logged Sudd Soils il<§ 09'"'Ectomorphic Soils &: '£'' c: -~ <2 Olayey Toich Soils ,aaDoe nO "0 0t: ~ '.;3 Seasonally F looded Toich Soils E3:6~ § .~ - -------«i'a £,-&(J E-<(J Sandy Toich Soils Inter-Regional Soils :;~ :.is ~~ c;~ Soils of Recent Alluvial Origin (J .~~ Q Endomorphic Soils -@'~.g -- (")'U- Skeletal Soils of Recent Origin TABLE 81 LATERITIC CATENA OF THE IRONSTONE REGION (based on Morrison et al.) Complex Phase I Symbol I Description ELUVIAL Eluvial Deep .. . E/ Dp. Relatively deep uneroded phase of eluvial complex. , A' horizon of reddish, light coloured soil, over , BI' horizon rich in pea-iron, merging into ' B2' ironstone of varying hardness. Eluvial Shallow E/ S. Shallower or partly eroded phase of eluvial complex, with a thin ' A ' horizon above' B, ' pea-iron. Eluvial Shallow-Redistributed E/ S-R. Shallower than E/ S. • A' horizon apparently mixed with ash and redistributed. Eluvial Shallow-Eroded E/ S-E. No ' A 'horizon. 'B, ' pea-iron on the surface. Eluvial (Ironstone) Outcrop ... E/ O. Exposed ironstone of 'B' horizon, or ironstone formed on parent rock. Eluvial-Colluvial Transitional E-Col. Transition between eluvial and colluvial, occurring where sufficient slope and rapidity of erosion have led to the accumulation of an appreciable quantity or ironstone blocks and fragments. CoLLUVIAL ... Colluvial Coarse ... .. . Colje. Coarser; shallower upper zone of the re-sorted colluvial material, light coloured, reddish or pinkish soil. Colluvial Fine I ... ... CoIjF-1 Finer, deeper, lower zone of the re-sorted colluvial material. Light coloured, reddish soil. By far the most extensive of colluvial phases. Colluvial Fine 2 .. . ... CoI/ F-2 Lowest colluvial, some with the finest of the re-sorted materials, occurring oaly under special conditions of topography. Dark coloured, chocolate-brown. Colluvial-Illuvial Transitional Col-II. Transition between CollF-1 and II/ Dr-lor I1/1d-1 in the absence of CoIjF-2. Soil grey-buff, often with rather large, reddish sand grains on its smooth surface. Colluvial-Alluvial Transitional Col-AI. Transition between the lower phases of colluvial complex and the alluvial materials. Coarse sand often mixed with fine material. Occasional signs of bedding. Dark grey or brownish grey. -----I------------I·---I-~------ - -- - - ------ ILLUVIAL Illuvial Drained I I1/Dr-1 Comparatively well-drained upper phase of iUuvial complex on similar topographical level to I1/1d-I but ground is flat not uneven. Brownish grey or grey, rather dark coloured soil, with considerable fine sand fraction. Illuvial Drained 2 I1/Dr-2 Similar to II/ Dr-I but on slightly lower topographical level "nd somewhat less drained. IIIuvial Impeded I II/Id-! Topographically the highest of the impeded phases. Uneven surface and impeded drainage; greyish-buff light coloured soil. IlIuvial Impeded 2 I1/1d-2 Similar to II/ Id-l but on slightly lower topographical level with drainage impedance nearer the surface, which is less obviously uneven. Brownish grey or grey, rather dark coloured soil. JIluvial Impeded 3 II/ld-3 Dark grey, fine textured soil with strongly impeded drainage occurring usually on 'partial basin sites ' . Reaction slightly acid . IIIuvial Vnd'rained II/ V. Dark cracking soil of the basin sites with no drainage. Very dark grey clod structure, reaction alkaline (cf. I1/ Id-3). IIIuvial-AIIuvial Transitional ... II-AI. Mixture of alluvial with iUuvial materials. May closely resemble one of the illuvial phases. Signs of bedding often present. Often texturally mixed. 128 TABLE 82 <:> A TYPICAL QOZ CATENA Examiner: T_ A_ JONES et a/_ Location: Qoz NGYOK ANALYSIS DESCRIPTION I Sand/ Silt + Clay I pH Salts % N p.p.m. ELUVIAL COMPLEX_ Albizzia-Combretum Zone 0-10' Red-brown light sand ; coarse particles admixed; many small surface roots uniformly distributed. Fairly compact. 0"- 6" 7·2 4-71 0-126 432 No iron redistribution. 6--12" 8-9 4-81 0-01S 396 IO-IS- Redder. more compact version of above. Herbaceous roots diminishing in number and large roots of trees fairly 12"-24" 4-8 4-95 0-019 319 abundant. No sign of iron redistribution. Slightly moist. 18-32- Red, compact coarse sand with considerable fine particles . Small channels less visible. More compact still, but 24"-36" 4-8 4 -95 0-020 139 structure uniform. No sign of aggregation. Moist. 32-3S- Somewhat paler, more buff colour, uniform compact sand with much fine material and no signs of iron redistribution. 36"-48" 3-5 5-20 0-021 206 Moist. 38-56- Slightly paler compact sand with fine material. Some white flecks increasing a little after third foot. CaC03? 48"-60" 3-2 6-66 0-034 250 Moist. 56-72- Compact sandy clay, moist and mouldablo. White flecks increasing in frequency and size. Some evidence of tongues 60"-72" 3-2 6-72 0-064 175 of reddish brown colour extending into seventh foot. 72"-84" 3-8 7-50 0-070 190 COLIoUVlAL COMPLEX_ Combretum-Anogeissus Zone ------ 0- 9- Chocolate brown coarse sandy loam with some fine particles. Compact and without structure. Very slight vertical 0"- 6" 6-1 7-12 0-027 323 ...... cracks and fine and coarse roots evident. A few small channels and no sign of iron redistr.ibution. 9-17" Ohocolate brown, coarser, less compact version of above. Slightly moist and a few white flecks 6"-12" 5-2 6-18 0-021 243 N \0 17-27- Red chocolate brown coarse sandy loam with some fine particles. Coarser than upper hori1:ons . Slightly moist 12"-24" 4-5 6-S0 0-022 196 and less compact· than previous horizon, with white and black flecks. Red tint may indicate some iron redistribution. 24"-36" 3-8 6-80 0-032 176 27-34- Clouded mottling on red and grey-brown horizon. Coarse sand, and more compact than above. Moist 34-4S- Greyish brown mottled and more clayey material. Considerable amount of coarse sand present still 48"-60" 3-8 6-92 0-034 16S CoLLUVIAL COMPLEX_ Acacia seyal Zone 0-10" Dark grey coarse sandy clay, almost black. Friable and porous . Many large and small roots. Well-established 0"- 6" 2-0 S-68 0-064 6S7 cracks. No carbonate. 10-24- More compact clay, but still abundant coarse sand crystals. Clod structure well preserved, slightly moist with some 6"-12" 1-2 9-92 0-111 390 carbonate concretions. 24-36- Sandy clay pocket but otherwise similar to above 12"-24- I-S 10-0S 0-171 2S0 24"-36" 1-6 '10-24 0-[91 208 36-4S- Taken by auger. Clayey coarse sand. Moist 36"-48" 2-1 10-24 0-190 154 4S-<>O- Slightly sandy heavy clay. Moist. Fine roots seem to end at about 18* depth and only coarse roots afterwards 48"-60" 0-7 10-14 0-307 ,164 ALLUVIAL COMPLEX_ Acacia mellifera Zone 0-18- Surface partiy worn, quartz pebbles and many gravel pebbles and numerous large spiral shells of land snails. 0"- 6- 0-6 9-74 0-142 No stones or shells in soil itself. 393 6"-12" 0-5 9-85 0-190 317 Black to dark grey uniform heavy clay with some embedded sand grains. CaCO~ eoncretions and specks frequent and of varying size increasing in number to IS". Moist at about 1', but not sufficiently moist to mould. Surface elod structure not particularly cracked, owing presumably to coarse river wash thinly overlying. 24--36- 0-3 9-86 0 -216 377 PART II. VEGETATION OF THE JONGLE! AREA 1. ECOLOGICAL FACTORS GOVERNING THE DISTRIBUTION OF VEGETATION Before any attempt can be made to describe the vegetation of the Jong1ei Area, it is necessary to classify the vegetation into smaller, more manageable groups, based as far as possible on closely related characteristics. A preliminary survey of the Jong1ei Area shows that it is possible to divide the vegetation up into major distinct entities, each entity having its own characteristic cOn;lposition. Within these larger units it is possible, although this is not always obvious at first sight, to make further divisions based on smaller differences in composition. The development of these differences in vegetation is milnating, but has larger areas of Phragmites communis. Behind the flood-plain there may be strips of mixed tall acacia dominated type forests or, as in the area south of Tombe on sandy soil!, palm. forests. Elsewhere and inland of the above two types there is usually a belt of short, open acacia scrub before the mixed deciduous broad-leaved woodland of the true imnstone country is reached. ALIAB AND CHICH DINKA COUNTRY These Dinka tribes occupy the country west of the Bahr el Jebel between Tombe and Shambe. Inland they penetrate as far as the ironstone country with i.ts mixed deciduous broad-leaved type of woodland vegetation. Permanent papyrus swamp occurs between approximately Minkaman, which lies in the Aliab Dinka country, and Shambe. It @ccurs as a strip of varying depth along the river, and in some of the deeper kb,ors of the flood-plain (see Vol. Ill, p. 832). Inland, grassland of the riverain swamp type is found along the River Gel, River Lau and Khor Waat. Bordering the flood-plain between Minkaman and Tombe there is a fringe of mixed tall acacia dominated type forest on sandy soil overlying clay. Northwards and westwards this forest gives way to open acacia scrub which in turn gives way (westwards) to the inixed deciduous broad-leaved type of woodland of the ironstone plateau on which the Chich Dinka have their permanent villages. BOR GOK DINKA COUNTRY High land occurs both north and south of Bor as a narrow ridge just inland from the river. There is also a triangular extension having its base between Malek and a point ap- proximately 15 miles north of Bor and with its tip eastwards towards Anyidi. Three forest types occur on this high land in relation to different soil types. The triangular extension mentioned above, which has soil with a medium clay wntent and is liable to seasonal flooding, is mainly covered by forest of the Acacia seyal-Balanites aegyptiaca type, which spreads on to adjoining intermediate land, thinning out into open parkland between Anyidi and Pengko and eventually into open grassland. Within the forests are Setaria incrassata and some Hyparrhenia spp., with a few areas of almost pure Hygrophila spinosa. South of Malek as far as a point approximately 40 miles north of Mongalla and mainly on the boundaries of Mandari country is palm forest, whtle south of this point the forest changes to the mixed tall acacia dominated type. In the last two forest types, which occur on the sandier soils, Hyparrhenia dissoluta is the dominant grass in the undergrowth. Eastwards of these forests, on the higher levels of the intermediate land, the forest changes to the Acacia seyal-Balanites aegypfiaca type and again eventually gives way to the open grassland of the intermediate land. This open grassland is in all cases of the intermediate land Hyparrhenia rufa type. In the Bor Gok country this stretches as one monotonous and continuous pure stand over the vast area of the Eastern Plain (see p. 261) between Pengko and the River Veveno. It also continues northwards and southwards of the Pengko-Pibor road as far as the eye can see. No grassland of the khor-bed type occurs, and riverain swamp grassland only occurs in limited amounts, especially between Bor and Malek. Here the high land runs close to the vertical banks of the river with a few depressions intersecting it which have a direct connection with the main river. In places where riverain swamp grassland does occur along this stretch, it is predominantly of the deep-flooded Echinochloa stagnina type. Most riverain swamp pasture is in the Aliab Valley on the other side of the main charmel of the Bahr el Jebel, an area shared by the Bor Gok with the Aliab Dinka and one which has been the subject for special investigation by the Jonglei Investigation Team (see Vol. III, p. 832 et seq.). 143 BOR ATHOICH, TWI, NYAREWENO AND GHOL DINKA; GAWEIR (MAINLAND) AND TRIANO (MAINLAND) NUER COUNTRY High land is limifed and is found along two main ridges, the first along the motor road from Baidit through Jalle and Paliau to Kongor, and the second lying about 15 miles east of it along the line Arielbek-Duk Faiwil. Forest is limited to these ridges and the higher intermediate land, and is of the Acacia seyal-Balanites aegyptiaca type only. It is found between the Bor District boundary and Duk Faiwil, and again south of Kongor where the forest opens out into Balanites aegyptiaca parkland. Between Duk Faiwil and Kongor the country is remarkably open with only occasional trees. Throughout the whole country the forest does not penetrate deeply east or west of the high land. Along the high land, besides annual grasses, Setaria incrassata and Sporobolus pyramidalis may be found; they also occur on the higher limits of intermediate land. Intermediate land covers the largest area in this country and the grassland is predominantly of the Hyparrhenia rufa type which extends eastwards as far as the Pibor River, whereas westwards it dominates the country until it is superseded by the riverain swamp grasslands of the Bahr el Jebel. These riverain swamp grasslands which lie behind the wide belt of permanent papyrus swamp actually bordering the Bahr el Jebel are dominated mainly by Echinochloa pyramidalis, except in the lower depressions where Echinochloa stagnina becomes dominant. At Duk Faiwil the series of sandy outcrops often referred to collectively as the Duk Ridge begin and extend over the boundary into Gaweir Nuer country as far north as Mogogh, a distance of about 70 miles. Along this ridge there is some poorly developed mixed deciduous broad-leaved type of forest, as well as palm forests in which both dom and doleib palms occur mixed with Ficus sp. and Tamarindus indica. Alternating with these and sometimes mixed with them are forests of Acacia seyal. East and west of the ridge is mostly open grassland with small forests of the Acacia seyal-Balanites aegyptiaca type, the Balanites aegyptiaca parkland formation becoming cornmon southwards. The country between rangak and Falagh is heavily forested with Acacia seyal, but from there eastwards to Mogogh it is mainly open grassland. Nortli of Mogogh Balanites aegyptiaca parkland is again common. Grassland throughout the area is mainly of the intermediate land Hyparrhenia rufa type, but within the Acacia seyal-Balanites forest there is in some areas, as between Fangak and Falagh, an appreciable amount of Setaria incrassata, while on the ridge Sporobolus pyramidalis is frequent. Along the Bahr el Zeraf and the many inland khors which join it in the upper reaches, there is the usual grassland of the riverain and khor-bed swamp type dominated by the Echinochloas, while along the upper Atar there is poor khor-bed pasture dominated by Vetiveria nigritana. East of the upper reaches of the Bahr el Zeraf, between the river and the higher ground around Awoi, there is a complex network of drainage-channels with Echinochloa stagnina at the lower levels and an abundance of Echinochloa pyramidalis on either side. Above these are areas of those grasses normally associated with intermediate land, Hyparrhenia rufa and also Setaria incrassata, which increase in frequency towards Awoi. There are also small forests of Acacia seyal and a few areas of palm forest (dom and doleib), notably at Rubciengdol, Fashir and along the Zeraf. High land is limited and widely distributed, the main areas and centres of permanent habitation being along the Duk Ridge and at Falagh and A woi. LAU NUER COUNTRY Lau Nuer country is well outside the Jonglei Area since none of its inhabitants are dependent on the flood-plains of the Nile for dry season grazing. It is nevertheless included in this description since it provides an example of a territory in which a much larger proportion of the animal population is able to subsist throughout the year away from the main rivers, though it is true that many herds are taken to the Sobat and Akobo Rivers in the dry season. There is practically no pronounced high land similar to that described in the previous areas, the high land areas being difficult to distinguish from the intermediate land. Forests are therefore of the Acacia seyal-Balanites aegyptiaca type and are found more in the northern part of the area than the southern part, which is mainly open country dominated by the Hyparrhenia rufa type intermediate grassland. In the more northerly parts, i.e. along the Waat-Faddoi-Akobo road, much of the area is open grassland alternating with .Acacia seyal-Balanites aegyptiaca forest. The open grassland is predominantly of. the Iiyparrhenia rufa type with the slightly raised area occupied by Setaria incrassata, interspersed with large areas of Hygrophila spinosa. Both east and west of the Khor Geni, south of the Waat-Akobo road, is a vast open plain of Hyparrhenia rufa with occasional open Balanites aegyptiaca parkland. Along the Khor Geni there is grassland of the khor-bed swamp type, but here 144 confined within the actual course of the river. Elsewhere on the Khor Nyanding and th~ many branches of the Fullus system there are flood-plains of varying width ov€r which this type of grassland dominates. Along some of the smaller drainage depressions, especially if they have a sandy bed, Vetiveria nigritana is common. LAK, THIANG AND GAWEIR NUER OF ZERAF ISLAND This area is bounded in the west and north by the Bahr el Jebel and White Nile and in the east by the Bahr el Zeraf, which together form the wedge-shaped area which is referred to as the Zeraf Island. The high land ridge which forms the backbone of th€ island, and which ruus from Berboi westwards to Wath Kech and then southwards to Malith in the Gaweir country, is similar in both vegetation and its distribution to the Bentiu-Ler ridge described later. Three types of high land forests occur: palm, mixed Acacia, and Acacia seyal-Balanites aegyptiaca (in which Acacia seyal predominates). The latter is found throughout the length of the island on either side of the ridge as far south as Gumbeil. Eastwards it extends almost down to the river, bu t westwards it is replaced after a few miles by open grassland. The major part of this type of forest is found in parts of Thiang country and in Lak Nuer country in the northern half of the island. Here in the northern part the forests continue to the Bahr el Zeraf, to the White Nile and westwards along the Ful Lita-Buffalo Cape road in isolated patches for a considerable distance right up to the banks of the Bahr el Jebel at Buffalo Cape and km. 60. Along the sandier parts of the central ridge, in Gaweir Nuer country, there are strips of mixed acacia forests, whereas farther south beyond Malith there are palm forests occurring on sandy ridges which appear as islands in an endless grass swamp which continues to the river. The distribution of grasses on the high land varies according to the type of soil. On clay soils open areas of Pennisetum ramosum are common; so also are areas of Setaria incrassata. On sandier soils Sporobolus pyramidalis is common and frequently spreads on to the less sandy intermediate land. In addition to Sporobolus pyramidalis, sandy high land carries a variety of small annual grasses. In some areas, particularly those lying between high land and intermediate land and on clay soils, a weed-Hygrophila spinosa-which is objectionable to cattle because of its spines, has become dominant, thus reducing the value of such areas as pasture. The grassland of the intermediate land is either of the Setaria incrassata or Hyparrhenia rufa type. The Setaria type usually occupies a position between the high land forests and the lower areas of intermediate land which carry Hyparrhenia rufa. It occurs widely in the north- eastern part of the island where in some areas it occupies the country from the high land ridge to the river, alternating with either forest or the Hyparrhenia rufa type. Elsewhere it is limited to within a short distance of the high land ridge, except around Gumbiel in the southern part of the island where it extends westwards in the direction of the Bahr el Jebel. The rest of the intermediate land is dominated by Hyparrhenia rufa which extends riverwards in both directions until it is superseded by the swamp grasslands of the flood-plain. These swamp grasslands occupy a position between the intermediate Hyparrhenia rufa grasslands and the permanent papyrus swamp which fringes the length of the Bahr el Jebel along the west side of the island and the Bahr el Zeraf as far north as Fangak along the south-eastern side. Typical khor-bed swamp pastures are found in the network of khors which dissect the higher ground and intermediate lan~ (K. Yirr Gwol, K. Tirru, K. Ful Ngera, K. Kweina, etc.) . NUONG, DOK, JAGEY AND W. JlKAING NUER COUNTRY These Nuer tribes occupy the country which is bounded in the north by the Bahr el Ghazal, in the east by the Bahr el Jebel, and in the west by the swamps of the Bilnyang system which separates them from the Dinka of the Bahr el Ghazal Province. On high land three types of forest are distinguishable: palm, mixed acacia dominated, and Acacia seyal-Balanites aegyptiaca. Palm forests are found south of approximately latitude 8° 45' N. along a well-defined ridge of higher, sandier ground which extends southwards from Bentiu to Nuong country but is broken by deep watercourses and tracts of open grassland. The central part of this ridge runs roughly parallel to Khor Waard from its mouth at Adok to Ngoing, then northwards along the Khor Gul to Kwernyang, Thorial, and Maar to the east of Dul, where, north of this point, palm forests give way to Acacia seyal-Balanites aegyptiaca type forest. The palm forests along this ridge do not occur as one continuous belt, but are broken by forests of either the mixed Acacia or A.cacia seyal-Balanites aegyptiaca types, the latter type being restricted to the heavier clay soils, and the former and the palm forests to the sandier soils. 145 10 Eastwards from Bentiu, between Khor Tiak and Khor Doleib, Acacia seyal-Balanites aegyptiaca type forest occurs. It also occurs in limited areas on either side of the Bentiu-Ler road, between Bentiu and latitude 8° 45' N. and then regularly between that latitude and latitude 8° 30' N., in which stretch it occurs eastwards and westwards of the palm forests of the high land as a fringe some few kilometres wide on the intermediate land before it gives way to open grassland. It also occurs along both sides of the Khor Padwar. The remainder of the intermediate land is open grassland either of the Setaria incrassata type or more frequently of the Hyparrhenia rufa type. The Setaria incrassata type occurs mainly in the triangle Chotyil- Bentiu and the point southwards along the Bentiu-Ler road where the branch road for Chotyil branches off. It also occurs in small quantities south of this, especially eastwards of the Bentiu- Ler road and in the vicinity of Ler. Areas covered by Setaria incrassata are marginal between what we have defined as high land and intermediate land and many parts have permarrent habitations, being relatively free from flood. The areas of open Hyparrhenia rufa grasslands, which are very extensive, do, however, occur on true intermediate land. The land west of the main high land ridge as far as the Khor Bilnyang is almost exclusively open grassland of the Hyparrhenia rufa type, although Echinochloa pyramidalis is often found and in increasing amounts nearer to the Khor Bilnyang. Elsewhere, to the east, it occupies the area between the high land fOrt~sts and Setaria incrassata grasslands and the swamp grasslands found along the Bahr el Jebel and Bahr el Ghazal. There is, in fact, a vast area of this type of pasture between the Khor Joknyang and the Bahr el Jebel, with Echinochloa pyramidalis as the dominant species. In many places it is over 20 km. in width. Inland khor-bed swamp grassland occurs in the many north-south channels which cut through this area, such as the Khor Bilnyang, K. Puragwier, K. Aradeiba, K. Tiak, and many other smaller khors. The main species within these khors are Echinochloa stagnina and Vossia cuspidata, with Echinochloa pyramidalis on the flood-plains. Along the main channel of the Bahr el Jebel from Adok to Lake No is a fringe of permanent papyrus swamp varying in depth, though rarely mor€ than two kilometres. It also occurs along the Bahr el Ghazal, although in smaller quantities and less luxuriant growth. There are also small amounts of papyrus to be found inland along the Khor Waard and Khor Aradeiba. LEIK NUER AND RUWENG (KWIL) DINKA COUNTRy-NORTH OF THE BAHR EL GHAZAL These people occupy the country north of the Bahr el Ghazal between a point west of Lake No and a point between Bentiu and the Bahr el Arab. To the north of Ruweng Dinka country there is the area of broad-leaved woodland on ironstone (approximately latitude 9° 50'), and north and west of Leik Nuer country there is an area of Acacia seyal-Balanites aegyptiaca with hariq grasslands where the country is transitional. Forests are generally of the Acacia seyal-Balanites aegyptiaca type and occupy what may be called intermediate land, since high land is scarce, especially in Kwil Dinka country. The area is more afforested in Leik country. In some places the forests extend right down to the river bank, but are mainly limited to the hinterland. Small areas of the mixed tall acacia dominated type forest occur in the neighbourhood of Yoynyang and beyond the flood-plain on the north side of the Bahr el Ghazal opposite Bentiu. Here the soil is a sandy brown loam. Alternating with the acacia forests, there are large areascof open grassland of the intermediate Setaria incrassata type with occasional areas of Hyparrhenia rufa. This is typical of the country between Bentiu and Riangnom where Andropogon gayanus is also an associated grass. Riverain swamp grassland occurs, but not in great quantity, only sometimes as a narrow fringe, the actual flood-plain north of Lake No producing more of the Hyparrhenia type intermediate land pasture than the usual swamp type. The inland waterways produce pastures of the khor-bed type with Echinochloa stagnina predominating, there being few khors with wide flood-plains on which the Echinochloa pyramidalis could develop. Papyrus swamp occurs as a narrow fringe along the Bahr el Ghazal and, as on the right bank of that river, is less luxuriant in growth than along the Bahr el Jebel. SHILLUK COUNTRY On the east side of the White Nile the Shilluk occupy the country north of the Sobat as far as the Khor Wol system and eastwards as far as Abwong. On the west side of the Nile they occupy the strip from just west of Tonga to approximately the Renk latitude. High land borders the main Nile and the major watercourses, and along these ridges only open palm stands occur. Inland from the river on the, higher parts of the intermediate land there are forests of the Acacia seyal-Balanites aegyptiaca type, there being more Acacia seyaZ 146 and Acacia fistula than Balanites aegyptiaca. Such forest also occurs between Malakal and Nagadiar. The inland grasses are predominantly of the intermediate land Setaria incrassata type, but there is a strong admixture of the tall hariq grasses north of Wau, while east of the Nile Hyparrhenia rufa and Andropogon gayanus occur in some areas beside the Setaria type. Grassland of the riv~rain swamp type occurs along the flood-plain of the White Nile, to a maximum width of about three kilometres. It also occurs as a narrow fringe along the Sobat. DUNJOL AND PALOICH DINKA COUNTRY These tribes live in the area which is vegetatively and pedologically transitional between the Semi-Arid Region and the Flood Region. It is difficult therefore to distinguish between the more obvious high land and intermediate land which form the basis of much of our classification in the Flood Region. Forests are of the Acacia seyal-Balanites aegyptiaca type with some Acacia fistula, while most of the grassland is of the Setaria incrassata type with a strong admixture of the tall annual grasses typical of the hariq grasslands of the Transitional Belt. Grassland of the inland khor-bed type occurs along the Khor Adar and along the Khor Wol system, while along the main Nile is a wide belt of riverain swamp grassland with Echinochloa spp. predominating. At the eastern extension of the area, where the Khor Adar and the Khor Wol system begin, lie the Machar Marshes. This area almost certainly contains large stretches dominated by Echinochloa pyramidalis as well as others dominated by Hyparrhenia rufa, but members of the Team have been unable to penetrate far into the Marshes and much of the data is based on information collected from peoples living on the periphery and reconnaissance flights. Small patches of higher ground and a few areas of forest land of the Acacia seyal-Balanites aegyptiaca type are found between vast tracts of open grassland. THE TRANSITIONAL BELT The area between the Flood Region and the Semi-Arid Region is transitional and is referred to in this report as the Transitional Belt (see Map 7). This belt borders on the Acacia mellifera thicket unit of the Semi-Arid Region in the north. On the west bank it extends from the latitude of Gelhak (latitude lioN.) south along the White Nile to a point opposite Melut (latitude 100 27' N.), where it turns away from the river to follow it at some distance inland to a point north of Tonga (latitude 90 30' N.). North of Gelhak it extends northwards and westwards following a line just south of the Kordofan-Blue Nile Province boundary. On the east bank it extends from Gelhak to the new Renk-Zarzura hafir road, and northwards until it passes into the kitr and qoz country of the Semi-Arid Region. Southwards it passes into the flat grass savannahs of the Flood Region at approximately a line drawn from the mouth of the Khor Wol to tbe Zarzura hafir. There is, in addition to the above areas west and east of the Nile, a smaller area north of the Bahr el Ghazal and west of Lake No (see Map 7). The vegetation within the area is of a transitional nature between that found in the Flood Region and that of the Semi-Arid Region. Forests are similar to the Acacia seyal-Balanites aegyptiaca type of the Flood Region but grasses are predominantly annuals, although the perennial grasslands found within the Flood Region occur. These grasses, though annuals, do not correspond to the predominant short grass annuals of the Semi-Arid Region, though naturally these occur in some areas, ~ut are tall annuals which grow to a considerable height in the rainy season. These grasses are commonly called hariq grasses since they are ideal for the hariq (from haraq: 'to burn ') system of cultivation in which the grasses are burnt to clear the land, a system which is described in detail later (see p. 329). We therefore refer to the vegetation of this area as Acacia seyal-Balanites aegyptiaca Forests with Hariq Grass- lands, as shown in Table 83. Only a few tree types are met within the belt, Acacia seyal being the most prominent and forming dense forest of considerable area. In addition, Acacia fistula is also very common, either as an associate of Acacia seyal or in pure forest of its own. Balanites aegyptiaca is also frequent either in association with these two acacias or as the single dominant in areas that have a parkland appearance. Alternating with these forest areas there are open grass plains in which tall annual grasses from five to nine feet in height predominate. Both the forests and the open grasslands are only found on the clay plains. These rich clay plains produce a heavy, thick stand of the hariq grasses if the areas are protected from the uncontrolled fires that sweep the belt in the dry season. The grasses of this belt are Sorghum purpureo-sericeurn, which is the best grass for hariq purposes, and Hyparrhenia pseudocyrnbaria and Brachiaria obtusiflora, which are also good. In the southern parts of this belt the rainfall is heavier and Setaria incrassata, Pennisetum 147 ramosum and Andropogon gayanus, grasses more frequently found in the Flood Region and not suitable for hariq cultivation, are more common. In the northern part of the area Sehima ischaemoides, Setaria pallidifusca, Cymbopogon nervatus, and Aristida mutabilis are the more prominent grasses. Other grasses occurring, but not very frequently, are Rottboellia exaZtata, Penniseturn mollissimum, Sorghum ZanceoZatum and Ischaemum brachyatherum. As already stated, large areas of the belt are swept by uncontrolled grass fires in the dry season. The established forest areas are not completely immune from these fires. They do, however, possess some fire-resisting powers through tfue nature of their bark, or, as in the thicker forests, are to some extent protected by their suppression of grass growth. The pattern of forests alternating with open grass areas is similar to that occurring in the Acacia mellifera thicket areas of the Semi-Arid Region described later. There is no definite evidence of the one advancing on the other (forests on grassland and vice versa), although it is probable that the open grassland occupies the largest area in this belt. This alternating pattern is probably due to two factors, fire and rainfall. Fire is the most important factor, but the effect offires is determined by the growth of grass which is in turn determined by rainfall. Where an area receives adequate annual rainfall year after year, there is sufficient grass growth to give a good burn each year. Under these conditions bush or tree seedlings are suppressed by the grass growth during the rains and later by burning in the dry season fires . When the total annual rainfall is below normal for a number of seasons, the growth of grass is affected and the seed supply may also be reduced, resulting in thinner stands. During such years tree and bush seedlings suffer less from competition with grass, and as a result their chances of becoming established in that area are enhanced. If they are not burnt during the following season, they should be firmly established before they have to compete again with grasses in conditions of normal rainfall. Once established, they are able to maintain their position by suppressing the growth of grass, which in turn reduces their chances of being burnt out by fire. Stands of Acacia seyaZ of nearly identical age are a common feature in this area, similar to the even-aged stands of Acacia mellifera found within the Semi-Arid Region. THE SEMI-ARID REGION The Semi-Arid Region covers the northern part of the Jonglei Area where the rainfall is generally less than 600 rom. The soils of the area vary from heavy cracking and non-cracking clays to pure sand which forms those rises of higher ground known in Arabic as qoz. The vegetation is less luxuriant than that of any of the oth€I Regions and has been divided into two main units according to rainfall and soil type (see Table 83, p. 167), a description of which is given below. . ACACIA MELLIFERA TmcKETs WITH OPEN AREAS OF MIXED TALL AND SHORT ANNUAL GRASSES This is often referred to as kitr bush, kitr being the Arabic for Acacia mellifera. Its development is essentially a product of the clay soils and it is seldom found on sands. There- fore the distribution of this type is limited to the clays, and further only to the clays receiving an annual rainfall between 400 and 600 rom. It does not extend in any quantity farther north and south of these two isohyets respectively. The actual demarcation lines are difficult to define, since there are penetrations from the surrounding units. For example, there are outcrops Qf the vegetation associated with the qozes farther north. Referring to the map it will be seen that on the west bank of the White Nile its southern and western limit of distribution has been drawn along the line passing west from Gelhak to Ugeiz, then along the line immediately south and west of the Kordofan-Blue Nile Province boundary line. Its northern limit of distribution runs approximately along the line Keri Kera-Er Rua'at, curving northwards and westwards, passing south of Abu Rukba, and then continuing westwards just south of the Kosti-El Obeid railway line through to Umm Ruwaba. On the east bank of the White Nile it is even more difficult to delineate the boundaries since the qozes farther north extend southwards in strips, the Acacia mellifera thickets occurring in the north-south bands on the clays between the qozes. Thicket areas do not, however, occur in quantity south of the latitude of Gelhak. The chief physiognomic characteristic of this unit is the great denseness of the thickets formed by the Acacia mellifera. They cover large strips of country in alternation with equally large, or larger, areas of open grassland. Travellers through kitr country are familiar with the impenetrable thorn thickets which bar their path. No other trees or bushes form such dense thickets over such large areas. The denseness of the Acacia mellifera thickets suppresses grass growth to a very considerable extent, only thin stands of annual grasses being able to tolerate 148 the conditions. However, during the dry season the only sign of any grass at all in some areas. is that under and within the Acacia mellifera thickets; grasses of the open plains are quickly consumed by the animals, but the thin stand of grass within the thickets is protected from them. This denseness of growth of Acacia mellifera and consequent suppression of grass growth within the thickets gives some degree of protection against fire. The grass is suppressed to such an extent that there is only a sparse growth in thin stands, and grass fires usually peter out within the thicket areas. Thus once Acacia mellifera becomes established there is little chance of its being burnt out, and the plants usually remain until they die out naturally. The initial establishment of the Acacia mellifera thicket areas depends primarily on there being a year when there is little or no grass growth. This occurs in a year when there has been little or no rainfall or in a year following a year when, because of the lack of rainfall, the plants (grasses and herbs) did not mature and produce seed. In the following rains season the Acacia mellifera seedlings have very little to compete against and are therefore able to assert themselves and thereafter become firmly established. It is a noticeable feature that in the large areas covered with Acacia mellifera the plants are nearly always evenly aged. These two factors are responsible for the two clearly demarcated sub-units, kitr thickets and open grassland areas. There is no definite evidence of the one advancing on the other and it is probable that the above two factors maintain the balance between the two units, since areas of the two appear to alternate in space and time. Acacia mellifera is the typical tree (bush) of this unit; other trees, however, do occur, but rarely within the thickets. They form stands apart from the Acacia mellifera. Towards the southern half of the zone Acacia seyal and Balanites aegyptiaca occur quite frequently, increasing in amount towards the south where annual rainfall is greater. Northwards, Cadaba rotundifolia appears in increasing amounts. The typical grass of the area, especially within the thickets, is Tetrapogon spathaceus, whereas SchoeneJeldia gracilis and Aristida sp. are more frequently met with in the open grass- land areas, especially in the northward extension of the Region (on the non-cracking clays). In the central area the grasses Sehima ischaemoides and Cymbopogon nervatus are more frequently met with in the grass areas. At the southern extension the tall annual hariq grasses are more in evidence-Hyparrhenia pseudocymbaria and Sorghum purpureo-sericeum being the most frequent. In this part of the area regular sandy outcrops (qozes) occur, but their distribution is more limited than in the northern part which we classify as' acacia-short grass '. On the west bank these qozes occur as southern outcrops of those farther north, but there is no definite or visible link. About twelve miles inland from the Hasoya wood station (on the banks of the river) is situated the qoz system known as Abu Seleih. Fifteen to twenty miles farther inland there is an even larger qoz system known collectively as Kohla. To the north of Kohla there is a smaller system called El Amoud, while to the south lies Telbeldiya. These qozes carry a similar vegetation to that of the ones farther north, Acacia senegal, Balanites aegyptiaca, Dalbergia melonoxylon, and Boscia sp. being present in varying amounts. They are well known to the Ahamda and Seleim Baggara not only as potential cultivation areas, if a regular water supply were to be installed, but also as a first-class grazing area to which cattle from the vicinity of Jebel Megeinis in the west are brought to graze, staying in the area as long as water supplies last. Their chief importance is that they supply grazing at that awkward period between the exhaustion of the pasture aroutld the permanently settled areas and the time when the riverain grazing becomes available when the flood-plain is exposed. In years of good rainfall there is sufficient water available for the people to remain and harvest good crops of dukhn (Pennisetum typhoideum), Jul Sudani (Arachis hypogea) and simsim (Sesamum orientale). On the east bank the southward extension of the qozes around Kosti reaches as far south as the Gelhak latitude, following the line of the Renk-Qoz Kash Kash hafir road. Other smaller qoz systems occur eastwards of this main system. Between the rise bordering the flood-plain of the White Nile and the qozes inland there is low-lying land carrying dense kitr thickets (Acacia mellifera), some Acacia seyal forests and some areas of grassland with a thin covering of the short annual grasses. Along the qozes of the Renk-Qoz Kash Kash road there are forests containing Anogeissus schimperi, Dalbergia melanoxylon, Terminalia sp. , Acacia seya/ and Balanites aegyptiaca. The rise bordering the flood-plain of the White Nile between Geigar and just north of Gelhak is of a sandy clay mixture which carries a fringe of forest, parts of which are dense. Also found in this area are Acacia sieberiana, Hyphaene thebaica (the dom palm) and large trees of Acacia seyal, the grasses being mainly short annuals-Chloris pilosa, Eragrostis sp. , Aristida sp. , Dactyloctenium aegypticum, etc., but with some Cymbopogon nervatus. 149 ACACIA-SHORT GRASS (QOZ CATENA) Although the vegetation of this area is termed acacia-short grass (because of the pre- dominance of short annual grasses, as compared with other Regions), the main feature is the lllumber of variations within this type because of the regular and close juxtaposition of different soil types. The different soils are the sands, those of the qozes, and the clays of the intervening plains. The qozes run in longitudinal banks from north to south, being southern extensions of the qozes north of the boundaries of the Jonglei Area. Between these qozes are the alluvial clay flats. Both types have their concomitant types of acacia-short grass vegetation, and the general feature is therefore a regular alternating variation in the acacia-short grass type. The extent of this development is difficult to assess, although in general it lies north of the Acacia melli/era thicket area previously described. On the west bank of the White Nile the boundaries of distribution are roughly the sides of the triangle between Umm Ruwaba, Kosti and Jebelein. South of this triangle only isolated outcrops of this pattern occur. On the east bank the development of the pattern is not so pronounced, and there is not the continuous repetition of the alternating sandy rise-clay plain conditions of the west bank. In fact south of Kosti qoz is almost non-existent UI1til south of Jebelein, where a rib runs N.W.-S.E. between 8 and 14 kilometres distance inland from the river towards Geigar and then continues southwards towards Renk. Nearly all the area north of Jebelein is non-cracking clay plain carrying Acacia melli/era, and the true alternating paUern described above rarely occurs. On the qozes, Acacia senegal, Balanites aegyptiaca, Boscia sp, Dalbergia melonoxylon, and Zizyphus spina-christi occur in varying frequencies, while tJhe most common grasses are Cenchrus bifiorus, Cymbopogon sp., Eragrostis tremula and Aristida species. Bordering the sandy qozes, in what may be in many cases a transition belt between the sands of the qozes and the clay of the flats (although there is often a sharp change from one to the other), Acacia mellifera finds its most suitable site. It spreads outwards on to the clay flats for some distance, but it usually gives way to open grassland composed mostly of Schoene/eldia gracilis and Aristida speci€S, with some Chloris pilosa and Dactyloctenium aegypticum. Often associated with the Acacia melli/era are Acacia seyal, Acacia or/ota and Cadaba sp., with an undercover of the grasses Tetrapogon spathaceus, Sporobolus marginatus, Chloris virgata, Chloris pilosa and Dactyloctenium aegypticum. Towards the centre of the clay flats it is usual to find a marked depression, more or less circular in outline, into which run-off water from the qozes and the surface water held on the plain (owing to the impermeable nature of the clay) collects to form a pool (Ar. rahad). These pools are usually recognizable from a distance of several miles since, because of ' the increase in period of moisture availability for plant growth in the immediate vicinity of the pool, the area supports a much higher type of tree growth than the surrounding plain. Tn'les common in the area fringing the pools are Tamarindus indica, Acacia sieberiana, Acacia campylacantha and Balanites aegyptiaca, mostly tall, handsome trees which stand above the other shorter trees of the plain. Within the pool the aquatic grasses Oryza sp. (breviligulata ?), Echinochloa colona, and sometimes Echinochloa stagnina, are present. After the rains the only water available is that present in the rahads and later in shallow wells dug in their beds, and the Baggara tribes inhabiting the area make the most of this supply. Many of these rahads provide sufficient water after the rains to enable the cultivators of the qozes to remain in the vicinity and harvest their crops and to water their cattle. By about the end of November, however, most of the pools dry out, and the inhabitants start on their annual trek in search of water, either to well-known well-centres or down to the White Nile. INTER-REGIONAL: SWAMP VEGETATION OF THE JONGLEI AREA THE INTER-RELATIONSHIP OF ENVIRONMENTAL FACTORS AND THE DISTRIBUTION OF PLANT SPECIES The Nile and its flood-plain is a feature common to all three Regions of the 10nglei Area. Along the flood-plain there has developed a type of vegetation peculiar to the conditions pro- duced by the present regime of the river. These conditions are not restricted to the Nile itself, but are also found on the flood-plains of other large rivers-the Bahr el Ghazal, Bahr el Zeraf and Bahr el Jebel- as well as those of smaller channels and drainage systems. The type of vegetation is similar throughout and is treated as one unit. This type of vegetation has already been briefly described in a previous section (see p. 140) dealing exclusively with the Flood Region. Its importance in the problem of the Equatorial Nile Project and its effects in the Sudan is so great that a separate description of the main constituents and the ecological factors which determine their growth and distribution is essential. 150 GENERAL OBSERVATIONS PLANT CONSTITUENTS It is not intended to repeat the list of speci~s plieviously given, but in order to simplify matters when attempting a quantitative study of dile ecological factors, only the major plant elements are considered. For example Cyperus papyrus is the dominant plant in th~ papyrus swamp, but there are many other plants such as the climbers (see p. 142); thes~, however, are ignored, since in general theilr distribution is so closely linked with that of papyrus that it is assumed that should, as a result of the Project, any radical ohange tak~ place affecting the growth and development of papyrus, it is likely that they too will be affected. We are therefore concerned primarily with the following speci~s, the development of which, we hope, represents the variations in environment : (i) Phragmites communis (ii) Cyperus papyrus (iii) Vossia cuspidata (iv) Echillochloa sfagllina (v) Oryza spp. (mostly O. barthii) (vi) Echinochloa pyramidalis. All the above constituents are rhizomatous, but differences arise in form and struoture and in the position of the rhizomes in relation to ground level (i.e. dry season ground level). Cyperus papyrus, Vossia cuspidata and Echinochloa stagnina hav~ superficial rhizomes which either creep along the soil surface of the flood-plain at low river or float on the surfac~ of the flood-water at high river. The other constituents have rhizom~s that ar~ buried at various depths below the soil surface. In all eases shoots and roots are produced from the nodes. In form and structure Echinochloa stagnina and Vossia cuspidata are stem-like, whereas Cyperus papyrus is strong, thick, woody, up to 10 cm. in diameter, and root-like. The buried rhizomes are root-like. The aerial shoots vary considerably in height, Cyperus papyrus and Phragmites communis reaching as much as three to four metres. The others are much shorter and vary only in degree; the upright shoots of Vossia cuspidata and Echinochloa stagnina (when floating) are about equal to Oryza barthii but shorter than Echinochloa pyramidalis, which often exceeds two metres. Root d~velopment also varies; the roots of papyrus are delicate and soft, either extending freely into water or, if in shallow water or in moist situations, penetrating the muddy sub- stratum or peat horizon. The roots of Vossia cuspidata and Echinochloa stagnina are stronger and in both cases may either hang freely in the water or penetrate the soil in shallow-flooded areas, or the moist soil in th~ case of areas exposed at low river. The roots of Phragmites communis, Echinochloa pyramidalis and Oryza spp. are stronger and longer, penetrating the solI to a depth of five feet or more, thereby firmly anchoring the plant. E~RONMENTAL FACTORS The environment in which these plants flourish is determined primarily by factors controlled by the river regime and secondarily by other factors such as climate and human activities. The hydrology of the White Nile and its tributaries has been explained in detail elsewhere (see Chapter 1), but we restate here iJil very brief form the main features which determine the growth and distribution of vegetation. THE NORMAL RIVER REGIME In all three Regions, when the rivers are at their minimum levels they are contained within their banks, except in the main part of the Sudd area along the Bahr el Jebel and upper waters of the Bahr el Zeraf where spilling may occur even at low level and produce permanent swamp. At maximum levels nearly all the flood-plain is covered, except between approximately Juba and Mongalla where large-scale spilling only occurs when levels are above the normal maximum. Flood-plains are neither regular nor flat, so that there are varying depths of flooding when the rivers are high. Moreover, since the rise and fall in water-level takes place gradually over several months, parts of the flood-plains receive different periods of flooding. In other words, both the depth and the duration of flooding at different sites are variable. Other factors also have their effect in determining the nature of the environment in which the plants can or cannot flourish. The first is the rate at which the river rises from low to high level; or in terms of the flood-plain, the rate at which the water reaches its maximum depth after flooding has begun. The second is the velocity of the current. 151 SOILS The soils of the flood-plain of the Nile, i.e. sudd soils and toich soils, have already been described in some detail. In general the soils of the toich (corresponding to the riverain swamp grassland) are predominantly clay in the central and northern stretches of the river, although along the southern reaches there is an increased percentage of sand. The soils of the Sudd (corresponding to the permanent papyrus swamp) are characterized by the presence of a surface organic horizon overlying soil with a variable percentage of clay and sand, but in general wit}} a gf(~ater percentage of sand than is found in toich soils except for those of the southern toiches south of Bor. HUMAN ACTIVITY Human activity, i.e. the annual burning off of areas of swamp pastures by the people to produce pasture (rtlgTowth) on which they can graze their cattle in the dry season, is important, since by the end of the dry season all the vegetation in such areas is grazed down to ground level. CLIMATE : RAINFALL As a generalization, the rainfall throughout the area is in phase with the period of river spill and flooding, but an important point, as will be seen later, is the total amount of rainfall which falls in an area prior to its being flooded by river spill. QUALITATIVE ASSESSMENT DEPTII OF FLOODING Phragmites communis, Echinochloa pyramidalis and Oryza spp. are firmly anchored in the gTound by their rhizomes and root-systems ; they are therefore unable to float on the rising flood-water. It is doubtful whether they could tolerate complete submersion and it is therefore considered that their distribution along the flood-plains is limited by depth of flood-water, since these species are unable to tolerate a site where the depth of the water exceeds the normal height for the species. This, of course, would allow a great deal of latitude for Phragmites communis which reaches heights between 2·5 and 6 m. , although it is never found gTowing in such depths of water for reasons explained later. On the other hand Cyperus papyrus, Vossia cuspidata and Echinochloa stagnina are not limited by depth to the same ext~nt, because their superficial rhizomes enable the plants to float on the rising flood-waters, thereby ensuring that the upright shoots are not completely submerged. DURA nON OF FLOODING Cyperus papyrus, Vossia cuspidata and Echinochloa stagnina possess superficial rhizomes and, in consequence of their semi-aquatic existence, have comparatively weak root d~velopment. According to Migahid (53), the roots of Cyperus papyrus can only penetrate moist soils because they cannot overcome the resistance offered by dry soils, and the clay soils of the Sudd (papyrus swamp) would need to be saturated, or nearly so, in order to btl easily penetrated by the roots ·of many swamp plants. The water content of the soils of the flood-plains under examination is determined primarily by the duration of flooding in the flood season. Rainfall is only an important contributor just prior to flooding. In some areas the presence of an underground water-table is also important. Therefore where the durati'bn of flooding is short, the resistance of the soils to root penetration will increase with the length of the exposed dry period, although resistance will naturally vary with the mechanical composition of the soil. As the duration of flooding becomes longer and the exposed period shorter, the degree to which soils dry out lessens, so that soil resistance to root penetration decreases until, when permanently inundated .o r saturated, resistance is at its lowest. Such areas where the latter conditions prevail are therefore suitable for Cyperus papyrus, whilst other areas with slightly less flooding are more suitable for Vossia cuspidata and Echinochloa stagnina. On the other hand, plant constituents with tougher root-development such as Echinochloa pyramidalis, Phragmites communis and Oryza spp. would be able to overcome the greater soil resistance met with in parts of the flood- plain subject to shorter durations of flooding. CuRRENT V ELOCITY Current velocity as a factor limiting the distribution of plants is only of importance along the main Nile and parallel channels, since current velocity on the flooded plain is negligible. Furthermore, current velocity is only of major importance to those constituents with floating rhizomes, because they .are much more subject to breakage than plants firmly rooted by under- 152 ground rhizomes. Plants with superficial rhizom~s, if they are to exist in the swamp, must either have their rhizomes firmly attached by their roots to the soil or have filexible rhizomes to withstand currents. They must otherwise limit themselves to sheltered areas where current velocity is slow. Cyperus papyrus, with a poorly developed root-system, is only actually anchored in shallower or expos~d sites (exposed for a short period). When the river rises the rhizomatous mass rises with the flood-water and at the deeper depths is eventually broken from its roots to become free floating. Such free floating masses are generally found lying behind the protection of the root-anchored papyrus along the deltaic bank and are ther~fore not subject to the swifter currents. Echinochloa stagnina, although rooted and possessing a more flexible rhizome, is more tolerant of stagnant conditions and is never found along the main river channel where it is subject to swifter currents. Vossia cuspidata, on the other hand, possessing a ' tougher' rhizome, is more frequently found firmly anchored by its roots to the bank with its rhizomes (or floating stems) extending outwards towards mid-stream, and is therefore subject to much swifter current velocities than the other constituents. It can thus be assumed that the resistance of its rhizomes to the water current is greater than that of other constituents. SOILS AND SOIL MOISTURE Sudd soils are characterized by a layer of organic material on the upper surface, which rarely dries out. Below this layer the mineral soil varies from clay to sand. Toich soils have been divided into the pn~dominantly clay type of the northern and central regions and the predominantly sand type found along the southern reaches of the Bahr el Jebel. In both areas the upper layer of the soil is strongly stained with humus. In the dry season this layer dries out below saturation point, although there is generally some moisture at some distance below. The. position and amount of soil moisture present, however, depend on how long the area has been exposed after the flood-waters recede, unless of course there is an alternatiw source on which the upper soillayeis can draw. Migahid, in his examination of sudd soils, has demonstrated that the higher th~ percentage of clay the slower the movement of water becomes, both downward, and vertically by capillary action. He has further shown that predominantly clay soils have a capiIJary rise of less than 3 to 4 feet. Predominantly clay toich soils, which are subject to longer periods of exposure and which have no water supply other than that which they receive during the flood season, probably dry out below this level or at least moisture does not become readily available. In such areas, plants such as Echinochloa pyramidalis and Phragmites communis, with persistent perennial rhizomes, are more likely to succeed than Echinochloa stagnina with its superficial, more delicate rhizomes. Furthermore Echinochloa pyramidalis, with its greater root-develop- ment, is able to continue growth over a longer period than Echinochloa stagnina, since its roots are able to penetrate further and make use of the soil moisture at lower levels. Echinochloa stagnina is therefore more likely to succeed at sites where soil moisture is more freely available. Such sites are those that receive a longer duration of flooding with a compensating shorter exposed period during which insufficient time is available for the moisture in the soil to drop below that which is essential for the continued existence of the plant. The above limits of distribution are, however, modified where there is an additional source of soil moisture. Where soils are mixed with alternating bands of sand and silt, or on predominately sandy toich soils, there is a water-table, the level of which is related to river level. On predominantly clay toiche!, owing to the high impermeability of the clay, any water- table drops quickly away from the river. Thus on flood-plains with a water-table related to river level, the distribution of species is not directly related to the duration of flooding, for, depending on the level of flood-plain above dry season river level, there will be sites at which water is available at all times near the surface, i.e. the surface layers are permanently saturated. At such sites the shallow-rooting, less persistent Echinochloa stagnina finds conditions tolerable. At other sites, however, where flood-plain levels are higher above river level and the water-table is lower in the soil, only the more persistent deep-rooted species find conditions suitable, e.g. Echinochloa pyramidalis and Phragmites communis. Such distribution of species is manifest on the flood-plains between Juba and Mongalla where the soil is of the predominantly sandy toich type. RATE OF RISE IN FLOOD-WATERS: RAINFALL : FOONG AND GRAZING The rate of rise in flood-waters , rainfall, grazing and firing, are all closely related in their effect on distribution of species. This is more so in the case of swamp grassland than of papyrus swamp, because deliberate firing and grazing of the latter is rare. Related to these factors is the growth rate of the different species. 153 Under normal river conditions the areas occupied both by papyrus swamp and by riverain swamp grassland are affected by fluctuations in river level. In the season of low river (which corresponds with the dry season) much of the area occupied by swamp grassland is exposed and much of the rank vegetation is burnt, grazed, or both, and by the end of the exposed period the vegetation-with the exception of Phragmites communis-has been grazed to ground level. At this time the rains begin and the river, being nearly in phase with the rainfall, begins to rise, eventually spilling over its banks and inundating the flood-plain. Moisture now becomes more freely available and plant growth responds to this stilnulus; almost simultaneously, in the southern and central regions at least, cattle are moved from the flood-plains so that growth has no longer to keep pace with continued defoliation. However in the Shilluk area and northwards the flood-plain pastures are grazed for considerably later periods in the rains. With the slow rise in the level of the flood-waters, there is a possibility of the plants being swamped unless their growth rate can keep ahead of the rate of rise. With the other limiting factors considered, and on the assumption that the different species have variable growth rates, there will be sites along the flood-plain which are favourable to some species and un- favourable, or unattainable, to others, depending upon the relation of growth rates to the rate of rise in flood-water. A sudden increase in water-level (say from 0 cm. to 50 cm.) when the river begins to rise would probably be unfavourable to most of the species (probably the open lagoons which are bare of vegetation are a result of this). Fortunately, however, under the present river regime the rise in river level is gradual and is also in rhythm with the rainy season, so that most of the area of the flood-plain has received a certain amount of rainfall before flooding commences. This stimulates the plants to growth before they are inundated by the flood-waters and, as long as the flood-waters continue to rise gradually, they are not adversely affected. On the other hand, where the rate of flooding is not gradual, such as low-lying sites which receive a sudden increase in water depth, the non-floating species Echinochloa pyramidalis and Oryza barthii may be adversely affected. Such sites are more suitable for the floating species-Echinochloa stagnina and Vossia cuspidata. COMPETITION We have shown that the various species occurring along the flood-plains are limited in their distribution by certain conditions. Within these limits they are subject to competition between themselves. Competitive ability is a difficult quality to assess in plants because it occurs not only between the aerial shoots, for light, but also between the underground root-systems, for soil nutrients . Within the area covered by permanent papyrus swamp it is reasona1>le to assume that water and nutrients are sufficient to meet all plant requirements (cf. its humus-rich layers and almost permanent flood conditions) and that competition is mainly for light. When competing for light it is the taller-growing plant that is generally successful-thus we find Cyperus papyrus as the sole dominant over vast areas, although sites within that area are suitable for other sp~cies, i.e. Phragmites communis. Migahid also attributes this success to the rapid growth rate of papyrus; he recorded growth rates as high as 7 cm. per day for the aerial branches. Migahid further contends that although it is possible for Phragmites com- munis to grow on sites suited to papyrus, it does not succeed because it is unable to compete successfully for light. It therefore favours sites which are outside the range of papyrus distri- bution, i.e. more elevated sites subject to shorter duration of flooding, but where its roots can still tap sufficient moisture at lower levels. On such ~ites, as along the flood-plains between Juba and Mongalla, it becomes dominant, but it is not a successful suppressor of the other species which also find the same site favourable. Echinochloa pyramidalis is, for example, found growing with Phragmites along this reach of the river. Vossia cuspidata and Echinochloa stagnina never need to compete against Phragmites communis because their limits of distribution are distinct. Similarly Vossia cuspidata is not often found competing with Echinochloa stagnina. Having outlined the main features of the environment which have their effect in the determination of plant distribution in the swamp, we must now attempt an estimate of their relative influence in more detail. QUANTITATIVE ASSESSMENT D EPTH AND DURATION OF FLOODING Extensive investigations and trials carried out by the Team have revealed the importance of depth and duration of flooding as measurable factors governing the distribution of plant species along the flood-plains of the Nile and other rivers in the Jonglei Area. Since the depth and duration of flooding will be radically altered under the Equatorial Nile Project, it follows that such factors must be examined in greater detail than others. 154 The figures given and examined below have been accumulated during two years' work along different reaches of the river by adopting different techniques. The reaches on which our results are based are: White Nile : Malakal-Kosti Sobat : Bilaiwal- Sobat Mouth Bahr el Ghazal: Yoynyang-Lake No Bahr el Jebel : Aliab Valley Mongalla-Gemmeiza Juba-Mongalla. It will be seen that data have been collected from almost every stretch of the river passing through the Jonglei Area. The necessity for wide sampling is obvious. In the first place, it is not unreasonable to assume that with the great variation of rainfall over the whole area, the relationship between species distribution and depth and duration of flooding also is yariable, e.g. the mean depth and duration figures are lower in the southern reaches, where rainfall is higher, than in the north, where rainfall is lower. Field observations, however, have not confirmed this assumption. Secondly, different reaches of the river flood-plain are dominated by different types of vegetation : Jebelein-Malakal-Lake No Riverain swamp grassland, Echinochloa dominated. River Sobat Riverain swamp grassland, Echinochloa dominated. Bahr el Ghazal Stunted Cyperus papyrus and riverain swamp grassland. Lake No-Bor Preaominantiy Cyperus papyrus with riverain swamp grassland. Juba-Mongalla Riverain swamp grassland, Phragmites communis dominated. Apart from more general observations and trials, the investigation of vegetation and soils in each area was undertaken where possible as a combined operation with the Team's surveyors who were already carrying out survey work in connection with hydrology. It was therefore possible to take soil and vegetation samples at different levels previously determined by precise levelling ; a method which has done much to support conclusions drawn from general observations in the field. The results are given below and in som€l cases are recorded in greater detail in a later volume (and in the maps and diagrams which refer to it) where data on special areas and experimental work are given (see Vol. III, Chaps. 1 and 6). The reader will note that the order of presentation in this section differs from that normally used in this report, since areas are here givtm roughly from north to south and not vice versa. This may appear inconsistent, but there is a definite reason for it. Investigations of thi.s type were first carried out in the Jebelein-Malakal reach and formed the basis of later work. The tech.nique used in this reach, however, differed from that employed in other reaches. Changes in technique were made deliberately in the hope that by these means more reliable figures could be obtained . THE WHITE NILE: MALAKAL-KoSTJ The flood-plain of the White Nile between Malakal and Kosti (Jebelein) is on the average two kilometres wide, limited on both banks by the pronounced high land ridge. The river itself runs in a well-defined channel with sfnall deltaic banks. At high level it spills to inundate the adjoining flood-plain, flood-water being limited in its spread by the gradual rise of ground level until it reaches higher ground, and at high river the flood-plain is inundated to a maximum depth of approximately 250 cm., the average being 60 cm. The flood-plain is not flat but is dissected by minor waterways, with small deltaic banks, running parallel to and receiving their water from the main river. In a normal year the river begins to rise in early April, and at Malakal spill usually starts in August (the gauge level at which spilling begins is 11 ·50 m. at Malakal and 11 ·70 m. at Jebelein) . In normal years the river completely covers the flood-plain at Malakal for five months between August and December. At Renk and northwards the duration of flooding is governed more and more by the operation of the Jebel Aulia Reservoir, and the flood-plain is normally covered for nearly seven months from the 21 st August to the 15th March, the average depth being 90 cm. Under normal conditions the river rises gradually, the area of flood-plain inundated increasing with each successive rise in gauge level, until at normal maximum it approaches the edge of the higher ground. The major grasses found along the flood-plain of this stretch are Vossia cuspidafa, Echinochloa sfagnina, Oryza barthii, and Echinochloa pyramidalis. 155 TECHNIQUE AND METHOD The investigation was limited to selected sites along the reach. The sites were Malakal, Melut, Renk, Jebelein and Kosti (Rabak). At these sites there are permanent river gauges, installed and read by the Egyptian Irrigation Department, and reliable figures relating to the rise and fall of the river are avaiJable over many years. The investigation was conducted when the river was high, just after it had passed its maximum peak. It was hoped that by choosing sites near to river gauges a close relationship could be obtained between water depth on the flood-plain and river level as indicated by the gauges, thereby enabling us to obtain maximum depths of flooding at selected places along the flood-plain for anyone year, or over an average period for a selected number of years. At each site, by disembarking from a steamer aNd wading (and sometimes swimming) across the inundated flood-plain from the bank of the river to the high land ridge, readings of water depth were taken at random. At each reading the grass species was recorded, thus giving a whole range of depths for each species at each site. The river level at each site as shown by the river gauge was recorded at the time. The data and results obtained from this investigation are given below. Table 84, p. 168, gives the data collected for each of the five sites; the date, the gauge- reading at the site at the time of the investigation, and the range of depths recorded for the four grass species. Table 85, p. 169, summarizes the information given in Table 84, giving the maximum, minimum and mean depths recorded for each species at each site. Table 86, p. 169, shows the relationship of the four grass specIes to the level of the flood- plain above river gauge datum at each site. These figures were obtained by subtracting the water depth of each species as shown in Table 85 from the river gauge-reading at the time of the investigation (see Table 84). . A prelimrnary analysis carried out by the statistician at the Wad Medani Research Farm showed that there were significant differences between species in the levels at which they grow, but that the differences between species were not consistent from site to site. Since these levels were obtained by using the recorded depths, it follows, theoretically, that the distribution of grass species at each site is related to depth of flooding, but that this relationship does not hold good between sites. It has been assumed in this investigation that the water-level in the flood-plain is at all stages the same as that in the main channel, which in this particular instance is given by the gauge-reading. This particular assumption is consistent with that already made in the analysis of the White Nile flood-plain as given in Chapter 1 of this volume. It was decided arbitrarily that the flooding criterion to be used in establishing the depth- duration relationship of the grass species should be the average of the previous five floods, with double weight given to the last flood (1950-51, the flood year of the investigation). This was necessary because it must be assumed that owing to different fluctuations in river level from year to year and hence variations in depth and duration, areas dominated by certain grass species extend and recede according to conditions which favour them, and their distribution is not constant or stabilized. Tables 87, 88, 89, 90 and 91 (pp. 170-2), show the number of lO-day periods in the floods of 1946-47, 1947-48, 1948-49, 1949-50, 1950-51 (the year being taken as May 1st to April 30th) in which th'e water-level exceeded selected gauge levels at the five sites. Weighted average is shown in the last column of each table. From these tables it is possible to interpolate for any particular flood-plain level (see Table 86) the depth and duration of flooding at that level (in the case of depth, maximum level is taken as that which is attained and held for two lO-day periods). This has been done for the mean and maximum and minimum range of flood-plain levels at or between which the various grass species occurred at each of the five sites. DEPTH OF FLOODING Table 92, p. 173, shows the mean, maximum and minimum range of depth for the four species at each of the five sites on the basis of the average previous five floods. It has already been stated that there are significant differences between the depths at which each grass species grows at each site, but that the differences are not consistent from site to site. Irrespective of this, and in order to have some figure on which to base future calculations for the reach, the last column in the table shows the over-all average of the maximum, minimum and mean depths for each species at the five sites. It is evident from this column and from each site that whereas Vossia cuspidata and Echinochloa stagnina readily attaiR depths greater than two 156 metres, Oryza barthii and Echinochloa pyramidalis never succeed at suoh depths. Equally evident is that whereas Echinochloa pyramidalis and Oryza barthii occur at the shallowest of depths (between 10 and 20 cm.), Vossia cuspidata and Echinochloa stagnina do not occur at depths less than 72 cm. (cf. Jebelein and Rabak). DURATION OF FLOODING We have already assumed that the water-level on the flood-plain is the same as that in the main channel at all times, and that the level of the flood-water on the flood-plain is related directly to the river gauge level. On this basis therefore any rise or fall in river level wm be followed by a similar rise or fall in water-level on the flood-plain. SinGe the interval of time between rise and fall is recorded at the river gauge, we are able to calculate the time period between rise and fall at any point on the flood-plain at the sites. Depth and duration of flooding are therefore, on this basis, related for anyone site. It should be mentioned, however, that because of the deltaic bank formation of the main river and minor waterways that run parallel to the main river, there is a time-lag before the rise in level recorded at the river gauge is translated to the flood-plain. There is also a time-lag before a fall in the river, recorded by a gauge read at the river, is translated when the river level has dropped below the spill level. We have had to assume that these lags are equal in time and that they cancel each other out. Consequently our original assumptions that water-level in the flood-plain is at all stages the same as that of the main channel is considered valid for the purpose of these calculations. Since it has been shown that then~ are significant differences between depths, it follows that there are significant differences between the durations of flooding received by each species at each site. Table 93, p. 173, shows the mean, maximum and minimum over-all raRge of flood duration for the four grass species at each of the five sites. The last column is the over-all mean of these figures. A study of all the sites shows that Vossia cuspidata and Echinochloa stagnina are never flooded for less than 180 days. On the other hand Oryza barthii and Echinochloa pyramidalis flourish at sites with shorter durations of flooding. A study of the means in each case shows that whereas at Malakal and Melut there are obvious differences in flood duration between species, these differences decrease northwards as the effect of the Jebel Aulia Dam becomes more pronounced. CONCLUSION From the above figures the following optimum conditions of depth and duration are put forward for the four species, although more careful and prolonged experiments will be necessary before they are precisely determined. It seems that 240 days' flooding at a depth of 175 cm. (approximately the over-all average of the means) favours the development of Echinochloa stagnina, but does not rule out Vossia cuspidata; that 175 days' flooding at a depth of 100 cm. favours Echinochloa pyramidalis, but also permits Oryza bar1hii. . These figures are based on the averages given in Tables 92 and 93 which also illustrate the wide range of distribution of the species in relation to depth and duratiotl. From the above data we have been able to obtain information about the distribution of the above grasses in terms of area between Malakal and Jebelein. The information is basically connected with the survey of the Jonglei Area, and since it arises out of the ecological study it is included within this section. FLOOD-PLAIN AREAS AND AREAS OF GRASS SPECIES Application of the results is limited to the reach between Malakal and Jebelein, using figures obtained for Malakal, Melut, Renk and Jebelein. We have shown that species distribution is related to the level, depth and duration of flooding at each site and that statistical analysis shows the differences between species to be significant for each site, although the differences are not consistent from site to site. In order to make the results applicable, we have assumed that the true distribution of depths for each species is a normal one, and that the standard deviation of the distribution is 0·28 m. for every species at every site (the standard deviation does, however, differ significantly at the 5% level). Furthermore, we have assumed that the area covered by each species at each site is proportional to figures obtained by estimating their proportions by linear transects at each site (in the case of Jebelein it is the average proportions of three such transects) . These proportions are given in Table 98, p. 176. By taking the mean gauge level for each species at each site and using the standard deviation of 0'28 m., we have calculated the proportions of the normal curve distribution that 157 lie between gauge intervals separated by successive steps of 0·25 m. for each species at each site (Tables 94, 95, 96 and 97, pp. 174-5). Tabl€s 99, 100, 101, 102 (pp. 176-7) have been obtained by multiplying the figures in Column 2 of Tables 94, 95, 96, 97, by the proportions of each species (as obtained by linear measurements) at each site, to give figures for the distribution of these proportions between gauge intervals separated by successive steps of 0·25 m. From these tables, Tables 103, 104, 105, 106 (pp. 177-8) have been prepared, which give the proportional distribution of the grass species over the area that would become flooded if the gaug€ level at each site were raised by successive intervals of 0·25 m. We next calculate the percentage distribution for the average gauges for the reaches Malakal-Melut, Melut-Renk, and R€nk-Jebelein by' taking the mean of the percentage dis- tribution at the same gauge intervals at the ends of each reach, i.e. Malakal, Melut, Renk and Jebelein. These figures are shown in Tables 107, 108, 109 (p. 179). From the id€alized cross-sections of the reaches, the actual area of flood-plain between gauge intervals for each reach is known (see Tables 208-9, pp. 447-8). From the percentage distribution for each reach as given in Tables 107, 108, 109, the actual area of each species between regular gauge levels has been calculated and summed for the total area of flood-plain between gauge intervals of 10·00 and 13·00 m. to give the total area occupied by each species for each reach. These are given in Tables 110, Ill, 112 (pp. 180-1). It is a simple calculation to obtain the area for each species for the whole reach Malakal-Jebelein. These figures are later used in calculating the present stock-carrying capacities of the riverain swamp pasture north of Malakal, but with a slight modification. In Tables 110, Ill, 112, the distribution of areas is given between 10·00 and 13·00 m. gauge level readings in each of the three n~aches, although it will be noticed that the actual toich (or riverain swamp) pasture grasses begin to fade out at the following levels : Malakal-Melut 12·25 Melut-Renk 12·25 Renk-Jebelein . 12·50 Furthermore, the maximum mean monthly gauge levels for normal river under Jebel Aulia for these reaches given in Tables 206 and 207, Vol. II, Chap. 2, are: Malakal-Melut 12·20 (November) Melut-Renk .. . 12·16 (November) Renk-Jebelein . .. 12·39 (November) The maximum areas inundated for the three reaches at these gauge levels are 224, 429 and 259 sq. km. respectively. The minimum mean monthly gauge levels for normal river under Jebel Aulia are: Malakal-Melut 10·03 Melut-Renk ... 10·12 Renk-Jebelein ... 10·14 At these levels areas inundated are I, 8 and 4 sq. km. respectively. Therefore, when considering the usefulness of the riverain grassland as grazing, only the areas inundated and exposed between the following gauge levels for each reach have been considered : Malakal-Melut 10·03 to 12; 20 Melut-Renk 10·12 to 12·16 Renk-Jebelein 10·14 to 12·39 The actual areas for each species worked out on this basis for each reach have been inter- polated from Tables 110, 111 and 112, and are given in Table 113 (p. 182). OTHER REACHES The field techniques and method used along all other reaches were basical\y the same, but different from that employed in the Malakal-Jebelein reach. Instead, as in the Malakal- Jebelein reach, of relating species, by random sampling, to depth of flooding, the distribution was related to levelled cross-sections across the flood-plain. The Team's surveyors were conducting special surveys of areas along the flood-plain and, where possible, the grass distribu- tion was related to the level\ed cross-sections in these areas (see Aliab Valley survey, Mongalla- Gemmeiza survey and Juba-Mongal\a survey, Vol. III, Chap. 1). In reaches where no special survey was being conducted, special individual traces were made and levelled, and the grass dis- tribution along the trace noted-as along the Bahr el Ghazal and the Sobat. The general procedure along these reaches was first to cut a trace from river to high land, or vice versa, often a laborious task, after which the surveyor took level readings. Soon after 158 this operation, the distribution of the vegetation along the line was recorded as accurately as conditions allowed. There were slight variations which will be recorded later when each reach is taken in turn. THE SOBAT RIVER (SOBAT MOUTH TO BILAIWAL) The reader is referred to Chapter 1 of this volume for a full account of the Sobat River. The main characteristics are its great variability of levels from year to year and the wide range in anyone year. The actual flood-plain is not very wide, but there are a series of comparatively large toiches. Two special cross-sections, one at Nagdiar and another at Bilaiwal (above Abwong), were levelled and the distribution of vegetation along the traces recorded. When recording the distribution, estimates of the percentage of each species were made. Diagrams of the cross-sections and a map showing their position are given in Fig. A 25. The distribution of cultivation is also given along the top of each section. The scale of the diagrams did not allow the distribution of all other types of vegetation (different grass species) to be shown. Along the Bilaiwal section swamp grassland and open water occurred only between 2·109 km. and 2·852 km., a distance of 743 m., the rest of the section being occupied by a grass mixture of Setaria incrassata, Pennisetum ramosum, Rottboellia exaltata, Sorghum sp., etc. (Setaria type grassland) alternating with cultivations of sorghum. When growing in water, the percentage proportions of the deep-flooded type grasses (Vossia cuspidata and Echinochloa stagnina) to shallow-flooded types (Oryza barthii and Echinochloa pyramidalis) were 54 : 46 approximately. Along the Nagdiar section (Fig. A 25) riverain grassland and open water occurred only between approximately 5·315 km. and 5·970 km.-a distance of about 655 m. The riverain swamp type of grassland is not continuous along this length but alternates and is sometimes mixed with Setaria sp. (probably S. lynesii) and Hyparrhenia sp. Further, of the 655 m., 121 m. is open water (i.e. the main channel of the river itself) and only about 45 m. is deep-flooded pasture, and that mostly Vossia cuspidata. Because of the few representatives of the deep-flooded types of grasses along the Nagdiar cross-section, our depth-duration of flooding relationship is based on the distribution along the Bilaiwal cross-section. By relating the distribution of grasses along the cross-sections to flood-plain levels and further relating these levels to a normal flood duration curve produced from the normal ten-day period river levels calculated for the site shown in col. 1, Table 114 (p. 182), we have obtained the figures given in Table 115 (p. 183) for depth and duration of flooding. It will be noted that in the table, under each heading of Depth and Duration there are two lists of figures; (a) for the over-all distribution, (b) for when the species is dominant. No statistical analysis was carried out on these figures, and furthermore, since there is no analysis of the Sobat flood-plain equivalent to that made of the White Nile flood-plain, the distribution of grass species according to gauge level intervals has not been attempted. THE BARR EL GHAZAL The survey along the Bam el Ghazal included two cross-sections (see Fig. A 2) : (i) At Yoynyang. • (ii) From the south bank of the Bahr el Ghazal to the west bank of the Bahr el Jebel. The field techniques and method of obtaining results were basically the same as those used along the River Sobat. YOYNY ANG CROSS-SECTION The reader is referred to Fig. A 21, which gives the distribution of the dominant species of the vegetation across the section and the water-level at the time of the investigation (February 1952). The cross-section is one kilometre in length, of which (approximately) high land and intermediate land grasses take up 10%, riverain swamp grasses 40%, papyrus 45 % and open water (the Bam el Ghazal) 5°/0. Vossia cuspidata is the predominant swamp grass, with some Echinochloa pyramidalis. Then~ is no Echinochloa stagnina. The papyrus is typical of that found along the Bahr el Ghazal, stunted and poor in growth. By relating the levels of flood-plain at which different species grow to a normal duration curve, based on the normal Yoynyang gauge levels in col. 2, Table 114, it is evident that the area dominated by papyrus is permanently flooded for 365 days a year. Table 116 (p . 183) gives maximum and minimum figure.s for depth and duration of flooding for each of the dominant species growing at the site. 159 BARR EL GHAZAL-BAHR EL JEBEL CROSS-SECTION The reader is referred to the relevant cross-section (Fig. A 21) which gives the distribution of the dominant species along the section and the water-level at the time of the investigation. It is assumed that spill from the Bahr el Ghazal is effective as far as approximately 5 km., and from the Bahr el Jebel as far as 6 km. Between 5 km. and 6 km. the effect is probably shared by both rivers. It will be noted from the distribution of species along the top of the section that the major part of the section is dominated by Echinochloa pyramidalis and Oryza barth ii, within which occur isolated clumps of Typha angustifolia. Towards the Bahr el Jebel end the line is dominated by papyrus, the growth of which is strong and vigorous, except for the growth between 6 km. and the maya which is similar to that found along the Bahr el Ghazal at Yoynyang. Vossia cuspidata occurs as a narrow fringe along the maya and the Bahr el Jebel. The section affected by the Bahr el Ghazal is therefore predominantly riverain swamp grassland and that affected by the Bahr enebel is papyrus. Figures for depth and duration given in Table 117 (p. 183) have therefore been based on the relevant duration curves, i.e. the Barn: el Jebel for papyrus and Vossia and the Bahr el Ghazal for river swamp pasture species, which were produced from the normal 10-day period river levels given in Table 114, cols. 4 and 5. BARR EL JEBEL: MALAKAL-BOR From just north of Bor to Lake No papyrus is a dominant species of the flood-plain . Beyond Lake No it continues for some distance eastwards along the south bank, but is super- seded by riverain swamp pasture along the northern bank. The investigation along this stretch included a section near Lake No, in addition to the Bahr el Ghazal-Bahr el Jebel cross-section from which figures for papyrus are applicable (see Table 117). LAKE NO CROSS-SECTION The section is situated just east of Lake No at the E.I.D. P.B.M. 8142. The reader is referred to Fig. A 21, whieh illustrates the section and gives an indication of the distribution of the dominant species '; it has not been possible to illustrate the frequent changes from one species to another which often occur within a few metres. The water-level at the time of the investigation is also given. It will be noticed that papyrus occurs only along the southern part of the section and that the northern part is dominated by riverain swamp and intermediate land grass species. Within, and sometimes as pure patches interspersed with, the papyrus along the southern part are found Echinochloa pyramidalis; Vossia cuspidata and Phragmites communis. A flood duration curve has been produced, based on the normal ten-day period river levels for the Lake No gauge as shown in col. 3, Table 114. ' It is immediately evident from the section that the distributio,l of papyrus is along sites lower than that of riverain swamp pasture. With particular reference to the southern part of the section it is further evident that the major part of papyrus distribution occurs at sites that are permanently flooded. It also occurs at a few sites which only receive 240 days' flooding. At both sites the growth is luxuriant. Depth of flooding varies between a maximum of 250 em. and a minimum of 40 cm. The normal range of river levels between maximum and minimum is 50 cm., Phragmites communis occurring at sites with a "ange of water depth between 40 cm. and.l50 cm. and of flood duration between 240 and 365 days; Vossia cuspidata between depths of 20 cm. and 225 cm. and flood duration of 120 days and 365 days; Oryza barthii between depths of 20 cm. and 140 cm. and flood duration of 120 and 365 days. Echinochloa pyramidalis is dominant over that part of the section which also carries forest , and other elevated sites, and is also found within the papyrus. Depths for Echinochloa pyramidalis vary from 0 (rain- flooding) to '90 cm. and duration from oC rain-flooding) to 365 days, but the major proportions occur at sites receiving between 0 (rain-flooding) and 210 days' flooding. It must be em- phasized that the figures given for the three latter species are not absolute because they probably grow at sites which are outside their normal distribution (Echinochloa pyramidalis was recorded growing in papyrus at greater depths than it normally does). This is explained by the fact that in such cases the plall-ts are rooted to the rhizomatal-humus mass and are therefore floating. For this reason the figures for the last three species should be treated with caution. With particular reference to the noriliern part of the section, Echinochloa pyramidalis is the predominant grass, often mixed with some Oryza barthii over most of ilie section. Vossia cuspidata occurs at the lower levels in the khors. At the higher levels intermediate grassland species-Hyparrhenia sp., Setaria sp. (lynesii) and Setaria incrassata-occur, sometimes mixed with Echinochloa pyramidalis or as pure isolated stands. 160 The most striking point about the northern part of the section is the high deltaic bank of the river, over which the river, even when at normal high level, does not spill. Th{) flood-plain behind is therefore protected. This does not mean that the flood-plain is completely free from flooding; it is very probable that flooding takes place owing to spill from the khors which run into it plus, of course, rain. However, since there is no direct river spill we are unable to relate levels at which species grow to river level on our previous assumptions. We are therefore unable to make calculations for depth and duration of flooding of the different species along this part of the section. THE ALIAB VALLEY The reader is referred to the detailed account of the Aliab Valley survey in Vol. III, especially to Figs. H I and H 3, which illustrate the distribution of grass and papyrus vegetation in relation to contour and flood duration; to the cross-sections Nos. H 4 to H 9, and the longitudinal section (H 10 and H 11) along the bank of the Bahr el Jebel. From these sections and diagrams, and from the information gathered in the course of the survey, a full account of the hydrology, general vegetation and soil distribution, and the utilization of the valley by its seasonal inhabitants in respect of grazing and fisheries has been prepared and recorded. With reference to Figs. HI and H 3, and the general map of the Aliab Valley and cross- sections, it will be noted that the valley is bounded in the east by the deltaic bank of the Bahr el Jebel, and to the west by the high land. Running through the valley is the River Aliab, which has deltaic banks effectively dividing the valley into two units . The eastern part receives direct flooding from the Bahr el Jebel by spill directly over its bank at the 90 mi d discharge (at Mongalla) in the northern part, and at all points at the 120 mi d discharge, and from spill through channels which branch off from the Bahr el Jebel between Tombe and Bor. Spill into these channels takes place at 50 mi d or 60 mi d at Mongalla. Another source-the major source-is from the Bahr el Jebel by backwater from the northern part of the valley. The western part of the valley is barred from flooding from the Bahr el Jebel and receives water only by rainfall and run-off and from Khor Gwir. It will be noted from Fig. H 3 that there is a marked difference between the grass vegetation of the western and eastern portions of the valley. Although the distribution of vegetation has been given for that part of the Aliab Valley surveyed, it is not possible to relate the distribution west of the Aliab channel to flood duration because it does not receive spill from the Bahr el Jebel. In addition, it has not been possible to relate the vegetation of a small part of the northern part of the valley and the two areas south of Khor Ker. From Figs. H 1 and H 3 it is possible to obtain a rough indication of the distribution of papyrus and grasses in the eastern portion of the valley in relation to the duration of flooding. The relationship is shown in Table 118, p. 183. It must be appreciated that these are only rough figures , because it has not been possible to illustrate the distribution accurately in the pictorial diagram, and the relationship is only general. To obtain a more accurate indication, one of the cross-sections (C.S. 13 : Fig. H 8) has been treated in detail in similar fashion to those farther north. The reader is now referred to C.S. 13 (Fig. H 8) between Minkaman and Jerkwot, and especially to that part lying to the east of the Aliab. From this section, maximum, minimum and mean (approx.) figures for depth and duration of flooding have been calculated and are shown in Table 119, p. 184. It will ~e noticed that the figures in Table 119 give a wider range of distribution in terms of duration than those in Table 118, and that the first few listed species all received a maximum flood depth of over 3 m. and a duration of 365 days. At the time of the survey the maximum depth of 340 cm. for Echinochwa pyramidalis was at sites where it was growing in association with Cyperus papyrus and Vossia cuspidata and was therefore probably rooted to the floating humus layer. A more detailed account of the Aliab Valley survey is given in Vol. III, where an estimate of the distribution of grass species on a proportional area basis related to flood duration(54) areas, similar to that made between Malakal and Jebelein, has been attempted. GEMMEIZA-MONGALLA-JUBA REACH The reader is again referred to the survey of these reaches which is given in detail in Vol. III, pp. 823-47. Over the reach Bor to Juba, as we have already explained, the papyrus which is dominant north of Bor gradually loses its dominance on the flood-plain south of Bor, and between Terakeka and Juba, especially on the east bank, Phragmites communis with Echinochloa pyramidalis becomes dominant. 161 11 For the purpose of determining the grass-water relationship an investigation of the vegetation of the Mongalla-Gemmeiza reach was carried out at the same time as the survey, and in Figs. H 12 and H 13 a vegetation map of this reach, as well as a contour map, will be found . From these two maps, together with the flood duration table for different Mongalla discharges-which is related to the contours-it has been possible to determine areas of flood- plain inundated at different discharges and in particular the depth of flood-water and the duration of flooding at the sites affected by a given discharge. The area between Mongalla and Gemmeiza ch.osen for more intensive surveys is truly representative of the whole of the reach. In the northern part of the flood-plain in this reach (i.e. , north of Gernmeiza), where flooding is for a considerable period, papyrus is still dominant, whereas between Mongalla and Terakeka southwards, where flooding is less prolonged, Phragmites communis becomes dominant. The area surveyed lies between the two extremes and contains in variable degrees the main characteristics of both. It has been possible to obtain an indication of the flood duration-grass relationship by simply referring to the contour and vegetation map with the flood duration figures given in Vol. III, but as in the Aliab Valley survey, it was decided to treat two cross-sections in detail. The reader is referred to C.S. 12 (Fig. H 16) which is situated just south of Terakeka and to C.S. 3 (Fig. H 14) which lies south of Gemrneiza and cuts across the lake just south of that point. The distribution of vegetation, other than forest, has not been shown, but in general terms C.S. 12 (Fig. H 16) is dominated by papyrus and C.S.3 (Fig. H 14) by Phragmites communis and Echinochloa pyramidalis jointly. Tables 121 and 122 (p. 184) give for C.S. 12 and C.S. 3 (Figs. H 16 and H 14) the maximum, minimum and mean levels at which species were recorded, the depths at these levels and the duration of flooding. The figures for depth and duration have been calculated from Table 120, p. 184, which gives the levels of flood-plain which become flooded at different Mongalla discharges, and the period of inundation at those levels. The figures are based on the Mongalla discharge figures for the period 1946--50, with double weight given to the 1950 flood (cf. Aliab Valley and Malakal-Jebelein surveys), on the assumption that inundation of the flood-plain starts at the northern end and works southwards until, at 90 mi d at the northern end and 120 mi d at the southern end, the river spills over its banks and the whole flood-plain becomes inundated (for further details see Mongalla-Gernmeiza toich, Vol. III, pp. 839-46). The first point to note from the tables is the difference in levels at which species grow at the two sites; secondly, in the case of Table 121, which refers to C.S. 12 (Fig. H 16), the differences in levels between Echinochloa pyramidalis and Phragmites communis, Cyperus papyrus, Vossia cuspidata and Echinochloa stagnina. This difference in ground levels is further illustrated in the difference in depths of water, but is not so pronounced in duration. It must be remembered, however, that although figures for Echinochloa pyramidalis and Phragmites communis have been given, these species only occur along the section in very limited amounts. Furthermore it must be pointed out that this part of the flood-plain is almost wholly and permanently flooded, and in fact where the figures (means) for Cyperus papyrus and Vossia cuspidata are merely given as greater than 293 days, it is reasonable to assume that the actual figures are nearer 365 days. From Table 122, which is based on C.S. 3 (Fig. H 14), the figures given for depth of flooding are not greatly different from those recorded along C.S. 12 (Fig. H 16) for similar species, but on the other hand there are considerable differences incthe case of duration, especially for Echinochloa stagnina. This requires explanation. The Echinochloa stagnina along this section is limited to the deeper channels and in particular to the deep depression that borders the high land (see Fig. H 13) where, in addition to backwater flooding and river spill, a considerable amount of water is received directly as rainfall and, in particular, as run-off from the high land. The average annual rainfall at Mongalla is 929 mrn. Furthermore the soils of these toiches contain a higher proportion of sand, which facilitates seepage from the main river, giving rise to an underground water-table which is not too distant from the surface (see Fig. H 2). The depth and duration figures given for Echinochloa pyramidalis and Phragmites communis may be assumed to be reasonably correct, but again rainfal1, soil type, and underground water-table must be borne in mind. This concludes the investigation into depth and duration of flooding; a summary table based on mean figures is given for each reach in Table 123, p. 185. CuRRENT VELOOITY No experiments have been carried out to determine the part played by current velocity in limiting the distribution of the plant constituents in the areas· under discussion. It may be 162 assumed, however, that it is only of importance to those constituents that are sited along ~he banks of the main river, since current velocity along the flood-plains during the flood penod is negligible. By observation it has been noted that Vossia cuspidata spreads out farther from the banks of the rivers into mid-stream than any of the other major constituents. Echinochloa stagnina is only found in the backwater khors where velocities are mNch less. It is suggested therefore that Vossia cuspidata can tolerate higher current velocities than other constituents. Migahid(55) has concluded: "The greatest velocity el1dured by Umm-soof (Vossia cuspidata) was 0·15 m/sec., that endured by papyrus was much less. The number of current measurements is, however, insufficient to justify the definition of current values that can be considered as limiting for the different elements of vegetation ". Professor Debenham(56), who carried out an investigation of the Chambeshi in Northern Rhodesia, has demonstrated that Vossia cuspidata can tolerate much stronger currents. In the Chambeshi River, belts of Vossia cuspidata up to 10 feet from the bank were found to be resisting currents up to 2 feet per second (0·68 m/ sec.), i.e., four times as strong as that recorded by Migahid. Table 124, p. 185, shows the average river velocities at four points along the main Nile, one along the Bahr el Zeraf, and one along the Bahr el Ghazal (between Yoynyang and Lake No). The first thing to note from the table is the decrease in velocity between Mongalla and Malakal along the White Nile, the slow velocity of the Bahr el Ghazal and the higher velocity of the Bahr el Zeraf, which approximates to that of the Bahr el Jebel as measured at Buffalo Cape. Along the Bahr el Jebel, Vossia spreads farthest from its banks between Bor and Malakal and north of Malakal. It also occurs as a wide belt along the Bahr el Ghazal. Along the Bahr el Zeraf Vossia cuspidata is found growing at greater distances from the bank than along the Bahr el Jebel. It appears then that the critical velocity is nearer to that recorded by Debenham than to that recorded by Migahid. A further point is that prior to the continuous navigation of the Bahr el Jebel north of Bor (in the Sudd area) by steamers, the main channel was blocked by the accumulation of masses of floating vegetation-mostly papyrus. Similar blockages occur today, but are rare. It is suggested that when blockages along the Bahr el Jebel were frequent, Vossia cuspidata spread farther from its banks towards mid-stream than it does today, and that open water in the actual river channel was narrower. Since then, however, clearing operations have frequently been carried out and the river is now a regular waterway for paddle-wheel steamers whose stern- wheels create considerable backwash, and in turning churn the vegetation at the side of the channels. This frequent backwash must have a considerable effect upon the spreading of the Vossia rhizomes outwards from the bank to mid-stream. By contrast, steamer traffic along the Bahr el Zeraf is less frequent and Vossia spreads farther out towards mid-stream, especially along the upper Zeraf, despite the fact that river velocity is greater than along the Bahr el Jebel as measur~ at Buffalo Cape. It seems clear that the spread is greater because of the absence of regular river traffic. This is borne out by observations on other less frequented channels, where figures for velocity are, however, nofavailable for comparison. OTHER FACTORS THE INCIDENCE OF RAINFALL IN RELATION TO FLOODING AND THE RATE OF RISE IN RIVER LEVELS It is reasonable to assume that if the grass species of the flood-plain were to be suddenly flooded to a depth of, say, 50 cm. before they had had sufficient time to recover from the burning and grazing they receive in the dry season, they would suffer a considerable setback in growth, which might even be fatal. It is possible in fact that they might be ' drowned '. However under the natural river regime they usually have time to recover since they are stimulated to growth by rainfall before they are flooded. In addition the rainfall brings on the inland grasses and the cattle graze increasingly inland, so that the grasses of the flood-plain receive decreasing defoliation. In most years sufficient rainfall has fallen to enable the plants to reach a growth stage at which they are able to ' keep their heads above water' when flooded, and afterwards to keep ahead of the gradually rising flood-water. The interesting and important factors are therefore how much rain is precipitated before flooding commences at the lowest site occupied by each species, and thereafter how the growth rate of the plant compares with the rate of rise of flood-water. 163 It is obviously not necessary to relate rainfall to sites that are permanently flooded; we therefore consider the relationship of the above two factors only along the White Nile flood- plain north of Malakal and the flood-plain between Mongalla and Gemmeiza. The meth.od given for sites north of Malakal is identical with that used along the Mongalla-GemmelZa stretch. In our study of depth and duration of flooding at sites occupied by the four dominant grass species of the White Nile flood-plain, we have related the grass species to the level of the flood-plain above river gauge datum for five sites. The maximum, minimum and average levels of distribution for each species are given in Table 86, p. 169. Here we are only concerned with those minimum levels, since it is at these levels that flooding first commences and they can therefore be regarded as the critical levels. The reader is now referred to Table 86; to the rainfall statistics given in the previous chapter (see p. 51) in which average monthly rainfall figures for Malakal, Renk and Kosti are recorded; and to the tabulated gauge level readings for normal river at Malakal, Renk and Kosti (see pp.170-2 ). From the above data, Table 125 (p. 186) has been compiled, which shows for each site -Malakal, Renk and Kosti-and for each species : (i) Minimum gauge levels for each species (Col. 1). (ii) The date (in IO-day periods) on which the rising flood water reaches the minimum gauge level for each species (Col. 2). (iii) The average rainfall precipitated at each site prior to the date given in Col. 2. Table 126 (p. 186) is related to species found along the Mongalla-Gemmeiza toich and is basically the same, except that Col. 2 gives the Mongalla discharge at which flooding at the levels given in Col. 1 commences. Col. 3 is the date (in 10-day periods) at which that flooding starts-i.e. when the Mongalla spill discharge is reached. Col. 4 is the rainfall (based on Mongalla figures) precipitated prior to the date given in Col. 3. From Tables 125 and 126 it is evident that a considerable amount of rain falls before even the species sited at the lowest level become flooded. Where there is permanent flooding, rainfall is not important-as along C.S. 12 Mongalla-Gemmeiza (Fig. H 16). It must be remembered that the occurrence of Echinochloa pyramidalis and Phragmites communis along this section is very limited and these particular figures are the least reliable. Along C.S. 3 (Fig. H 14) Mongalla-Gemmeiza, it will be remembered that spillover the bank takes place at Mongalla at a discharge of 90 mi d and over, so that almost the whole of this part of the flood- plain is inundated simultaneously-hence the similarity in rainfall figures in Col. 4·for each of the species occurrrng along the section. Furthermore, it will be remembered that Echinochloa stagnina is limited to the fringe of the high land where it receives moisture in addition to direct precipitation by run-off. However, in general the tables illustrate that the non-floating grasses- Echinochloa pyramidalis and Oryza bclfthii-are flooded later than the floating types- Vossia cuspidata and Echinochloa stagnlna- and in consequence they receive a greater amount of rainfall prior to flooding. This gives them time to reach a growth stage which will enable them to tolerate flood conditions. This is confirmed by observation at Malakal, where in 1951 Echinochloa pyramidalis had reached the height of 1 to It m. on rain precipitation alone before flooding from the river began. We have shown above that at the lowest sites already favoured by the different species, sufficient rainfall is precipitated prior to flooding to ell'able them to reach a growth stage sufficiently advanced to keep ahead of the rising flood without being flooded out. At present therefore the rate at which the normal river and the water on the flood-plain rises is of no great importance--except possibly in years when rainfall is extremely late followed by a steep rise in river level, which would simulate the conditions already existing in the open lagoons at present found in certain parts of the flood-plain. .A t present, however, the rate of rise in level is gradual. At Malakal the river starts to rise about the end of April and continues to rise until the beginning of November. At Malakal, too, the highest rates of rise occur in May, June and July, i.e. about or just after the time when flooding begins at the lowest site occupied by Vossia cuspidata and Echinochloa stagnina. When flooding starts at the lowest sites occupied by Oryza barthii and Echinochloa pyramidalis in July, the rate of rise is decreasing. A similar trend is shown at Kosti- the rate of rise in June is low, 1·06 cm./day, but in July and August, the period at which flooding begins at the lowest site of Vossia cuspidata and Echinochloa stagnina, the rate is 3·2 and 5·06 cm./ day respectively. In September and October the rate falls to 2·30 and 0·5 cm./ day, at which time the flooding begins at the lowest sites of Oryza barthii and Echinochloa pyramidalis. In respect of these two latter factors we may therefore draw the following general conclusions : 164 (i) Sites favoured by Vossia cuspidata and Echinochloa stagnina flood earlier than those favoured by Oryza barthii and Echinochloa pyramidalis. (li) Vossia cuspidata and Echinochloa stagnina have a considerably lower total amount of rainfall precipi- tated at their lowest site before flooding commences than do Oryza barthii and Echinochloa pyrami- dalis. (iii) Once flooded, sites occupied by Vossia cuspidata and Echinochloa stagnina are subject to a quicker rise in flood level than sites occupied by Oryza barthii and Echinochloa pyramidalis. CONCLUSIONS A qualitative and quantitative study has been attempted of the factors governing the distribution of individual plant species found along the flood-plain of the main rivers in the 10nglei Area, and we have given for the major species certain ranges of figures between which they may develop. It has also been shown that although the depth of water in the flood season in the case of papyrus, Vossia cuspidata and Echinochloa stagnina is not a limiting factor, Oryza barthii and Echinochloa pyramidalis rarely occur at sites receiving depths greater than 180 to 200 cm. under normal conditions, with an average depth of 120 cm. Further, it has been shown that although Phragmites communis may occur at such depths, or occasionaUy even deeper, it is in fact more frequently found growing at sites which are flooded to a lesser degree. We have seen that soil moisture availability throughout the year is of importance, and that the amount and availability is determined by anyone, or combination of more than one, of the following factors: river spill (depth and duration of flooding), climate (particularly rainfall), soil type and underground water supplies. In general, moisture received by direct rain precipitation alone is insufficient for any of the species found growing along the flood-plain (even with poor drainage), and they are dependent upon the other additional sourC6S. Where the soil of the flood-plain is predominantly clay there is no underground water-table beyond a certain short distance from the bank of the river, and, because of the impermeability of this type of soil, downward penetration of the initial supply by rainfall and river spill and vertical rise by capillary action from a water-table, where present, are limited. In clay soils rich in organic matter the surface penetration is deeper. Shallow-rooted plants such as Cyperus papyrus, Vossia cuspidata and Echinochloa stagnina are suited to such sites (clay flood- plain), provided moisture in the surface layers is available. Such conditions are met with only in areas where the supply is not withheld for long periods, i.e. where the combination of rainfall and river spill is for a long duration. In most cases river spill alone provides the necessary duration, e.g. for papyrus. However, it is a feature of the flood-plain that where the necessary duration period is provided by river spill, it follows that there is a considerable depth of flooding, which limits other species, i.e. Echinochloa pyramidalis and Oryza barthii, from competing at such sites. Thus we find that Cyperus papyrus will develop only on those areas of flood-plain which are supplied with moisture in the surface layers for almost the whole year. The minimum period of duration of flooding from the river as recorded in the Aliab Valley is 175 days, but this is supplemented by rainfall, and in general the period of flooding is more frequently in the region of 300 to 365 days. Similarly, in the case of Vossia cuspidata and Echinochloa stagnina, which have the lowest minima of 172 days and 167 days respectively (Sobat), the means are approximately 263 and 230 days respectively. The deeper-rooted plants, Echinochloa pyra- midalis, Oryza barthii and Phragmites communis, can reach moisture supplies that are deeper in the soil and do not rely upon moistuie in the surface layers only. Thus they are able to tolerate sites that are exposed for longer periods, i.e. where moisture supply by rainfall and river spill is cut off for approximate mean periods of 152 .and 150 days for Echinochloa pyramidalis and Oryza barth ii, and longer for Phragmites communis. Where the soil of the flood-plain is of a mixed or sandy nature it frequently happens that there is an underground water-table (dry season), i.e. Aliab Valley and Mongalla-Gemmeiza toiches. Where the height of flood-plain level is over a metre or two above river level (low season), shallow-rooted plants such as papyrus and Ecinochloa stagnina are unable to benefit from the supply, but deeper-rooted plants may do so. At such sites inundation from the river does not occur or occurs only rarely, and supply is only by precipitation in the rainy season and by capillary rise from a water-table in the dry season . Soil moisture in the upper layer is not permanent and often dries out. At such sites the shallow-rooted plants (Cyperus papyrus, Vossia cuspidata and Echinochloa stagnina) cannot exist, but suitable conditions do occur for the deeper-rooting Phragmites and Echinochloa pyramidalis. Thus on the southern toiches of the Bahr el Jebel, where such conditions exist, Phragmites and Echinochloa pyramidalis are dominant species, and in general Echinochloa stagnina, Vossia cuspidata and Cyperus papyrus are absent. Where, however, the level of flood-plain is less than a metre above river water- level and the water-table is near the surface, which usually coincides with longer duration of 165 flooding from the river, there is a further prolongation of moisture supply in the upper layer and the shallow-rooted plants can establish themselves. Thus behind the deltaic bank of the Bahr el Jebel in the Aliab Valley and Mongalla-Gemmeiza flood-plain, where a high water-table exists in the dry season and flooding is more prolonged in the wet season, Echinochloa stagnina has become established. Current velocity is a limiting factor in the spread of some plants. Firmly anchored plants, such as Phragmites communis, Echinochloa pyramidalis and Oryza spp., are not affected by velocities, since they are sited well within the flood-plain and not subject to the swifter currents of the river as are the floating types which often extend from the bank along the surface of the water towards mid-stream. Vossia cuspidata has been found to stand greater current velocities than either papyrus or Echinochloa stagnina. The importance of a sluggish velocity in relation to sudd blocks and steamer traffic has also been stressed. Grazing, firing, rainfall, and the rate of rise in flood-waters have been examined in relation to the distribution of species of riverain swamp pasture plants, and certain general conclusions have been reached. Finally, competition has been considered in general terms, and it has been stressed that competition within the papyrus swamp is mostly for light, and that in general it is the taller-growing plants that are successful at sites suitable for more than one species . .' 166 TABLE 83 ECOLOGICAL CLASSIFICATION OF THE VEGETATION WITlllN THE AREA AFFECTED BY THE EQUATORIAL NILE PROJECT Rainfall Soil Main Vegetation Units within Regions Major Regions miIlimetres I I I 1,300 LATERITIC a. Mixed deciduous broad-leaved woodland (with tall perennial grasses) IRONSTONE REGION to 1,000 CATENA b. Forests CLAY FORESll a. Palm type b. Poorly developed deciduous broad-leaved type and HIGH LAND c. Mixed acacia dominated type d. Acacia seyal-Balanites aegyptiaca type SANDS GRASSLAND 1,000 FOREST Acacia seyal-Balanites aegyptiaca type l/REDOMINANTLY FLOOD REGION to INTERMEDIATE LAND GRASSLAND a. Hyparrhenia rufa type CLAY b. Setaria incrassata type 700 ...... - PREDOMINANTLY a . Permanent papyrus swamp 0\ --.l SWAMP CLAY and (SEASONAL AND b. Riverain and inland khor-bed PERMANENT) grassland (see Inter-Regional) PEAT 700 TRANSITIONAL BELT to CLAY Acacia seyal-Balanites aegyptiaca forests with hariq grasslands 600 600 to CLAY Acacia mellifera thickets with open areas of mixed tall and short annual grasses 400 SEMI-ARID REGION Under QOZ Acacia-short grass vegetation 400 CATENA INTER-REGIONAL 1,100 PREDOMINANTLY COMMON TO to Main riverain swamp grassland and inland khor-bed grassland ALL REGIONS 400 CLAY TABLE '84 RECORDED DEPTHS OF FLOODING-MALAKAL TO KOSTI (Flood depths recorded for four grass species along the White Nile flood-plain between Malakal and Kosti) Site Date Gauge Reading Species Depth in centimetres I I I Malakal ... ... 29.12.50 12-40 Vossia cuspida la ... .. . 180, 130 Echinochloa slagnina ... ... 180, 130, 120, 120, 100 Oryza barlhii ... ... .. . 130, 100, 100, 100, 80, 70, 70 Echinochloa pyramidalis ... 100, 100, 90, 90, 80, 70, 70, 70, 70,70,70,60,50,50,40,40,30 Melut ... .. . 16. 1.51 11 ·86 Vossia cuspidata ... .. . 280, 250, 220, 210, 210, 200, 200, 200, 170 Echinochloa slagnina .. . . .. 200, 180, 160, ISO, 160, ISO, 140, 130, 130, 130, 90, 90, 80 Oryza barlhii ... ... .. . 130, 11 0, 11 0, 90, 80, 80, 70, 60, 60, 20, 10, 0, 0 Echinochloa pyramidalis ... 160, 130, 110, 110, 100, 100,90, 80,80,70, 60, 60,60,40, 40, 40, 40, 30, 30, 10, 0, 0, 0, 0, 0, 0 Renk .. . ... 17. 1.51 12·03 I Vossia cuspidala .. . . .. ISO, ISO, ISO, ISO, 120 I Echinochloa slagruna ... .. . 180, ISO, ISO, 140, 120, 120, 120, 11 0, 110, 95, 90, 90 I I Oryza barlhii ... ... .. . 150,130,130,120,110,110,110, I 100,95, 95, 9$, 90, 90, 90, 90, 90, 90, 85, 85, 80, 80, 80, 80, I 80, 75, 75, 65, 60, 50, 50, 50, I 50, 40, 40, 25, 25, 20, 15 I I Echinochloa pyramidalis ... I 95, 90, 90, 90, 85, 80, 80, 80, 50, 50, 50, 50, 40, 40, 25, 25, 20, 20, 20 I Jebelein ... 18. 1.51 12·53 Vossia cuspidala ... .. . 200,190,175,170,170, 160, ISO, .. · 1 I ISO, 145, 140, 130, 130, 130, 120, 120, 110, 100, 50 L Echinochloa slagnilla ... ... 185, 170, 170, 170, 170, 170, 160, 160, 160, ISO, 150, 130, 130, 130, 11 0, 11 0, 100, 50 Oryza barthii ... ... ... 160,150, ISO, ISO, 140, 140, 130, 130, 130, 120 Echinochloa pyramidalis ... o to 30 ------ -. -- --- - - - Kosti (Rabak) ... 20. 1.51 13·93 Vossia cuspidala ... .. . ISO, 110, 90, 85, 80, 65, 65 Echinochloa slagnina ... ... 120 . Oryza barlhii ... ... .. . 80 Echinochloa pyramidalis ... 110, 90,65,60, 55, 45, 40, 3D, 3D, 30, 25, 20, 20, 15 168 TABLE 85 MAXIMUM, MINIMUM AND AVERAGE DEPTHS RECORDED Vossia cuspidata Echirzochloa stagrzirza Oryza barthii Echinochloa pyramidalis Gauge Site Date Reading Maximum IM inimum I Average IM aximum I Minimum I Average Maximum I Minimum I Average Maximum I Minimum I Average I I Malakal ... ... .. . .. . . .. .. . 1 29.12.50 I 12·40 180 130 I 155 180 100 128 130 70 92 100 30 70 Melut ... ... ... ... ... .. . 16. 1.51 11 ·86 280 170 215 200 80 138 130 0 62 160 0 55 Renk ... .. . ... .. . .. . .. . 17. 1.51 12·03 150 120 144 180 90 123 150 15 79 96 20 57 lebelein ... ... .. . ... .. . .. . 18. 1.51 ~ 12·53 200 50 142 185 50 143 160 120 139 30 0 15 Kosti (Rabak) ... ... ... .. . ... 20. 1.51 13-93 150 65 90 120 120 120 80 80 80 110 15 45 I I 0-, \0 TABLE 86 RECORDED HEIGHTS OF GRASS SPECIES ABOVE RIVER GAUGE DATUM Malakal Melut Rerzk Jebeleirz Rabak Minimum IM aximum I Average Minimum IM aximum I Average Minimum IM aximum I Average Minimum IM aximum I Average Minim~ IM aximum I Average Vossia cuspidata ... ... .. . 10·60 11·10 10·85 9·06 10·16 9·71 10·53 10·83 10·59 10·53 12'03 IHI 12-43 13'28 13·03 Echinochloa stagrzina .. . .. . 10·60 11-40 11 ·12 9·86 11·06 10'48 10·23 11 ·13 10·80 10·68 12·03 11·10 12·73 12·73 12·73 Ory7P barthii ... .. . .. . 11-10 11 ·70 11·48 10·56 11 ·86 11·24 10·53 11 ·g8 H ·24 10·93 11 ·33 11-14 13-13 13·13 13-13 Echoinochla pyramidalis . .. ... 11'40 12'10 11 ·70 10·26 11'86 11·31 11 ·07 11·83 11 ·46 12·23 12·53 12·38 12-83 13·78 13-48 I I TABLE 87 GAUGE-DURATION MALAKAL Number of ten-day periods in which gauge equalled or exceeded T.D.M. Weighted Gauge Average 1946-7 1947- 8 1948-9 1949-50 1950-51 9·50 36 36 36 36 36 36 9·75 34 36 36 36 36 36 10·00 33 36 36 36 34 35 10·25 29 30 33 31 31 31 10·50 26 26 27 27 27 27 10·75 26 29 25 25 23 24 11·00 23 22 24 23 21 22 11 ·25 21 21 22 21 20 21 11 '50 20 19 19 20 18 19 11·75 19 16 17 17 16 17 12·00 18 14 15 15 14 15 12·25 17 11 12 13 11 12 12·50 16 0 0 10 7 7 12·75 13 0 0 0 0 2 13·00 o o o o o o NOTE: (I) Yean an: May !sl-April30th. (2) 1950-51 is given doubJe weight in averaao. TABLE 88 GAUGE-DURATION MELUT Number of ten-day periods in which gauge equalled or exceeded T.D.M. Weighted Gauge Average I 1946-7 I 1947 -8 I 1948-9 I 1949-50 I 1950-51 9·25 36 36 36 36 36 36 9·50 35 36 36 36 36 36 9'75 34 36 36 36 36 36 10·00 32 34 35 35 34 34 10'25 32 31 29 29 31 31 10·50 27 26 26 , 26 26 26 10·75 26 24 25 25 22 24 11·00 23 22 23 22 21 22 11·25 22 19 21 20 18 20 11 ·50 20 17 18 16 17 18 11·75 19 16 15 15 15 16 12·00 19 13 13 14 12 14 12'25 17 9 0 II 8 9 12·50 15 0 0 0 0 2 12·75 1 0 0 0 0 0 13·00 o o o o o o NOTE: (1) Years arc May 1st-April 30th. (2) 19,0-51 is given double weight in the avera,e. 170 TABLE 89 GAUGE-DURATION RENK Number of ten-day periods in which gauge equalled or exceeded Weighted Gauge Average 1946--7 1947-8 1948- 9 1949-50 1950-51 9·75 36 36 36 36 36 36 - ~ 10·00 33 36 36 36 36 36 10·25 32 34 33 34 34 34 10·50 30 31 31 30 31 31 10·75 28 27 28 27 27 27 11'00 24 24 25 24 24 24 11 ·25 23 20 22 20 20 21 11'50 21 17 18 18 17 18 11·75 19 15 14 16 14 15 12·00 -17 12 11 13 11 12 12·25 16 1 0 0 0 3 12·50 12 0 0 0 0 2 12·75 0 0 0 0 0 0 NOTE: (\) Years are May 1st-April 30th. (2) 1950--51 is given double weight in the average. TABLE 90 GAUGE-DURATION JEBELEIN Number of ten-day periods in which gauge equalled or exceeded Weighted Gauge Average 1946--7 I 1947-8 [ 1948-9 I 1949-50 I 1950-51 9·50 36 36 36 36 36 36 9·75 36 36 36 36 36 36 10·00 33 36 36 36 36 36 10·25 31 35 33 34 35 34 10·50 29 32 31 30 29 30 10·15 28 28 . 28 26 27 27 11·00 27 25 26 25 24 25 11·25 26 24 25 . 23 23 24 11 ·50 25 22 23 22 22 23 11 ·75 23 20 21 21 20 21 t2'00 21 17 18 18 17 18 12·25 19 15 16 IS IS 16 12'50 17 12 10 11 9 11 12·75 12 0 0 0 0 2 13·00 0 0 0 0 0 0 NOTE : (l) The year 1950-51 IS gIven double WClaht m the 8veraSe. (2) Years arc taken as May lst-April 30th. 171 TABLE 91 GAUGE-DURATION RABAK Number of ten-day periods in which gauge equalled or exceeded Gauge Weighted 1946--7 1947-8 1948-9 Average 1949-50 1950-51 9'50 36 36 36 36 36 36 9·75 35 36 36 36 36 36 I I I 10·00 33 36 36 36 36 36 10·25 31 36 33 34 36 34 10·50 30 36 31 32 34 33 10·75 29 34 30 30 30 30 11·00 28 31 29 29 27 28 11·25 28 30 29 27 27 23 11·50 28 29 28 26 25 27 11 ·75 27 28 27 26 25 26 12·00 27 27 27 25 25 26 12·25 27 26 25 24 24 25 12·50 26 23 25 23 23 24 12·75 26 22 23 23 22 23 13·00 24 21 21 21 21 21 13·25 22 18 19 19 18 19 13·50 20 16 17 17 17 17 13·75 18 13 15 15 14 15 14·00 8 4 0 0 I 0 2 14·25 0 0 0 Ie 0 I 0 0 I NOTE: (1) Tho year 1950-51 IS given double weight in the average. (2) Years are taken as May 1st-April 30th. 172 TABLE 92 DEPTHS OF FLOODING OF GRASS SPECIES MALAKAL TO JEBELEIN I MALAKAL MELUT RENK JEBELElN RABAl{ OVER-ALL AVERAGE I I Max. I Min. I Av. Max. I Min. Av. Max. I Min. Av. Max. I Min. Av. Max. I Min. Av. Max. Min. Av. I I I I I I Vossia cuspidata ... ... .. . 215 I 165 190 344 234 280 197 167 191 222 72 163 I 157 72 99 227 142 184 Echinochloa stagnina ... .. . .".'.1 215 I 135 165 264 144 202 227 137 170 207 72 165 127 I - 127 208 122 165 Oryza sp. (barthii) ... ... .. . 165 105 128 194 RF 127 197 62 126 182 144 162 87 - 87 165 78 121 Echinochloa pyramidalis ... .. . . .. 135 65 • 103 224 RF 119 142 67 104 52 RF 37 117 I 22 52 134 31 82 I R.F. = Rain Flooding . ..... -.J w TABLE 93 DURATION OF FLOODING OF GRASS SPECIES MALAKAL TO JEBELEIN I MALAKAL MBLUT I RENK I JEBELElN RABAK OVER-ALL AVERAGE I I Max. I Min. Av. I Max. : Min. Av. Max. I Min. I Av. I Max. I Min. Av. I I Max. I Min. Av. Max. I Min. Av. I 1 I I I Yossia cuspiliata ... ... ... .. . 260 220 230 365 325 360 310 260 290 290 180 250 240 190 210 I 293 233 268 Echinochloa stagnina ... .. . .. . I 260 200 210 360 220 260 350 230 260 280 180 250 230 - 230 296 207 242 Oryza sp. (barthii) ... .. . ... 220 180 190 250 RF 200 300 140 210 260 240 250 200 - 200 246 140 210 Echinochloa pyramidalis ... .. . .. . 200 140 180 310 RF 200 , 240 140 190 160 100 1\10 220 [50 170 226 106 '170 - - R .F. ~ Rain Floodiug. TABLE 94 NORMAL CURVE DISTRIBUTION OF GRASS SPECIES MALAKAL Vossia cuspidata Echinochloa stagnina I Oryza barthii \ Echinochloa pyramidalis Gauge I Col. I I Col. 2 Gauge I Col. I I Col. 2 Gauge I Col. I I Col. 2 Gauge I Col. I I Col. 2 9·75 10·00 ·500 10·00 ·016 10·25 ·484 10·25 ·500 10·25 ·090 ·013 10·50 ·394 10·50 ·487 10·50 ·500 10·50 ·254 ·081 ·005 10·75 ·140 10·75 '406 10·75 ·495 10·75 ·500 --- - - - ---- ·241 ·038 ·006 11·00 ·165 11 ·00 ·457 11·00 -494 ·163 ·048 10·85 ·379 11·25 ·294 11·25 ·446 --------- ·182 --------- 11·12 ·344 11 ·50 ·264 11·48 ·323 --------- 11·00 ·239 ·184 11·70 ·336 11·25 ·423 11·25 ·179 --------- ·067 ·234 11·50 '490 11·50 ·413 11·50 ·029 --------- ·010 ·081 ·297 11·75 ·500 11·75 ·494 11 ·75 ·326 11·75 ·072 --------- ·006 ·142 ·285 12·00 ·500 12·00 ·468 12·00 ·357 ----------- ·029 ·118 12-25 ·497 12·25 '475 ·003 ·023 12·50 ·500 12·50 ·498 --------- ·002 12·75 ·500 TABLE 95 NORMAL CURVE DISTRIBUTION OF GRASS SPECIES MELUT Vossia cuspidata [ Echillochloa stagnina .[ Oryza barthii __ I Echinochloa pyramidalis " Gauge I Col. 1 I Col. 2 Gauge I Col. I I Col. 2 Gauge I Col. I I Col. 2 I Gauge I Col. 1 I Col. 2 -------- 8·75 ·500 ·006 9·00 -494 ·045 9·25 ·449 ·176 9·50 ·273 9·50 ·500 --------- ·005 9·75 ·495 ·038 9·71 ·330 10·00 ·457 '163 10·25 '294 10·25 ·500 10·25 ·500 --- --------------- ·004 ·002 10·50 ·496 10·50 ·498 9·75 ·057 ·021 10·48 ·322 ·036 ·293 10·75 '460 10·75 ·477 10·00 ·350 ·156 ·1 II ·123 --------- 11 ·00 ·304 11·00 ·366 10·25 ·282 '473 --------- ·024 11 ·25 ·084 10·50 ·497 10·50 ·028 11'24 ·320 - - - I- ·003 ·304 10·75 ·500 10·75 ·332 11·31 ·336 ------ ·136 ---------11 ·00 ·468 ·029 --------- 11·25 '497 11·25 ·016 ·003 ·307 11·50 '500 11·50 ------ ·323 11·50 ·252 - - - '142 '189 11'75 '465 11·75 ·441 ·032 ·052 12·00 '497 12·00 ·493 ·003 ·007 12·25 '500 12·25 ·500 174 TABLE 96 NORMAL CURVE DISTRIBUTION OF GRASS SPEOIES RENK Vossia cuspidata Echinochloa slagnina Oryza barlhii Echinochloa pyramidalis Gauge I Col.I I Col. 2 Gauge Col. I I Col. 2 Gauge Col. I Col. 2 Gauge Col. I Col. 2 ---- --------- -- 9·50 ·500 ·001 9·75 '499 9'75 ,500 ·016 ·002 10·00 ·483 10·00 '498 I ·094 ·023 10·25 ·389 10·25 ·475 10·25 ·500 ·263 '118 ·004 10·50 '126 10'50 ·357 10·50 -496 10·50 ·500 10·59 ·343 ·286 ·036 '006 10·75 ·217 10·75 ·071 10·75 ·460 10·75 ·494 ·210 10·80 I ·334 ·156 ·045 11 ·00 ·427 11·00 ·263 11·00 ·304 11·00 ·449 - - - - ----- ·064 ·183 11 ·24 ·318 ·176 - -------- 11 ·25 ·491 11·25 '446 11·25 '014 11·25 ·273 -.. - - ------ ·009 ·048 ·309 1\046 ·330 ---~ --- 11·50 ·500 11·50 ·494 11 ·50 ·323 11·50 ·057 ·006 ·142 ·289 11-75 ·500 11·75 ·465 11·75 ·346 ·032 ·127 12·00 '497 12·00 ·473 ·003 ·024 12·25 ·500 12·25 ·497 ·003 ------ - -- 12'50 '500 TABLE 97 NORMAL CURVE DISTRIBUTION OF GRASS SPECIES JEBELElN Vossia cuspidala Echinochloa slagnilla I Oryza barthii Echinochloa pyramidalis - - - -------_. - Gauge Col. I Col. 2 G auge Col. I Col. 2 Gauge [ Col. I Col. 2 Gauge Col. I Col. 2 --------- 10·00 10·00 10'00 10·25 ·500 10·25 ·500 10·25 ·500 ·015 ·106 ·011 10·50 '485 10·50 ·484 10·50 -489 ·084 ·090 ·072 10·75 ·401 10·75 ·394 10·75 -417 ·249 ·255 ·226 11·00 ·152 11·00 ·139 11·00 ·191 -------------------- - ----- - -- 11·11 ·343 11·10 ·343 11·14 ·343 --- --------- 11 ·25 '191 11·25 ·204 11·25 ·152 ·226 ·220 ·249 11·50 ·417 11·50 ·424 11 ·50 ·401 11·50 ·500 ·072 ·063 ·084 11·75 ·013 '489 11·75 ·487 11·75 ·485 11·75 ·487 ·011 ·013 ·005 12·00 ·500 12·00 ·074 ·500 12·00 ·500 12·00 '413 ·235 12·25 ·178 ------- -- 12·38 ·343 --------- 12·50 ·165 '241 12·75 ·406 ·080 13 ·00 ·486 ·014 13·25 '500 175 TABLE 98 RECORDED PROPORTIONS OF GRASS SPECIES MALAKAL TO JEBELEIN Site Vossia Echinochloa 1 Oryza Echinochloa cuspidala slagnina barthii pyramidalis MALAKAL 19 3 570 703 MELUT __ _ 39 240 28 428 RENK 88 57 336 667 JEBELEIN __ _ 10 68 21 TABLE 99 PRODUCT OF NORMAL CURVE DISTRIBUTION AND RECORDED PROPORTIONS OF GRASS SPECIES MALAKAL Malakal Gauge Vossia Echinochloa Oryza Echinochloa < cuspidata stagnina barthii pyramidalis __1 Sum ----- 10-00 \., -304 -304 10-25 1-710 -039 1-749 10-50 4-826 -243 2-850 7-919 10-75 7-201 -723 21-660 4-218 33-802 11-00 3-496 1-032 92-910 33 -744 131-182 11-25 j -273 -702 184-110 127-946 314-031 II -50 -190 -243 169-290 236-208 405-931 11-75 -018 80-940 200-355 281-313 12-00 16-530 82-954 99-484 12-25 1-710 16-16~ 17-879 12-50 1-406 1-406 12-75 TABLE 100 PRODUCT OF NORMAL CURVE DISTRIBUTION AND RECORDED PROPORTIONS OF GRASS SPECIES MELUT Melut Gauge Vossia Echinochloa Oryza Echinochloa clispidaia stagnina barlhii pyramidalis Sum 8-75 -234 -234 9-00 1-755 1-755 9-25 6-864 6-864 9-50 12-870 1-200 14-070 9-75 12-427 9- 120 20-547 10-00 4-797 39-120 10-25 43-917 -936 77-280 -112 -856 79-184 10-50 -117 72-960 1-008 8-988 83-073 10-75 32-640 4-368 47-508 84-516 11-00 6-960 8-960 120-696 136-616 -.. 11-25 -720 8-596 143-808 153-124 II -50 3-976 80-892 11 -75 84-868 -896 22-256 23-152 12-00 -084 2-996 3-080 12-25 176 TABLE 101 PRODUcr OF NORMAL CURVE DISTRIBUTION AND RECORDED PROPORTIONS OF GRASS SPECIES RENK Renk Gauge Vossia Echinochloa Oryza Echinochloa cuspidata I stagnina barthii pyramidalis Sum 9·50 ·088 I I I ·088 9·75 I 1'408 ·114 1'522 10·00 I 8·272 1·311 I 9'583 10·25 I 23-144 6'726 1·344 I 31 ·214 10·50 I 30·184 16'302 12·096 I 4·002 I 62'584 10·75 ! 18·480 19·038 52-416 30·015 I 119·949 11·00 5·632 10·431 106·848 117·392 I 240·303 11·25 ·792 2·736 103-824 I 220'110 I 327-462 11·50 I ·342 47·712 192·763 240'817 11·75 I 10·752 I 84'709 I 95-461 12·00 1·008 16·008 17·016 12·25 2·001 2·001 13-00 TABLE 102 PRODUCT OF NORMAL CURVE DISTRIBUTION AND RECORDED PROPORTIONS OF GRASS SPECIES JEBELEIN Echinochloa Oryza lebelein Gauge ,V ossia Echinochloa cuspidata stagnina barlhii pyramidalis Sum 10·00 10·25 ·15 1'09 ·01 1·25 10·50 ·84 6' 12 ·07 7·03 10·75 2-49 17·34 ·23 20·06 11·00 3-43 23·32 ·34 27·09 11·25 2·26 14·96 ·25 17·47 11·50 ·72 4'28 ·08 ·27 5·35 11·75 ·11 ·88 ·01 1·55 2 ·55 12·00 4·94 4·94 12-25 7-20 7-20 12·50 5·06 5·06 12·75 1·68 1·68 13-00 ·29 ·29 TABLE 103 PERCENTAGE-DISTRIBUTION OF GRASS SPECIES MALAKAL Malakal Gauge Vossia Echinochloa Oryza Echinochloa cuspidata stagnina barthii pyramidalis 10·00-10'25 100% 10·25- 10·50 98% 2% 10·50-10·75 61 % 3% 36% 10·75- 11 ·00 21 % 2% 64% 13% 11·00-11·25 3% 1% 71 % 25% 11·25- 11 ·50 59% 41 % 11·50-11·75 42% 58% 11·75- 12·00 29% 71 % 12·00-12-25 17% 83% 12·25- 12·50 10% 90% 12·50-1 2·75 100% 177 12 TABLE 104 PERCENTAGE DISTRIBUTION OF GRASS SPECIES MELUT Vossia Echinochloa Oryza Echinochloa Melut Gauge cuspidata stagnina I barthii I pyramidalis 1 8·75- 9·00 100% 9·00- 9·25 100% 9-25- 9·50 100% 9-50- 9·75 92% 8% 9·75-10'00 56% 44% 10-00-10'25 11% 89% 10·25-10·50 1% 98% 1% 10·50-10·75 88% 1% 11% 10·75-11·00 39% 5% 56% 11 ·00-11 ·25 5% 7% 88% 11 -25-11 '50 6% 94% 11'50-11 -75 5% 95% 11'75- 12-00 4% 96% 12·00-12·25 3% 97% TABLE 105 PERCENTAGE DISTRIBUTION OF GRASS SPECIES RENK Renk Gauge Vossia Echinochloa Oryza Echinochloa 1- cuspidata slagnina barthii pyramidalis 9·50- 9-75 100% 9-75-10-00 93% 7% 10·00-10·25 86% 14% 10'25-10-50 74% 22% 4% 10·50-10·75 48% 26% 19% 7% 10·75-11·00 15% 16% 44% 25% 11 -00-11 -25 2% 4% 45% 49% 11·25-11 ·50 1% 32% 67% 11-50-11 -75 20% 80% 11 -75-12'00 11% 89% 12·00-12·25 6% 94% 12·25- 12·50 100% TABLE 106 PERCENTAGE DISTRIBUTION OF ORASS SPECIES JEBELEIN Jebe1ein Gauge Vossia Echinochloa Oryza Eclrinochloa cuspidata stagnina I barthii pyramidalis 1 10'25- 10-50 12% 87% 1% 10'50-10·75 12% 87% 1% 10'75-11·00 12% 87% 1% 11·00-11 ·25 13% 86% 1% 11 ·25- 11 '50 13% 86% 1% 11·50-11 ·75 13% 80% 2% 5% 11·75-12·00 4% 35% 61 % 12·00-12·25 100% 12·25-12·50 100% 12-50-12'75 100% 12·75-13·00 100% 178 TABLE 107 MEAN PERCENTAGE DISTRIBUTION OF GRASS SPECIES MALAKAL-MELUT Vossia Echinochloa Oryza & hinochloa Av. Gauge Interval cuspidata stagnina barthii pyramidalis I Mkl. Mit. Av. I Mkl. Mit. Av. Mkl. Mit. Av. I Mkl. Mlt. Av. 8·75- 9·00 100 50 I 9·00- 9·25 100 50 9·25- 9·50 100 50 9·50- 9·75 92 46 8 4 9·75- 10·00 56 28 44 22 10·00-10·25 100 \I 56 89 44 10·25-10·50 98 I 49 2 98 51 I 10·50-10·75 61 31 3 88 46 36 I 18 \I 5 10·75-11·00 21 10 2 39 20 64 5 35 13 56 35 11·00-\1·25 3 2 I 5 3 71 7 39 25 88 56 I I ,25-\1·50 $9 6 32 41 94 68 \1'50-\1'75 42 5 23 58 95 77 \1·75-12·00 29 4 16 71 96 84 12·00-12·25 17 3 10 83 97 90 12·25-)2·50 10 0 5 90 45 12·50-12·75 100 50 Abbreviations: Mkl.= MaJakal Mlt.=Melut Av.=Averaae for reacb. TABLE 108 MEAN PERCENTAGE DISTRIBUTION OF GRASS SPECIES MELUT-RENK Vossia Echinochloa Oryza Echinochloa Av. Gauge Interval cuspidata stagnina barthii pyramidalis MIt. Rk. Av. i MIt. Rk. Av. MIt. Rk. Av. MIt. Rk. Av. 8·75- 9·00 100 50 9·00- 9·25 100 50 9·25- 9·50 100 50 I 9·50- 9·75 92 100 96 8 4 9·75-10·00 56 93 75 44 7 25 10·00-10·25 \I 86 48 89 14 52 10·25-10·50 1 74 37 98 22 61 4 2 1 0 10·50-10·75 48 24 88 26 57 I 19 10 \I 7 9 10·75-11·00 15 7 39 16 28 5 44 24 56 25 41 11·00-\1 ·25 2 1 5 4 4 7 45 26 88 49 69 1\·25-11 ·50 1 0 6 32 19 94 67 81 11·50-11·75 I 5 20 12 95 80 88 11·75-12·00 4 11 7 96 89 93 12·00-12·25 3 6 4 97 94 96 12·25-12·50 100 50 12'50-12·75 I I Abbreviations: Mlt.-=Melut. Rk.-Renk A v. = A verago for reach. TABLE 109 MEAN PERCENTNJE DISTRIBUTION OF GRASS SPECIES RENK-JEBELEIN Vossia &hinochloa Oryza Echinochloa Av. Gauge Interval cuspidata stagnina ! barthii pyramidalis \ Rk. Jeb. Av. Rk. Jeb. Av. I Rk. Jeb. Av. I Rk. Jeb. Av. 9·50- 9·75 100 50 9·75-10·00 I 93 47 7 3 10·00-10·25 86 43 14 7 10·25-10·50 I 74 12 43 22 87 55 4 1 2 10·50-10·75 I 48 12 30 26 87 57 19 I 10 7 3 10·75-11·00 15 12 13 16 87 53 44 1 22 25 12 I 1·00-1 I ·25 2 13 7 4 86 46 45 1 23 49 24 11·25-11·50 I 13 6 1 86 44 32 I 16 67 34 11·50-11·75 I 13 6 80 40 20 2 11 80 5 43 11·75-12·00 4 2 35 18 11 5 89 61 75 \2·00-12·25 I 6 3 94 100 97 12·25-12·50 100 100 100 12·50-12·75 100 50 12·75-13·00 100 50 13·00-13·25 100 50 Abbreviations: Rk.= llcnk Jeb, = lebelein AT .... J\.VefaIC for reach. 179 TABLE 110 AREAS OF FLOOD-PLAIN OCCUPIED BY GRASS SPECIES MALAKAL-MBLUT Vossia cuspidata Echinochloa stagnina Oi"yZa barthii Echinochloa pyramidalis TOTAL Av. Gauge Interval Increment Total Increment Total Increment Total Increment Total Increment Total sq. kIn. sq. kIn. sq. fan. sq. km. sq. kIn. sq. fan. sq. km. sq. kIn. sq. kIn. sq. km. 10·00-10·25 4·8 4·8 3·8 3-8 o 0 o 0 8·6 8·6 10·25-10'50 6·3 11·1 6·5 10·3 o 0 o 0 12·8 21·4 10·50-10·75 3·3 14·4 4·8 15'1 1·9 1·9 0·5 0·5 10·5 31·9 10·75-11·00 1·1 15·5 2·2 17·3 4·0 5·9 4·0 4·5 n'3 43·2 11·00-11·25 ·3 15·8 ·4 17-7 5·4 11 ·3 7-8 '12·3 1309 57-1 11 ·25-11·50 o 15·8 o \7·7 5·9 17·2 12·5 24·8 18·4 75·5 11 ·50-11·75 o 15·8 o 17·7 13-2 30·4 44·3 69·1 57·5 13·3 11 ·75-12·00 o 15·8 o 17-7 8·0 38·4 42·0 111·1 50·0 183·0 12·00-12·25 o 15·8 o 17·7 5·0 43-4 45·0 156·1 50·0 233·0 12·25-12·50 o 15·8 o 17·7 1·8 45·2 15·7 171·8 35·0 268·0 12'50-12·75 o 15·8 o 17·7 o 45·2 11·0 182·8 22·0 290·0 12·75-13·00 o 15·8 o 17·7 o 45 ·2 o 182·8 113·0 303·0 SUMMARY OF TABLE CI) Total Area of Flood-Plain ?n:q:~ (3) The remaining 41'S sq. km., since it arises o ut of the petering-out (2) Area of Yossia cusp/data of the total grass distribution above level 12-25. is assumed to be " Echinochloa slagnlna 17·7 sq. km . occupied by intermediate land grass species. .. Oryza ba,/I,,'; ... . .. 45·2 sq. Rm . - .. Echillochloa pyramidalis 182·8 sq . km. TOT:"L 261~~ sq. km. e00 TABLE III AREAS OF FLOOD-PLAIN OCCUPIED BY GRASS SPECIES . MELU'T-RENK ! Vossia cuspidata Echinochloa stagnina Oryza barthii Ecllinochloa pyramidalis I TOTAL Av. Gauge Interval. I Increment I Total Increment Total Increment Total Increment Total Increment Total sq . kIn. sq. km. sq. km. I so. km. so. km. I sq. km. I sq. km. I sq. km. sq. km. I sq. kIn. 10·00-10·25 8·9 8·9 9·7 9·7 0 0 0 0 18·6 18·6 10·25-10·50 10·7 19-6 17·6 27 ·3 ·6 ·6 0 0 I 28·9 47·5 10·50-10·75 5·8 I 25·4 13·7 41·0 2·4 3·0 2'1 2·1 24·0 71 '5 )0·75-11'00 2·2 27-6 8·8 49·8 7·6 10·6 12·9 15·0 I 31·5 103'0 11'00-11·25 ·4 28·0 IA 51·2 9· 1 19"7 24·1 39·1 35·0 138·0 11 ·25-11'50 0 I 28·0 0 I 51·2 9·5 29·2 40·5 79·6 50·0 1 J ·50-11·75 188'0 0 28·0 0 51·2 7·2 36·4 52·8 [132·4 60·0 11·75-12·00 248·0 0 28·0 0 I 51·2 8·9 45·3 I n8·1 250·5 127·0 375·0 12·00-12·25 I 0 28 ·0 0 51·2 3·0 48·3 72·0 322·5 12·25-12·50 I 75·0 450·0 0 28·0 I 0 51·2 0 I 48·3 27·0 349·5 12·50-12·75 54·0 504·0 0 28·0 0 51·2 0 48·3 0 349·5 12·75-13'00 I 38·0 0 542·0 28·0 I 0 , 51 ·2 I 0 I 48·3 I 0 349·5 50·0 592·0 SUMMARY OF TABLE (1) Total Area of Flood-Plain 592'0 sq. kID. (3) The remaining liS sq. km .• since it arises out of the petcring·out of the (2) Area of Yossla cusp/dala """ 28·6 sq. km. loich grasses above level 12'25, is assumed to be occupied by inter· ., Echinochloa slagnina 51 '2 sq. km, mediate grass land species . .. O,yza banhi/ "". 48'3 sq. km. .. Echlnochloa pyramidalis 349'5 .q. km. TOTAL 477'0 SQ. km. ~ TABLE 112 AREAS OF FLOOD-PLAIN OCCUPIED BY GRASS SPECIES RENK-JEBELEIN Vossia cuspidata Echinochloa stagnina Oryza barthii Echinochloa pyramidalis TOTAL Av. Gauge Interval. ! I I Increment Total Increment Total Increment Total Increment Total Increment I Total sq. km. sq. km. sq. km. sq . km. sq. km. I sq. km. sq. km. sq. km. sq. km. sq. km. I 10·00-10·25 3-9 3·9 ·6 ·6 9·0 9·0 10·25-10'50 7·3 I 11·2 9·4 10·0 ·3 ·3 17·0 I I 26'0 10·50-10·75 I 4·6 15·8 • 8·8 18·8 1·6 1·9 ·5 ·5 I 15·5 41·5 10·75-11·00 1-4 17-2 5·6 24·4 2·3 4·2 1·2 1·7 10·5 52·0 - 11 ·00-11 ·25 1·0 18·2 6·6 31',0 3·3 7·5 3·4 5·1 14·3 66'3 00 11·25-11 ·50 1·2 19·4 8·8 39·8 3·1 10·6 6·8 11·9 19·9 86·2 11·50-11·75 I 2·4 21·8 16·0 55·8 4·4 15·0 17·0 28·9 39·8 I 126'0 11·75- 12·00 1·1 22·9 10·1 65·9 2·8 17·8 42·0 70·9 56·0 182·0 12·00-12·25 0 22·9 0 65·9 1·6 19-4 52·4 123-3 54'0 236·0 12·25-12·50 0 22·9 0 65·9 0 19·4 40·0 163'3 40·0 276'0 12'50-12·75 0 22·9 0 65·9 0 19·4 15·5 178·8 31·0 307'0 12·75- 13·00 0 22·9 0 65·9 0 19-4 8'5 187'3 17·0 324·0 SUMMARY OF TABLE (I) Total Mea of Flood-Plain 324 sq. km. (3) The remaining 28·5 sq . km., as it occurs mostly above gauge Jevel12'SO (2) Area of Vossla cuspldata 22'9 sq. km. where the faich grasses begm to peter out, is assumed to bo occupied (3) .. .. Echlnochloa stagnina 65'9 sq. km. by intermediate land grass species. The area of 4·5 km. between .. Oryza barthil ... gauge levels 10·00 and 10'25 which is Dot included in the grass 19·4 sq. km. distribution is considered to be open water. .. Echlnochloa pyramldlliu 187·3 sq. km. TOTAL 295'$ sq. km. TABLE 113 TOTAL AREAS OF SWAMP PASTURE GRASSES NOR MALLY FLOOD ED AND EXPOSED (Interpolated from Tables 11 0, 111, and 112) Areas in square kilometres Open Water Total REACH Vossia Echinochloa Oryza Echinochloa cuspidata stagnina I barthii pyramidalis or Bare Area of I Ground I F lood-Plain Malakal-Melut I 10,03-12-20 ... 15 ·2 17·3 42·5 148·0 223·0 Melut- Renk 10·12-12'16 ... 24·2 47·0 47·5 302-3 421·0 Renk-Jebe1ein 10·14-12'39 ... 21·2 65·6 I I 19·4 146·3 I 2·5 I 255·0 TMOaTlAakL al-Jebelein .. . I 60·6 129·9 109·4 596·6 2 ·5 899·0 TABLE 114 NORMAL RIVER LEVELS AT SITES OF CROSS-SECTIONS CoL 1 CoL 2 CoL 3 CoL. 4 COL. 5 IB ahr el Ghazal-Bahr el Jebel e.S. 100day Periods Sobat C.S.1. Yoynyang e.S. Lake No C .S. I I Ghazal End I Jebel End I January 1 .. . ... 386·59 386·90 I 386·29 I 386·42 386·75 2 ... ... 385-92 386·89 I 386·25 386·39 386·72 3 ... ... 385·39 386·88 386·21 386·35 386·69 February 1 ... .. . 385·00 386·86 I 386·17 I 386·31 386·66 2 ... ... 384·80 386·84 386·13 386·28 386·62 3 .. . ... 384·70 386·82 I 386·09 I 386·24 386·59 March 1 .. . .. . I 384·64 386·79 386·06 386·21 I 386·56 2 .. . .. . 384·58 386·76 386·03 386·18 386·54 3 ... ... 384·52 386·71 I 385·99 386·14 386·50 April 1 ... ... I 384·41 386·67 I 385-97 I 386·12 386·48 2 ... ... 384·66 386·62 I 385·95 I 386·09 386·46 3 ... ... 384·75 386·57 385·95 386·08 386·46 May 1 ... ... 384·92 386·53 385·95 386·07 386·46 2 .. . ... 385·28 386'51 385·95 386·07 386·46 3 ... .. . 385·73 386·49 385·98 386·09 386'48 June 1 ... ... 386·24 386·49 386·02 386·12 386·51 2 .. . ... 386·65 386·49 386·05 386·14 386·54 3 .. . ... 386·99 386·51 386·09 386·18 386·58 July 1 ... .. . 387·27 386·53 386·12 386·21 386·61 2 ... ... 387·53 386·56 3S-6'14 I 386·23 386·63 3 ... ' " 387·77 386·61 386·19 386·28 386·67 August 1 ... ... 388·09 386·67 2 .. 386·23 386·32 386·71 , ... 388·17 386·73 386·27 I 386'37 3 .. . 386'73 .. . 388·34 386·77 386·31 386'41 386·77 September 1 ... ... 388·49 386·82 ... 386·34 ... 386·44 386·79 2 ... 388·62 386·85 ... 386·37 386·47 386·81 3 388·74 386·88 386·40 386·50 386·83 October 1 ... .. . 388·86 386'90 386·42 ... 386·52 2 .. . 386·85 388·95 386·92 386·44 .. . ... 386·54 386·87 3 389·03 386·93 386·44 386·54 386·87 November 1 . _. ... 389·08 386·93 386·42 386·53 386·85 2 ... .. . 389'09 386·93 386·41 .. . 386·52 386·84 3 .. . 388'99 386·92 386·39 386·56 386·82 December 1 .. . .. . 388·74 386'93 386·38 ... ... 386·50 386·82 2 388·30 386'93 .. . 386·37 ... 386·49 386·81 3 387048 386·93 386'34 386'46 386·79 182 TABLE 115 DEPTH AND DURATION OF FLOODING OF GRASS SPECmS RIVER SOBAT CROSS-SECTION DEPTH (CENTIMETRES) DURATION (DAYS) Over-all Dominant Over-all Dominant D istribution Distribution Distribution Distribution Min. I Max. Min. I Max. I Mean Min. I Max. Min. I Max. - I Mean Vossia cuspidata ... 145 230 180 220 200 150 200 165 180 172 Echinochloa stagnina ... 128 230 140 215 177 150 195 150 185 167 Oryza barthii ... .. . 10 140 108 116 112 30 150 125 145 135 Echinochloa pyramidalis R .F . 180 R.F. 100 50 0 135 0 135 67 NOTE: Figures arc based on normal river. R.F. = Rain Flooding. TABLE 116 DEPTH AND DURATION OF FLOODING OF GRASS SPECIES BARR EL GHAZAL CROSS-SECTION ____D_ E-PT,H-_(C_EN_TlM_ET, R_BS)_ ___ l DURATION (DAYS) Species 1 Minimum I Maximum I Average Minimum I Maximum I Average Cyperus papyrus .. . ... . .. 40 230 135 365 365 365 Vossia cuspidata ... ... .. . 20 160 90 180 365 270 Oryza barthii ... . .. ... 20 110 65 210 365 287 Echinochloa pyramidalis .. . ... R .F . 50 25 R.F. 270 135 NOTE: The figures are based on normal nver. R.P . = Rain Flooding. TABLE 117 DEPTH AND DURATION OF FLOODING OF GRASS SPECIES BARR EL GHAZAL-BARR EL JEBEL CROSS-SECTION DEPTH (CENTIMETRES) DURATION (DAYS) Over-all Dominant Over-all Dominant Distribution Distribution Distribution Distribution Min. I Max:. Min. I Max:. I Mean Min. I Max. Min. I Max. I Mean Cyperus papyrus(') ... 30 210 70 210 140 280 365 310 365 337 Vossia cuspidata(') .. . 10 200 40 140 90 60 365 270 365 317 Oryza sp . barthii(2) ... 10 120 40 120 80 60 365 135 365 250 Echinochloa pyramidalis(2) ... 10 80 10 40 25 75 365 75 240 157 NOTE : (1) Based on normal river (Bahr el Jebel). (I) Based on normal river (Bahr cl Ghazal). TABLE 118 DEPTH AND DURATION OF FLOODING OF GRASS SPECIES BARR EL JEBEL ALIAB VALLEY Discharge in Flood mi d at Duration Species Mongalla Days 40 347 Open water; Cyperus papyrus; Vossia cuspidata 50 293 Cyperus papyrus; Vossia cuspidata; Echinochloa stagnina and Echinochloa pyramidalis 60 227 Vossia cuspidata ; Echinochloa pyramidalis 70 173 Echinochloa stagnina; Vossia cuspidata; Echinochloa pyramidalis 80 108 Echinochloa pyramidalis and Echillochloa stagnina 90 71 Echinoclrloa pyramidalis 100 37 Phragmites; Echinochloa pyramidalis 110 20 Phragmites; Eclrinochloa pyramidalis 120 7 NOTE: F1I,UfCS based on 1946-50 (Wlth double weight liven to 1950 Ooods) . 183 TABLE 119 DEPTH AND DURATION OF FLOODING OF GRASS SPECIES BAHR EL JEBEL ALIAB V ALLEY e.S. 13 DEPTH (CENTIMETRES) DURATIO:-l (DAYS) SPECIES Max. Min. Mean Max. Min. Mean Cyperus papyrus .., ... ... 450 160 220 365 175 210 Vossia cuspidata ... .. . .. . 450 170 250 365 180 230 Echinochloa stagnina .. . .. . 340 170 240 365 180 215 Echinochloa pyramidalis ... .. . 340 10 130 200 23 120 Phragmites communis ... ... 200 10 130 200 25 120 NOTE: Figures arc based aD Mongalla discharges for the period 1946-50, with double weight given to the 1950 period. TABLE 120 LEVELS OF FLOODING ACCORDING TO MONGALLA DISCHARGES AND DURATION BAHR EL JEBEL MONGALLA-GEMMEIZA TOICH LEVELS OF BACKWATER MongaJIa Discharges Duration (Days) Cross-Section 12 Cross-Section 3 50 mi d 428·28 293 60 mi d 428·62 227 70 mi d 488·89 173 80 mi d 429·70 108 90 mi d 429·94 432-87 71 100 mi d 430· 10 435-96 37 110 mi d 430·21 437·53 20 120 mi d 430·30 437·72 7 NOTE : FJood duration figures are based on Mongalla discharges for the period 1946-50, with double weight given to 1950 flood . TABLE 121 RECORDED LEVELS AND DEPTH AND DURATION OF FLOODING OF GRASS SPECIES BAHR EL JEBEL MONGALLA-GEMMEIZA TOICH C.S. 12 I :~TH DURATION (DAYS) SPECIES (CENTIMEIRES) I Mean I Max. I Mean Min. I Max. I Mean Cyperus papyrus 428·08 427-16 ' 1 427'65 222 314 265 293+ 293+ 293+ Vossia cuspidata . .. 428 ·50 427-16 427·78 180 314 252 227-293 293 + 293+ Echinochloa stagnina 428·40 427·16 42H8 190 314 252 227-293 293 + 293 + Echinochloa pyramidalis 429·90 427-80 1 428'40 40 250 190 71-108 293+ 227- 293 Phragmites communis .. . 429·96 427-84 428·44 34 246 186 37-71 293+ 227-293 TABLE 122 RECORDED LEVELS AND DEPTH AND DURATION OF FLOODING OF GRASS SPECIES BAHR EL JEBEL MONGALLA-GEMMEIZA TOICH C.S. 3 LEVELS DEPTHS (CENTIMETRES) DURATIC)N (DAYS) SPECIES Max. I Min. Mean Min. I Max. I Mean I Min. I Max. I Mean Echinochloa stagnina ... .. . 436·05 434·74 435·5 67 298 222 20-37 37-71 37- 71 Echinochloa pryamidalis ... 437-48 435·44 436·5 24 228 122 20-37 37-71 20-37 Phragmites communis ... .. . 437'55 435·45 436·5 17 227 122 7-20 37- 71 20-37 184 TABLE 123 MEAN DEPTHS AND MEAN DURATIONS OF FLOODING OF GRASS SPECIES WHITE NILE, RIVER SOBA T, BAHR EL GHAZAL AND BAHR EL JEBEL I I Cyperus Vossia Echinochloa Oryza Echinochloa Phragmites I papyrus cuspidata stagnina bar/hii pyramidalis communis Site Col. 1 I Col. 2 Col. 1 I Col. 2 Col. I I Col. 2 Col. I I Co1.2 Col.I [COl. 2 Col. 1 j col. 2 WHITE NILE I Rabak(l) ... ... . .. - - 99 210 127 230 87 200 52 170 - - Jebelein(l) ... ... ... I - - 163 250 165 250 162 250 37 110 I - - Renk(l) ... - - 191 290 170 260 126 210 103 190 - - Melut(l) .. . ... . .. - - 280 360 202 260 I 127 200 119 200 - - Malaka\(I) ... - , - 190 230 165 210 128 190 103 180 - - I SOBAT I Bilaiwal C.S.(·) '" ... - - 200 172 177 167 I 112 135 50 67 - -- I BAHR EL GHAZAL I Yoynyang C.S.(') ... 135 365 9Q 270 - - 65 287 25 135 - --I Bahr el Ghaza1 to Bahr el Jebel C.S. CNorth)(') I - - 90 317 - - 80 250 25 157 - - BAHR EL JEBEL Lake No C.S. (') .. . 145 365 - - - - - - - - - - -- Bahr el G hazal to Bahr el Jebel C.S. (South)(') 140 337 - - - - - I - - - - - -_ .. Aliab C.S. No. 13(1) ... 220 210 250 230 240 215 - - 130 120 130 120 Mongalla C.S. No. 12(1) .. 265 293 252 293 252 293 - - 190 227- 186 227- 293 293 MongaUa C.S. No. 3(1) ... - - - - 222 37-71 I - - - 122 20 37 122 20- 37 1 - I Col. J = Depth in centimetres. Col. 2 = Duration in days. NOTES: (1) Based on 1946-50 floods. double weight given to 1950-51 flood. (I) Based on normal river. TABLE 124 AVERAGE RIVER VELOCITIES Sites Average River Velocity BAHR EL JEBEL Mongalla ... 1·IOm/ sec. Bor 1·10 m/ sec. Buffalo Cape 0·75 m/ sec. Malakal ... 0·50 m/ sec. BAHR EL ZERAF 0·80 m/ sec. BAHR EL GHAZAL 0·02 m/ sec. 185 TABLE 125 RAINFALL AND FLOODING WHITE NILE MALAKAL RENK KOSTI Species I I Column I Column 2 Column 3 Column 1 I Column 2 Column 3 Column 1 Column 2 Column 3 I I I I I I Vossia cuspidata ... .. . ... .. . .. . .. . .. 1 10·60 June (I) I 120+ 10·53 June (2) 42++ 12·43 Aug. (1) 177+ Echinochloa stagnina ... ... ... ... . .. .. . 10·60 June (I) 120+ 10·23 May (3) 42 12·73 Aug. (2) 177++ Oryza barthi; ... . .. ... .. . ... .. . . .. 11·10 July (1) 147 + 10·53 June (2) 42 13·13 Sept. (1) I 314+ Echinochloa pyramidalis ... ... ... .. . .. . ... 11 ·40 July (3) I 416 11·13 Aug. (I) 261 + 12·83 I Aug. (3) 314 I NOTE: Column 1. The minimum level at which each species is recorded at each site. Column 2. The dates (in IO-day periods) at which the levels in Column 1 arc attained by normal rising river, i.e., the date ftooding commences . Column 3. The total rainfall precipitated before flooding. The rainfall figures are based on those given under Olimate and are the totals up to the end of the month prior to the IO-day period given. The additional rain during these odd IO-day periods is indicated by + sign. -00 TABLE 126 Q\ RAINFALL AND FLOODING MONGALLA-G'EMMBIZA I ~ CROSS-SECTION No. 12 CROSS-SECTION No. 3 Species t I Column 1 I Column 2 Column 3 I Column 4 C0lumn 1 Column 2 I Column 3 Column 4 I I , I I I Cyperus papyrus .. . ... .. . ... 427·16 I Below 50 Pennanently Not Measurable - - - - Flooded Vossia cuspidata ... ... .. . ... I 427·16 - " " " " " - - - Echinochloa slagriina ... .. . 427·16 434·74 " " 90-100 Aug. 3 " " " 667 Echinochloa pyramidalis ... .. . . .. 427·80 I 435·44 Aug. 3 " " " " " 90-100 667 Phragmites communis ... .. . .. . 427-84 I 435·45 90-100 Aug. 3 " " " " " 667 I Column t. The minimum Hood-plain level at which each species is recorded. Column 2. The Mongalla discharge at which, when reached, ftooding at the levels in Column 1 commences. Column 3. The dates the discharges in Column 2 are reached. Column 4. The amount of rainfall precipitated before the date in Column 3, Rainfall is based on Mongalla figures. APPENDIX NOTE ON FAUNA OF THE JONGLEI AREA An account of the ecology of the Jonglei Area and of the Flood Region as a whole would not be complete without a description of the fauna. It is not possible here to give a compre- hensive account of all animals found in this region of the Sudan and we must confine ourselves to the larger and more common varieties. The main species are as follows : Elephant (Loxodonta africana): found in herds throughout most of the area south of Jebel Ahmed el Agha (latitude 11 ° N.). Rhinoceros: the White Rhinoceros (Ceratotherium simum cottoni), an animal which is now all too rare in Africa, is found in considerable numbers, notably between Shambe and Tombe on the left bank and also on the left bank in the Nimule reach. The Black Rhinoceros (Diceros bicomis) is comparatively rare and is found only on the right bank in the southern reaches, except in two places in the Bahr el Ghazal Province, north of Tonj and west ofWau. Buffalo (Syncerus caffer aequinoctialis) is common throughout the area south of about latitude 10° N. Giraffe (Giraffa camelopardalis) is common in all regions with the exception of the Kosti area, where it doubtless once existed but has long since been wiped out by the Baggara Arabs. Zebra (equus quagga granti) is common in some regions, notably in the East~rn and South-Eastern Plains. Tiang (Damaliscus korrigum tiang) is one of the most common antelopes throughout the region, while the Lelwel Hartebeeste (Alcelaphus buselaphus lelwel) is also common, but not in the Sudd area. Other antelopes are Roan (Hippotragus equinus bakeri), White-eared Kob (Adenota kob leucotis), Uganda Kob (Adenota kob thomasi)<57), and Waterbuck (Kobus defassus). Eland (Taurotragus oryx) is fairly common in the south-east, though generally well outside the boundaries of the Jonglei Area, while the Oryx (Oryx beisa) is confined to the plains east of Kapoeta. The lesser Kudu (Streps- iceros imberbis) is also found in the same region, but is largely restricted to areas of heavy acacia bush. Of the smaller animals, Bushbuck (Tragelaphus scriptus bor) is found in bushland and scrub, and Reedbuck (Redunca redunca) is common. Dik-Dik(5s) and Duiker, parti- cularly the Abyssinian Duiker (Sylviacapra grimmi abyssinica), are not common except inland and in the south. Oribi (Ourebia montana) is to be found throughout the area. The Red- Fronted Gazelle (Gazella rujifrons laevipes) is found in large numbers north of lat. 9° 30'(59), and its counterpart, the Mongalla Gazelle (Gazella albonotata), is extremely common in the region we have referred to as the ' Eastern and South-Eastern Plains '. Warthog (Phaco- choerus aethiopicus bufo) is found along the southern reaches, but the Red River-Hog (Choiro- potamus porcus) is only to be seen inland in areas of dense bush and forest. Situtunga (Limno- tragus spekei) is to be found in the swamps and on the edge of the toich in more remote areas, though it is not common, while the Nile Lechwe (Mrs. Gray: Onotragus megaceros) is common in the swampier areas of the toich as far south as Bor. Of the carnivora, Lions abound in all regions where smaller game animals are found and also take a heavy toll of cattle. Leopards are also common in the afforested areas and in the hills, and of the smaller cats the Serval is the most common. Other cats, e.g. Lynx, are only found in the high parts of the Ironstone Region. Both the spotted and striped Hyaenas are common. In the rivers, Crocodiles are numerous as far north as Kosti, and Hippopotamus is found in almost all reaches south of Renk, with occasional individuals even as far north as Khartoum. The Hippopotamus plays an interesting part in the changing formation of the swamps; its tracks through the papyrus and other swamp vegetation are often the beginning of new channels in the river. This list of animals is not comprehensive and does not include the numerous smaller animals, including the rodents. Nor have we attempted to list any of the birds of the area, some migratory, some permanent inhabitants, and some peculiar to the area like the Shoe-Bill Stork (Balaeniceps rex). As far as snakes are concerned we can only refer the reader to published works. Without examining in detail the characteristics of these animals, it is worth while stressing a few points which are of particular interest in the general ecology of the region. (i) The distribution of wild animals, which is largely determined by physical conditions, also has its seasonal aspects, most animals being compelled to move in search of pasture and water during the dry season and again to escape the floods during the rains. (ii) As a general classification we have already defined the different types of land found in the Jonglei Area, each with its particular kind of vegetation and grass associations. Briefly restated they are as follows: Sudd, permanent swamp with papyrus, Phragmites and other heavy vegetation-a tangled mass which rarely if ever dries out; toich, riverain marshes which are inundated or exposed accord- ing to the seasonal fluctuation of water-levels in the river. Above this level is intermediate 187 land, of which the Eastern and South-Eastern Plains are typical examples. Here the grasses are largely tufted perennials like Hyparrhenia rufa and Setaria spp. and, at a slightly higher level, the area is sparsely covered with trees (heglig: Balanites aegyptiaca) and in some areas dense bush (Acacia seyal and Acacia ./istula, etc.). Slightly above this level is high land where the grasses are again largely perennials with a few annuals, and where acacia bush is also found as well as larger trees (Ficus, etc.) and palms (Barassus aethiapium Mart. and Hyphaene thebaica Mart.), particularly on the sandier outcrops. As far as wild animals are concerned we must add, at one extreme, open water in rivers, lakes and pools, and, at the other, hilly country which is not really found in the Jonglei Area except south of Mongalla. (iii) These classifications, which are amplified in the previous sections of this chapter, are the very broadest possible and are more precisely applicable to what we call the Flood Region, but the distribution and habits of the fauna can be roughly related to them. Quite apart from the very obvious fact that Crocodiles and Hippopotami cannot move far from open water, permanent swamp and river banks, some animals are specially equipped to exist more or less permanently within the confines of one type of land. The Situtunga, with its elongated hoofs which distribute its weight, is able to move freely over the top of floating vegetation and is confined to permanent swamp in the Sudd, though it may move into the taiches when these are inundated during the rains. The Nile Lechwe, which also has elongated hoofs though much less so than the Situtunga, exists on the edges of permanent swamp and in the taiah, moving up to the lower limits of intermediate land when both the former areas are covered with water. At the other extreme the Greater Kudu appears to be largely, though not entirely, confined to the hills. Other animals move in between, few of them being limited to one type ofland. The Bushbuck, for example, is found on intermediate land as well as high land where there is plentiful acacia bush, and in areas of broad-leaved forest. Elephant and Buffalo on the other hand have an extreme range, b~ing able and often compelled to move from the edges of permanent swamp to the hills and back. (iv) This range of movement is conditioned by seasonal variation. The Sudd dries out to some small extent during the dry season and some animals are able to enter it, notably the Nile Lechwe whose more usual habitat is the taich. Most other animals, with the exceJ)tion of Elephant (and to some extent Buffalo) which can crash through the thick vegetation, provided there is not too much water or mud, are unable to penetrate true Sudd at any time of the year. Taich land, at other times of the year inundated and only accessible to some animals, provides grazing during the dry months for many types of animal. This also applies to intermediate land immediately after the rains and just after they begin to fall again, while in between it is generally too dry. It is then only habitable by such animals as arll equipped to go for long periods without water, like the Mongalla Gazelle, or able to range over long distances from one watering centre, like the Giraffe or Zebra. During the rains these animals withdraw to higher ground, giving way to others retreating from the toich who are better able to move about in country which is at times a morass of clinging mud. High land is also inhabited by animals during the rains, but in the Flood Region high land is limited and therefore densely populated by the human inhabitants at that time of the year. This naturally discourages game and is, no doubt, why many animals move much farther inland to regions which are both high and free from mankind. In hill country similar movements are necessary owing to changing seasonal conditions, though in some areas-notably the Imatong Mountains-many animals (Elephant, for example) exist throughout the year. From this very generalized account, we see that the wild fauna of the region is for the most part compelled to move with the seasons, just as man, dependent on his domestic animals, is forced to move, his economy being based on a balanced utilization of all kinds of land according to seasonal climatic variations. As an example it is worth while examining in slightly more detail one particular area and the migrations of some of the animals in it. The Eastern and South-Eastern Plains (see p. 20) are, roughly speaking, bounded by the line Duk Fadiat-Waat-Akobo in the north and the line of the south-eastern mountains in the south, and limited by the Bahr el Jebel on one side and the Ethiopian foothills on the other. Some animals, perhaps animals of all kinds, are permanently resident within the confines of this area, but others only enter it during the dry season. Most of them seem to come from the south-east, many of them from as far away as the Ilembi triangle and also from the Boma plateau. The almost unbelievable concentrations of game at certain points of this migration have often been described. Writing of the Illigration in 1932, Captain R. C. R. Whalley says: "Game was present in its thousands, in many varieties, which will be detailed later, and to give an illustration, lest the writer be suspected of exaggeration, one day an attempt was made to estimate the numbers of game by counting them passing a tree. They passed at the average rate of twenty to thirty a minute on one game trail and the procession passed the camp unceasingly and solidly for four and a half hours. The game that passed was composed of Tiang, Giraffe, Lesser Eland, Oryx beisa and Grant's and Mongalla Gazelle and a few Zebra "(60). This pheno- menon, which is also reported elsewhere(61), does not occur within the plains, for by that time it seems that the concentrations have dispersed over a wide area. Many animals, however, clearly enter the Eastern and South-Eastern Plains and some, perhaps, reach the river. Two examples are worth examining in more detail, especially as air reconnaissance flights made by various members of the Team tend to confirm observations made from the ground. During the dry season the Mongalla Gazelle is found in tens of thousands in the South-Eastern Plains as far south as the Badigeru swamp and as far north as the Pengko area. It seems likely that its distribution at this time of the year is. largely determined by the frequency of Hyparrhenia rufa regrowth. The short and rather delicate green shoots are admirably suited as fodder to these animals where larger animals cannot obtain the necessary bulk. During the rains the Mongalla Gazelle retreats to higher ground, probably because it cannot thrive on the coarse and luxuriant growth of this kind of grass and also because its small sharp hoofs, though cloven, are not well adapted to the heavy clay, by then moist and sticky. Tiang are also found in this region, but not in large concentrations until much farther north. The reason for this distribution seems to be that the Tiang cannot survive in the southern areas, which are particularly waterless during the dry season. It can roam 188 far and wide, but must come back to a water centre every two or three days. The Mongalla and Grant's Gazelle can survive without water almost indefinitely and therefore (with a few Reed Buck) occupy this region almost without other intruders. Farther north the Tiang is able to water (usually at night since the area is in a populated r~gion) in the upper reaches of the Fullus system and the Khor Nyanding. Tiang are found, though not in large numbers, along the Veveno and Lotilla, and we attribute their lack of concentration in this vicinity to the Murle, who are particularly indefatigable hunters and force th~m to move farther northwards. (v) It has been suggested that a more complete knowledge of game distribution and migration might throw light on to the availability and value of pasture areas hitherto unexploited for animal husbandry. In the present state of our knowledge we consider that no conclusions could safely be reached on these lines. We know all too little about the grazing requirements of wild animals in grass types, bulk, nutritive value, palatability, etc. Moreover the movements and numbers of game animals are no indications of carrying capacities of pasture land where cattle are concerned. We consider, however, that the grazing requirements of wild animals might repay further study. (vi) We have already described briefly, and will later describe in detail, how man in this region is also compelled to move with the seasons, his economy being conditioned by the utilization of toich, intermediate land, and high land for both crop production and pasture for his stock. In present circumstances wild animals are a competitive factor in pasture utilization, since to some extent they live off grasses which would otherwise be of use to cattle and other domestic stock. This competition is obviously not an important matter, for otherwise the Nilotic herdsmen would take more serious steps to liquidate the game, and it applies largely to intermediate land. Intermediate land has been classified as such according to levels and grass associations, but it is also intermediate in the sense that its main function is to provide grazing in the periods between the main migrations from high land to toich and vice versa-i.e. it provides pasture at a time of year when the toich is inaccessible owing to inundation from the rivers and high land has already dried out and is waterless. Pasture at this period depends on the palatable and nutritive regrowth from perennial grasses after burning or heavy cropping. Many wild animals are also dependent on this type of pasture, and in some areas the people allege that they would be able to graze their cattle in certain intermediate areas until much later in the year, were it not that much of it is reduced by the large herds of game at the beginning of a dry season. This is a factor to be noted when we come to consider (in later volumes) the availability of intermediate pastures as an alternative rather than a supplement to toich. (vii) Wild animals are also an important and much more adverse factor in animal husbandry, since many diseases are transmittable. For example, many animals-Buffalo, Tiang, etc.-are extremely susceptible to rinderpest, and are not only an immediate source of contagion (through grazing, etc.) but also an uncontrollable reservoir of such diseases. TIus is a serious matter because the final objective of total immunization campaigns, the elimination of the disease for ever, can never be achieved. Foot and mouth is also transmitted in this manner. Finally trypanosomiasis is the most serious problem of all. Much has been written on this subject and we do not propose to do more than draw attention to the particular problems of the Jonglei Area. As a broad generalization, tsetse fly has not been found in this area, except in the reach south of Juba and on the edges of the region occupied by cattle-owning tribes. Its incidence largely determines the linllts of their country. Trypanosomiasis is, however, widespread and is a major factor in the depletion of stock. It appears that the disease is transmitted by direct infection through biting-flies (Tabanus, Stomoxys, etc.), i.e. from cattle already infected (see p. 317). Campaigns for the eradication of the disease have already been carried out, with only limited success. They cannot in any case be completely successful so long as game animals-which have no respect for the limits of the safe area-move to and fro in the vicinity of cattle. The damage done to crops by wild animals is mentioned later (see p. 352). It is far from our intention to advocate the extermination of game, and though something may be said on this subject when we come to consider the Equatorial Nile Project, its effects and repercussions, we must stress that we are here only stating the facts. (viii) Finally, some mention must be made of the value of game animals in the diet of the human popula- tion of the area. Their value is sometimes exaggerated. Some people, like the Murle and Latuko (who are outside the Jonglei Area), are more energetic hunters, but on the whole the Nilotics of the Jonglei Area do not hunt a grdh deal except in years of famine, and the meat of wild animals forms only a minor item in their diet. This no doubt accounts for the very considerable numbers of wild animals still to be found in the area. By contrast the Baggara of the Semi-Arid Region, who are particularly skilled and enthusiastic hunters of elephants and giraffe, have already more or less exterminated the animals wluch must have been common there in the past. They would have made greater inroads into the game population had not the administration controlled their move- ments. Enterprise in the hunting of wild animals can also be related to lack of animal proteins in certain tribal areas. Cattle-owners not only live in the open grasslands where hunting is, perhaps, more difficult and the game less vulnerable, but are usually well off for meat. Meat from domestic animals is in short supply or not available at all in the afforested regions where tsetse abounds, and here the people are bound to expend more energy in hunting and trapping the wild fauna. This is particularly noticeable among the Madi, Moru and Zande tribes and others living in a similar environment. These generalizations are not derived from any scientific investigation of the wild fauna of the Southern Sudan. The Team has had no qualified zoologist to carry out such a study and it is doubtful whether such a post would have been justilied solely in connection with the Equatorial Nile Project and its effects. The subject is, however, one of intense interest and throws much incidental light on related problems. In recording our observations we wish to draw attention to some of these problems, but realize that much will be open to ~riticism. 189 NOTES AND REFERENCES PART I-SOILS (') All soil samples collected by the Team were analyzed by the Soil Section, Ministry of Agriculture, Research Division, to whom we are indebted fo r this help. (') For more detai led information see Andrew, G., ' Geology of the Sudan ' and Bibliography in Tothill, J. D. (Ed.), Agriculture in the Sudan, O.U.P., 1948. (') " Recently obtained evidence shows that at least 20 m. of' Gezira clay' has been removed from the Khartoum area. T his general lowering 0\ the surface implies a proportionate erosion of the plains farther. south (the amount can be only g,!ess~d In any partlcular slle). Therefore the clay plains in the northern part of the Jonglel Area (but not In Its centre which IS sull bemg upgraded) must also have suffered from the normal sheet erosion." Andrew, G., prIvate commUniCatiOn. (.) Tothill, op. cit., p. 117. (6) For explanation of the term' Buchanan's laterite' see Russell, E. J., Soil Conditions and Plant Growth, 8th ed., Longmans Green & Co, 1950, p. 507. (.) Egyptian Ministry of Public Works, Veveno-Pibor Scheme, Cairo, 1932, p. 18 et seq. (,) Called by Tothill ' Sudd marginal area'. See Tothill, op. cit., p. 139. (.) Russell, op. cit., p. 354. e) For explanation of the • catena conception ' see Milne, G .,' Some suggested units for classification and mapping, particularly of E. African soils', Soil Research 4, 1935, p. 183. (10) For the purpose of this report, the hydrological factor is defined as soil moisture conditions and aeration as a pedogenetic force. (n) Hurst, H. E., The Nile Basin, Vol. V, Cairo. See also Figs. A 7 and A 15. (n) Mohr, E. C. J., Be/wopt Vers!' Hanwal. Boden. Congo Djocja. 61-69. (") Pearsall, W. H., ' The investigation of wet soils and its agricultural implications ',Emp. JOllr. Exp. Agric. 72,1950, p. 289 et seq. (") ibid, p. 295. (n) ibid , p. 296. (") Hegh, E., Les Termites . Partie Generale, Bruxelles, 1922. Also Marais, E., The Soul of Ihe White Alit, 4th ed., London, 1939. Also Troll, c., ' Termiten Savannen. Studien zur Vegetations und Landschaftskunde der Tropen II', Landeskundliche Vorschrift . Festschrift fur Norbert Krebs, Stuttgart, 1936, p. 275. (n) Marbut, C. F ., 'A scheme for soil classification', Proc. lsI Mt. Congo Soil Soc ., 1927. Com. V, p. 1 et seq. See also Woifanger, L. A., Major Soil Divisions of the United States No.7, pp. 15 and 16. (") Tothill, op. cit., pp. 102-103. ( " ) Greene, H ., ' Einige Boden der Aequatorial Provinz des Anglo-Aegyptischen Sudans " Soil Research 6, 1939, p. 325 et seq. (~O) Morrison, C. O. T ., et al., 'Tropical soil-vegetation catenas and mosaics', Jour . Ecology 36, 1948. (n) Tothill , op. cit ., p. 122. (") ibid , pp. 122, 160 (footnotes) . ( ,,) ibid, p. 147 et seq. (U) Morrison, op. cit., p. 9. (16) Jewitt, T. N., private communication. (") Tothill, op. cit., p . 443. (") A theory that the dark colour of the upper layer may be due to manganese staining is unlikely when the pH of these soils is considered. (n) Nikiforoff, C. C., and Alexander, L. T., 'The hardpan and the cla'Ypan in a San Soaquin soil " Soil Sc. 53, 1942, p . 157. Winters. E., Proc. Soil Soc. Amer. 7, 1943, p. 437. ( n) Marshall, C. E., and Whiteside, E. P., Missouri Agric. Sta . Res. Bull. No. 386, 1944, and No. 387, 1945. (") Duff, A. K., ' Mechanism of aggregation of clay minerals by soluble silicates', Soil Sc. 65, 1948, p. 309. (n) Russell, op. cit., pp. 529-30. (") Salter, R. M., and Green, T. c., Jour. Amer. Soc. Agron. 25, 1933, p. 622. Also Myers. H. E., ef al., Kansas Agric. Exp. Sta. Tech. Bull. No. 56, 1943. (" ) Tothill, op. cit., p. 140. (n) ibid. , p. 134. (") For description of two typical profiles see Tothill , op. cit. , pp. 131-132. (n) Milliequivalents of sodium expressed as percentage of 100 gm. of clay. (") Jones, T. A., 'Preliminary Report on Soil and Vegetation Survey undertaken in connection with the O.U. Soil Survey Expedition', 1948. Unpublished Sudan Govt. Report. (U) Migahid, Mohd. Ahm., Report on a Botan;cal Excursion to the Sudd Region, Fuad 1 Univ. Press, Cairo, 1948. (n) The analytical results quoted in this section should be taken with caution. The samples, after collection, had to be dried for transport to the laboratory. This drying may have altered irreversibly some chemical properties of the soil in the field. (") Russell. op. cit., pp. 51-52. PART II- VEGETATION (H) Andrews, F. W., in Tothill, op. cit. Smith, J., • Distribution of Tree Species in the Sudan in Relation to Rainfall and Soil Texture', Min. 0/ Agr;c. Bull. No.4, Khartoum. (") Tothill, op. cit., p. 40. (U) Smith, op. cit. (") ibid ., p. 26 . (n) ibid. , p. 26. (") ibid., p. 27. (") ibid., p. 23. (") Harrison, M. N ., ' Report of a Survey of the Grazing Areas of tbe Sudan " unpublished Sudan Govt. Report. (oo) Smith, op. cit., p. 52 . ('0) Morrison, op. cit. (61) Harrison, op. cit. (") Information kindly supplied by the Asst. Conservator of Forests, U.N.P. ( U) Migahid, Mohd. Ahm., An Ecological Study of the Sudd Swamps of the Upper Nile, Cairo, 1947. (") Flood duration figures are based on Mongalla discharges for the period 1946-50, with double weight given to 1950 (cf. Malakal-Jebelein). (U) Migahid (1947), op. cit. (") Debenham, F., Study of an African Swamp, H.M.S.O ., 1952. APPENDIX-FAUNA {&7) Vaughan's ~ob (Adenola kob vaughani) is found in the Bahr el Ghazal Province, but not, as far as we know, in the Jongiei Area. A slightly smaller variety of the Uganda Kob (Adenola kob alurae) is found in the extreme south. (U) RhY.llch.otr'!gus guelllheri smith;. It is rarely found within what we have defined as the Jonglei Area, but is common in Tont district. (") I.e. approximately the line north of the Sobat and BahI el Ghazal rivers. (") Whalley, R. C. R., ' Southern Sudan game and its habitat " S.N. & R. 15, 1932, p. 263. (") See for example Zaphiro, Bimbashi D., . Notes on LoelU game', Sudan Wild Life and Sport , 1, 2, 1949. 191 CHAPTER 3. THE INHABITANTS 1. INTRODUCTION This volume of our report is written as a general and factual survey of the present position. We are not concerned with suggestions or recommendations, nor do we attempt to predict changes in physical or other conditions as they may affect human society under the Equatorial Nile Project. Yet the definition of the area and the particular aspects with which we are concerned are bound by just those considerations, and in connection with the human occupants must be limited to those peoples who will be diroctly or indirectly affected. By , directly' we mean those peoples who will have to face definite losses in pasture or other existing economic assets as a result of changed physical conditions and also a reduction in potential assets not at present exploited. By' indirectly' we mean those who, through economic and social contacts, will obviously feel the effects; what may be called the g0) Luxmoore, H. R., ' Final Veterinary Report on Equatoria Province' (unpublished). (") Tothill, op. cit., p . 634. (") Bisschop, J. H. R., ' Report to Director, Veterinary Service', 1952 (unpublished). (II) Mason, I. L., The Classification of West African Cattle, Commonwealth Agricultural Bureaux. (") J.I.T., op. cit., p. 24. (U) Tothill, op. cit., Chap. XXII. (") Tothill, op. cit. (17) Findlay, J. D" Hannah Dairy Research Institute: Bulletin No . 9, 1950. 321 CHAPTER 5. CROP HUSBANDRY PART I. GENERAL 1. THE INFLUENCE OF ENVIRONMENTAL, ECONOMIC AND OTHER FACTORS ON CROP HUSBANDRY SYSTEMS IN THE JONGLE! AREA An agricultural husbandry system, including crop husbandry, d(;)velops under the combined influence of physiological, economic, technological, social and historic forces. It is possible not only to trace the general structure of the system to the action of these forces, but also to discover in it the reasons for more particular problems, such as the choice of crops and cultivation sites, the division of labour, and the failure or success of att€mpts at improvement. Understanding of the correlation between the factors affecting the agricultural system and the system itself, with the diversity of reasons underlying native practice, is essentia,l in the study of problems involving changes in environment and consequent changes in the economic and social structure; problems which will arise in connection with the Equatorial Nile Project. The agricultural system in the Jonglei Area, though primitive and by modem standards inefficient, is in equilibrium with the forces under the influence of which it has developed. It is therefore static, and satisfies the simple needs of the inhabitants, providing them with a livelihood at subsistence level. It is based on the balanced utilization of the natural resourc€s of the area by stock and crops, and aims at self-sufficiency in all essentials. The Equatorial Nile Project, by affecting grazing resources, will seriously disturb this equilibrium. Before ways and means of restoring it by counterbalancing the losses on one side by improvements in other directions can be designed, the present balance must be understood. Local stock and crops, and the methods of th€ir husbandry, must be studied, and their environmental, economic, social and historical foundations €stablished. At the beginning of this part of our survey, let us review briefly the different factors influencing crop husbandry in this area, and show how the cultivator has adapted his crops and the ways and means of raising them to the often very difficult conditions prevailing. We start with a summary of limiting factors of the physiological environment, and their effect. This environment, itself a resultant of climatic, physiographic and edaphic factors, provides the physical framework for the development of a system of crop husbandry. The cultivator, given the necessary knowledge and materials, may modify this environment to a certain extent, but to be successful he must adapt his practices to it. FACTORS OF PHYSIOLOGICAL ENVIRONMENT MOISTURE CONDITIONS These are undoubtedly the chief limiting factors in the greater part of the Jonglei Area. They are the outcome of : (i) The seasonal rainfall, which is extremely erratic and characterized by torrential downpours. (ii) The flatness of the country and the inadequacy of drainage-channels, which spill over annually during the rainy season. (iii) The prevalence of heavy, structureless, impermeable clay soils with a high wilting point. These conditions influence crop husbandry differently in different parts of the area. In the Ironstone Region the rainfall is better distributed, the country more undulating, the soils generally lighter though poorer, and the drainage is definitely better than in the Flood Region. Therefore moisture conditions, though of importance, cannot be regarded as of so great a significance as in the Flood and Semi-Arid Regions. The ecological habitat is here suitable for a considerable variety of crops. In the Flood Region, inhabited by Nilotes, moisture conditions only too often alternate between flood and drought, producing the most difficult agricultural circumstances. The crop husbandry system must be carefully balanced, so as to be at least partly successful; total success is very seldom attained. It must include crops resistant to both flood and drought, qualities which are limited and only se~dom found together in one species or single variety, though some local sorghum varieties combine them to a surprising extent. The cultivator usually resorts to planting a mixture of varieties, hoping that subsequent moisture conditions 323 21 will leave at least one of them undamaged. He also spreads the risk by planting both quick and slow-maturing varieties, by sowing at different times of the season, and by cultivating both on high land, unlikely to be flooded but susceptible to drought, and on intermediate land, liable' to flooding but much less so to drought. He spreads his fields over as wide an area as' possible, becaus€ rainfall is very often localized. He also tries, though with little success, to improve drainage and protect his crops from flooding. In the Semi-Arid Region, lack of moisture and the short duration of the rains must be regarded as the chief limiting factors . Crops are therefore confined to those which are quick- maturing and drought-r€sistant, and the cultivator aims at using the limited amount of water to the b€st advantage by siting his cultivations, wherever possible, in places where run-off will supplement the rainfall, by utilizing to some extent during the dry months the moisture left by the floods on the riverain flood-plain, and by irrigating wherever possible. TEMPERATURE, HUMIDITY, WIND, LENGTH OF DAY(') From the point of view of temperature, the Jonglei Area may be regarded as one unit. Nowhere in this area is temperature a direct limiting factor of great importance. It does, however, affect crop husbandry indirectly by influencing rates of evaporation and transpiration, and therefore moisture conditions, limiting crops to those suitabl€ for tropical or sub-tropical climates, and eliminating a priori those plants requiring vernalization. Finally, it affects the cultivator himself, his ability and inclination to work, and his need for food, clothing and shelter. Humidity in the Jonglei Area is mainly determined by the rainfall, and must be regarded as a limiting factor of secondary importance. The main direct effect of high humidities combined with high temperatures during th€ rainy season is to create an ideal environment for fungoid diseases of crops. Thus resistance to these diseases is one of the conditions of success, and obviously a factor limiting the choice of crops. Humidity affects indirectly rates of evaporation and transpiration, i.e. the hydrological factor of €nvironment. Wind, though never a main limiting factor, affects crop husbandry both directly, by mechanical impact of flowing air masses, and indirectly, through its effect on general climatic conditions. Though the arrival of the north wind at the end of the growing season could be expected to create desiccating conditions, the resulting cooler and slightly more humid weather often leads to an improvement in rain-grown sorghum crops in the northern part of the J onglei Area. In fact, the success of the late sown hariq crop largely depends on the early arrival of the north wind and cold weather. Length of day is another important factor limiting the range of crops which can be grown in the area. As will be seen from p. 8, length of day at Malakal in the Jonglei Area varies between 11 hours 37 minutes and 12 hours 39 minutes. It is well known that plants are subject to photoperiodism, and their development is affected by the day length, or rather its variations. The crops of the Jonglei Area must be 'short day plants', otherwise they are unable to complete the development cycle, i.e. to flower at all or in time before the end of the rainy season. On the other hand some of the' short day plants' cannot be sown in the winter as they come into flower before making an adequate growth. SOIL As we have already pointed out, soil is one of the most important factors determining moisture conditions, and therefore the agricultural system. The physical characteristics of the heavy clay soils predominant in the area affect crop husbandry both directly and indirectly. They result in mechanicallinlitation of root develop- ment and therefore favour crops with a strong root-system. Further, the friability range of these soils, being strictly limited, causes tfle difficulties encountered by the cultivator in tilling his fields. Hence there arise difficulties in the improvement of tillage methods, and there is a lack of advance in this direction. As a result the acreage under cultivation is limited, there are attempts to spread labour throughout the season, and the double cropping of cleared fields and the reduction of yields by insufficient cultivation are common features . Finally, the sandy soils, though generally less fertile, are more suitable for some crops and easier to cultivate. In consequence, where they occur we find an attempt to balance the system by including both principal kinds of soil in the fields of each cultivator. The chemical characteristics of the soils also affect the crop husbandry system. First, the fertility of the soils depends largely on their nutrit-nt status. The clays of the 10nglei Area must be regarded from this aspect as highly fertile. They are therefore able to support a mono-cropping system and give encouraging yields for a number of years, thereby reducing 324 the need for frequent shifting of fields. The sands and lateritic soils are much poorer; conse- quently their system of cropping must be more diversified, so as to make full use of available nutrients and even to improve their status. Furthermore these soils are more quickly exhausted under cultivation, and the shifting of fields is correspondingly more frequent and the resting period longer if the area is large enough. The soil nutrient status may be regard6'd as the chief limiting factor, which largely determines the crop husbandry system in the southern region of lateritic soils. Secondly, reaction and salinity also influence the agricultural system in so far as they limit the range of suitable crops. As we have seen in Chapter 2, the clay soils in the Jonglei Area are often highly alkaline and contain a considerable proportion of soluble salts. In consequence a number of crops which require acid conditions cannot be grown successfully on these soils. The alkalinity may also cause some of the nutritional shortages in these soils, e.g. phosphorus deficiency. ECONOMIC, HISTORICAL AND SOCIAL FACTORS ,. The influence of economic, technological, social and historical forces on the agricultural systems of the Jonglei Area is more complicated and obscure than that of the physiological environment. To be fully understood it would require much further study of the anthropological, economic and historical background as well as of the technical aspects. We have had no time, opportunity or means to carry out such a study of so large an area in the detail required, and though a certain correlation between these forces and the agricultural system has become evident, we cannot even attempt to present a complete and clear picture of the interdependence of present agricultural economy and these factors. We can only point out a few more obvious instances of the interrelationship. As may be seen from the description of physiological environment, the cultivator in this area faces considerable difficulties and hazards resulting from the operation of different environmental limiting factors. The law of diminishing returns begins to operate at a com- paratively low level, that is to say, additional labour effort above a certain minimum cannot be expected to produce worth-while returns, and the cultivator's opportunities for advancement by industry are limited. Moreover the climate of the area is not conducive to hard effort. The operation of the law of diminishing returns and the necessity for struggling with nature for a bare existence leaves little chance for accumulation of the capital necessary for advance- ment. Consequently, in an environment difficult for agriculture, the local peasant has little opportunity to progress without outside help. The physical environment, though difficult and uncertain from the point of view of agriculture, is not in itself exacting enough to provide incentives. Clothing is unnecessary for survival, the simplest shelter is sufficient, and nature provides a considerable part of the necessary diet with comparatively little effort on the part of the inhabitants; only limited exertion is needed to supplement natural products to reach a minimum dietary level. Adaptation to present circumstances necessitates a migratory way ofIife and this in itself discourages virtuosity or the desire fcir more pennanent establishments. Incentives to harder work are therefore small; so also are incentives to accumulate capital, as the amount of capital necessary to over- come environmental limitations is in any case outside the scope of the saving of the local population. At present the inhabitants of the area are generally satisfied with their moderate achievements, and consider the additional effort involved in labour, saving, or even learning, not worth while. Both opportunities and incentives are subject to changes by outside economic, social and technological influences. Until recently, however, such influences have been only the minimum. On the one hand natural difficulties in communication have prevented extensive contact with the outside world; on the other the resources of the area have been generally insufficient, in the face of these difficulties of communication, to attract outside influences. As a result the agricultural economy of the Jonglei Area has remained almost entirely outside the orbit of world economy, and even of the economic life of surrounding countries. Exchange of goods has been very limited, so that the internal economic-and hence agricultural-system has developed to satisfy only internal needs. Isolation has also prevented the spread of general knowledge, which usually, by increasing demand, awakens incentive. It has in particular prevented the spread of the technical knowledge and skill which would improve the present primitive agricultural system. These circumstances have also prevented the influx of foreign capital, necessary for economic progress. In recent years contact with the outside world has increased and agriculture has begun to develop; but this development is inevitably a slow process. The agricultural system still aitps at self-sufficiency down to the family unit, methods are rudimentary, and output is low, being barely enough to satisfy minimum internal needs. Moreover public security is essential for progress in agriculture, particularly in crop husbandry. 325 Until recent times the history of the area has been unsettled, and this is undoubtedly largely responsible for the lack of advance in agriculture. As a rule the peasant is a most conservative individual, and tradition is often responsible for the ueculiarities of agricultural practice. In some cases the reasons for practices handed down from generation to generation have been forgotten, and have now to be rediscovered by investigation or conjecture based on knowledge of agronomic principles. In others where the practices originated in different environmental conditions and were brought by the people to their new habitat, the original reasons are no ~onger valid; the practices, however, still continue. The locall peasant has discovered ways of raising his crops and tending his animals through centuries of trial and error; he knows that any sudden departure from established crops and methods in most cases spells disaster. Hence his conservatism and deep suspicion of anything different from what is customary. The preoccupation of so many peoples of the area with cattle and the social and ritual values involved have already been mentioned in various parts of this report. Here we must consider these attitudes again from the point of view of their influence on crop husbandry. The limiting factors of the physiological environment of the Ionglei Area affect crop husbandry • to a far greater extent than they do animal husbandry. The area is thus more suitable for stock-raising than for crops, so that the economics of the people's life are concentrated largely on the former. This concentration on stock, and the fact that the growing of crops requires harder work which often produces disproportionately low returns, has doubtless created a one-sided attitude towards these two main sources of livelihood. While domestic stock, and especially cattle, are objects of love and deep interest, crop husbandry is often regarded as a necessary evil. A man may be extremely proud of his cattle, but excessive pride in his crops would be very unusual and might even be considered' bad form ' . A minimum amount of crops has to be grown, but once this is done, no further interest is taken in them. This attitude is very largely responsible for the lack of progress in cropping practice. Moreover the local systems of animal husbandry involves seasonal migrations and the great majority of the inhabitants of this area lead a semi-nomadic existence, which is never conducive to advanrement in crop husbandry. 2. CROPPING SYSTEMS INTENSIVE CULTIVATION The lands especially suitable for the cultivation of crops in the Ionglei Area are restricted, and therefore subject to more or less intensive cropping. In the poor ironstone country round Yirol and in parts of the Bari and Mandari area, though land naturally suitable fQr cultivation is less limited, intensive cropping of homestead cultivations is common. Accessibility from the homesteads is one of the reasons for this, for apart from the natural disinclination to walk long distances to cultivations scattered in the bush, damage to crops by wild animals and birds is a constant danger and cultivations near the homesteads are more easily protected. Finally, this area is usually covered with bush, and the difficulty of clearing new fields is largely responsible for intensive cultivation on sites concentrated near the homesteads. In the Flood Region the homestead cultivations surround the wet season habitations on the high land, as it is less susceptible to flooding; suitable flood-free land in this Region is very restricted (see Fig. E 1). The intensity of the cultivation also naturally depends on the pressure of population. This pressure is undoubtedly greatest on the Shilluk ridge, where some fields have been cultivated, with only occasional faRow years, throughout living memory. In the rest of the area the high land is a little more plentiful; it is nevertheless cultivated very intensively, though more fallowing is generally permitted than on the Shilluk ridge. The Nilotic and other peoples of the southern parts of the Ionglei Area generally divide their fields into , homestead ' cultivations and' out-cultivations' (outlying cultivations), each family unit having plots of various sizes of each type. It is the homestead culti vations which are as a rule subject to intensive cropping. In the northern, semi-arid part of the Ionglei Area, two kinds of land are considered specially valuable by the semi-nomadic Baggara and Abialang Dinka; the sandy qoz soils are particularly suitable for some crops, such as groundnuts and sesame, and where they occur only as isolated outcrops they are generally cropped intensively. The clayey soils lying at the bottom of qozes and jebels are also often subject to intensive cropping with sorghum, as the run-off from the adjacent slope supplements the scanty rainfall. As well as semi-nomadic cattle-owners, we find in this Region a more settled population with permanent villages, for the most part near to the river. The land adjacent to these villages is usually also intensively cultivated year after year. The cropping system on the new pump schemes in this area must also be classed as intensive cultivation, but this is ' discussed in more detail elsewhere (see p. 331). 326 Unless the farming system is very well balanced, intensive cultivation always leads to deterioration of yields, until a new man-made balance is established between the crop and its environment. Usually, b~fore such a balance is reached the cultivator is forced to abandon the field because the yields are not comm~nsurate with his efforts. Mono-cropping, only too common in the Jonglei Area, is particularly inclined to lead to deterioration of yiekls under intensive cultivation. Double cropping, that is the growing of two (n~ar1y always sorghum) crops on the same field in the same season, puts an even greater demand on soil fertility. Deterioration of yields is usually ascribed to soil exhaustion ; this, however, is not neces- sarily true. Various factors of a cumulative character can be responsibile. It is true that on the lateritic soils of the southern and south-western parts of the Jonglei Area crop deterioration is probably due mainly to soil nutrient exhaustion. Rainfall here is not a main limiting factor, and the soils of this type are generally poor. Both nitrogen and available phosphorus, and on lighter soils even potassium, can be exhaust~d by prolonged cropping. Experiments carried out in this area confirm this. As far as the Flood Region is concerned, it is much more difficult to lay down a rule con- cerning any single factor causing deterioration in yields. Rainfall is so variable here that, be- cause of frequent periods of drought, itis doubtful whether the crop is given a ohance of exhausting the soil nutrients, except on the poor, s8Jndy soils. Experiments carried out at Malakal on heavy clay soil previously cultivated for many y~ars, and actually abandoned by the local cultivators because of deterioration in yields, showed no striking response to manuring, except where irrigation was applied. This supports the view that nutrient exhaustion cannot be the main reason for deterioration in crop yields. Structural deterioration of the soil following intensive cultivation is probably responsible, at least in part, for reduction in yield. However, the chief factor in this combination is almost certainly the accumulation of weeds, parasites, and diseases on the constantly cultivated fields. Striga hermonthica is the most important and spectacular agent of yield reduction. In the Semi-Arid Region the availability of moisture in the soil is of paramount importance. Consequently intensive cultivation, often leading to increased weed infestation, results in greater competition between crop and weeds for limited moisture and in a drop in crop yields. Increased infestation with Striga hermonthica, and to a lesser extent with crop pests and diseases, is also a very important factor causing deterioration of yields on such land. The lack of response to fertilizers in manurial trials on unirrigated clay soils in the Semi-Arid Region suggests that some factors other than the availability of soil nutrients determine crop yields on intensively cultivated land. On the other hand if moisture shortage is eliminated by irrigation, and infestation with Striga hermonthica etc. is controlled, the exhaustion of soil nutrients (mainly nitrogen and phosphorus) may lead to yield deterioration(3). On much poorer, sandy soils deterioration of nutrient status may be responsible for poor rain-grown crops, especially if combined with intensified competition between crop and weeds for the plant nutrients in impoverished soil. Crop rotation is the usual method of counteracting the deterioration of yields due to intensive oultivation. Yet no orthodox rotation is practised anywhere in the Jonglei Area, with the exception of pump sohemes in the northern parts, though intercropping of at least a certain part of the intensively cultivated fields is common throughout the area. In the northern region mixtures of sorghum and cowpeas are often found on clay soils, and mixtures of groundnuts and sesame on the sandy outcrops. In the Flood Region sorghum and maize grown on homestead cultivations are usually interplanted with cowpeas. In the southern region a considerable variety of crops mixed together are planted in one field, including sorghum, maize, and finger millet, as well as cowpeas, groundnuts and cucurbits. Interplanting is a form of simultaneous rotation and is generally found in densely populated areas in the tropics, such as India, China, and certain parts of Africa(4). Mixed cropping, if rightly applied, can in suitable conditions replace orthodox successive rotation. Unfortunately in the Jonglei Area it is not practised extensively enough, and considerable improvements could be made (see Vol. II, p. 581). Intensive cultivation, especially in conditions where nutrients are the main limiting factor, can be carried out successfully only if a sufficient level of manuring is applied. In fact through- out the area under study a certain amount of manuring takes place on the more intensively cultivated fields. It is usually incidental, and caused by the fact that during the wet season the animals are concentrated round the homesteads. However, the rate of manuring must be low and for that reason response is probably insignificant. Experiments carried out in the Gezira led to the conclusion that the value of animal droppings in that area was insignificant(S). Though the concentration of animals per intensively cultivated feddan in the Jonglei Area is probably much higher, it must be remembered that a large proportion of dung is burnt nightly and never reaches the soil. This burning of dung accumulated at night in the cattle-byres is 327 considered by the local population a,s essential for keeping insects, mainly mosquitoes, off man and beast. The fact that the dung is not returned to the soil is no doubt one of the reasons for poor yields. Some of the western Dinka tribes have developed a system of tethering their cattle in homestead fields before cultivation(6). This practice is, however, unknown among most Nilotics inhabiting the Ionglei Area. Its exttmsion would undoubtedly be beneficial, as confirmed by experiments carried out by us at Malakal and by previous experiments in Bahr el Ghazal Province. The Shilluk practice of path is another method of counteracting loss of fertility in intensively cultivated fields. At the beginning of the rainy season the women and older children of the household cut considerable quantities of toich grasses-Oryza barthii being the first choice- and carry it to the homest(:ad cultivations, where it is spread thickly over the ground and burnt. The practice of burning stubble to increase fertility was known to the Romans, and Virgil put forward various possible explanations. Heating of the soil has also been practised from time immemorial in India, and is known as rab(7). The beneficial effects of path are probably due to four causes, but further scientific investigation is required to confirm these hypotheses: (i) It is well known that partial sterilization of soil by heat or chemical compounds increases the amount of nitrogen(8). Hence the practice of path may benefit the succeeding crop by its sterilizing effect on soil micro-organisms. (ii) Path is regarded by the Shilluk as almost indispensable for intensive cultivation of non-cracking sandy clay (dethworo) soils. The main limiting factor of these soils is their physical inferiority. Heating induces a certain amount of cracking of the surface layer, and consequently improves the physical condition of the seed-bed, giving better germination and early growth. (iii) The manurial value of the minerals added in the form of ash cannot be disregarded. (iv) Finally, and probably most important, the effect of path is to produce a clean seed-bed. The burning is done after the early rains have caused the germination of weeds, and it kills this young growth. From this aspect, path may be regarded as a modified form of hariq cultivation, common farther north, which we describe in detail later. In spite of these ways of maintaining fertility, both accidental and deliberate, a system has not evolved which would permit intensive cropping over long periods without considerable reduction in yields. Indeed, with the exception of pump schemes in the northern parts of the Ionglei Area, the husbandry practices on fields subjected to extensive cultivation do not essentially differ from those applied to the intensive shifting cultivation, where long periods of fallow allow the restoration of fertility. As the pressure of the population on the land increases, problems of the maintenance of fertility become more urgent. In some parts of the Ionglei Area, such as the southern Shilluk district, over-population, caused by the disproportion between the amount of high land available and the number of inhabitants who apply extensive methods to intensive cropping, leads to a chronic shortage of grain. SHIFTING CULTIVATION Shifting cultivation must be regarded as the most important cropping system in this area at present, as in this way is produced the greater part of the total crop output. This system is practised on the largest part of the cultivation area and there is no clear-cut distinction between this method and intensiv~ cropping. As the pressure of population increases, more and more land, in order of suitability, becomes cultivated more and more intensively. In some regions, such as the Shilluk district already mentioned, shifting cultivation may now be considered as a system of farming marginal land. In other districts, espedally where there is little difference between high land and intermediate land and all the land may be regarded as marginal, very nearly all the crops are grown by this system, and even the homestead cultivations, together with the homesteads themselves, are shifted once th~ yield drops below the economic level. Shifting cultivation is a primitive form of rotation, consisting of periods under crop followed by long periods of fallow during which the land has a chance to regain fertility under indigenous vegetation. The relative length of th~se periods varies, but in the Jonglei Area where, apart from sites particularly suited to crop growing, there is almost unlimited marginal land, the cultivation period is fixed by the rate at which fertility deteriorates, while the resting period is as a rule long enough for the natural vegetation to reach its ecological climax. On the sandy soils and some poor soils of the Ironstone Region, soil nutrient exhaustion, causing deterioration in yields, induces people to abandon old fields and shift to fresh ones. On the strong clay soils, Striga hermonthica infestation is usually the main reason for-shifting cultivation on to new land. Hariq cultivation, described below, is also essentially a shifting cultivation system, with frequency of shifts determined by the success of 15urning as a cleaning operation, and by the period necessary for the recovery of hariq grasses. 328 The main drawback of the shifting cultivation system is its wastefulness of land and labour. In general there is fortunately still enough land for everyone, but clearing of new fields, especially on land covered with bush, requires very much more effort than the preparation of fields already cultivated. If, as is the case on some poor, sandy soils covered with bushes and trees, it is necessary to move the cultivations every few years, a disproportionately large arr~unt of labour is expended in clearing fresh fields, which could probably be used more profitably in other cultivation work. It should be noted that many clay soils, able to support cultivation for long periods, are covered only with grass, and this involves less work in clearing. It is well known in this area that a freshly cleared field does not produce its maximum yield until the third season under cultivation. The explanation for this probably lies in the different equilibrium existing between soil and natural vegetative cover and between soil and crop, and the necessary lapse of time for the establishment of the new equilibrium. Field observations here confirm similar observations made elsewhere, that during the transition period the first crop suffers from nitrogen deficiency. Rounce(9) offers two possible explana- tions: (i) Addition of a large amount of carbon stimulates bacterial activity and causes fixation of nitrogen in organic forms. (ii) Low level of organic material-i.e. lack of carbon in virgin soil-causes a low level of free-living nitrogen bacteria. The addition of the roots ' of the cultivated crop leads to an increase in carbon, a subsequent increase of nitrogen-fixing bacteria and an increase in the amount of nitrogen in the soil. Experiments in the United States of Amedca and India show that in arid soils these bacteria playa major part in supplying the soil with nitrogen. Further investigation into the cause of low fertility and particularly nitrogen deficiency in newly opened land in our area is needed. If the main cause is nitrogen fixation in an organic form, the addition of carbon must be considered harmful, and a method of opening new fields should be designed to avoid mixing plant debris with the soil, e.g. by burning all natural vegeta- tion, early ploughing, etc.(lO). If, however, low population of nitrogen-fixing bacteria is responsible, then the addition of carbon, which would stimulate their activity, is desirable, and all wastage of organic carbon, by burning, etc., must be avoided. In any case it is generally advisable to plant the late crops on the newly opened land so as to allow time for the re-establish- ment of carbon-nitrogen equilibrium in the soil. In other areas in Africa it is a common practice to plant crops other than sorghum in the first season of cultivation; leguminous crops, sweet potatoes, cassava and even finger rnillet(u). In the Bari and Mandari districts groundnuts and cowpeas, with or without finger millet, are planted as a rule. Elsewhere in the Jonglei Area sorghum is planted, as no other crops are suitable for the type of land where shifting cultivation is practised. HARlQ CULTIVATION " Hariq (from the Arabic haraq, ' to burn ') cultivation makes use of old grass growth t6 get rid of new by burning the old and new together, leaving clean land for sowing(12)". It is widely practised on the grass plains of the central Sudan lying in the 450-700 rnm. isohyet belt. Hariq can be practised only where annual grassland predominates and forms a mat thick enough to ensure a burning sufficiently intense to kill young, sprouting plants. Rainfall must therefore be sufficient for good growth of annual grasses and not heavy enough to promote extensive growth of perennial grasses, especially the tufted species. The main hariq grass in the 450-600 mm. rainfall belt is Sorghum purpureo-sericeum (Ar. anis). In the 600-750 mm. rainfall belt this grass, together with Hyparrhenia pseudo- cymbaria (Ar. ansora), provides the hariq mat. Brachiaria obtusiflora (Ar. um cher), Pennisetum mollissimum (Ar. dukhn misikhat), Aristida mutabilis (Ar. danbalab), Cymbopogon nervatus (Ar. na'al) and Rottboellia exaltata (Ar. um bellila) are also suitable for this method of cultiva- tion, but less common and less universally distributed(13). Though the main purpose of hariq is to obtain clean ground for sowing, it has also the advantage of conserving soil moisture. The grass mat is not burnt until there has been sufficient rain to induce germination of seeds. Consequently, during the early rains the soil structure is protected against the puddling effect of raindrops and the moisture is able to penetrate deeper into the soil, while at the same time evaporation is reduced. It is not yet possible to decide how far burning itself is beneficial owing to sterilization of the soil and improvement of its physical characteristics, a problem already discussed in detail in connection with the practice of path among the Shilluk (see p. 328). In the heavier rainfall belt, especially after a season of good rain, the accumulation of grasses may be sufficient after a single ytar to produce a good hariq mat. In general, however, a minimum of two years and sometimes three to five is necessary for the best results. The 329 recovery of annual grasses after a period of cultivation depends largely on the effectiveness of the original burning and subsequent weedillg of the crop. It depends further on rainfall and, of course, on the duration of cultivation, though this, in real hariq farming, is not more than one season. The"ti.min.g of the burning is most important. If the mat is burnt too early, before the early rains have induced germination of the majority of the grass seeds, these will germinate later and little benefit will be derived from the clearing effect of the fire. If it is burnt too late the hariq may will remain unburnt in patches, and eVen in plac~ where it has been burnt some weeds will survive since the fire is less powerful and the growth of weeds more advanced. With good judgl'Ilent and good IUGk dean land can be obtained, and only a Gomparatively small amount of subsequent hand-weeding will be necessary. Hariq cultivation is not a new introduction to the Ionglei Area. It is, and was, used on suitable land as a method of opening up new GUltivations or extending existing ones. If no further hariq land is available, cultivation is continued by ordinary weeding. Hariq is still plainly carried out in this area on a limited scale by family units, and must therefore be regarded as a method of land preparation for sowing rather than as a definite farming system. Three conditions are necessary for hariq to become a farming system : (i) An extensive area of suitable grassland. (ii) Capital for fire protection of land recovering from cultivation, especially in the season preceding controlled burning, amI for the employment of a considerable amount of seasonal labour ; though for the peasant this may be found within the extended family. Often capital is also necessary for the provision gf drinking-water supplies and the ojilening of roads, which are seldom found in hariq areas. (iii) Managerial ability to make proper use of land and capital resources. Hariq cultivation as a farming system needs therefore a capitalist farmer or an organized peasant co-operative, and this is too difficult for a primitive and economically de-centralized system based on the peasant family unit. The hariq farming system is essentially suitable for extensive grain production, provided natural conditions are favourable. In recent years the government and a few of the more enterprising and richer northern merchants have begun to provide the necessary capital and managerial skill; this has resulted in the rapid development of crop production by hariq methods, especially in the Paloich area. The government has concentrated on fire-lining, extension of domestic water supplies and the road system, so that large tracts of potentially suitable but hitherto unexploited .hariq land have been opened. The fire-lining campaign has been largely responsible for the increased crop output. Fire-lining is still done by manual labour, though in the 1952-53 season tractor-drawn machinery was used, considerably increasing the protected areas (see Table 159). The following method is usually employed to clear fire-lines by manual labour. At the beginning of the dry season an area of suitable grassland (generally a stand which has not been burnt accidentally in the previous year) is chosen. Fire-lines dividing this area into a number of large blocks (2 km. square to 7 km. square) are made by hand-cutting grass in bands 5-8 m. wide and approximately 100 m. apart. Additional cross-bands to control the fire are cut every 100 m. Subsequent burning of the grass between the cleared bands, against the wind and under strict supervision, produces a fire:line 100 m. wide, which provides sufficient protection if the people take the least care to control their fires. Unfortunately at present this is not always done, and as a result some protected hariq areas are burned accidentally every year. Table 159 gives the areas protected against fire (chiefly in the Paloich area) during the last four seasons. TABLE 159 AREAS OF HARIQ GRASSLAND PROTECI'ED AGAINST FIRE UNDER GOVERNMENT AUSPICES Season IA rea in feddans I Remarks 1949-50 2,200 1950-51 9,000 1951- 52 11,000 1952- 53 70,000 Increase due to partial mechani- 1953- 54 40,000 zati&n (Figures kindly supplied by the Inspector of Agriculture. Melut.) 330 The inadequacy of drinking-water supplies in the hariq areas and lack of access to roads have recently become the main factors limiting further extension, and a large-scale programme based on haftr digging and the provision of roads is being carried out. This should permit further development in hariq cultivation, but will also tend to tum some of the hariq areas into areas of permanent cultivation. ,. The districts in which hariq cultivatioN has been developing most rapidly in recent years (Paloich, Maban and northern Shilluk), though lying on the southern extremity of the true hariq belt, are specially suitable for this type of farming. They are within the 650-700 mm. rainfall belt, so that rainfall is sufficient to promote rapid development of both a good hariq mat and the crops themselves, while at the same time it is insufficient to induce rapid change from annual to perennial grassland under conditions of regular burning. In areas farther north, where this type of cultivation is also widely practised, rainfall is often light and generally unreliable and yields of crop are often disappointing. In areas farther south hariq land is more limited, there are more perennial grasses in the sward, necessitating subsequent weeding, and the change in the ecological character of the sward from annual to perennial grasses following periodic burning is more rapid. In this area, as far as 90 N., patches of Hyparrhenia pseudo- cymbaria suitable for hariq are often found on old cultivation sites, as a transition stage in the process of regeneration to the natural climax of predominantly perennial grasses. How far these could be used for hariq cultivation depends on the infestation of the ground at this stage with dormant Striga hermollthica seeds. Two crops are mainly grown on hariq-cJeared land: sorghum and rain-grown cotton. Sesame and cowpeas are also sometimes found on this land. This type of cultivation therefore permits large-scale development of the main food and cash crops. The economic importance of the extension of hariq cultivation in the Paloich, Maban and Shilluk country cannot be over-estimated. Since its inauguration the hungry areas farther south have been, to a very considerable extent, fed with the surplus produced in the hariq area. Upper Nile Province as a whole, thanks to this form of cultivation, has become largely self-sufficient in grain and at present, in a good year, may produce some surplus. The main advantage of hariq cultivation as an agricultural system is its suitability for the Semi-Arid Region and the Transitional Belt between this and the Flood Region. It appears to be a well-balanced system involving one season under crop and a resting period under natural grass protected from fire, a definite rotation system often lacking in this part of the Sudan. Under true hariq farming the land cannot be over-cultivated, and Striga hermonthica has little chance to become a serious limiting factor. The value of the protection of natural vegetation for some period against fire, and the influence of this practice on soil fertility, especially soil organic matter status and structural development, have yet to be investigated; the value of hariq as an easy method of preparing land in a country where the chief agricultural implement is the simple hoe is obvious. IRRIGATED CROP-GROWING The need for irrigation and the engineering aspects are mentioned elsewhere in this report (see Vol. I, pp. 41-4, and Vol. II, pp. 619-22). It only remains to discuss here the agricultural side of irrigated crop hu~bandry. Irrigation in the Jonglei Area is at present used for two purposes: (i) The growing of vegetables and fruit. (ii) The growing of cash crops-mainly cotton with sorghum as a subsidiary crop. Irrigation is obviously necessary for the production of fresh vegetables and fruit throughout the year. Apart from the use of pumps in government, mission, and a few private gardens, a certain amount of dry season garden irrigation is done by shaduf, a primitive device for lifting water(14). These can be found in considerable numbers in the northernmost part of the Jonglei Area, while in the rest of the area, inhabited by Nilotics, they are limited to the vicinity of larger urban centres, where the non-Nilotic inhabitants create the demand for fresh vegetables throughout the year. It should be noted that the technique of using even a shaduf is very inferior in the Nilotic area, where it has been comparatively recently introduced. The fulcrum is often wrongly placed, the balancing weight is either too light or too heavy, and the length of rope or stick on which the water-container hangs is nearly always badly adjusted. It should also be recorded that, with a few exceptions, the gardens in the whole area are not well managed and are only too often too small to supp1y a sufficient quantity of either European or Sudanese vegetables. 331 The production of irrigated cotton and sorghum is at present confined to the Semi-Arid Region. There are two types of enterprise engaged in this sort of agricultural production: private pump schemes and saqia (water-wheel) schemes. The.development of private pump schemes in the J onglei Area started after the first scheme at Geigar, opened in 1946, had proved a success. The cotton boom of 1950-52 gave a strong stimulus, and development has been so rapid that it is almost impossible to give an up-to-date picture, as changes are taking place all the time. The future depends largely on cotton prices, since it would app€ar that some of the schemes established in the boom period do not rest on very sound economic bases. The urge for cash in the boom made some owners hurry with the installation of pumps, irrigation channels, the clearing of land, etc., almost regardless of cost and without a sufficiently detailed preliminary survey and investigation ofland values and other matters of primary importance. All present schemes are on a share-cropping basis, with the capitalist providing the land (on lease from the government), water, as a rule equipment such as ploughs, and the major part of the working capital. The tenants provide labour, som€: of the hand implements and the remainder of the working capital. The running of these schemes is governed by the Nile Pumps Control General Regulations 1951, and specifically the Blue Nile Province Tenancies Regulations 1947 and the Upper Nile Province Tenancies Regulation 1953. No duty on water is at present charged to landlords or tenants by the government. In theory all pump schemes in this area are run on either a three-course rotation (cotton- t sorghum + t leguminous crop-fallow) or a four-cours€: rotation (cotton-sorghum- fallow-fallow). In practice, however, the following rotations are employed: TABLE 160 ROTATION EMPLOYED IN PRACfICE ON PUMP SCHEMES IN JONGLEI AREA 1952-53 SEASON Number Number of Pump of Details of rotation Schemes courses on which employed Cotton-t Sorghum + t Fallow-Fallow Cotton-2 Fallow ... 2 4 Cotton--Sorghum-2 Fallow 6 4 Cotton-t Sorghum + t Fallow-2 Fallow 8 4 Cotton -3 Fallow ... 13 4 Cotton-t Sorghum + t Groundnuts- 2 Fallow ... Apart from cotton and sorghum, some fruit and vegetables are usually grown in the irrigated gardens belonging to the owners of pump schemes. (In the 1952-53 season 28 out of 31 pump schemes possessed such gardens.) Some further numerical details concerning the pump schemes in this area in the 1952-53 season, gathered in the course of a survey carried out by Abu Bakr Elf. Osman Arbab under the direction of the agricultural member, are given in Table 161. The average tenancy acreage on the pump schemes with three-course rotation is 15 feddans gross area, while on the pump schemes with four-course rotation it is 16 fed dans gross area. A long staple cotton is grown on these pump schemes. The area north of Jebel Ahmed Agha has been declared a long staple cotton region and no rain-grown short staple cotton is allowed to be cultivated there. This legislation is intended to control the spread of pests and diseases and the mixing of different types of cotton, which would endanger the reputation of Sudan cotton on international markets. A 25-mile cotton-free belt is imposed between the long staple and short staple cotton regions. The technique of cotton-growing is similar to that employed in the Gezira area(15). The yields vary between 3 and 10 Kantars (315), and 5 to 6 Kantars (315) may be regarded as average. The late sowing is undoubtedly one of the main factors which reduce yields. 332 TABLE 161 SOME NUMERICAL DETAILS CONCERNING PRIVATE PUMP SCHEMES IN THE JONGLEI AREA 1952-53 SEASON .. Quantity Kosti District I Kosti District I Upper Nile West Bank East Bank Province Total I Pump Schemes with pump inlet: diameter 20' .. . ... ... No. 1 - - 1 n 10' ... ... ... - - 2 2 n 8" .. . ... ... n I 1 6 8 n 6' .. . . .. .. . n 5 12 - 17 n 4' .. . ... ... n n - 3 - 3 Total ... ... .. . n 7 16 8 31 Total area canalized .. . Feddans 10,955 4,510 7,285 22,750 ···1 1 1 1 Area cropped with: Cotton ... ... . .. ... Feddans 3,588 1,123 1,800 6,511 Sorghum ... ... ... . .. 176 354t I 900 1,43ot Groundnuts ... ... ... n - - 100 100 Vegetables and fruit ... ... " 44 131 85 260 " Total ... ... .. . " 3,808 1,608t 2,885 8,301t Average yields: Cotton ... ··· 1K antars (315) I 5·0 4·8 IA ve~age 5·0 Cultivation Implements, etc. : Ploughs (all types) .. . ... No. 35 44 36 115 Draught bulls and oxen .. , 36 94 90 220 Tractors " '" '" 6 - 2 8 Pump Schemes on which tractors " were used '" '" '" " 3 - I 4 NOTB: This table is based on the information provided by owners and the Team cannot guarantee the accuracy of the figures given, The standard of land pr(lparation and after-sowing cultivation is generally low. It is generally held by pump scheme owners and managers that cotton derives little benefit from late (March) irrigation. This question is important from the point of view of water economy and the economics of crop production. It is discussed in some detail in Volume II of this report (pp. 604-5). The technique for growing sorghum is also similar to that employed in the Gezira(16), though in practice less care is given to the crop. Yields are about 3 ardebs (450 kg.) per feddan. Apart from private pump schemes, irrigated cotton is grown on a considerable scale on saqia schemes. In the 1951-52 season the estimated number of these saqias was 150, with an irrigated cotton area of about 3,000 feddans. Metal low-lift Middleton saqias are used almost exclusively. Most saqias are owned by single individuals (merchants, etc.) and are run by direct labour. Tenants are permitted on the same shar(l-cropping basis as on pump schemes, each tenant having 5 feddans of cotton. Most of the benefit to local people from the rapid development of these saqia schemes is in the provision of a source of employment, not in a direct share of the profit. • Brief details of this type of cultivation are as follows: Saqia licence: Annual, allotted preferably to natives of the district. Water used is not debitable against the Sudan quota. Land: Mainly government-owned and rented to saqia owners on an annual tenancy at a rent of loom/ m per feddan. Gross area of each saqia scheme is 60 feddans. Rotation: Cotton---t Sorghum or Fallow+t Fallow-Fallow, i.e. 20 feddans cotton , 10 feddans sorghum and 30 feddans fallow per annum. Saqia lift: Maximum 2 m., optimum 1.80 m. Cropping : X 1730 A variety of cotton is sown throughout; it is usually established on rain and receives irrigation only in later stages. Ammonium nitrate is commonly applied, and spraying against pests (mainly jassids and thrips) often practised. The standard of cultivation is very low, but good yields are nevertheless attained. Very little attention is paid to sorghum ; it is sown-if at all- on rain and later irrigated. Saqia owners claim that it is quite uneconomic. Onions, which used to be grown on a large scale and could be regarded as the backbone of the saqia system, are at present almost entirely replaced by cotton as the chief cash crop. 333 Costings: . The net cost of establishing a Middleton type saqia in 1952 was estimated at £EI,200, including purchase, canalization and erection of the saqia (£E230), and cost of draft animals and incidentals. Two pairs of camels or three pairs of bulls are required and the wastage rate of these animals is high. Ne~rly all the saqias depend on the early filling and late emptying of the Jebel Aulia Reservoir, which enables the cultivator to sow the cotton reasonably early and to continue watering until most of the crop is picked; in the 1951-52 season, when the dam was opened unusually early, many of these saqias suffered considerably. Yields were low and the quality of the last pickings very poor(17). DRY SEASON RIVERAIN FLOOD-PLAIN CULTIVATION This type of cultivation, though supplementary, plays a very important part in the economy of the people inhabiting the northernmost and southernmost parts of the Jonglei Area, especially the former. In the central parts inhabited by Nilotics it is of little importance. In the region inhabited by Bari and Mandari, the cultivations on the riverain flood-plain are found both on areas dominated by Phragmites and on the deeper-flooded parts, usually dominated by Echinochloa species. Phragmites dominated toich, subject only to sporadic floods from the river, is cultivated to some extent throughout this area. The general practice is to plant maize and, in the southern part of this area, quick-maturing sorghum in the early rains, and to harvest in early July before the rising river inundates the land. If flooding is not serious, a second crop of maize and quick-maturing sorghum is often planted in October or November. Green gram and tobacco are also sometimes planted at the beginning of the rainy season, and the latter is transplanted later on to higher ground. The cultivation of deeper-flooded toiches is mainly limited to the area south of the latitude of Terakeka. A considerable variety of crops, including maize, cowpeas, green gram, tobacco, pumpkins, lady's fingers, and cassava (which seems resistant to flooding) are planted as soon as the flood recedes. Some of these crops, especially tobacco, may be hand-watered at the end of the dry season. These cultivations are of considerable value to the local popUlation, assuring them of a steady supply of food throughout the year. Their importance is especially great for the cattleless sections. Gur/us) cultivation in the Kosti area is carried out along the belt of toich land exposed after the emptying of the Jebel Aulia Dam, and on small inland areas flooded from the river by khors and known locally as amara (i.e. not registered). The local population regards this type of cultivation as very valuable; it supplies fresh vegetables at a time when otherwise there would be none, and in addition provides a source of cash income, since many of the vegetables are sold in the markets of Kosti, El Jebelein, Geigar, etc. The main crops are cucurbits, of which pumpkins and marrows (Cucurbita spp.) and sweet melons (Cucumis melo) are the most iinportant; water-melons (Ciirullus vulgaris) and agur (Cucumis sativus) are also grown to a lesser extent. Maize is sometimes planted, but it is alleged that it requires irrigation in the later stages of growth, although very quickly maturing varieties are planted. Maize is usually cultivated in any quantity only in years when the sorghum has failed, and only in areas where shadu/ irrigation is possible. Cultivation begins with clearing the land of toich grasses in late February, and cattle are kept off the land to avoid puddling of the soil. Sowing begins as early as possible after the Jebel Aulia Dam is opened, usually in the first halt of March. If the ground is soft and moist, sowing is done with a sowing-stick (seluka), but if it becomes too dry and hard, a digging- hoe (toria) is used to make deeper holes, and the seeds are covered with fine soil brought from elsewhere. Weeding is frequent and, as much attention is paid to the crop, yields are generally good. The commercial value of the crop is probably at least £EIO per acre. The total area of this type of cultivation in the Semi-Add Region is estimated at 3,250 feddans. It is alleged that the new regime created by the erection of the Jebel Aulia Dam has greatly reduced the areas suitable for gUlf cultivation, and that in years when the late emptying of the dam coincides with the early rise of the river, crops are ruined by flooding before they mature. The effect of the Equatorial Nile Project on the future of these toich cultivations is discussed in Volume II, pp. 530-3. 3. IMPLEMENTS Agricultural operations in the Jonglei Area are carried out entirely by manual labour, with the help of the most rudimentary tools. An\mal and mechanically-drawn implements have appeared in the northernmost part of the area on private pump schemes and saqias oniy during the last few years. The low produotion of crops is largely due to the use of primitive 334 tools, which, inefficient as they are, represent a serious handicap to increase in areas of cultiva- tion, better preparation of fields and improved weeding. On the one hand the cultivator faces one of the least friable soils, most difficult to till; on the other hand the implements at his disposal are most inefficient. Though there appears to be a great variety of different shaped and sized tools, this fs largely accidental. Traditional shapes and forms are now to some extent modified by the fact that most iron implements are imported from the North. There is hardly any specialization of implements for different tasks, each cultivator usually owning only a simple axe, a hoe, a sowing- stick, a harvesting implement, and occasionally a primitive rake. (1) Weeding-hoe (Ar. maloda; Shilluk kwer; Nuer pur; Dinka fur). This is found in various forms throughout the area, and is the main agricultural tool of the inhabitants. It usually consists of a metal head and wooden handle; the shape of the head may vary very considerably from district to district, but the cutting edge seldom exceeds 5 in. The metal is as a rule inferior soft steel, which will not retain an edge and wears very quickly. The handle also varies in length. The Baggara in the north, the Bari, Mandari, and some Dinka tribes in Bahr el Ghazal Province generally use a long handle, often well over 6 ft., while the Nilotics use a very short one, seldom over 2 ft. long, and adopt a sitting or kneeling position when using it. The handles are very roughly made, showing how little care is taken of the most important agricultural implement in use. A typical Nilotic hoe consists of a strip of metal about 3 in. wide and 6 to 9 in. long when new, though the blade wears away very quickly, bent at one end round a piece of wood which forms the handle. This type of hoe is used for preliminary preparation of the field and for subseqmmt weeding; only the most superficial cultivation can be achieved with it. When somewhat deeper cultivation is required the Nilotic, who either does not possess or does not like the heavy digging-hoe (toria, described below), sets the metal head at right angles to a curved handle, and chops the soil with this. This form of hoe is known as dikok in Shilluk, kocok in Nuer, and atoktok in Dinka. A similar form, i.e. a metal head at right angles to the haft, is used by some Baggara tribesmen for sowing, but in this case the handle is longer. (2) Digging-hoe (Ar. toria; Shilluk hamarab; Nuer toria or tuktuk). This implement is composed of a large metal head, usually of European manufacture, fastened at right angles to a wooden handle 3 to 4 ft. long. It is essentially similar to the small weeding-hoe described above, but longer and considerably heavier, the average head weighing about 3 lb., while the average head of a weeding-hoe weighs only lIb. The digging-hoe is used for deeper cultivation and it penetrates local soils to a depth of 2 or 3 in. Unlike the weeding-hoe it is a double- handed implement, and much more energy is expended by its user. Though it is found through- out this area, it is used extensively only by the riverain cultivators in Kosti District. The Nilotics in particular dislike it, regarding the heavy labour involved as not worth while. (3) Sowing-stick (Shill uk dikagi; Nuer cop). Different forms of this implement are found throughout the area. In Baggara country a seluka is usually employed. This is a wooden stick 3!- to 5 ft. long with a slightly flattened and curved end. A foot-rest is attached to it, and with this it is forced into the ground and rotated to form a hole. It is now also used by some Nilotics, and by them also is called a seluka. The Nilotics generally use a much longer stick, often as long as 10 ft., and as a rule it is made from the wood of Ziziphus spina-christi. The cultivator stands in his field and makes holes round himself as far as his stick will 'reach, thus avoiding the constant alteration of his position. . (4) Harvesting knife. Though 'both European manufactured and locally made sickles are in use in the Nilotic area, an old spear-head is the main instrument used for cutting dura heads. The Shilluk often use a small harvesting knife made of a strip of metal with a cutting edge on one side (palo). Similar knives are sometimes used by other Nilotic, Bad, and Mandari cultivators. Among the Baggara a two-sided blade with a hilt is universally carried in a sheath strapped to the left arm above the elbow, and is commonly used for cutting dura heads. (5) Grass-cutting knife . This is a most primitive instrument, consisting of a strip of metal 1t to 2 ft. long with a cutting edge and one end rolled to provide a handle. It is common and is used with a slashing action, similar to that of a machete, for cutting grass. It requires a great deal of effort with disproportionately small results. (6) Axe (Ar. las; Shilluk dora; Nuer jup; Dinka yup). This is used for clearing the bush before cultivation. It is generally made locally of soft iron, and is seldom heavier than 2 lb. The more advanced types have a socket in the head which takes the handle, but the majority are merely a wooden handle through which a hole is bored to hold the head. (7) Animal-drawn implements. These have only recently been introduced by pump scheme owners in the Kosti District. The Ransome' ET ' composite plough, ridger and hoe is most frequently found. The Ransome' Levant' plough, though probably more suitable for local 335 condltiotls, is unfortunately difficult to obtain now, and only a few are found in the area. In spite of the large number of potential draught animals present, animal-drawn implements are completely uNknown in the rest of the Jonglei Area. 4. CULTURAL OPERATIONS Cultural operations and the amount of care bestowed on crops vary from tribe to tribe and individual to individual, so that there are no general rules applicable to the Jonglei Area as a whole. By and large, labour rather than land is the factor which limits extension of crop production. Centuries of trial and error have taught the cultivator the right labour: land ratio in his particular conditions. Though to the outside observer it lIlay seem that the local cultivator does not give enough labour per unit ar,ea of crop, until his tools and husbandry methods are improved it is doubtful if he could increase total output by increasing the intensity of his oultivation, at any rate without reducing the acreage. This basic problem requires further study. LAND PREPARATION The main object of the local cultivator in preparing the land for sowing is to reduce to the minimum subsequent competition from weeds, and therefore weeding. Improvement of soil tilth, its physical properties, moisture conditions, aeration, consistency, etc., is given little thought. The actual method of preparation depends on whether the field has been sown before or requires opening up, and on the type of indigenous vegetation which must be eradicated. In the open grassland plains, apart from the hariq methods widespread in the Semi-Arid Region, which we have desoribed, the most common method of land preparation is simple weeding with a weeding-hoe or, less often, a digging-hoe. The roots of the grass are cut about t in. to 1 in. below the soil surface and the soil is seldom disturbed deeper than this; this softens the soil sufficiently so that sowing holes can be made with an ordiaary sowing-stick. In fields previously cultivated, this preliminary weeding is done after the first rains have induced germination of weeds. The more provident cultivator generally tries to clear as much virgin land as possible either at the beginning of or during the preceding dry season, as this not only makes for better distribution of labour throughout the year but also provides cleaner fields, since the uprooted grasses dry more thoroughly. Though after this early grass clearance a second pre-sowing weeding may be necessary to rid the field of annual weeds, post-sow.ing weeding is reduced, and the practice should be encouraged. Opening of new fields during the rainy season is more common, though more difficult and laborious. It has to be done at the same time as the weeding of earlier sown crops. It is, however, a usual process among improvident people and where the agricultural season is short because drinking-water runs out and seasonal migration to grazing grounds is necessary. Fields opened during the later rains are generally sown with a late sorghum crop, which during the first season often gives very disappointing yields. In the afforested areas, which in some districts are preferred for grain-growing, preparation of the field usually starts in the preceding rainy season. Trees and bushes are cut and either burnt on the spot or left to decompose and be eaten by termites. If any rubbish remains, it is burnt at the beginning of the next rainy season. The tree stumps are often left in the ground until they rot, and in some places they may be used during the first cultivation season for growing varieties of climbing bean. Some trees which produce edible fruit, especially Balanites aegyptiaca (Ar. heglig), or those which provide welcome shade, are preserved. The initial weeding of newly opened forest fields is done in the early rains. The Jonglei Area abounds in lazy farmers, and' short cuts' in pre-sowing weeding are often attempted. Seed may be sown without any initial weeding being done, even on virgin land, if the grass and herb stand is poor. Weeding is not done till gerplination is well advanced, and as a result young seedlings at a particularly susceptible stage have to compete with weIl- established weeds. Competition for water can be especially great during the early rains, which are often light. Another lazy method is to weed small patches only, in the centre of which the seeds are sown. This e1iminates initial competition for light and air, but competition for moisture and nutrients is reduced only to a small extent. Furthermore it is doubtf\ll if these methods save labour at all, as, though initial clearing is less, subsequent weeding is more laborious and less effective. Some Western Nuer tribes (Bul, Leik and Western Jikaing) have a somewhat more specialized method of land preparation, which is combined with their method of Striga hermonthica control. Whea weeding is done, the weeds, together with the soil attached to their roots, are carefully put together to form low banks, dividing the field into a number of 336 small basins. These banks make it possible to carry out controlled rain-flooding, considered to be a method of controlling Striga hermonthica (see p. 355). They also form a simple com- post heap, which is spread over the fields in the succeeding season. In other areas similar banks are often found serving simply as boundary lines between fields belonging to different owners. On sandy soils in the Semi-Arid Region they might be of some value as "an anti- erosion measure. Methods of preparing land for sowing on the pump schemes in Kosti and Renk Districts are in general little more advanced, except on pump schemes when') tractor and animal-drawn implements are used. SOWING Two methods of sowing are employed in this area; sowing in individual holes, made by sowing-stick or weeding hoe, and the broadcasting of seed. When the former method is employed, spacing is almost always irregular and only roughly adjusted to the requirements of climate, crop, and variety. Similarly the number of seeds placed in each hole varies a great deal. As a result, initial stands are very uneven and the cultivator spends considerable time on remedying this by thinning, transplanting, and resowing. . Though it is not generally appreciated, the thinning and transplanting is harmful to the young crop as it disturbs the root-system and often causes a loss of moisture in the soil round it. The seed-rate employed is also higher than necessary. Irrigated cotton is usually sown with greater care. In some areas certain varieties of sorghum are commonly sown by broadcasting, and some primitive form of harrowing the seed into the soil, either with a weeding-hoe or simply by pulling a tree branch over the sown field, is employed. Small seeds such as sesame and finger millet are almost invariably sown by this method. The standard of broadcasting seed is generally low, so that either too dense or too open, and usually patchy, stands result. Seed selection is usually practised, the best heads being chosen for next season's sowing. It is interesting to note that though varieties of sorghum belONging to one type are always sown in mixed stands, the heads chosen for seed are tied in bundles of separate varieties and stored separately, only to be mixed again at the time of sowing. This is probably a method of preserv- ing the right proportion of different varieties, established by many seasons of trial and error. In this connection, we must consider the whole question of sowing stands of mixed varieties of sorghum, a rule in the Jonglei Area, and the intersowing of one crop with another. Natural pure stands of anyone species are almost never found; in mecharuzed or semi-mechanized farming, pure stands are preferred for ease of post-sowing cultivation and harvesting. In primitive husbandry, where a great number of plants are grown and harvested separately rather than as a field, the advantages of mixing varieties and crops certainly outweigh the disadvantages of uneven growth, maturation, etc. A number of varieties grown together, and the mixing of crops, makes better use of soil moisture and nutrients, light and air. The parts of plants above and below the ground in a mixed stand are better distributed in the soil volume and the space above the soil, and a mixed stand is usually able to make use of greater volume than one of a single variety in which all plants are morphologically similar. The demands for moisture of a mixed stand are also spread more evenly through the season, as the periods of maximum demand vary according to the plant. Optimum and critical conditions vary from variety to variety and crop to crop, so that sowing of a mixed stand is a form of insurance in areas where climatic c(mditions may vary considerably from season to season. Cultivators in this area have found over the years that some varieties of sorghum are better able to survive drought, while others may be better suited to periods of flood. Consequently they plant a mixture and hope that at least a few varieties will survive the critical period, and then, free of competition from other varieties, will produce a reasonable yield, even though the others have failed. WEEDING The importance of weed competition as a factor limiting crop yields in this area should not be under-estimated. Poor tools and the limited stamina of cultivators are responsible for weeding being generally slow, ineffective and not frequent enough. Especially the first weeding, but often subsequent ·ones also, is carried out too late, causing serious competition at critical stages of crop growth and development. The late weeding of crops about to mature is often practised, but this gives little benefit to the crop and the labour involved could be used more profitably in the preparation of new land for next season's sowing. It is interesting to note that the sandy soils, generally requiring less weeding, are in fact weeded more frequently than the clay soils. The simple explanation of this is that weeding on lighter soils is mechanically easier and less expenditure of energy is required. . 337 BIRD-SCARING This is a universal, and often a major, operation in crop growing in the 10nglei Area, and it is possible for birds to reduce yields by more than three-quarters. The normal method of protectiQn in most parts of the area is to build a scaffold 10 to 15 ft. high with a platform on top ind on this to post children, whose job it is to scare away the birds. From dawn to dusk someone is constantly on duty, throwing mud pellets, shouting, cracking whips or pulling one end of a rope, to the other end of which is attached a jangling can, or simply banging a petrol can. These methods of protectmg the crops can be most effective(19). Un- fortunately it is usually difficult to protect those cultivations which are far from the homesteads, and damage by birds on these may be considerable. Similar methods are used for scaring wild animal pests, including elephants and hippopotami. Another method of protecting grain from damage by birds is partly to break the stalks of the sorghum and lay them on the ground to mature. This method can also be surprisingly effective. HARVESTING AND THRESHING Harvesting is invariably carried out by slow and laborious hand-cutting with a knife, spear-head, or much less frequently with a sickle, either of local or European manufacture. With larger crops the ears are cut individually, with smaller crops they are cut in handfuls. All cereals are cut just below the ear; only sesame is cut with the straw, and is tied in bundles which are hung on special frames to dry before threshing. The ears of other crops are either carried straight to the threshing-floor or, more usually, placed on specially built platforms, covered with grass and left till ready for threshing. Threshing is done on specially made floors of mixed clay, grass and dung, the last preventing cracking and consequent loss of grain. It is generally done with sticks or spear-shafts, and sometimes laboriously by hand. Grain intended for seed is usually kept in unthreshed heads until sowing time. As a rule the straw is not harvested but is left in the field, where some is eaten by domestic animals, mainly goats and sheep, while the rest rots in the field throughout the dry season, a certain amount being incorporated into the soil by termites. It may be worth investigating the fertilizing value of this process. Most Nilotic tribes (notably the Shilluk) use the stalks of the larger varieties of millet for fences round houses, screens for cattle-byres and cattle- camps, and for making mats. In the northern part of the area some dura stalks may be cut and stacked to provide fodder for domestic stock, or be used for screening the households, or, by some tribal sections, even for building. STORING Over-year storage of grain is almost unknown in the Nilotic parts of the area, even if there is enough grain to store. Only too often, after a bad harvest, the cultivators sell their dura to a local merchant at the beginning of the dry season, and later, when themselves short of grain, buy back their grain from the merchant, who charges a considerable percentage for his storage services. The causes and the results of this improvident practice are further discussed on p. 358 in connection with trade in agricultural commodities. The excuse usually given by the Nilotic for over-selling is difficulty of storage. However, if properly dried, sorghum grain keeps well throughout the dry season, even in uncovered heaps, though heavy losses may sometimes result from. attacks by storage pests; Sitophilus oryzae L., Apate herebraus Pall., Anomala obscuripes Faim., Xanthochelus vulneratus Bob., Pyenodactylus albogilvus Gyll., and probably a number of others not identified, as well as rats and termites. A more valid reason for over-selling is the difficulty of carrying large stocks of grain . during the annual migration, often for long distances, to the toich grazing grounds. Fear of pilfering is also an obvious reason for the absence of storage in villages which are completely abandoned during the dry season, though pilfering is, in fact, very unusual. In the northernmost part of the area grain is often stored in holes dug in a suitably well- drained site; the northern Dinka, the Shilluk, and Nuer store grain in baskets inside their huts or cattle-byres, or in earthen containers which are kept either inside the hut or on specially made stands outside. Among some of south-eastern and western Dinka, as well as among the Bari and Mandari-, the great part of the harvest is stored in large baskets made of palm leaves, reeds or grass, often holding five or more sacks of grain. These baskets are often kept outside on special platforms raised well above ground level and provided with a thatched roof. Methods of storage vary from tribe to tribe, and even from section to section. Tradition undoubtedly plays a great part in determining the melhods used. There is, however, a general trend towards more and more aeration as one goes from north to south. 338 5. CROPS CEREALS SORGHUM (Sorghum spp.; Ar. dura) Sorghum is the backbone of agriculture in the Jonglei Area, as in the rest of tile Sudan. 'The entire crop husbandry is concentrated on its production to such a degree that the farming system suffers considerably from mono-cropping. The Ironstone Region is an exception, with more diversified cropping, though even there sorghum is by far the most important crop. ADAPTATION The fact that in the major part of the J onglei Area sorghum occupies more than nine-tenths of the cultivated land is not surprising, as there is no crop better adapted to the conditions of both Semi-Arid and Flood Regions. ' The drought-resistant qualities of sorghum are many(20); they may be summarized as follows; (i) Ability to draw efficiently on soil moisture reserves, due to a deep and very well developed root- system, tapping a considerable volume of soil ; and to high osmotic pressure which may lead to ability to utilize soil moisture at higher pF values than other plants. (ii) Xerophytic features such as small cells, waxy cuticles, and a comparatively small leaf surface, which lead to a well-regulated moisture absorption/ transpiration ratio and eliminate dangers of physio- logical drought. (iii) Well-developed ability to survive seasons of drought, due to: slow rate of growth in early stages, and subsequently when 'drought conditions prevail (growth of sorghum may stop completely); lack of profuse tillering, and resulting comparatively small transpiring area; and ability to produce rapid growth from dormant basal buds if the main stalk has failed. Though sorghum is particularly well equipped for produotion in dry areas, a period of drought at the maturation stage can reduce yields considerably(21), and unfortunately this happens only too often in the Jonglei Area. This crop can also withstand a certain amount of flooding, so that it is particularly suitable for the Flood Region. This may be explained by the plant's deep root-system, even in the clay soils in which it is generally grown. As already described, the early rains cause the cracks to close, and though later in the season the top layer of the soil is waterlogged and there may even be standing water on the surface, the lower layers are probably still not completely anaerobic, some air being trapped in the soil pores. Sorghum roots penetrate to this layer, and are able to remain active. They can feed the crop, provided its upper parts are not killed by too prolonged flooding, until this supply of oxygen is exhausted. For this reason it is often noticed that sorghum is less affected by flooding on clays than on sandy soils. The ability to withstand flooding on the clay soils is also possibly due to less free carbon-dioxide, because these soils usually contain a considerable quantity of calcium and often have a alkaline nature, and consequently can ' fix ' CO, in the form of carbonates. Sorghum will also tolerate high temperatures, soil alkaline reaction, and salinity, which exclude many orops from this area; it is, in fact, able to grow on almost all types of soil in the area, whereas other crops are much more selective. CULTIVATION Methods of field preparation, sowing, weeding, etc. for this crop have already been described in general terms. As far is sorghum is concerned the following details of cultivation should be noted. Spacing for short-strawed, quick-maturing varieties is usually between 6 in. and 2 ft. , while for tall varieties it may be as much as 4 ft. or more. A number of seeds, varying between 2 and 10 or more, is dropped into each hole; later, the seedlings are often thinned to leave 2-4 plants in each hole, the surplus seedlings being transplanted to even out the stand. Only too often weeding is inadequate. Harvesting continues for weeks on end, as the individual ears mature very unevenly. The stalks of an early crop are often cut down to provide a ratoon crop. Those stalks which are cut are usually left on the ground, a practice which unfortunately encourages stem-borer. TYPES AND V ARIBTIES The number of varieties of sorghum grown in the Ionglei Area is considerable. Even the individual cultivator grows many varieties in one area. From the botanical point of view, these varieties almost certainly belong to more than one species of the genus Sorghum. A few varieties of the sweet sorghum type are known, and one or more is cultivated by nearly all peoples of this area. The collection, d~cription, and botanical identification of these varieties :would take years of specialized work, though in the long run it would prove well worth while. 339 22 The difficulties are considerable; as well as varieties known in different districts for a long time, the recent increase in the grain trade and government hunger relief measures have brought grain from outside, mainly from the Northern Sudan, so that new varieties are constantly introduced. More settled conditions and the increase in inter-tribal intercourse have made it possible for some better local varieties, formerly known only in limited areas, to spread widely. A good example is agono (Shilluk) or jak (Nuer and Dinka), a type of sorghum which may have originated in Shilluk country and which is now universally planted in the open grass plains from Paloich southwards to Nyareweng country. The local method of planting mixed varieties, even when they are recognized as such, leads to constant crossing and the formation of hybrid strains. In fact none of the so-called local varieties tried at the Malakal Experimental Farm bred true, even if seed from a single head was used. The local cultivators certainly recognize several types and varieties of sorghum, though only too often their ideas on this subject are not clear. Local names cannot be taken as a basis for classification ; the same variety or type may be called by different names in different places, or diffenmt varieties may be known by the same name. The Nilotics are not precise in their use of crop names. Crop varieties are given either attributory names, based on the colour of the grain or some other characteristic feature, or are known by the name of their place of origin in a particular area, or by the name of the person who probably either imported them or bred them by accident. Consequently the investigator has to exercise caution in dealing with sorghum varieties, or those of other crops. Some attempt has been made to collect data concerning sorghum types and varieties. Apart from the general classification given below, further details of this aspect of our work will be found in later sections of this chapter where we describe agricultural practice in various localities in the Jonglei Area. The many kinds of sorghum grown in the Jonglei Area can be grouped in two classes, quick-maturing, usually short and often coloured, and slow-maturing, tall varieties, though there are intermediate varieties between these two. Each class contains several types of dura suited to local conditions, and each type is, as a rule, composed of several varieties and hybrids. Thus, for example, the agono or jak type belongs to the slow-maturing, tall class of sorghum and is grown widely on open intermediate grasslands but is considered unsuitable for forest cultivations. In this type are several more or less recognized and definitely distinct varieties, differing in colour of seed, glumes, character of infioresence, etc. Varieties belonging to one type are, as a rule, sown together in a mixed stand, even if, as mentioned on p. 337, they are harvested and stored separately. Types, however, are very seldom mixed, though at least two types are usually planted. A most interesting point, which should be noted in connection with the comparison between quick-maturing varieties in the north and quick and slow-maturing varieties in the south, is the difference in yields. In the south the heaviest yielders are the slow-maturing, tall and leafy varieties; the quick-maturing varieties are definitely poorer grain producers. In the north, however, the yields of short, sturdy, quick-maturing varieties compare favourably with those of slow-maturing varieties in the south, and are generally higher than those of quick-maturing southern varieties. The advantages of the longer season and the better moisture conditions in the heavier rainfall area seem to be used by the crop for increasing its vegetative parts rather than its productive parts. That is to say, with two varieties of similar grain-producing capacity, the grain/ straw proportion will be much higher in the northern variety than in the southern. As straw is not extensivdy used as fodder in the south, the plants' productive capacity is consequently wasted. Of the great number of varieties and types grown in the Jonglei Area, some are suitable for varying conditions and have therefore spread farther than others. The general trend observed as one 'goes from the Semi-Arid to the much more humid Ironstone Region is most interesting, and illustrates the theory of adaptation of plants and farming systems to environ- ment. To quote a few examples; while the ' close head' varieties predominate in the north, the majority of southern varieties have the open type of infioresence, less susceptible to fungoid attack and therefore suited to more humid conditions. Closely spaced, broadcast varieties are generally grown in the Ironstone Region where competition for moisture is least. The number of varieties within one class and type is greatest in the Flood Region, where climatic and hydrological variations from season to season are greatest. DISEASES The most serious disease of sorghum in the Jonglei Area is undoubtedly smut (Ar. 'sueid; Shilluk ocolo; ' Nuer buk; Dinka tuk) caused by 'Sphacelotheca spp. S. sorghi (Link.) Clint. (' covered smut') is most common, but isolated cases of S. cruenta (Kuhn) Potter (' loose 340 smut') and S. reiliana (Kuhn) Clint. ('head smut') have also been observed at the Malakal Experimental Farm. S. sorghi can fortunately be controlled by seed treatment with copper carbonate, and sporadic attempts have been made in this area from time to time to organize a smut control campaign, but never on a large enough scale or for a long enough time to have had much effect. The inhabitants of the Jonglei Area consider that smut is especially-prevalent in dry years and on late sorghum crops. Leaf spot is very common, particularly at the end of the growing season, when the majority of fungi causing it may be regarded as saprophytic rather than parasitic. The following parasitic species have been identified at Malakal Experimental Farm: Cercospora sorghi Ell. and Ev., Colletotrichum graminicola (Ces.) Wils., Helminthosporium sp., Phoma sp., Gleocercospora sorghi Bain and Edgerton. A curious disease in which sooty blotches appear on the leaves has also been noted, the sooty appearance being due to innumerable small, apparently superficial, black fungi fructifications. The symptoms resemble those of ' sooty stripe' of sorghum in America, caused by Ramulispora sorghi (Ell. and Ev.) Olive and Lefeb., but the fungus itself in part resembles Gleocercospora sorghi Bain and Edgerton, which causes zonate leaf spot of sorghum in America. Judging by the symptoms, this disease might become severe(28). The majority of pests of sorghum are not confined to this crop alone, and are discussed in a separate section (see p. 352-3). MAIZE (Zea mays Linn.; Ar. esh er rif or dura shami.) Maize is planted in small quantities by almost every cultivator in the Jonglei Area. It is valued mainly for its quick maturation, being an important contribution to diet in the hungry, pre-harvest gap. Slower-maturing types are also planted, adding variety to the food. In the Nilotic area, with the exception of parts of Eastern Nuer District, planting of dry season maize is uncommon, but (as already mentioned) this crop is important in toich cultivations in Kosti District and also in the Bari and Mandari area. On the whole, maize makes only a limited contribution to the total year's diet of the inhabitants. Maize is definitely not as well adapted to the environmental conditions of the Jonglei Area as sorghum. It requires more moisture and is less drought-resistant(24), which limits its cultivation in the Semi-Arid region. It is not flood-resistant, and consequently can only be grown on high land sites in the Flood Region, where a number of other crops of greater value can be planted. It requires fertile soils abundant in readily available nutrients, including a fair supply of phosphorus, and a high nitrogen content(25), so that on the majority of soils of the Ironstone Region yields are generally poor. Its poor storage qualities also limit its popularity. Resistance to bird devastation is its only advantage over sorghum. Maize is usually planted in the immediate vicinity of homesteads, and interplanted with beans, pumpkins, gourds, and in some cases sesame. Since maize out-crosses very easily, distinct varieties are not found in undeveloped areas where the farmer is not interested in purity of stock or breeding. In the Jonglei Area there is a considerable number of hybrids, generally divided into two types, short, quick-maturing and taller, slower-maturing. In some districts more than two types can be distinguished when maturing time is taken as a basis for classifica- tion. Fungoid leaf diseases are common in maize, and leaf stripe caused by Helminthosporium sp. has been identified at the Malakal Experimental Farm. BULRUSH MILLET (Pennisetum typhoideum (Burm.) Stapf. and Hubbard; Ar. dukhn.) This crop is of little importar:ce in the Jonglei Area, though in Kordofan, just to the west, it is grown extensively on sandy soils. In the Jonglei Area some is grown in the Semi- Arid Region, and a little is occasionally planted by Dinka in Yirol District and by Bari and Mandari in the Ironstone Region. The grain of bulrush millet is generally considered, especially by Nilotics, inferior to that of sorghum, and in suitable conditions the latter is grown in preference to the former, though in experiments carried out at the Malakal Experimental Farm bulrush millet has given yields which compare favourably with those of sorghum. Bulrush millet, however, is very susceptible to flooding, and therefore unsuitable for the majority of soils in the Flood Region. With the exception of awned varieties, it is also much liked by birds. Fungoid leaf rusts and spots are common, and Puccinia penniseti Zimm. and Cercospora fusimaculana Ask. have been identified at the Malakal Experimental Farm. FINGER MILLET (Eleusine coracana Gaertn.; Ar. telabun) This crop is most important in certain districts in Equatoria, where conditions are too wet for sorghum during its maturation period. It is grown in the Jonglei Area on a small scale by Bari, Mandari and some Yirol Dinka. It is usually broadcast mixed with sorghum, 341 maize, or sesame, being sown by some cultivators twice in the season if the rains permit. It requires a lot of moisture if it is to do well, but is susceptible to waterlogging. It is tolerant of thin, poor soils, stores well and is considered highly palatable. There is some evidence that it is spreading slowly northwards; it has been grown successfully on an experimental scale as far north as Malakal. LEGUMINOUS CROPS GROUNDNUTS (Arachis hypogaea Linn. ; Ar. ful sudani) Groundnuts are, or can be, grown throughout the 10nglei Area, but while in the Semi-Arid and Ironstone Regions this crop occupies an important place in the agriculturai economy, in the Flood Region only occasional patches are grown and nowhere in this Region can ground- nuts be regarded as important. The reasons for the lack of popularity of groundnuts in the Flood Region are clear. In the first place this crop cannot tolerate waterlogging, and consequently sites suitable for its cultivation are limited. Secondly, though heavy soils can produce high yields (yields well over a ton per feddan were harvested at the Malakal Experimental Farm), cultivation on clay soils is difficult and laborious. Groundnuts require a deeply stirred seed-bed which, with only primitive tools available, demands disproportionately heavy labour. While on light soils they can be harvested simply by pulling them out together with the roots, on heavy soils they have to be dug up. MoreGlVer on lighter soils in the Flood Region, e.g. Bor District, the constant danger of damage by elephants (see p. 353) is a major limiting factor. Finally, among some Nilotic peoples of this area there is a prejudice against groundnuts (as against all the recently introduced crops), though this prejudice is already disappearing. On the other hand, in the Ironstone Region groundnuts are a staple crop and an important item of diet; they are found in some quantity in the fields of almost every cultivator. There is a certain amount of trade in this produce, and since the organization of the Yirol Co- operative Society and the establishment of the Yirol Oil Mill this trade has been increasing rapidly, though in the Region as a whole groundnuts are still grown mainly for subsistence purposes and not as a cash crop. In the Semi-Arid Region a considerable acreage of groundnuts is grown on sandy outcrops west of the White Nile and between Geigar and Gelhak on the east bank of the river. Here groundnuts are regarded largely as a cash crop, and are exported through merchants from Renk, lebelein and Kosti; only a proportion of the crop is consumed locally. In the Semi-Arid Region groundnuts are usually grown in a pure stand, though they may occasionally be found interplanted with bulrush millet or sorghum. In the Ironstone Region they are often used as a ' pioneer crop', in which case they are planted alone. In subsequent years groundnuts, self-sown or interplanted, are usually grown with sorghum, bulrush millet, maize, etc. Spacing in pure stands varies from 6 in. to 14 in. according to climate, soil, variety and the inclination of the cultivator; spacing is generally closer in the Ironstone than in the Semi-Arid Region, and for upright varieties than for surface creeping ones. From 2 to 4 seeds are planted in each hole and covered with earth or, if planted on uncleared land, the field is weeded after planting and the weeds spread over to form a mulch considered to benefit the crop in its early stages. The seeds are not always shelled before sowing, so that uneven germination results and damage is done by pests before sowing. On the light soils the crop is harvested by pulling the plants out of the soil, roots and pods togetlt,er, though, as mentioned above, in heavy soils digging is necessary. This is often done too late, and secondary germination results. Immature pods are shelled and eaten immediately, while mature pods are either sold or stored for future consumption. There are considerable numbers of strains and varieties adapted to conditions in different localities; they belong to two main types, upright and surface creeping, with some showing intermediate growth habits. The general trend is towards creeping varieties in the north and upright varieties in those areas where the rainfall is heavier. Leaf spotting is common in ground nuts at the stage approaching maturity, at which time some of the fungi responsible may be regarded as saprophytic rather than parasitic. Cercospora personata (Berg and Curt.) Ell. and .Ev., has been identified as one of the fungi, and may cause some damage to yields. COWPEAS AND BEANS (Vigna spp. and Phaseolus spp.; Ar. lubia, fasulia, etc.) Several strains of Vigna unguiculata and a number of species and strains of Phaseo/us are grown throughout the 10nglei Area. These crops have a well-established place in local agriculture and a small patch of cowpeas and beans'inay be found in the vicinity of almost every homestead. 342 Cowpeas and beans are usually grown on homestead cultivations, interplant~d with maize, early sorghum and cucurbits. Climbing varieties are planted either under trees or at a late stage in the maize and dura fields, where the stalks of the cereal crop are left after the heads have been harvested and thus serve as supports for climbing beans. Occasionally the main crop of sorghum, planted on ' out-cultivations', is interplanted with cowpeas or bealis, though these legumes, being susceptible to bad drainage conditions, suffer in years of heavy rainfall on intermediate land. Cowpeas and beans constitute a considerable part of the diet of the inhabitants of the Jonglei Area, especially in the autumn. Unfortunately storage pests often make it impossible to conserve th~m successfully. The green leaves are generally used as a vegetable, and it is said that leaf-pruning stimulates pod formation . The Nilotics r~ognize several varieties of beans and cowpeas but, disregarding botanical features, group them according to their habit of growth into three classes, creeping, upright and climbing. Bacterial blight, probably due to Xanthomonas phaseoli (E. F . Sm.) Daws., rust caused by Uromyces vignae Barel. , and leaf spotting due to Cercospora spp. (C. canescens Ell. and Mart.) and Phyllosticta phaseolorum Sacco (?), have been identified on cowpeas and beans at the Malakal Experimental Farm and Kodok Experimental Farm. Leaf spotting on these legumes is common throughout the area, and may be responsible for considerable loss of yields. PIGEON PEAS (Cajanus cajan Millsp.; Ar. ads sudani) This perennial legume is occasionally grown in the Ironstone Region by Bari, Mandari and Dinka cultivators. It is rather susceptible to waterlogging, which limits its success in the Flood Region. In the Semi-Arid Region the rainy season is not long enough for it to reach maturity and it can be grown only under irrigation. In the north it is fairly CO)IlIl1only grown along the irrigation channels, where it serves as a wind-break as well as producing nutritious seed. With the recent expansion of pump schemes in the Kosti Area it may be expected to spread into that region. BAMBARA GROUNDNUTS (Voandzeia subterranea Thou. ; Ar. Jul Abu Ngawi) This legume is grown on a small scale by Dinka in Yirol District. It is consumed locally and is of minor importance. ' The crop is hardy, and can be stored for up to two years. Leaf spotting caused by fungi is common. Cercospora canescens Ell. and Mart. has been noted as a cause of leaf spotting at the Malakal Experimental Farm. OIL CROPS SESAME (Sesamum orientale; Ar. simsim) Sesame is grown throughout the Jonglei Area, with the exception of some Nuer sections where, in spite of attempts to introduce it, the crop has not been successful with the local population. The main areas where it is cultivated almost universally are the Ironstone Region and the Semi-Arid Region with the exception of some of its northernmost parts. In these two areas sesame is to some extent regarded as a cash crop, while in the Flood Region, if grown, it is consumed entirely by the cultivator and his family and is of very minor importance. The geographical distribution ~f this crop is based largely on its requirements and con- sequently on its comparative yields, which in the Flood Region are usually extremely low. Because of its deep root-system sesame, once established, can grow successfully even in places of light rainfall. It is, however, very susceptible to waterlogging, and requires a well-drained type of soil for the free development of its root-system. The environmental conditions of the Flood Region are therefore not suitable. A considerable number of varieties are grown in the Jonglei Area, and are usually classified by the local population, according to the colour of the seed, into white, red and black. The varieties also differ in maturation period, but this characteristic is not linked to seed colour. As might be expected, varieties grown in the Semi-Arid Region are usually quick-maturing, short, and with few lateral branches, while those grown in the south are slower-maturing, taller, and more branching. In Bari and Mandari country, where the rainy season is long enough, two crops of quick-maturing sesame are often grown, as well as one crop of a slow-maturing variety. Within the different varieties there are also variations of secondary characteristics such as shape of inflorescence, shape of seeds, etc. A high degree of uniformity in the maturation period has, however, been achieved, a fact which is of great importance in a crop which is harvested en masse and not by individual picking. The red type of sesame grown on Kordofan sands matures less uniformly and its branches are harvested separately. 343 Though in the Semi-Add Region pure stands of sesame are often sown, in the south it is \>Isually interplanted with sorghum, maize, beans, etc. In the north it is sometimes sown in rotation with sorghum or bulrush millet, being regarded as a ' resting crop' which feeds at greater depth than cereals and therefore does not exhaust the upper layer of soil to the same extent as fallow weeds. The crop is harvested before it starts shedding and is dried on special frames; it is easy to thresh but difficult to winnow. The seeds contain about 50% oil. In the Semi-Arid Region, where it is regarded partly as a cash crop, a consideraple number of primitive presses are in use to extract the oil. In the south the seed is generally (laten without extraction of th(l oil, though an attempt is sometimes made to extract it by crushing the s(led and boiling it. The construction of a government oil mill in Yirol and the organization of trade through the Yirol Co-operative Society have given considerable assistance to the dev(llopment of trade in this produce. Leaf disease caused by Cercospora sesami Zimm. is widespread though not serious. Bacterial' blood disease' (Ar. marad ed dam), ascribed by the Government Plant Pathologist to the attack of Pseudomonas sesamicola Malkoff., is common and may, in bad years, be responsible for a very considerable reduction of yields. MELON SEED (Citrullus vulgaris Schrad.; Ar. battikh) A certain area of melons is grown along the river in Kosti District during the dry season, and on the qozes during the rains. A conside.able proportion of th&l crop goes to Kosti as fruit, but there is also a little trade in the oil-bearing s&led, and some is consumed locally. CASTOR (Ricinus communis Linn.; Ar. khirwa) An occasional plant of this species can be found growing by the huts in the Nilotic areas, and the oil is used for anointing. HYPTIS (Hyptis spicigera Lam.) This oil seed, important in Equatoria, is occasionally grown on a small scale by Bari and a few riverain Mandari, usually mixed with sorghum. SHEA NUT (Butyrospermum parkii var. niloticum Kotschy.) This is a large tree yielding useful timber, the kernels of its nuts containing 50% stearine, and is indigenous in the Ironstone Region. The oil is used locally for food and .anointing. The government oil mill at Yirol is capable of extracting shea butter, and as a result of its construction trade in this indigenous product should develop. COTTON (Gossypium spp.) HISTORY AND ECONOMIC ASPECTS The history and &lconomic aspects of cotton in the Jonglei Area are worth study, as they provide many useful lessons on policy and organization in connection with the introduction of a cash crop to a primitive peasant area. CENTRES OF PRODUCTION Odd plants of cotton, usually of one of the indigenous perennial species, can be found throughout the Flood and Ironstone Regions of the Jonglei Area. "In the early twenties, a considerable amount of wild cotton flourished along the Pibor and Anuak villages. The crop was tended and a demand for the lint existed in Abyssinia. The presence of these heavy bearing bushes influenced considerably the agricultural pioneers in the south in their search for a cash crop(26)". Even now in the Nuer dis1lricts one or two cotton bushes are almost invariably grown in the vicinity of the homesteads, the lint being woven into strings and made into short skirts which are worn by the married women. As far as proper commercial production is concerned, however, cotton--growing in the Jonglei Area was, before the last war, centred largely on the northern parts of Shilluk country, which throughout that period produced by far the greatest portion of the crop. In those days a certain amount of cotton was grown in s.outhern Abialang, Paloich, Maban, and Dunjol country, and in post-war days this became the most important rain-grown cotton area. In the mid-I920s Yirol District was the southern centre of cotton-growing, but in the following decade this centre moved to Western Nuer District. Sporadic attempts were made to introduce the crop to Bor and Central N uer Districts, but the amount produced was negligible. From 1935 to 1939, cotton was also grown by the Bari and Mandari. The geographical distribution and relative importance of these centres of 344 production are worth nothing. North~rn Shilluk country, the most important, was the most accessible and nearest to the ginnery; the moving of the southern production centre from Yirol to Western Nuer District meant also bringing it nearer to ginning facilities. While no irrigated long staple cotton was grown in the Jonglei Area before the end of the war and only a small quantity of rain-grown cotton was cultivated in the Semi-Arid Region, at present Kosti and Renk Districts, with nearly 8,000 fed dans of irrigated cotton, constitute the most important centre of production in the area. Rain-grown cotton is at present mainly produced in the Paloich and Dunjol area, and Shilluk production has become of little importance. The present location of cotton-growing centres should be noted; cotton is now grown under normal economic stimuli and without government pressure, and therefore tends to be concentrated in the most accessible areas, nearest to markets and ginneries, and also, as far as peasant-grown short staple cotton is concerned, in areas where the development of hariq cultivation permits the extension of both cotton and grain production so that, apart from cotton, a considerabl~ quantity of grain is available for export every year. For the period from 1923 to the end of the war, when cotton was grown under government auspices, numerical data concerning its production are available in the various annual reports of the agricultural and administrative authorities. TABLE 162 PRODUCTION OF RAIN-GROWN corrON IN THE JONGLRI AREA, 1923-45 (Unginned cotton in small kantars : 1 kantar = approx. 100 lb.) Northern Area(') Southern Area(') Bari and Season Shilluk, Yirol, Mandari Total N. Dinka and Area(') Kaka W. Nuer, etc. I 1923-24 1,165 - - 1,165 1924-25 4,122 234 - 4,356 1925-26 9,955 4,818 - 14,773 1926-27 10,968 5,245 - 16,213 1927-28 3,939 3,151 - 7,090 1928-29 7,291 8,293 - 15,584 1929-30 10,383 8,702 - 19,085 1930-31 5,714 4,675 - 10,389 1931-32 8,218 - - 8,218 1932-33 8,698 - - 8,698 1933-34 10,753 - - 10,753 1934-35 16,395 429 - 16,824 1935-36 15,090 1,074 3,524 19,688 1936-37 13,927 1,773 7,120 22,820 1937-38 8,096 1,010 4,539 13,645 1938-39 10,857 754 2,508 14,119 1939-40 10,896 1,533 - 12,429 1940-41 935 563 - 1,498 1941-42 5,687 1,135 - 6,822 1942-43 719 407 - 1,126 1943-44 539 775 - 1,314 1944-45 400 - - 400 (1) Based on Upper Nile Provmce (Ministry of AgncuJture) reports: Annual Reports, 1923-45 . ( I) Based on data in the Equatoria Handbook. 1936-48. Data from the post-war period when the growing and marketing of cotton is done largely without direct participation of govttrnment authorities, are scarcer and less exact. We include the latest estimate: TABLE 163 ESTIMATED AREAS AND TOTAL YIELDS OF corrON IN THE JONGLEI AREA 1952-53 SEASON Rain-grown short Irrigated long staple cotton staple cotton Area 7,862 feddans 1,500 feddans "' 1 Total production ... 6,000 kantars (100) ... 33,000 Kantars (315) " ' 1 Estimates by Inspectors of Agriculture, Melut and Kosti . As may be seen from the above tables, pre-war production varied considerably from year to year. This variation was due partly to climatic and other natural factors affecting the yields, but mainly to the amount of I1ressure applied by government authorities in any given season to induce the local peasant to grow this crop, much against his inclination. Then, 345 during the war, when it was decided to remove this pressure, cotton practically disappeared from the area after twenty years of effort to make it the main cash crop of the local peasant. " Cautiously conceived in 1922 and born in 1923, cotton growing as an economic factor in the province, after reaching its maximum development in 1935, then fell into a decline, and its final passing hlf.3 not invohd a single tear(·7)" in Upper Nile Province. In Bari and Mandari areas " no persuasion was used ..... and from that date (about the same period) none has been grown(·8)" . High prices after the war, reaching their peak in 1951-52, caused the revival of cotton- growing in the area, and the introduction of long staple irrigated cotton. No persuasion by the authorities was necessary, and in some areas this crop had to be discouraged, as its production might have affected grain-growing which at present is considered of primary importance. ORGANIZATION OF COTION GROWING In pre-war days the lion's share of the efforts ofInspectors of Agriculture and a considerable part of the work of the administrative officers lay in persuading and cajoling the local inhabitants to grow cotton. After cotton seed was issued, pressure had to be applied to make the local cultivator plant it, do a little weeding, and pick the cotton before the lint fell off. At the end of the season, a marketing organization had to be set up, and disposal to ginneries organized. The following example gives an idea of the amount of work which cotton-growing meant for agricultural and administrative officials for what nowadays must be regarded as very meagre returns. TABLE 164 NUMERICAL DATA CONCERNING COTTON-GROWING CAMPAIGN IN V.N.P. 1936-37 SEASON SEED DISTRmUTED District or area Number of Cotton bought Cash paid to Quantity cotton markets kantars (100) cultivators By (sacks) £E. m/m I Renk ... .. . .. . D.C. 170 5 556.86 195.260 Kaka ... ... .. . I. of A. 66 I 657.64 ·227.375 N. Shilluk ... ...... I. of A. 3,051 23 12,176.27 I 4,236.585 S. Shilluk .. . I. of A. 300 13 587.00 203 .130 W. Nuer ... .. . ... D .C. 680 10 1,172.53 443.130 Zeraf Island ... ... I. of A. 2 - - I - Total 4,269 52 15,150.30 5,305.480 · .. 1 The methods used in introducing and maintaining cotton production were essentially agricultural extension methods applied to local peasant farming. The local capitalists- merchants- were, in those pre-war days, not interested in cotton, though they were occasionally used by the Department of Agriculture as specially paid demonstrators. Since the end of the war, however, the organization of cotton-growing has changed considerably. Private pump schemes-essentially capitalist enterprises- have sprung up in the northern part of the area. The local merchants have started growing cotton in very large quantities, either by direct labour or on a share-cropping basis. ECONOMICS OF COTTON PRODUCTION The economics of cotton production merit further study, since there are reasonable doubts as to whether cotton-growing in this area in pre-war days was based on sound principles. The unpopularity of cotton among the local cultivators in that period and the spontaneous revival of cotton-growing since the war were no doubt caused by the entirely different economic con- ditions in these two periods. Unfortunately numerical data are lacking, and the economic a~pects can only be discussed in general terms. It may be said that cotton cultivation was introduced and' pushed' in the pre-war period primarily for the benefit of the local inhabitants and secondly as a source of government revenue. The local inhabitants did not appreciate the benefits of cotton cultivation. It competed with grain production for the limited amount of labour, especially at sowing time, and for the best agricultural land, while the area was chronically short of grain. In years of hunger cotton 346 could not be eaten, while the prices obtained for this crop, which were based on world prices, were not high enough to pay for the purchase of grain in a bad grain year, the price of grain being determined by local supply. In fact, though cotton was intended as a cash crop for the local cultivator, cash returns per feddan were often higher for sorghum than for cotton (see Table 165). There was thef(~fore no price incentive for cotton-growing. Apart from' this, in an area where historical causes, enormous distances and virtual lack of communications have resulted in an economy of self-sufficiency at the family unit level, and where trade is still undeveloped, the need for cash is very slight and consequently a cash crop, though desirable for general reasons of economic and social development, was not welcomed by the cultivator. The small amount of money he ne€ded could be more economically obtained by the sale of grain or an occasional animal. Furthermore, cotton cultivation involved more work than grain-growing. It is therefore not surprising that it was difficult to persuade the local cultivator to give his time and best land to cotton rather than to sorghum, and that when he was persuaded to do so he gave it only the minimum attention necessary to satisfy the authorities. As a result standards of cultivation and hence yields were very low. TABLE 165 COMPARISON OF PRICES AND RETURNS PER FED DAN OF COTTON(') AND SORGHUM 1935-39 AND 195<1--52 1935-39 1950-52 Cotton I Sorghum Cotton Sorghum I Estimated yield in kantars (100) ... 2 I - 3 -in ardebs ... ... - It - It Price in £E. per kanlar (100) ... 0·350 - 2·800 - per ardeb ... ... - 0·700 - 1·800 Estimated return per feddan in £E. 0'700 1'050 8'400 20'100 (1) Rain-grown short staple cotton. N .B.: Estimated average yield of cotton increased by 50% in post-war period because of better care given to this crop as a consequence of change of attitude towards it. In the post-war period changes have taken place which have been responsible for the revival of cotton-growing, though so far only in the northern parts of the Jonglei Area, and there is no doubt now that a price incentive exists. Development of communications, trade and other contacts with the outside world have made the inhabitants of these parts much more • cash conscious'. Finally, it should be noted that cotton has become popular in these northern parts, which are also the most accessible areas, and where an exportable surplus of grain is now produced. Moreover cotton-growing is now organized on different lines, with the active participation of local capital, and therefore it competes less directly with grain Prodl1ction. The industry as a whole rests on a much sounder economic basis and thus the attitude towards this crop has changed. In the southern districts of the Sudan, cultivation of cotton before the war was regarded as a source of government revenue, both direct and indirect. Policy was founded on the premise that the inhabitants of the area had to pay at least in part for government services ; exportable produce was therefore necessary, and cotton was considered the best form. The choice, however, is debatable, though no definite ' conclusion can be reached in the absence of numerical data ; neither can any assessment be made of the contribution of cotton to government finances . Under-developed inland communications, together with the great distance from markets and ginneries, militated against cotton being a very profitable crop. It is doubtful if the difference between the price paid to the cultivator and the export price was sufficient to offset the costs of transport and ginning, as well as those of the annual cotton-growing campaigns and market organization, and still leave an appreciable margin of profit. Cotton cultivation and handling in this area could not have provided the government with a substantial source of direct revenue. Indirectly, however, this crop provided revenue through taxation, though the proportion of this revenue attributable to cotton is again difficult to assess. It must be remembered that expenditure on hunger relief and other necessary measures was at least partly due to the effects of cotton-growing on grain production, and must therefore be subtracted from the revenue from cotton. The balance, if any, cannot have been large. By contrast, in post-war days cotton-growing is often discouraged by the authorities, as it is alleged that grain production is more desirable from the point of view of the economy of the area as a whole. Unfortunately in this area, which is largely cut off from outside economic influences, 347 where exchange economy is virtually non-existent, and where different parts are divided by lack of communications, comparative prices do not provide a measuring-rod, so that present policy, like that of pre-war days, is largely based on conjecture, and investigation of the economics of agricult•u. ral production here is badly needed • COTTON CULTIVATION Standards of cotton cultivation before the war were very low, as the cultivator took no' interest in this crop. The fields were badly prepared, and cotton was sown at variable spacings and usually too late in the season, sorghum and other food crops being given priority; only too often it was left un thinned and weeding, if anx, was generally insufficient. The crop was often left ungathered too long, and quality was indifferent. TABLE 166 AVERAGE PERCENTAGES OF DIFFERENT GRADES OF COTTON, 1935-40 Grade II III Ungraded I Percentage ... ... , 56·2 31·1 H % Yields also were low, and the estimated average yield per fed dan during the 1935--40 period was just over 2 kantars (100). Since the war, as cotton is now grown by people who are spurred on by the desire for cash in contrast to pre-war days in other areas farther south, standards of cultivation and hence yields, if not quality, have increased considerably. In 1950-51 and 1952-53 yields were estimated at 4 kantars (100) per feddan. Standards of cotton on pump schemes in the Semi-Arid Region, though lower than in the Gezira, are generally higher than those of rain-grown cotton. Until 1937-38 the' Webber' variety of American upland cotton was grown in Upper Nile Province. In the 1938-39 season it was replaced by' Pump Scheme' variety which, in the post-war period, was superseded by , SP 84' brought from Uganda during the war; '511 D' was cultivated in the Bari and Mandari areas. 'X 1739 A ' is the standard variety at present grown on pump schemes in the Semi-Arid Region. DISEASES Blackarm disease, caused by Xanthomonas malvacearum (E. F. Sm.) Dawson, is usually found on cotton in this area. On the Malakal Experimental Farm is has been observed on both Egyptian and American types of cotton, including BAR varieties. Leaf curl is sometimes found, but does not appear to be common. OTHER CROPS TOBACCO (Nicotiana rustica, N. tabacum) Tobacco is grown throughout the Jonglei Area, 'with the exception of the Semi-Arid Region. Both Nicotiana tabacum and N. ,rustica are cultivated, and several varieties of each are found in the area. Varieties of N. rustica predominate on dry season cultivations and are used, especially in the northern part of the area, for chewing as well as smoking. This species is regarded by the Nilotics as ' strong tobacco' and is predominant in local trade. Tobacco is grown both as a rain crop round the homesteads and, in some areas at least, as a dry season crop on the riverain flood-plains. The rainy season crop is usually sown round cattle-byres where, in addition to some shade, it receives extra water which drips from the byre roof. The ground round the byre is usually well manured by dung ash, dung, and urine. A separate seed-bed is often prepared, incorporating quantities of dung, ashes and even superior soil imported from elsewhere, such as Sobat riverain silt. Tobacco is usually sown in the seed-beds at the beginning of the rains, though some Shilluk simply scatter the seed as soon as it is harvested in November or December, and let it lie in the ground throughout the dry season until it germinates in the first rains. Nursery seed-beds are weeded almost continuously, and, if sown too closely, the seedlings are thinned. The seedlings are transplanted 4 to 6 weeks after germination into specially prepared plots, though in some cases, e.g. among the Bul Nuer, they are transplanted on to an area where early m~ze has been grown and where the stalks are left to give a certain amount of protection from wind and heavy rain. The transplanted 348 seedlings are generally shaded for the first few days, being uncovered at night. As a nile tobacco is the best weeded crop in the area. Flowering at too early a stage is usually prevented by the pinching out of buds, and harvesting starts about three months after transplanting, maturity of leaf being judged by colour, smell and taste. • In addition to tobacco grown during the rainy season, in some areas, notably those of the Twi and Nyareweng Dinka, some Nuer, the Shilluk, and the riverain Bari and Mandari, 'tobacco is grown during the dry season on riverain flood-plains with the help of hand-watering from the river. Only the dwarf type is grown in this way. Winter-grown tobacco is sown in October, transplanted in December to a fine seed-bed which requires much preparation, and is watered every three or four days, special care being taken not to over-water it. The first leaves are picked about a month after transplanting, and steady picking goes on until the beginning of the rains. This tobacco is mainly grown for sale or barter, and the areas under a winter crop are small, as few people are prepared to face the work which its cultivation entails. Yields of locally grown tobacco are impossible to ass€ss and it is generally of very poor quality. Methods of preparing tobacco for smoking, though varying from tribe to tribe and section to section, can be divided into two classes, according to the final product. Loose tobacco is prepared in two ways. The first method is to pick the leaves, wilt them in the sun for an hour or so and, before they become brittle, put them in a wooden mortar. After thorough pounding, the tobacco is placed in earthenware pots and fermented from two to five days. When fermentation is judged, by smell and taste, to have reached the required stage, the tobacco is taken out and spread in the sun to dry thoroughly. The second method is more generally used. After picking, the leaves are placed in a basket inside the hut and are allowed to dry for a few days. When still moist they are cut into small pieces and put in a bag made of calf- skin, then hung on a pole to ferment and dry slowly for a month or two, after which the product is considered ready for smoking. Some people preserve the tobacco by covering the skin bag with mud, blocking the pores of the skin, and preventing complete drying until the beginning of the next rains, by which time it has achieved considerable strength and, the people say, helps them to forget the empty stomachs usual at this time of the year and gives them the stimulus necessary for heavy work in cultivating their fields. Another method of preparing tobacco for smoking is to form it into large cones or blocks. Here again details vary, but the basis of this method is to wilt the leaves, pound them thoroughly in a wooden mortar and knead the almost homogeneous mass into ' heads', which are then slowly dried and fermented in the shade of a hut. After 4 to 6 weeks the tobacco is ready for smoking. This method is generally used when preparing the tobacco for sale or exchange. Except for the more sophisticated popUlation of towns and larger administrative centres, the local man is satisfied with his home-grown, home-cured tobacco, and foreign brands are imported only on a small scale. There is, however, some importation of locally grown tobacco from the Maban and Eastern Jikaing to the Paloich, Dunjol and Abialang Dinka, and to Kosti District. ROOT CROPS AND VEGETABLES SWEET POTATOES (Ipomoea batatas Lam. ; Ar. bambei) Sweet potatoes are grown occasionally in the Ironstone Region, especially along the river in Bari and Mandari country, but their contribution to diet may be considered of very minor importance. The wild species of Ipomoea produce edible tubers which are sometimes consumed by Nilotics in times of hunger. CASSAVA (ManihoJ utilissima Poh!. ; Ar. bafra) This crop is found cultivated in small patches in the Ironstone Region, but is usually considered as a stand-by rather than a major item of diet. Attempts to introduce it into the Flood Region have been made several times, but apart from a small patch in the garden at Bentiu, it has not survived more than one season. It is doubtful if it could be grown in the Flood Region without dry season irrigation, which is not warranted by its nutritive value. PUMPKINS AND MARROWS (Cucurbita spp. ; Ar. qara, etc.) Universally grown throughout the Jonglei Area, this crop is of considerable importance. It is usually planted in homestead cultivations, mixed with maize, beans and early sorghum. It is eaten fresh as a vegetable, and when in season constitutes an appreciable part of diet. 349 MELONS AND CUCUMBERS (Cucumis spp.) Melons and cantaloups are an important crop in Kosti District where they are grown during the dry season on gurf cultivations. A large quantity is exported annually to towns and to the Gezira. Various types of cucmribers are also grown in this area, but on a small scale. ONIONS (Allium cepa L.) This crop was until recently, when partly displaced by cotton, one of the most important cash crops of Kosti District. It was generally well cared for and very creditable yields were produced. At present its cultivation is carried out on a comparatively small scale on saqia and pump schemes, and on numerous small plots irrigated by shaduf An appreciable quantity of onions is exported annually to the southern provinces, especially Bahr el Ghazal, where they command comparatively high prices. OTIIER VEGETABLES A number of other vegetables are grown on a small scale in this area. Jew's mallow (Corchorus olitorius Linn.) is frequently cultivated in the Kosti area, and is indigenous farther south. Its green leaves are eaten cooked as a vegetable. Lady's fingers (Hibiscus esculentus Linn.), occurring as a weed in the Sudan rainlands, is also an important vegetable. It is often cultivated under irrigation and is also grown extensively on dry season toich cultivations in Bari and Mandari country. A number of other vegetables, both Sudanese and European, are also grown in government and a few private gardens in this area. They are, however, very seldom cultivated by Nilotics who, on the other hand, include a considerable variety and appreciable quantity of fruit, leaves and roots of indigenous flora (for details see Chap. 3, p. 245 et seq.) in their diet. 6. WEEDS,· PESTS AND PARASITES WEEDS Weeds are a serious problem in the agriculture of the Jonglei Area, mainly because of inadequate and superficial cultivation both before sowing and while the crop is growing. We have had no opportunity to investigate this problem fully. During the dry season weeds are either dead or unidentifiable, while inconsistency of local nomenclature makes enquiry by the questioning of local cultivators difficult and often misleading. During the rains only very limited areas can be covered, and identification of a large number of species is still difficult as only a proportion of the weeds are in flower or in a stage of development at which they are recognizable to a person not wholly familiar with local flora. It can be generally said that the severity of weed competition increases from the Semi-Arid Region to the Ironstone Region, as does the complexity of plant associations. It is difficult at this stage to specify which weeds are the remains of original vegetation not fully cleared before sowing, and which are species typical of cultivated land; that is plants better able to compete with a crop and with the cultivator's efforts than with natural vegetation. A number of grass species indigenous and common in the natural climax of any part of the Jonglei Area if found on cultivated land can be regarded as belonging to the former category; e.g. Setaria incrassata, Hyparrhenia pseudocymbaria, Sorghum dimidiatum, etc. are often present in cultivations round Malakal, because of insufficient initial and subsequent clearing. On the other hand, an interesting trend was observed in connection with graminaceous weeds; many grasses, especially annuals (e.g. Eragrostis spp., Aristida spp., Digitaria spp., Panicum sp., Eleusine spp., Brachiaria spp., Dactyloctenium spp., Setaria pallidifusca, etc.), while occurring as important members of the grassland climax in the northern part of the Jonglei Area, farther south are found as weeds of cultivated land, apparently being better able to survive on cultivated land than in competition with locally predominant species. Two species require special mention; Cyperus rotundus and Imperata cylindrica. The former is found throughout the Jonglei Area, but is of considerable importance only under irrigation in the northern part of the area and in districts of heavy rainfall (over 1,000 mm.) in the south. The latter is an im- portant weed of cultivated land in the Ironstone Region, though in the Flood Region it is still, fortunately, uncommon. . Dicotyledonous weeds also occur throughout the Jonglei Area and may definitely be regarded as the main weeds of cultivated land. They belong to a number of families, and a considerable number of species is found. In the Flood Region, particularly round Malakal, the following may be considered the most important: 'Evolvulus nummularius, Ipomoea cordofana, Euphorbia sp., Corchorus spp., Borreria sp., Cassia sp., Commelina sp. 350 The main weeds of the irrigation channels and their edges are Typha angustata, Cyperus rotundus, Ceratophyllum sp., Cynodon dactylon. In Tables 167 and 168 the results of identification of weeds collected at Malakal and Arielbek, in Bor District, are summarized. It ml1st be noted, however, that these lists ,are very far from complete. TABLE 167 WEEDS FOUND ON CULTIVATED LAND ARIELBEK COLLECTION. SPECIMENS IDENTIFIED BY THE ROYAL BOTANICAL GARDENS, KEW Family Species Turneraceae Wormskioldia biviniana Tul. Ficoidaceae Trianthema pentandra L. Amarantaceae Celosia trigyna L. Onagraceae Jussieua suffruticosa L. var. linearis Ludwigia multiflora (Guill. et Perr) Walp. Euphorbiaceae Euphorbia hypericifolia L. Compositae Ethulia conyzoides L. Eclipta alba (L.) Hassk. Chrysanthellum americanum (L.) Vatke Convolvulaceae Ipomoea eriocarpa R. Br. Scrophuiaraceae ... Striga hermonthica Benth. Striga geanoroides (Willd.) Vatke Labiatae ... Odmum americanum L. Commelinaceae ... Commelina nudifiora L. Commelina benghalensis L. Cyanotis axillari. Roem. et Shultes Cyanotis (?) foecunda Hassk. Amaryllidaceae ... Pancratium trianthum Herb. TABLE 168 WEEDS FOUND ON CULTIVATED LAND MALAKAL COLLECTION PROVISIONAL IDENTIFICATION Family Species Amarantaceae Celosia argentea Cyathula (?) prostata Lythraceae Ammannia (?) baccifera Cucurbitaceae Cucurbita sp. Tiliaceae ... Cor chorus (?) fasciculari. Malvaceae Abutilon glaucum Euphorbiaceae Euphorbia sp. Phyllanthus niruri Phyllanthus maderaspatensis Cephalocroton cordofanus Acalypha indica Tragia cannabina Tragia sp. Papilionaceae Crotalaria sp. Phaseolus sp. Rhynchosia memnonia Composi/ae Chrysanthellum procumbens Picridium tingilanum Rublaceae ... Borreria sp. (Spermacoce ? compacta) Solanaceae Solanum dubium Physalis sp. Convolvulaceae Evolvulus nummularius Ipomoea (1) aquatica Ipomoea cordofana Acanthaceae Hygrophila spinosa Dicliptera sp. Commelinaceae Commelina sp. Aneilema sp. Cyperaceae Cyperus sp. Gramineae Eleusine indica Digitaria marginata Panicum hygrocharis Dinebra arabica Setaria incrassata Ischaemum brachyatherum Pennisetum ramasum 351 PESTS Pests in the Jonglei Area range from the minute sorghum midge to the elepham. The relative importance of different species vari()s from locality to locality and from crop to crop. Dama~ done by crop pests is universal, and losses of yield very considerable. Birds are a major pest in the Jonglei Area, and the damage to crops which they cause is very considerable, in some cases perhaps reaching 75% or more. The Sudan Dioch (Quelea que lea aethiopica) is the most serious pest; large swarms of this species are a constant menace to cultivations, especiaJly in the northern half of the area. In 1952 the first attempt was made to control these pests by exploding gelignite in the roosting areas. Though thousands of birds were killed, it is doubtful if the four operations carried out had any lasting eifeGt on the Sudan Dioch population in the ar()a, especially as it is recorded that the swarms invade the northern part of the area from Blue Nile Province and Abyssinia. As in the latter country no measures are taken to control these birds, they will provide a constant source of infestation even if control measures in the Sudan are successful. Roosting sites in the Jonglei Area are generally smaller than those farther north and therefore more difficult to control by explosion, and it is doubtful whether this method could be successfully employed. The interrelation between the nesting sites, roosts, and general geographical distribution of these birds, and the water available during the dry season should be studied, as it may provide important information. The theory has been advanced that there is a general southerly migration of these brrds from October or November onwards, the migration being caused by the shortage of water in the northern inland areas. As this is variable according to the year, dates of migration vary. The birds can eat the crop only at the milky stage, as the grain is too large for them to swallow whole. Before being eaten the grain has to be broken, and the damage done to the crop is much in excess of the grain actually consumed by birds, as more is spoilt than eaten. Once the grain hardens it is safe from attack from small weaver birds. Consequently the state of the crops at the time of arrival of weaver birds is of great importance. In some' years the birds happen to arrive in a cultivation belt just as the crops are maturing and the damage done by them is considerable. In other years they arrive late, but may still damage a crop in a more southerly belt. Wilson(30) has noted that these enormous concentrations of Sudan Dioch do not, curiously enough, seem to attract any large numbers of predators or to show any sign of epidemic diseases. The ecological balance of nature seems to be in favour of these pests. The Nile Masked Weaver (Sitagra taeniopterus taeniopterus) is common throughout the Flood Region of the Jong1ei Area, and is also responsible for damage to grain crops at the pre-harvesting stage. There are many other species of weaver birds, but their identification, distribution, and economic importance as pests require further specialized enquiry. The Red Bishop Bird or Dura Bird (Euplectes franciscana franciscana) is very conspicuous because of its striking rainy season plumage, and in some localities may ,be a major pest. The Sudan House Sparrow (Passer domesticus arboreus) is widespread throughout the area and, though not conspicuous, is probably responsible for a lot of damage. The Sudan Mourning Dove (Streptopelia decipens decipens), the Pink-Headed Dove (Streptopelia roseogrisea roseogrisea), and even the European Turtle Dove (Streptopelia turtur turtur), are common and are responsible for some damage. These and many other birds attack grain at the pre-harvest stage, while some species cause damag() at both pre-harvest and pre-germination stages. The Pied Crow (Corvus albus) and Black Crow (Heterocorax capensis capensis) are the greatest nuisance in this respect, but the Golden Crested Crane (Balearica pavonina ceciliae) is often guilty too. Guinea Fowl (Numida sp.) and Francolin (Chapinotryx sp.) do a good deal of damage at pre-germination stages. . Bulrush millet is especially liked by birds, and in some areas it is not planted for this reason. The awned varieties are less vulnerable, and may be the solution to the problem of damage by birds. Sorghum, of course, suff()rs most; groundnuts and simsim are also subject to the attacks of these birds. M AMMALS The mammalian fauna of the Jonglei Area have been described elsewhere (see Chap. 2, p. 187) and the majority of its members are guilty of some damage to crops, the chief culprit varying according to district. Oribi and Red-fronted Gazelle in the northern half of the Jong1ei Area and MongaUa Gazelle in the southern half, together with Tiang, common in most 352 parts of the area, are responsible for a good deal of crop damage; Bushbuck, Reedbuck; Waterbuck, and even White-eared Kob, Giraffe, Buffalo and Roan Antelope are known to cause occasional damage. All these animals attack the crop at an early stage, grazing the succulent young shoots. Baboons cause a lot of damage in some parts by pulling crops, such as groundnuts and oth~r legumes, out of the ground, leaving them half consumed.~ In the southern reaches of the river Hippopotami ascend from the water to devour maize and other crops planted on the flood-plains. Warthogs are accused by the Bari and Mandari of eating . groundnuts. Elephant deserve special mention. While other species may check the crop, but only seldom ruin it completely, a herd of elephant can, in a few minutes, completely destroy the fruits of an entire season's labour, leaving behind them bare earth covered with the remains of the trampled crop. In localities where elephant abound they are a nightmare to the cultivators. The damage these animals do is more than just the destruction of growing crops; they discourage the extension of cultivated areas and cause the cultivator to limit his cultivation to the immediate vicinity of his homestead, so that these fields are overcultivated. They also prevent the introduction of some new crops, such as groundnuts for which elephant have a special liking ; the failure of attempts to introduce this crop to the Bor area was largely due to constant damage done by this pest. Rodents- rats and mice-are also responsible for a lot of damage to crops at n~arly all stages of growth. INSECTS There is practically no crop in the 10nglei Area which is not subject to insect attack. Grasshoppers, crickets, millipedes (e.g. Daphis nervii L.) attack young seedlings of all crops. Fortunately since the 1932-33 season there has been no plague of locusts. Reports before that date record both Hairy-Chested Locust (Locusta migratoria migratorioides Rch. and Frm.) and Desert Locust (Schistocerca gregaria Frost) as very serious in some seasons. Termites and ants cause damage to most crops at different stages of growth ; 'physiological wilt ' is caused almost entirely by these insects. Stem-borer (Sesamia cretica Led.) attacks sorghum, maize and bulrush millet; in bad years it is responsible for damage estimated at over 75% of the yield. The practice of the local inhabitants of leaving cereal stalks on the field until they rot or are consumed by animals undoubtedly increases the infestation of fields with this pest. Probably also the staggering of grain crops, which is so common, has the same effect. Stem- boring Beetle (Ceroplesis irregularis Har.) has also been identified on sorghum at the Malakal Experimental Farm. Dura Aphis (Aphis sorghi Theob.) is common throughout the 10nglei Area, especially in dry years. Damage done by this insect is difficult to assess. Cotton Aphis (Aphis gossipyii) has also been observed, while pentamoid bugs (Agonoscelis versicolor F.) have been found on sorghum, sesame and castor. The Sorghum Midge (Contarinia sorghicola Coq.) is responsible for considerable damage to this crop in some seasons. The following pests have been recorded on cotton crops : Thrips (Hercothrips spp.), lassids (Empoasca libyca de Berg), Stainer bugs (Dysdercus sp.), Stem-borer (Sphenoptera gossypii Cotes), Sudan Bollworm (Diparopsis castanea Hampson), Egyptian Bollworm (Earias insulana Boisd.), Pink Bollworm (Platyedra gossypiella Saunders), Flea Beetle (Podagrica puncticollis Wse.), Sudan Cotton Worm (Xanthodes graellis Feist). This list is far from complete and further entomological studies are required in order to identify, assess the seriousness of, and devise remedial measures against insect pest attacks. CONTROL M EASURES Control measures employed by the inhabitants are not efficient or generally effective. Bird-scaring is a major part of the wet season work of small boys (cf. p. 338), but though they undoubtedly save a lot of grain in the vicinity of the homesteads, the area protected in this way constitutes only a small fraction of the total area of crops. Scaring is also used to protect grain against wild animals. In afforested areas, especially in the Bari and Mandari districts, fences are constructed of thorny bushes, which give the crop some protection. Nowhere is any attempt made to grow hedges, though some suitable indigenous plant could undoubtedly be found for the southern part of the area, where rainfall is heavier and the dry season comparatively short. The cotton crop is the only one for which some attempt is made to control insect damage, by burning all debris including roots and, in the case of pump schemes, spraying with insecticides. Insects attacking other crops are regarde~ as an unavoidable evil. The whole problem of pest control requires further investigation and experimental work. 353 PARASITES WITCHWEED (Striga hermonthica BeRth.) (Ar. buda; Shill uk motto; Nuer wal bar; Dinka dai; Mandari laliok; Bari laUom). o ThB semi-parasite warrants a section to itself, as it is one of the main scourges of cultivation in this area. Striga hermonthica and its near relative Striga asiatiaca (or lutea) occur throughout the· tropics and sub-tropics of the Old World(31). In the Sudan only the former is of importance, though Broun and Massey(32) have recorded eight species of Striga genus occurring in this country. " Striga hermonthica Benth. is a branched herb 1-2 ft. high with bluish-green linear to lanceolate leaves, 1-3 in. long, hispid and scabrid. The spikes are terminal, approximately 6 in. long; calyx 5-toothed, upper smaller; corolla pink to light violet, tube i in. long, curved and inflated, upper lip 2-10bed, t in. long, lower 3-10bed, t in. long; capsules ellipsoid, t in. long "(33). It is a semi-parasite on sorghum, bulrush millet, finger millet and a number of other C€lreals and wild grass species. There is also evidence that it can parasitize some leguminous plants, though rather feebly(34). Striga is well adapted to its semi-parasitic mode of life, and is consequently very difficult to eradicate. Its seeds are distributed in the soil to at least 15 in. below the surface. The great majority of these seeds can only be germinated by the secretions from roots of certain plants (see Vol. II, p. 589) not all of which can act as hosts(35). The seeds require a considerable period for ripening before they will germinate. Comparatively few germinate in the first season after shedding, the rate of germination increasing in the subsequent seasons. The seeds remain viable in the soil for at least 10 years(36). After stimulated germination the radicle of Striga seed attacks the host's root and forms hostorium, and eventually a parasitic connection with the xylem of the host is established. Unless Striga becomes attached to a host plant at an early stage, its seedling is unable to survive. In the early stages of growth the development of Striga is concentrated on the formation of underground parts, swollen with starchy material acquired from the host plant. During the first four or six weeks no aerial parts appear on the soil surface, but the main damage to the host is probably done in these early stages. Two weeks after its appearance Striga is flowering, though only a small number of the Striga plants parasitic on the roots of the host ever reaches the surface and flowering stage (in Andrew's experiments only approximately 5%). This, however, is not a great disadvantage to reproduction, as the same author found that each plant was able to produce ·over 40,000 seeds. As far as is known Striga does not perennate even if parasitic on perennial plants. The main crop affected by Striga is sorghum, the chief agricultural product of the Jonglei Area. Maize, bulrush millet, and finger millet are also susceptible, but economically they are less important. Striga infestation is cumulative, i.e. the longer the land is under sorghum cultivation, the more serious the intestation. Consequently the parasite causes the most serious damage in the central part of the area where the better-drained land is limited and hence intensively cultivated with sorghum. Further investigation is necessary to establish what conditions favour attack by Striga. However, it is certain from field observations that conditions favourable for the growth of host plants reduce the severity of infestation and damage to crop or both. In years of low rainfall damage by Striga is particularly serious. Andrew's(37) experiments in the Gezira confirm this; light irrigation increases the intensity of attack, while heavy irrigation decreases it. The effect of Striga infestation can be extremely serious as it may reduce yields of sorghum to practically nothing. In 1950 we measured the effect on some local <:ultivations: TABLE 169 EFFECf OF INFESTATION WITH STRIGA HERMONTHICA ON YIELDS OF SORGHUM Appro". % Reduction in Degree of infestation Sorghum plants Yield in kg.p.f. yields affected (estimated) % Very slight ... ... 10 216 slight Slight ... ... 20 . 176 20 Heavy .. . .. . 70 72 3S 354 , In a year of light rainfall (e.g. 1952) Striga attack in many places caused complete failure of the crop. On the basis of field observations, enquiries, and experiments carried out at Malakal, we estimate roughly that the average loss of crops in the area due to Striga attacks may be as much as 30%. Homestead cultivations suffer most, while in ' out-cultivations' Striga is one, and in some districts the main, reason nec~ssitating fr~quent shifts in cuJtivation grounds. Jikaing, Leik and Bul Nuer have a method which they allege controls Striga. Small banks . are erect~d in the fields by means of which flooding can be controlled. Striga appears above the ground when the sorghum is 12-18 in. high and able to withstand a certain amount of flooding. At this stage after heavy rain the wat~r is held by the banks and the field subjected to waterlogging for 5 to 10 days, which does more harm to theStriga than to the crop (see also p. 590). Further investigation of this difference in susceptibility is required, as it might provide a means of Striga control. The only other method of control at present attempted by the majority of the inhabitants of the area is weeding. Th~ effectiveness of this method depends mainly on the timing of the weeding operations and the persistence of cultivators in carrying it out season after season. It is further discussed in Volume II, Chapter 7, p. 590. 7. ECONOMIC ASPECTS LABOUR One of the main reasons for the small use made of the natural agricultural potential of the Jonglei Area is the limited labour available for crop production. As can be seen from Table 128 (p. 230) and p. 232, the density of population in this part of the country is low, varying between approximately 60 persons and 7 persons per square mile. The effectiveness of the labour is also low, as can be seen from Table 170 in which we compare data from the Malakal Experimental Farm, employing Nilotics, and the Gezira Research Farm, wh~re mainly northern and western labour is employed. It should be noted that the labour requirements summarized in this table are higher than the amount of labour expended by the average cultivator for similar tasks in raising his own crops. Not only do experimental crops generally receive more care, but also the cultivation of small experimental plots cannot be compared with the cultivation of large fields . The figures in Table 170 below show that the Nilotic is generally a less effective agricultural worker than the northern and western labourers employed in the Gezira. The reasons appear to be both physical and psychological. We have found on the Malakal Experimental Farm that no more than 4 or 5 hours of hard work can be expected of him; lengthening the daily working period does not necessarily lead to any increase in his daily output. This applies particularly to heavy work like using a digging-hoe, excavating earth, etc., operations which are not part of his normal methods of cultivation. On the psychological side, as we have already pointed out, the Nilotic generally regards crop husbandry as merely a necessary evil and gives it only a minimum of interest. Those few Nilotics who migrate northwards to work on pump schemes and other projects do not seek a permanent livelihood there; their main aim is to earn cash with which to buy cattle, and the majority return to their homes after only a short absence. The annual migration in search of pasture or because inland water supplies are dry, which often takes almost the wholt' able-bodied population a considerable distance away from the cultivation areas, is another factor lirrllting the amount of labour available for crop production. Migration to distant cattle-camps means that usually only a little preparatory cultivation is done from late December to mid-April, and therefore nearly all of it has to be done within approximately eight months of the year. Crop output suffers particularly from the limitations imposed by the short time available for field preparation and sowing. The labour available for cultivation is further decreased by the amount of time spent by many cultivators in travelling between their homesteads and ' out-cultivations' on the higher patches of land, which are often a considerable distance from the homesteads. For example it was found that in the Shilluk village of Lelo the average time required to walk from home- stead to puothe wak (see Chap. 3, p. 222) was 1 hour IS minutes. To summarize; in most parts of the Jonglei Area the amount of labour available for the production of crops both for subsistence and for cash seriously limits the output. The reasons for this are: (i) Low density of population. (ii) Low effectiveness of labour. (iii) Annual migration and other operatio~s necessary in animal husbandry, limiting the time available for crop production. (iv) Wastage of time involved in walking between homesteads and cultivations. 355 23 TABLE 170 HAND LABOUR REQUIREMENTS OF DIFFERENT AGRICULTURAL OPERATIONS AT THE MALAKAL EXPERIMENTAL FARM. AND THE GEZIRA RESEARCH FARM REQUiREMENTI> IN LABOUR-DAYS(') Type of Work Malakal Gezira Experimental Farm Research Farm(') per feddan per feddan A. LAND PREPARATION 1. Initial clearing of perennial grass 30 2. Land previously cultivated : (a) digging with loria 10 (b) weeding with maloda ... 5 B. SOWING Spacing 50 cm. x 80 cm. 8 Spacing 100 cm. x 100 cm. 4 Broadcast 3 C. THlNNJNG(3) Cotton (to 3 stalks) 5 12 Sorghum (including transplanting) 2 D. WEEDING Heavy ... 20 8 Medium 7t 6 Light 5 4 Fallow weedi-\lg per season (October-May) 8 E. HARVESTING GRAIN, IITC. Sotghum-n ridges 12 Rice- 6, Equatoria Province 5, Blue Nile Province 3) : • EXPORTS FROM THE SUDAN Year Metric Tons Value £E. 1940 ... 716 20,700 1941 1,120 35,200 1942 1,923 67,300 1943 1,827 70,500 1944 1,710 99,704 1945 1,619 108,803 1946 1,939 145,324 1947 1,924 116,621 1948 1,242 82,832 1949 1,336 62,787 1950 . .. 1,583 84,356 19.51 2,027 131 ,65 1 BIBLIOGRAPHY Boulenger, G . A . .. . The Fishes of the Nile, 2 Vols., London, 1907. Catalogue of the Freshwater Fishes of Africa, London, 1916. Bloss, 1. F . E . 'The Sudanese angler', S.N. & R., 1945. Girgis, Sabet , A list of common fish of the Upper Nile with their Shilluk, Dinka and Nuer names " S.N. & R. 1948. Graham, Michael .. The Victoria Nyanza and its Fisheries, Crown Agents, 1929. Irvine, F. R. The Fish and Fisheries of the Gold Coast, Crown Agents, 1947. Jonglei Investigation Team First Interim Report, 1946. Second Interim Report, 1947. Third Interim Report, 1948. Progress Report, 1948-1949. , Problems of fisheries in the area affected by the Equatorial Nile Project " ed. H. Sandon, S. N. & R., 1951. Sandon, H. An Illustrated Guide to the Freshwater Fishes in the Sudan, Khartoum, 1950. Stubbs, J. M. , Freshwater fishes of the northern Bahr el Ghazal " S.N. & R., 1949. Weiman, J. B. Preliminary Survey of the Freshwater Fisheries of Nigeria, Lagos, 1948. Worthington, S. and E. B. Inlalld Waters of Africa, Macmillan, 1933. Worthington, E . B. A Report on the Fishing Survey of Lakes Albert and i(ioga, Crown Agents, 1929. Science in Africa, C.U.P., 1938. Sci~nce ill the Middle East, H.M.S.O., 1946 . • 397