‘O.~-vnt .- v .0 THE SOMALI BOUNDARY: DISPUTE AND FUNCTiONAL EVOLUTION AND AREDITY AND LAND use m NORTHEAST AFRICA . Research Papers for the Degree of M. A. MICHIGAN STATE UNIVERSiTY HARRY E. comsrocx m 1972 L131 RARY Mid)! ga an State Il‘I-UHIV VCrs “7 III II IIIIII III IIII III IIIIII IIII III III III IIII III IIII III IIII II 3 350g I?” I) THE SOMALI BOUNDARY: DISPUTE AND FUNCTIONAL EVOLUTION BY I. Harry E: Colestock III A RESEARCH PAPER Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF ARTS Department of Geography 1972 TABLE OF CONTENTS Page LIST OF TABLES . . . . . . . . . . . . iv LIST OF ILLUSTRATIONS. . . . . . . . . . v INTRODUCTION. . . . . . . . . . . . . 1 Chapter I. AN ANALYSIS OF THE SOMALI BORDER DISPUTE FROM A HISTORICAL—POLITICAL GEOGRAPHIC VIEWPOINT . . . . . . . . . . . 6 Causes Somali Conflicts: Pre-colonial History Menelik Somali Holy War and Movements into Kenya Colonial Partition of the Horn British Somaliland French Somaliland Italian Somaliland Menelik and the Colonial Powers Treaty of London Italo-Ethiopian War Post-War Politics Trigger Action Ethiopian Nationalism Nationalism in French Somaliland Somali Nationalism The Northern Frontier District and Kenyan Nationalism ii Chapter Page Aims of the States Concerned The Preservation of Ethiopia Somali Politics Peace and Status Quo in Africa French Neo—colonialism Geographic Arguments Economic Geography and Nomadism Tribal Boundaries and Islam Results of the Dispute Detente in the Horn of Africa Recent Deve10pments in the Somali Republic II. FUNCTIONAL EVOLUTION OF THE SOMALI BOUNDARY SINCE 1960 O O C O I O I O O O O O 49 Boundary Functions and Their Evolution The Boundary of the Somali Republic Regional Politics and Somali Boundary Functions Existent in 1960 Foreign Policies in the Horn The State Functions Evident at the Somali Boundary Correlation between Foreign Policies and State Functions at the Boundary after 1960 Somali-Ethiopian Boundary Somali—Kenya Boundary Boundary between the Somali Republic and the French Territory of Afars and Issas Conclusions Regarding Correlation between Policy and State Functions Applied at the Boundary CONCLUSION . . . . . . . . . . . . . . 69 BIBLIOGRAPHY. . . . . . . . . . . . . . 74 iii LI ST OF TABLES Table Page 1. Air Passengers During Selected Months Flying between Ethiopia and the Somali Republic. . 60 2. Somali Trade with Neighboring Countries. . . 65 iv LIST OF ILLUSTRATIONS Map Page 1. The Horn of Africa . . . . . . . . . 7 2. Political Boundaries in the Horn 1888—1972 . 12 3. Regional Political Units 1950-1960. . . . 26 r‘.‘ ,. -_..,._....-..._ —v-.. .5...“ fiw—u‘r’VH—qw INTRODUCTION During the 1960's armed conflict occurred in the Horn of Africa between the Somalis and the troops of Kenya, Ethiopia and French Somaliland. The conflict arose ostensi— bly from disagreement concerning the political boundaries separating the Somali peoples. The irredentist calls for Somali unification and self-determination have provided much material for students of boundary disputes. Boundary studies in disputed areas have tradition— ally been a favorite research subject for political scien- tists, historians, international lawyers and political geographers, "but the facility of geographers with maps and their understanding of regional characteristics, have given them an advantage in such studies."1 Among geogra- phers "there seems to be no standard methodology for study— ing disputed areas, although the backgrounds of the areas in terms of the bases of the conflicting claims are always deemed very important."2 1J. R. V. Prescott, The Geography of Frontiers Egd Boundaries, Chicago, 1965, p. 110. 2J. V. Minghi, "Boundary Studies in Political G€3\ hock / f I Q - OIIbouII I l gulf of Aden I. Zeila / Zladeh .\ ~ \ Barbara L-|O° \. (J '09 \ W V\‘ \ AddIs Aboba Howr‘ \ \l \_\V H 0‘ '\‘\ Q U \ d P I o 50 I00 I50 \\ miles 45° +4- DIIbouII-Addrs Abobo rarlrood LL]. Limit 0‘ Somalr-Inhobned area FTAI French Territory of Mars and Issas \\ l / \x ,’ Obea , _ __\< 5.” I‘ ’ ’ y \m ‘\ | I 5 fi,./ \0\ ‘ ‘ d I d m n ,I o n I V :. I O c e a n I V‘ / I \\‘ - -~ Worshelkh .,., .J I, c O _ / oqadishu ,4 d , ’ orca . o /, .‘I _/.J . I” (I rovo I - — Internohonol b0undary 200 HEC '72 I A ~|50 I- I0’ L50 ,0. Jar — — .— _..~ — . -— —»—- —. —.. A MAPZI A FIIKEBI.A N P'E N IN SIJI.I\ POLITICAL BO UNDARIES IN THE HORN I888-l972 Ade" IO.‘ Indian Ocean Anglo-French Agreement at IBOBIZonee at Influence) “Anglo-Itahan Treaty of I89! (Zone: at Influence) -+- Anglo-Iraluan Protocol at I894 (Zones of influence) - - Franco—EthrapIan Convention at l897 -—- Anqla— Etnrapuan Treaty at I897 —- Ital-an Interpretatuon at l908 Conventron -o- Etnlopuan Interpretatran at l908 Conventran 44— The "Somali Lune" Jubalana tranetered to Italy In l925 -- Bntreh Pravmonol Adrmntetratwe lune at l950 - - International boundary at Kenya 0 50 IOO ISO 200 . -. __._—l miloe 4?’ HEC'72 l3 extremely significant as the British began the partition of the Horn. British Somaliland A combination of two factors were probably re— sponsible for the small amount of exploration along the Somali coast by EurOpeans: The first was the presence of hostile tribes, and the second was the lack of economic gain for the Europeans from this barren land. Ironically though, it was the former of these factors which eventually led to the British occupation of the coast. British interest in the Somali coast was originally motivated by a concern for the security of trade and communications with India. Thus, in the early 1800's the British made treaties with several of the coastal tribes to keep them from endangering British shipping. The event which neces- sitated further British security in the Red Sea area was the opening of the Suez Canal in 1869 and the increased British trade through this avenue. The Britons were also interested in protecting the caravan trade from northern Somaliland which supplied the important garrison at Aden.22 The strategic importance of this area influenced Britian in the 1870's to go into negotiations with Egypt to secure safety for British operations in the Sea along the Somali coast. Egypt had become (since 1860) the 22Touval, pp. 32—33. 14 "de facto" ruling force along the Somali coast, but since the Turkish government exercised nominal suzerainity over Egypt itself, the treaty reached by Egypt and Britain in 1877 was meaningless in law because of Turkey's rejection of it. However, this treaty points out the extent to which the Egyptians had attempted to be influential in the Somali coast area (and a probable reason for continued Egyptian attempts to influence the Somalis today). Naturally, this treaty rejection was not the end of the British colonial fortunes in the Horn. As a result of the Mahdist revolt in the Sudan in the 1880's, Egypt decided to withdraw its commitments from the entire Somali coast. This decision did not go unnoticed in Europe. In 1881, Great Britain realized that if she were to protect her interests in the Horn, she must act quickly. As it became clear that the Egyptians were going to withdraw, Britain informed Turkey that she would recognize Turkey's claim over the coast as far as Zeila as long as Turkey would replace Egypt and maintain its authority in the area. Even as the Turks were belatedly contemplating the situ— ation, Britain "notified them that in view of information regarding the imminent outbreak of disturbances in the area, British forces would be there at once to maintain order, unless the Turks did so. . . ."23 With no clear 23Ibid., p. 36. 15 reply from the Turks, the British quickly moved into Zeila in 1884. During the years 1885 and 1886, the British ob- tained new treaties with the Somali tribes along the coast. In these pacts several tribes pledged "never to cede, sell, mortgage or otherwise give for occupation, save to the British government, any portion of the territory presently inhabited by them or being under their control."24 These agreements were to be cited in the dispute when the Somalis perceived their territory given to their enemies, the EthiOpians, in the years that followed. French Somaliland In the formation of French Somaliland, France had three main objectives in mind: "(1) her need for a base and coaling station along the route to Madagascar and Indochina, (2) her desire to develop trade, and later (3) the exigencies of the competition among the powers."25 On March 11, 1862, a treaty initiated the colonial influ- ence on the French in the Horn of Africa. The French, as the British, had seen that the Egyptian withdrawal from the coast opened up new possi— bilities for colonial expansion. During the years 1884 24British and Foreign State Papers, Vol. 76, 1884-85, pp. 101-107. 25Touval, p. 37. l6 and 1885, the French obtained additional territory from the Issa Somali and other nearby tribes. Unfortunately, both the French and the British had concluded treaties with the Issa tribe and a boundary dispute arose between the colonial powers. The possibility for open conflict was great, but conflict was averted by direct negotiations. The result of this Anglo—French Agreement of 1888 was that the Issa tribe was divided between two colonial countries. This became one of the first of many examples where little thought was given to the distribution of ethnic groups in the partitioning of the Horn of Africa. Although the original motives have relatively little relevance in today's world, France still maintains her foothold in the Horn. Italian Somaliland The Italians ventured into the Horn in the 1870's, motivated by aspirations for imperial grandeur and nationalistic fervor. Italian activity first became evi- dent in Eritrea and the Red Sea coast. Then in 1884 the Italian government decided to send an exploratory mission to the Benadir region (in present-day southern Somali Republic). The Italians received many rebuffs from the SUltan of Zanzibar when they made known their desires for land.along this coast. Nevertheless, on November 18, 1889, Witfln the help of Britain, Italy acquired control over the towuas of Brava, Merca, Mogadishu and Warsheikh. Italy 17 acquired these towns under the terms and conditions of the Sultan's concession, which meant that Italy was entitled only to the hinterlands of these towns within a radius of from five to ten miles. "But the Italian government assumed a protectorate over the portions of the coast lying between the aforementioned towns and so notified the European powers on November 19, 1889."26 In that same year the Italians gained concessions in the North at Obbia and all the coast to Bender Ziadeh. Thus, by 1889, the Italians had claimed the Somali coast to the approximate extent that Italian Somaliland would eventually have. Menelik and the Colonial Powers Menelik began his invasions of Somali terri- tory in 1886; and, by 1891, he had advanced his claims over a large amount of Somali territory. He then decided to send a circular letter to the Powers defining the boundaries of Ethiopia. The Somali territories that were claimed were the province of Ogaden, the tribes of the Habar Awal, the Gadabursi, and the Issa (the last three being peoples with whom Britain, and in the case of the Issa, France, had treaties of protection). These claims seem to have suited the Italians and the Ethiopians, the 26Raffaele Ciasca, Storia coloniale dell'Italia cC’ntem oranea (Milan: Ulrico Hoepli, 1938) , pp. 237—238 found in Touval, p. 42. 18 former of which had become recognized as the "protectors of Ethiopia." But the other Powers took little notice of this letter or had no knowledge of its existence.27 It was in this role as "protector" that Italy entered into negotiations with Britain concerning the boundaries in the Horn. In March, 1891, an Anglo—Italian protocol was signed which defined the boundary between East Africa and Italian Somaliland, and the Italians proceeded to press for further definition between her and Britain's Somali territories. At this time Britain was reluctant to negotiate, but in May, 1894, a protocol was signed defining their respective spheres of influence (see Map 2). The British reluctance to negotiate was justified when Menelik invaded areas of the British sphere. But since the Italians insisted that they should be the party consulted about Ethiopia, the British cries were heard by those who could do little with the Ethiopian raiders. Nevertheless, the outcome of the negotiations with Italy "amounted to a recognition that the Ogaden lay within the 28 Italy's Italian Sphere, and the Haud in the British." later claims, therefore, derived mainly from her "pro— tectorate" over Ethi0pia and the raids of the "protected" people in the Ogaden area. In 1896, with Italian defeat by the Ethiopians at Adowa, it was clear that Ethiopia was not a protectorate 27 28 Touval, p. 48. Ibid., p. 56. 19 of Italy. It was then that Britain realized that she must deal directly with the EthiOpians. The French had already recognized this fact, and, a few days before the British arrived in Addis Ababa, the French reached an agreement with Menelik which established the boundaries between French Somaliland and Ethiopia. The price for this con— cession to the French was a railroad from Djibouti to Addis Ababa.29 Thus, by this treaty the French violated their earlier treaty with the Issa Somalis never to cede their territory. This fact contributed to the resentment of the colonial policies in the French sphere in the Horn. The British, too, had to pay dearly for their territorial claims in the Horn. The British Envoy, Rennell Rodd, went to Addis Ababa and talked with Menelik about the boundaries and some "other important matters." These other matters included an assurance from Menelik that he would prevent the passage of arms to the revolting Mahdists in the Sudan. With this favor in mind, the British were told by Menelik to negotiate with Ras Makonen rather than himself, for he himself had little knowledge of the local situation in the disputed area. Thus, the negotiations were held in Harar with Ras Makonen, one of Menelik's ablest military leaders and skillful diplomats, with the result being a line about halfway between the 29O. Homberg, "Future of Djibouti," Asiatic Review, October 1936, 32:850-854. 20 coast and the 1894 Anglo-Italian boundary (see Map 2). In an analysis of this Treaty of 1897, one author sees simply that "the Somali territory was bargained away to stop Ethiopia from sending help to the Mahdists in Sudan."30 This boundary, more or less dictated by Makonen, was again determined with little respect to the wishes to the in- habitants of the territory. However, a provision in the Treaty provided that tribes on either side of the border would have the right to use the grazing grounds on the opposite side. After these negotiations and an Italo-Ethiopian peace treaty, the Italians attempted to discuss their boundaries with the Ethiopians. The vaguely-defined sphere of the Italians was approximately 180 miles from the coast running from the boundary of the British Somali— land Protectorate to the Juba River. The 1908 convention between Italy and Ethiopia did little to clear up the boundary definition (see Map 2). In later years this ambiguity resulted in a "tradition of uncertainty'and conflict which eventually led to the WalWal incident and Italo-Ethiopian war. . . ."31 306. W. Waterfield, "The Trouble in the Horn of Africa? The British Somali Case," International Affairs, Vol. 32, January,l965, p. 54. 31Lewis, History, p. 62. 21 Treaty of London The vagueness of the Italo-Ethiopian agreements was not duplicated when Britain divested itself of 33,000 square miles of Jubaland to Italy in 1920. "Britain's commitment arose out of a secret undertaking (Treaty of London), between Italy and Britain in 1915, that 'in the event that Great Britain increased her territory in Africa at the expense of Germany, Her Majesty's Government would agree in principle that Italy might claim some equitable .n32 compensation. (See Map 2.) Italo-Ethiopian War The events leading up to the war in the Horn point out the great potential for conflict that the Somali boundary area has held. Initially, events centered around the wells of WalWal, which, from 1930, had been permanently occupied by authorities from Italian Somaliland. Although documents of the time indicated that these wells were over 100 kilometers west of the boundary of Italian Somaliland, the EthiOpian government had not lodged a protest with Italy during the four years after 1930. Then, in November 1934, we observe a force of 600 EthiOpian troops moving into WalWal as a protective escort for an Anglo-Ethiopian boundary commission, which was 32J. Drysdale, The Somali Dispute, New York, 1964, p. 38 quoting J. Ramsey MacDonald, Housefiof Commons, March 3, 1924. 22 ascertaining information about which tribes had the right to graze on either side of the British Somaliland- Ethiopian border. There was a sizable contingent of Italian Somalis at WalWal and after some minor incidents, reinforcements were sent in by both sides. On December 5, fierce fighting broke out and the Ethiopians were routed when the Italians brought in tanks and aircraft. The WalWal incident and some political maneuvering by Italy gave that country a pretext for the invasion of Ethiopia on October 3, 1935. England and France seem to have stood idly by while Italy embarked on her conquest of Horn. During the Second World War, nationalism arose in Ethiopia, and as a counterweight, the Italians paid special attention to the Moslem population there. This favoritism caused an even greater rift between the Somalis and the EthiOpians. After the Italians had taken over Ethiopia in 1936 and British Somaliland in 1940, the British took the offen— sive and in a few months had taken both Italian and British Somalilands as well as Ethiopia. Thus, some of the tribes of the Horn had come under three separate administrations in the space of five years. The British victory was accomplished with many British Bantu troops, whom the Somalis regarded as in- ferior. However, these Black Africans were literate and skilled in their jobs, and the Somalis then realized their 23 own "backwardness." The realization produced many demands for education after the war and aided in the rise of Somali nationalism and awareness. Another boost to the Somali nationalistic fervor during the war was probably the Allied propaganda. This influence was not overwhelming, "but the prOpaganda could hardly have failed to raise in their minds the question whether 'freedom' and 'the right of all peoples to choose the form of government under which they will live' might not have some relevance to their own problems."33 Post-War Politics In 1946, the Haud and the Ogaden were kept under British Military Administration rather than being turned over to the Ethiopians. The British had apparently planned to unify the Somali territories after placing them under British trusteeship. Concerning this plan of the British Foreign Secretary: "There was a great outcry in EthiOpia against Ernest Bevin's plan to unite all the Somalis under a British protectorate, and the prOposal was immediately exploited by the Russians as an indication of the British imperialistic tactics in Africa and the Near East."34 33Touval, p. 78. 34C. J. Jaenen, "Whither Somalia?" Middle East Affairs, April, 1957, 8:135. 24 In 1948, a commission composed of American, Russian, French, and British officials formally took up the Somali question. British Secretary Bevin again pro- posed "that British Somaliland, Italian Somaliland and the adjacent part of Ethiopia, if Ethiopia agreed, should "35 But the be lumped together as a trust territory. Russians were against the British trusteeship and the French (who controlled Djibouti) were against any sug— gestions of a Greater Somalia. The result was the Longrigg solution whereby Italy assumed trusteeship over former Italian Somaliland, while Britain remained in British Somaliland. It should be clear by now that the causes for the Somali border dispute lie in the history of colonial rule over the Horn and the lack of accurate boundary definition by the parties involved. This framework then overlies the complex relationships of the Somali tribes, their move- ments in the Horn, and age-old animosities between the Somalis and the other peoples of Northeast Africa. These causal factors set the stage for the "trigger action" which marks the beginning of the modern history of the Somali boundary dispute. 35Parliamentary Debates, House of Commons, fifth series, 432:1840-1 in Touval, p. 79. 25 Trigger Action The "trigger action" is defined by Prescott as that action "which creates a favorable situation for a claim to be made."36 In the case of the Somali boundary dispute, the situation is more complex than to suggest a single trigger action. In any analysis of this dispute, confusion is created by the mere number of disputants: EthiOpia, Italy, Kenya, Great Britain, and the Somalis. Ethiopian Nationalism If one had to select the most important factor as a "trigger action," it would have to be the rise of nationalism within Africa. Ethiopia was the first part of the region to attain its former autonomy after the War, while the rest of the Horn remained under a British Mili- tary Administration. Ethiopian nationalism initially took the form of a desire to unify the Horn under one adminis— tration and consolidate the greatest area possible into a nation-state. The first steps that Ethiopia took were to push Britain for a settlement of the boundaries in the Horn, and in so doing started a series of events which have brought dispute and disagreement in the region. Britain settled on a "provisional" boundary that was somewhat a compromise between the two interpretations of the 1908 Italo—Ethiopian Convention (see Map 2). It was fixed in the Anglo-Ethiopian Protocol "as being without 36Prescott, p. 110. 26 MAP 3 REGIONAL POLITICAL UNITS I950" |960 ( Ade" I (3qu of \.__,,..... ' v .I I < I ‘. I I \ . o I ! .4 l0 ‘5‘ B RITI SH I '0 SOMALILAND I ."\.\.‘. .l’ \_\.‘ / .\‘. / x,‘. ./ ,1 II V \ E T H I o P I A x I9 [I \ v V / \ /’ \ar w>' / V)" .’l 9° 5° ......... / V 5'- /".' Q I Y I, ’./"\‘-‘/P'\I"/ Q Yo . -~-.--’ I, 0 \« Ind/an / Ocean /" 91% I ‘9 04“ I Q I \ I KENYA I I I I. . ' 0 I -—---Pa|ltical boundariee O‘I I a so I00 I50 aoo \_ miles 40: 49° 29' HEC'72 27 prejudice to the international frontier between Ethiopia 37 The Italians took over and former Italian Somaliland." the administration of Somalia on April 1, 1950, and in- herited this boundary. Although the British wished to remain in the "Haud and Reserved Area" after the War, the Ethiopians pushed for a settlement of the boundary in this area, too. After some years of delaying the action, Britain finally decided to uphold the 1897 Treaty previously mentioned. This meant that Ethiopia would be sovereign in Somali-inhabited areas even through the tribes on either side of the boundary had grazing rights on the other side. The Somali were outraged by this agreement, for they said their territory had been given to Ethiopia without their knowledge or consent. Despite the efforts of Somali political leaders to urge Britain to reconsider, the 1954 agreement was upheld.38 The problems associated with the new Anglo- Ethiopian agreement were (1) it vagueness on certain points, and (2) its lack of consideration of the con- ditions which had arisen since 1897. Perhaps the main reason which demonstrates the two faults of the Treaty is the seemingly purposeful failure to mention the 37Drysdale, p. 71. 38M. Perham, The Government of Ethiopia, North- western University Press, 1969), pp. xxxiv-xxxv. 28 c:itizenship of the nomadic inhabitants concerned. This (Dinission led Ethiopia, not unnaturally, to claim that any person spending more than six months in the Haud was an Ethiopian. Coincidentally, the increase in Somali pOpu— ;L.éastion since 1897 had caused the Somalis to spend more t:.;i.zme in the Haud in search of food. The seasonal migration c>m1§? thousands of Somali herders was usually in search of p a sture, "but what had been happening is that many British £3<::>Imalis had been staying in the Reserved Area longer than they used to in order to sow and reap a crop."39 The Ethiopians argue that these people were no longer nomads, that they were Ethiopian subjects, and that the provisions (Diff 'the 1954 agreement did not apply to them. Ethiopian nationalism and the successes in the 51‘:=¢ Hw>flo HMGOflumcumucH ”condom mew Hem mm» one «on cow 0 o fleece mom ace NHH ace see c o o panama senacnooz am «NH NOH «ma es o o o senaonooz humana emm HAN cam emm mum com o o manna mecca someonooz mom mom one mom mmm mmm o o acnecnooz manna manna some some come some some coma some come 09 some mom can no: new um: mom no: new .oaanommm Hamfiom men was mamownum cmmzuwn mcflmam mnucoz Umuomamm mcflusa mummcommmm “Hm .H magma 61 In the years following the detente, Somali- Ethiopian relations have only been marred by sporadic incidents which occur in the Somali-inhabited Ogaden of Ethiopia. These incidents have little influenced the basic policies of the two states, and a cautious spirit of c00peration has replaced the antagonism of the early 1960's. Somali—Kenya Boundapy The boundary between the states of Kenya and the Somali Republic has existed only since Kenyan independence in December, 1963. However, independence did not change the position of the political boundary. It is possible to consider the boundary of Kenya Colony from 1960 through 1963 as serving the purposes of a valid international boundary. Although the British were the policymakers during the pre-independence years, their decisions were largely based on the desires of those who were to ultimately receive power in Kenya. Therefore, foreign policies be- fore and after independence were decidedly similar. Britain was interested in maintaining a viable Kenyan state during the years preceding their departure from their East African colony. To this end they devoted much time to ignoring any demands for secession from the colony. The Somalis of the Northern Frontier District (N.F.D.) were the most outspoken group desiring separation 62 from Kenya. In order to temporarily appease the Somalis, in October, 1962, the British appointed a "Commission to ascertain and report on public opinion in the N.F.D. re— garding arrangements to be made for the future of the area. . . ."88 But the British chose to disregard the finding of the Commission that an overwhelming majority of the population of the N.F.D. desired secession from Kenya and union with the Somali Republic. It was announced in March, 1963, that the N.F.D. would remain part of Kenya when independence arrived in December. The Somalis immedi— ately severed diplomatic relations with Great Britain, although the Somalis still attempted to persuade the new African leaders of Kenya to accept their claims. However, the leaders were committed to a policy of prohibiting any fragmentation of the new country. No resolution of the policy conflicts occurred prior to independence, and hostility between the two states began to increase. Shortly after freedom was granted, Kenya proclaimed a state of alert in the N.F.D.89 The Kenyans were unable to create the "effective boundary" they desired between themselves and the Somali Republic, and much subversive trans—boundary activity occurred. Although the Somali government denied that military 88Somali Republic, 1964, p. 54. 89Somali Republic, Ministry of Foreign Affairs, 23 Peuple Somali a 13 Recherche de Son Unité, London, September, 1965, p. 34. 63 assistance was given to rebels in the N.F.D., the weapons and training of the "shifta" guerrillas there almost conclusively originated from within the Somali Republic. The Somalis "conducted a campaign against the Kenya police and army for more than four years."90 The campaign in— cluded hostile radio propaganda from the "Voice of Somalia radio in Mogadishu which was able to greatly influence the ."91 In reference to the level of guerrilla activity. . . radio broadcasts, Kenya was also unable to establish its "effective boundary" where it desired. The state function of suppression of the entry of revolutionary materials into the country was only partially performed, even though overall movement across the border had been drastically reduced by the border closure in 1966-67. To further inhibit the infiltration of Kenya, the government "not only demarcated its border with Somalia more clearly, but also cleared a belt of adjacent land of all settlement."92 Not until the change of Somali governments in 1967 was the situation altered. The policy shift of the Somalis resulted in agreement to end the feud between the two states. The joint communique issued in October, 1967, also Specified that the parties were to "refrain 90Kaplan, 1969, p. 228. 91 Ibid. 92de Blij, p. 205. 64 from conducting hostile propaganda through mass media . . . ,"93 suspend emergency border regulations, re- establish diplomatic relations, and consider encouraging trade and economic cooperation. The previous closure of the border was suspended and new functions replaced the right of states to inhibit all movement across their boundaries. The abrupt alteration of policy has been followed by a period of relative tranquility and normal relations between the states. The changes in policy and boundary functions may be correlated by observing the trade between the Somali Republic and her neighbors illustrated in Table 2. Major decreases in trade occurred (1) in the years following the border war with Ethiopia, and (2) during the closure of the Kenya border in 1967. The detente in late 1967 is reflected by increases in all categories except the export trade to Ethiopia, which does not reflect an increase until 1969. Boundary between the Somali Republic and the French Territory of Afars and Issas In 1960, the F.T.A.I. was called French Somali- land, but the name change in December, 1966, has done little to alter the character of French colonialism in the former colony. The policy makers from 1960 have 93Thurston, back cover. 65 .mmIommH memo» mnu How .moeumHuMpm opens HmcowumcnwucH mo xoonumww .mcowumz popes: "mousom .emme ou HOeum oxemcmmcma one weapon mo mmonu sues woumuomuoocm mums mwcmx How mowumwumum ovmeam oee me me em em eee eem eem eee New neooeeem emoe eeoe eee meme II II II II II II custom ecoeuoceunoee meme eeme eeme eeme eeme eeme meme meme eeme eeme ennooxm eeome emeoe eeme meem meme eeee eeem eeee omen eeae eeooeeum eemee eeme Nee emee II II II II II II enscox Acwmwuov meme eeme eeme eeme eeme eeme meme meme eeme eeme nonoose .emmceaeenom HHmEom oooev mmeuucsou mnemonsmeoz £ue3 memes meEOm .N manna 66 continued to be the French, despite the fact that colonial status was supposedly terminated in the territory. The maintenance of the French position of power in the F.T.A.I. has nurtured a reactionary policy which includes tactics of suppression and extra-legal means to achieve goals. This policy was in direct Opposition to the rise of Somali nationalism within the territory. Nationalism reached its peak in the riots of late 1966 during French President de Gaulle's visit. "After the riots, in which a demand for an immediate end to French rule was the main issue, all Somali who could not prove their citizenship were temporarily expelled from French Somaliland. A plebiscite was held under French supervision in March, 1967, which resulted in a victory for those supporting 94 The pursuance of the French continued ties to France." policy necessitated measures restricting the registration of voters who might favor independence in this referen- 95 "The vote [was] denied to thousands of Somalis by dum. alleging they [were] 'foreigners' not born in the terri- tory. The existence of documentary proof of place of birth is very rare in these areas of the world, so that it 941. Kaplan, et 31., Area Handbook for EthiOpia, Washington, D.C., 197T: p. 315. 95Somali Republic, 1967, p. 10. 67 [became] virtually impossible for the Somalis to counter this allegation."96 The French maneuvers were calculated to deprive from voting those Somalis who traditionally crossed the border frequently and whose allegiance was primarily to the Somali Republic. The state functions of restricting immigration and establishing requirements for citizenship were effectively utilized for the purposes of the French. Not only did the French disenfranchise Somalis who were considered illegal aliens, but many who were probably citizens of the territory were incarcerated prior to the referendum. "The French . . . rounded up 4,000 Somalis and trucked them out to a prison camp in the desert out- 97 Although the French governor had wanted side Djibouti." to deport the Somalis over into the Somali Republic, Somali border officials refused, for they wanted to keep the problem in the French territory and before the world as an example of French oppression. The situation soon stabilized after the referendum, and the election of the Egal government later that year furthered a lessening of 96Somali Republic, Ministry of Foreign Affairs, Memorandum on French Somaliland Submitted by the Govern- ment of the Somali Republic to the United Nations Spec1al Committee on the Situation with Regard to the Implemen- tation of the Declaration on Jthe Granting of Independence to Colonial Territories and: Peoples, Mogadishu, June 1, 1965, p. 6. 97New York Times, March 26, 1967, p. 2. 68 tensions and a return to functions of a non-restrictive nature at the Somali border. Conclusions Regarding Correlation between Policy and State Functions Applied at the Somali Boundary The Somali boundary forms a case characteristic in many areas inhabited by nomadic peoples. Historically, few functions have ever been applied at these boundaries, and specific border crossing points are few. However, the divergent policies promoted by the countries in the Horn following 1960, and the dramatic changes in those policies, resulted in a rapid evolution of the state functions found at the Somali boundary. A close correlation exists be- tween policy and the nature of functions at a boundary. The policies of adjoining states effect the degree to which the state functions are applied at the boundary, and the amount of agreement between the two states on the geo- graphical realities of the region. CONCLUSION The nature of the analytical framework in boundary studies will obviously determine the types of observations made concerning the boundary. The conclusions reached through the utilization of the two different approaches in this research are of varying degrees of usefulness to the political geographer. In the historical political geographic approach, the Horn of Africa is seen as a region which, for several centuries, has contained a zone of confrontation between the Somalis and the area's other inhabitants. The present Somali political boundary may be viewed as "superimposed" upon the actual dividing line between the conflicting cul— tures. Such "superimposed" boundaries are characteristic in many parts of Africa, where the colonial legacy of boundary-making has been a source of irredentism. The basic causes of the Somali border dispute still exist, but the catalytic factors (so-called "trigger actions") have since abated and active opposition to the position of the boundary has almost been eliminated. The rise of nationalism in Africa following World War II has been largely replaced by a realization of the monumental 69 70 tasks required in building a nation—state within the frame- work of colonial boundary definition. The decline of nationalistic fervor has served to focus attention upon the internal economic and political problems of the indi- vidual states. The nature of these problems may be traced, in part, to the political boundaries separating the various states in the Horn. The economic consequences of (or threat of) being deprived of traditional grazing areas have been adverse on the Somali herders. Such a problem is only one of many facing the Somalis. The massive loans from the Soviet Union and the lack of a resource base have created a situation from which the Somalis will be hard-pressed to extricate them— selves. The dependence upon outside aid for development of the economy has plunged the country into enormous debt. From the Somali standpoint, the integration of Somali lands would greatly improve the economic viability of the state. The drawing of the political boundaries of the Horn has affected not only the economic interrelationships in the region, but also the power structure of the coun- tries defined by those boundaries. The multitude of tribal and clan units within all the states of Northeast Africa provides the challenge of unification to any governmental structure. 71 The future of the states of the Horn will depend heavily upon the effectiveness with which they deal with their internal problems and the degree to which regional cooperation overshadows national differences. The recent past has shown that negotiation may replace confrontation, and that the development of the region may now take place, unimpeded by armed conflict. The above approach reveals several political geo— graphic relationships in a framework of history. This view emphasizes the time element98 in boundary disputes. In fact, a possible criticism of this traditional approach is that historical aspects of the problem are treated more thoroughly than geographical factors. Although one section of the analysis specifically treats geographic operables, the bulk of the approach is usually only indirectly re- lated to a geographic view. The historical-political geographic view provides one of the most comprehensive examinations of boundaries in disputed areas. The framework suggested by Prescott can be useful if the political geographer realizes its limitations. The facts concerning individual boundaries will not always fit neatly into the categories of Pres- cott's approach. The uniqueness of each boundary tends 98See Minghi, p. 414, for detailed discussion. 72 to invalidate many of the generalizations concerning them.99 The second of the approaches involves unique examples of political geographic processes in reference to boundaries. The state functions applied at the Somali boundary are viewed as directly related to the foreign policies of the countries of the Horn. The conflict be- tween the Somalis and the EthiOpians and Kenyans in the early 1960's resulted in policies restricting movements across the border. The restriction on movements included not only peOple but also goods and ideas. The policy of detente in 1967 was reflected by an easing of the regu- lations imposed at the Somali border and a return to the pre-existing cross-boundary movements. This view of the functional evolution of the boundary introduces the concept of the "effective bound— ary." Such a notion may be applied not only to inter- national boundaries, but also to ideological boundaries. The division of the world into ideological blocs has given rise to a supranationalism that has "transferred conflict from the boundaries of territorially organized states to the minds of man."100 998. B. Jones, Boundaryémaking, a Handbook for Statesmen, Washington, 1945, p. vi. Found in Prescott, p. 24. 100N. J. G. Pounds, Political GeogrdpQXI New York, 1963, p. 95. 73 This type of study is capable of greatly augment— ing our understanding of boundaries in disputed areas. The dynamic factors in the problem are emphasized and interrelationships between the political situation and the conditions at the boundary are readily apparent. Al- though numerical correlations between policies and the state functions applied at the boundary are often diffi— cult to obtain in developing areas, intuitively related factors reveal the processes involved with disputed boundaries. The productivity of this approach will un— doubtedly increase as the research data base is improved and as methods are devised to effectively measure the necessary variables. BIBLIOGRAPHY BIBLIOGRAPHY Ayele, Negussay. "The Politics of the Somalia-Ethiopia Boundary Problem: 1960-1967." Ph.D. Dissertation, University of California at Los Angeles, 1970. Bayne, E. A. "Birthday for Somalia." American Universi- ties Field Staff Reports (A.U.F.§; Reports), Northeast Africa Series, Vol. VIII, No. 1, August, 1961. . "Brinkmanship on the Horn." A.U.F.S. Reports, Northeast Africa Series, Vol. X, No. 1, March, 1963. . "Chiaroscuro on the Horn." Part II: "Origins of Detente." A.U.F.S. Reports, Northeast Africa Series, Vol. IV, No. 2, December, 1968. "Somalia on the Horn." Part III: "Politics." A.U.F.S. Reports, Northeast Africa Series, Vol. VII, No. 8, March 8, 1960. . "Somalia on the Horn." Part IV: "The Ethi- opians." A.U.F.S. Reports, Northeast Africa Series, Vol. VII, No. 9, March 11, 1960. . "Somalia's Myths are Tested." A.U.F.S. Reporpe, Northeast Africa Series, Vol. XVI, No. 1, October, 1969. . "The Issue of Greater Somalia." A.U.F.S. Reports, Northeast Africa Series, VoI. XIII, No. 2, March, 1966. Boggs, Samuel W. "Boundary Functions and the Principles of Boundary Making." Annals, Association of American Geographers, Vol. 22, March, 1932. . International Boundaries: A Study of Boundary Functions and Problems. Columbia University Press, 1940. 74 75 Braine, B. "Storm Clouds Over the Horn of Africa." International Affairs, Vol. 34, October, 1958. British Society for International Understanding. "The Somali Republic." The British Survey, No. 203, February, 1966. Castagno, Al. "Somalia Goes Military." Africa Report, Vol. 15, No. 2, February, 1970. Ciasca, Raffaele. Storia coloniale dell'Italia con— tempoanea. Milan, 1938. Czeslaw, Jesman. The EthiOpian Paradox. London, 1963. de Blij, Harm J. Systematic Political Geography. New York, 1967. Drysdale, J. The Somali Dispute. New York, 1964. . "Somali Frontier Problems." World Today, Vol. 20, January, 1964, pp. 1-5. Great Britain. British and Foreign State Papers, Vol. 76, 1884-1885. . Parliamentary Debates (Commons), Fifth Series, Hartshorne, Richard. "The Polish Corridor." Journal of Geography, Vol. 36, May, 1937, pp. 161-176. Held, Colbert C. "The New Saarland." Geographical Review, Vol. 41, October, 1951, pp. 590-605. Homberg, 0. "Future of Djibouti." Asiatic Review, October, 1936, 32:850—854. International Civil Aviation Organization (ICAO). Digest of Statistics, 1966—1970. Jaenen, D. J. "Whither Somalia." Middle East Affairs, April, 1957, 8:135. Jones, Stephen B. Boundary—making, A Handbook for States- men. Washington, D.C., l9 5. Kaplan, I. et el. Area Handbook for Ethigpia. Washifigton, D.C., 19717 . Area Handbook for Somalia. Washington, D.C., 1969. 76 Kasperson, Roger B. and Julian V. Minghi, eds. The Struc- ture of Political Geography. Chicago, 1969. Lewis, I. M. The Modern History of Somaliland. London, 1965. . Pegples of the Horn of Africa. London, 1955. MacKay, J. Ross. "The Interactance Hypothesis and Boundaries in Canada." Canadian Geographer, No. 11, 1958, pp. 1-8. Mariam, Mesfim Wolde. The Background of the Ethio— Somalia Boundary Dispute. Addis Ababa, 1964. Minghi, Julian V. "Boundary Studies in Political Geography." Annals, Association of American Geographers, Vol. 53, 1963, pp. 407—428. Moodie, A. E. The Geography Behind Politics. London, 1957. New York Times. 1947-71. Perham, M. The Goyernment of Ethigpia. Northwestern University Press, 1969. Pounds, N. J. G. Political Geography. New York, 1963. Prescott, J. R. V. The Geogrephy of Frontiers and Boundaries. Chicago, 1965. Randall, Richard R. "Political Geography of the Klagen- furt Basin." Geogrephical Review, Vol. 47, July, Shaudys, Vincent K. "Geographic Consequences of Estab- lishing Sovereign Political Units." Professional Geographer, Vol. 14, March, 1962. I "Somali Republic: Desert Fix." Economist, Vol. 206, March 2, 1963, p. 792. _ Somali Republic. Ministry of Foreign Affairs. French Somaliland: A Classic Colonial Case. Mogadishu, March, 1967. Le Peuple Somali a la Recherche de son Unité. London, September, 1965. 77 Somali Republic. Ministry of Foreign Affairs. Memorandum on French Somaliland Submitted by the Government of the Somali Republic to the United Nations Special Committee on the Situation with Regard to the Implementation of the Declaration on the Granting of Independence to Colonial Territories and Peoples. Mogadishu, June 1, 1965. . The Somali Republic and the Organization of African Unity. Mogadishu, 1964. Thurston, Raymond. "Detente in the Horn." Africa Report, Vol. 14, No. 2, February, 1969, pp. 6;13. Times (London). 1956—71. Touval, Saadia. Somali Nationalism. Harvard University Press, 1963. United Nations. Yearbook of International Trade Sta- tistics, 1960-70. . U.N. Document. A/C.4/SR.1001. Waterfield, G. W. "The Trouble in the Horn of Africa? The British Somali Case." International Affairs, Vol. 32, January, 1965, pp. 52-60. Wilkinson, H. R. "Yugoslav Kosmet: The Evolution of the Frontier Province and Its Landscape." Institute of British Geographers, Transactions and Pepers, No. 21, 1955, pp. 171-193. ARIDITY AND LAND USE IN NORTHEAST AFRICA BY Harry E. Colestock III A RESEARCH PAPER Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF ARTS Department of Geography 1972 TABLE OF CONTENTS Page LIST OF FIGURES. . . . . . . . . . . . iv LIST OF MAPS O O O 0 O O O O O O Q 0 I V Chapter I. INTRODUCTION . . . . . . . . . . 1 Definition of Study Area. . . . . . 2 Physiographic Regions. . . . . . . 3 II. THE CLIMATE OF NORTHEAST AFRICA WITH SPECIAL EMPHASIS ON THE DEGREE OF ARIDITY O I O O I O I O O O I 6 General Circulation Pattern. . . . . 6 Temperature and Precipitation . . . . 9 Definition of Aridity. . . . . . ll Climatic Patterns of Northeast Africa According to Thornthwaite' 5 Water Balance Method . . . . . . . . 15 Problems of Water Balance Calculations in Northeast Africa . . . . . . 16 Moisture Need. . . . . . . . . 18 Moisture Surplus. . . . . . . . 21 Moisture Deficit. . . . . . . . 23 Moisture Index . . . . . 25 Elevation-Moisture Index Relationship. 27 III. LAND USE PATTERNS, NON—CLIMATIC FACTORS, AND THE REGIONAL WATER BALANCE . . . . 30 Bases for Land Use Data and Definitions. 30 Non-Climatic Influences on the Land Use Pattern. . . . . . . . . . . 32 ii Chapter Page The Impact of Culture . . . . . 32 Political Factors in Somali Areas . . 35 Spatial Interrelationships between Land Use Patterns, Ethnic Areas, and the Regional Water Balance in Northeast Africa . . . . . . . . . . . 37 IV. CONCLUSION . . . . . . . . . . . 43 BIBLIOGRAPHY. . . . . . . . . . . . . 46 APPENDIX . . . . . . . . . . . . . . 49 iii LIST OF F IGURES Figure Page 1. The Water Balance at Selected Stations in Northeast Africa . . . . . . . . . l4 2. Elevation—Moisture Index Correlation . . . 28 iv LIST OF MAPS Page Physiographic Regions and Weather Stations of Northeast Africa . . . . . . . . . . 4 Mean Monthly Pressure and Wind Distribution, January 0 O O O O O O O O O I O O 7 Mean Monthly Pressure and Wind Distribution, April 0 O O O O O I O O O I O O 7 Mean Monthly Pressure and Wind Distribution, July 0 O O O O O O O O O O O O 7 Mean Monthly Pressure and Wind Distribution, October . . . . . . . . . . . 7 Moisture Need. . . . . . . . . . . 20 Moisture Surplus. . . . . . . . . . 22 Moisture Deficit. . . . . . . . . . 24 Moisture Index . . . . . . . . . . 26 Land Use, Critical Climatic Zone, and Somali Regions of Northeast Africa . . . . . 39 I . INTRODUCTION ' The pattern of land use in a specific area is largely a consequence of the actions taken by the inhabi- tants of the region. These actions are the collective results of decisions based upon the individual's perception of his environment. In Northeast Africa, the most signifi- cant influence upon perception is usually the availability of water. The lack of water profoundly affects the life styles of the peOples of the region. The paucity of adequate water supplies is directly related to the regional climatic distribution, and this relationship is essential for the attainment of the objectives of this study. The major purpose of this essay is to investigate climate and its influence on the land use pattern of North— east Africa. The basic hypothesis states that there exists a critical climatic zone which overlies the division between non-irrigated agriculture and herding. Within this zone, decisions concerning the utilization of land depend primarily upon the degree of aridity. A corollary Problem considered involves the influence of non-climatic factcms upon the land use pattern. It is theorized that analysis of the cultural and political elements of the societies of the region will explain irregularities in any correlation between the patterns of climate and land use. Since aridity is the central subject of this research, its accurate definition is a necessity. The water balance method of Thornthwaite (and Mather, 1955) has been selected as the basis for judging the aridity of the region in relative numerical terms. There are few studies, such as the works of Garnier (1956) and Sibbons (1962), which effectively evaluate the application of the Thornthwaite system to an arid area. Northeast Africa provides unique evidence to test the usefulness of the water balance method in a predominantly arid region. Definition of Study Area The arbitrary western boundary of Northeast Africa used in this paper is represented on all maps by a dashed line. This line forms the border line from the Red Sea along the Western side of the rift valley south of Lake Rudolf. From Lake Rudolf, the border is parallel with and east of the Matthews Range until the Tana River, where it parallels the Tana to the coast. The Eastern boqudary is defined by the coasts of the Indian Ocean, the Red Sea, and the Gulf of Aden. The area thus defined encompasses the territory 0f the Somalis. In order to show more contrast of climatic and cultural phenomena, the study area was ex— tended beyond the Somali—inhabited lands to include the wetter parts of Northeast Africa, where the population is primarily Galla. Physiographic Regions The regions used in this study are adaptations from authors (Kendrew, 1961; Tato, 1964) who used physio- graphic criteria as a basis for explaining the rainfall distribution in Northeast Africa (see Map l). The coastal plain is a narrow strip which is extremely hot with little or no rainfall. Elevations are normally below 400 feet. The rift valley is also hot and dry in its lower elevations in the north, but as altitudes rise in the south, it becomes cooler and wetter. Zones of higher elevation along the sides of the rift also receive high amounts of rainfall, both in the north and the south. In this study, the highlands are con- sidered as land above 3500 feet, and land between 1000 and 3500 feet has been arbitrarily selected as territory ‘within the plateau regions. The Ogaden and Northern plateau slopes gently toward the east, but there is a steep escarpment to the Gulf of Aden. The Southern plateau slopes southeasterly toward the Somali coast, and elevations are generally lower than in the Northern plateau. The alluvial plain is found in the lower reaches of the southeasterly flowing rivers before they reach the 3'5- 4‘0» .2- 56- -2o- 20--l MAP | Physiographic Regions ”.3 and Weather Stations 5“ of Northeast Africa -I5° I" O. W1 \ “‘ -. “\““I I I “I \‘J I ,J ‘I ‘L. _.I' I ."‘J\.\ I I I' l ' 7 ( I l ’/ 3 e ‘51 9 I ‘: ’F’J; ,—-;,'-‘-’.’-~ I ‘ -l0° i" I” 2 XII‘S} :~;—v;(_"5 46- .‘7 ("I l0‘-‘ ( ’1’ X35 .'\,, '76,“ ,9?53 ,5; .‘9 I’ll, i8 ’5; 025 ’ “- .37 \IP”/ 5 Us? '5‘ so I’ll” r~." '27 /’ V""~~-.se so Mr“ 5‘ l/ .2.”/’ 3 ".324 "' ._.._..;.. (‘1 ll, "~._ {On/I, I ’fl .60 " x’ I \ .32 ' r’ {‘2‘ [’1'3”:3' '33 .34 1“ .62 4 _./':{’:7' I!” a? I (:30 ______________ e’;\“~\ i ”I l,’ \. I, ‘-" \\\ \\\‘ '7’: J r— 5. ‘-I l, .63 \\\ [TIE I, 6 ’fel” ‘— ..... —. L, (I fall”. ’. r __“_\'_':'..72 71”,! 5.— .‘I I: “\.\I. 5 J \ 'I . ’{7 s. I I '1, 'I ‘ .,-_.- .1" "mt-"Pix ' I 7‘ "I '3 ‘\ 6.. I". -' I I C. ’1’ x, \. ‘x. I” I' I 39’ .273 ‘ \\\ 65 .i ’1” 0771* ’ " “Isl/I; .|.2 ) “2A” ,4: '4 / I .'I \\\ {I i ’.\‘I I] 6 Lo. ‘3 “ x l 510 TI 437:! it" I I” z - 0 O" \ _\ \. ‘\‘:-__,‘J<’ .\\ x ‘2‘} \‘ 2 '~.. \_ (I ‘4 \-. x“.- -s° 5'1 I Coastal Plum 2 Rift Valley 3 H|gh|onds O 50 IOOsc“.ZOO 300 4 Ogaden and Northern Plateau ‘ mm, m 5 Southern Plateau 6 In'erior Auuvia' PIOin See Table I for Key to StotIoM 35' 0° 1 49° - 4f _ _________ 5 l“ __ HEC '7I coast. Elevations here are between 400 and 1000 feet above sea level. The plateaus and the alluvial plain show a marked decrease in precipitation as altitude de- creases from west to east. These regions are only approximations of the physiographic areas of Northeast Africa. Their value lies in their convenient grouping of the weather stations, and they represent a framework to judge the climatological data. II. THE CLIMATE OF NORTHEAST AFRICA, WITH SPECIAL EMPHASIS ON THE DEGREE OF ARIDITY General Circulation Pattern A key to the understanding of the variation of rainfall is found in the general circulation pattern (Maps 2-5). The circulation pattern in Northeast Africa is dominated by the dry Northeast monsoon in winter and the wetter Southwest monsoon in summer. The dry season begins in late October-early November, and the effects of the Northeast monsoon are evident until March. The mean surface air features dur- ing the period include two anticyclones, one centered over the Sudan, and the other overlying Arabia and pro- jecting southward into EthiOpia (Tato, 1964). The other major surface feature is a Low over Eastern Sudan which moves generally north-south during the year. In the be- ginning of the dry winter, the mean air flow is southerly. As the season progresses, the winds abruptly change to easterly and northeasterly. The dry continental tr0pica1 air masses prevailing at this time are derived from the .Asiatic winter High and can be considered as belonging to the planetary trade flow. They provide little moisture for precipitation in Northeast Africa. Mean Monthly Pressure and Wind Distribution ’.._..-.-_-- MAP 2 1" January MAP 4 July Home. VIM Distribution. and Geography Northeast Africa and the mute tut. lutltute up, Jun, 857. use “It Transitional periods between the monsoons usually last only a few weeks. In regard to precipitation, the most important transition occurs in the spring. A mean surface Low, which has overlain the Eastern Sudan during the Northeast monsoon, begins its movement northward in February or March. Wind directions shift generally from northeast to the east and southeast as the Low moves over Ethiopia during late April and early May. At the juncture of the Northeast monsoon and the Southeast trades is the intertrOpical convergence zone (ITC), which produces what are locally called the "little rains." By June, the low pressure cell has moved over Northern EthiOpia, and the flow over the region has become southwesterly. Trewartha (1960) observes that most of the precipitation in Northern East Africa is derived from the southwesterly air stream. However, in Northeast Africa east of the mountains of Ethiopia, the lack of precipi— tation during the Southwest monsoon is attributed mainly to divergence (Flohn, 1964). According to Tato (1964), the general winds in EthiOpia from June to September are also from the southwest, but the air from the South Atlantic is sufficiently moist to produce large amounts of rainfall in the Ethiopian highlands. Trewartha (1960) has summarized both the main causes of precipitation in Northeast Africa and the reasons for modest amounts of rainfall in the region. He states: There are developed two major zones of wind dis- continuity and convergence, (1) an east—west zone, the ITC, separating hot, dry, anticyclonic Saharan— Arabian air to the north from the cooler, moister southwesterly air of maritime origin to the south, and (2) a north-south zone separating relatively moist westerly Congo air from the South Indian Ocean. Since Congo air rarely is able to penetrate east of the Abyssinian barrier, its influence on the weather of the Somalia area is slight and this no doubt is a partial explanation for that region's drought. The rainfall in Ethiopia, on the other hand, is to a much greater extent derived from Congo air. His reasons for the low amounts of precipitation include: (1) the divergent character of both monsoons over extensive land areas; (2) the modest depth of the Southwest monsoon, especially over the highlands; (3) the strongly meridional flow in all but the transitional seasons, a feature that limits the advection of sea moisture and reduces orographic effects; and (4) the stable stratification of the air aloft, including a marked decline in moisture content. Temperature and Precipitation The small annual temperature ranges in the trOpical regions are characteristic of Northeast Africa. The mean annual range is about 10°F for most stations, with only a few parts of the interior exceeding 30°F. Most of the variation from station to station is a consequence of elevation. The mean monthly temperatures vary from a high of 96.5°F in Djibouti (2)1 in July to a low of 50.7°F at Goba (33) in December. The elevations for these two stations are 33 feet and 9100 feet respectively. 1Values in parentheses indicate station numbers in Table l and on Map 1. 10 The small temperature ranges along the Eastern Somali coast are partly explained by the upwelling of cold waters. Besides lowering the annual temperature range (Walton, 1969), the cold current helps create high humidity and a high incidence of fog. The fog restricts insolation and in turn reduces temperature. But little of this moisture is precipitated because of the stable or even inversion structure of the air along the coast. The precipitation of Northeast Africa is highly variable and most of it falls in the hottest months. Although rainfall, like temperature, is related to ele- vation, the general circulation pattern affects the relationship between precipitation and altitude. The general decrease in precipitation from west to east in Northeast Africa modifies the elevation—rainfall corre- lation. The range of annual rainfall is from 1mm at Guardafui (7) to over 1300mm at Cencia (30), but annual values may vary considerably from the long term averages. Rainfall maxima occur at different times in different parts of the study area (refer to Map 1). In Region 1, almost no rain is recorded, and the rainfall regime shows little periodicity. Region 2 has a single summer maximum and, like Region 1, is almost without rain from October through February. Precipitation is fairly evenly distributed through the year in Region 3, with a summer maximum in some locations. In Regions 4, 5, and 6 11 there are double maxima, usually April—May and October, with April-May being the larger. Definition of Aridiry The lack of moisture in certain areas of the world produces environments which have traditionally been called arid. Early attempts to define aridity utilized average annual rainfall as a simple indicator of aridity. However, rainfall alone does not reveal accurate values of water available in the soil. Temperature, evaporation, and plant transpiration form the other major factors which are needed in calculating the water balance of an area. Penck (1910) recognized the need for a better definition of aridity when he defined the arid boundary as the line where evaporation and precipitation are equal in amount. The difficulty of obtaining evaporation sta- tistics, however, prevented development of this notion on a worldwide basis. Koeppen (1931) linked temperature and rainfall in an attempt to better define climatic boundaries. Koeppen took into account the increased water losses in areas of higher temperature. The definition of the boundary of aridity required that an increase in mean annual temperature be offset by an increase in mean annual precipitation. Further revisions of the work took into account the seasonality of rainfall and the coastal desert phenomena of frequent mist and fog. Koeppen, and his 12 contemporaries Lang (1920) and de Martonne (1927), ex- pressed indices of aridity in terms of relatively simple ratios between precipitation and temperature. The amount of evaporation was always approximated by considering it directly proportional to temperature. The climatic classification proposed by Thornth- waite (1948) includes a definition of aridity based on the water balance. This concept refers to the balance between availability of atmospheric water, its storage, and its expenditure at a given place (Thornthwaite and Mather, 1957). The important factors in the Thornthwaite system are precipitation and evapotranspiration. Thornth— waite and Mather (1955) defined potential evapotranspi- ration as the amount of water which would be lost from a surface covered with vegetation if there was sufficient water in the soil at all times for the use of vegetation. This value is computed by means of an empirical formula involving mean monthly temperature. A vital concept utilized in computing the water balance is soil moisture storage (Thornthwaite and Mather, 1954). When potential evapotranspiration (PE) is com- pared with precipitation (P) and allowance is made for the storage (ST) of water in the ground and its subsequent use, periods of moisture deficiency (D) or moisture sur- plus (8) are revealed. These values indicate the relative :moistness or aridity of a climate. The seemingly complex 13 formulae of the system have been put in tabular form, which greatly simplifies working with the classification (Thornthwaite and Mather, 1957). Although there are drawbacks to any determination of the water budget by an empirical formula, the Thorn— thwaite method has attained widespread popularity, both because it requires only records of commonly accessible parameters (i.e., temperature and precipitation), and be- cause it is probably the best approximation of climatic- biospheric relationships on a worldwide basis. The climatic classification displays a general correlation to the world distributional patterns of soil and vege— tation (Chang, 1959). Thornthwaite's classification is based on the calculation of the moisture index which is obtained in the following manner:1 I = 100 (S-D) m PE .All positive values are considered humid climates, and all negative values arid climates. 1The original Thornthwaite formula read I = 100 (S—0.6D) m PE frhe six-tenths weighting factor had been justified in terms Iof soil moisture availability, but improvements in the Iestimation of soil moisture by Thornthwaite and Mather (1955) eliminated the need for this factor. 14 The Water Balance at Selected Stations in Northeast Africa Hosaina are lest-rem stereos: 300" m leathers stereee: too am W 200 Molsture '5‘; l ...........o .00 my \ O- O JFIAIJJASONDJ JFIAIJJASONDJ Daloh an Iexlrnurn "“000: 300 mm nun Innis-en store's: I50 an 2 200 (“L goo Moisture I50 Index-~27 tOO loo ........ L. muffiiimiih 0 O JFUAIJJASONDJ JFflAIJJACOIOJ Ionte nun Iexlnurn stereos: 300 or ISO nun — Precipitation 20c ”’ Potential evapotranspiration “"”‘ Actual evapotranspiration lac flolsture Moistus deficit '““""7¢9 [ED Moisture surplus Soil moisture utilization Soil moisture recharge Figure 1 JFMAHJJASONDJ “EC '72 15 Climatic Patterns of Northeast Africa According to Thornthwaite's Water Balance Method Carter's study (1954) represents perhaps the most thorough application of the water balance method to class- ify the climates of Africa. Although the same author (with Mather, 1966) revised the classification system, the cli- matic maps of Africa in the latter work appear essentially unchanged. The diagrams in Figure 1 represent water balance relationships at selected stations in Northeast Africa. The effect of the selection of differing maximums of soil moisture storage may be compared in these graphic dis- plays. The maps on the following pages were constructed utilizing Carter's original work as a base. Climatic values for 78 stations (see Map l and Appendix 1) were calculated using the formula of the 1966 revision and climatic isolines were redrawn where there were conflicts with the Carter maps.1 1In areas of no water surplus, the 1966 formula simply increases the moisture index by a factor of 1.667. Thus, the boundaries of the arid and semiarid climates remained essentially unchanged except where more accurate 'values were found. But where water surplus occurs, the climatic boundaries may be greatly altered. For example, Bishoftu (27) had a climatic index of +3.6 from the original formula, while the revised calculations reveal a value of —lO.2. Almost all areas with humid climates ‘were reduced in size when the new formula was used. 16 Problems of Water Balance Calculations in Northeast Africa Any investigation of climate in Northeast Africa suffers from a general lack of data. Also, the data that are available are often inaccurate and the periods of observation are not uniform. For instance, the climatic records used in this study range from 4 to 40 years of observation. Nevertheless, averages of temperature and rainfall, even for the shorter periods of observation, displayed only small deviations from mean values, facili- tating the drawing of the isolines on the scale of the maps used here. Few, if any, inconsistencies due to climatically unusual years were observed in the distri- butional patterns. In cases where a needed climatic parameter was missing, records were interpolated, utilizing nearby locations of similar altitudes.1 Although measurements of precipitation were available for all stations, mean monthly temperatures had not always been recorded. Even though temperature values required estimation, the small ranges in Northeast Africa, in addition to temperature's close relationship with elevation, greatly reduce the margin of error in these estimations. Interpolation was 1In emphasizing the importance of altitude and climate, Carter (1954, p. 46) states that "The topographic map is to some degree a guide to interpolation among cli- matic values." l7 kept to a minimum insofar as no gross generalizations have been based on estimated data only. Lack of data also causes some speculation on the values of soil moisture storage, an essential ingredient for water balance calculations. In this study, soil mois— ture is estimated to reach a maximum storage of 300mm, although some observations have shown this figure to be much too high.1 However, the selection of the maximum storage value effects only the magnitudes of the annual moisture deficit and surplus, and not the moisture index. This statement may be explained through examination of the Thornthwaite method of water balance accounting. The basic scheme requires that the sum of all monthly changes in soil moisture must equal zero. Thus, no matter what the maximum storage value at a given place, the net annual value of water supplied from the soil moisture for evapo— transpiration is zero. Water deficits (D) and water surpluses (S) are computed in terms of potential evapo- transpiration (PE), precipitation (P), and the change in 1Thompson (1943) observes that soils have a low permeability in most areas. After rains and resultant surface flooding, moisture was not found at depths greater than 6 inches (150mm). Mud from surface floods would flow into the interstices and the sun would bake these surfaces into even more impenetrable barriers. Although other studies support the fact that there is little penetration in areas of surface floods, an increase in obstacles to the floods greatly improves soil moisture storage. Studies show that in areas of vegetation, the plant life may impede the massive surface runoff so that amounts well over 300mm can be stored (Glover, 1950). 18 soil moisture (Ast). Since the net annual Ast is zero, this term may be omitted in an equation which summarizes the yearly water balance relationships: S-D = P-PE If both sides are now divided by PE and multiplied by 100, we obtain 100(S-D) — .13.- _ PE — 100[PE 1] _ I m Therefore, for any given P and PE, the moisture index (Im) is a constant. The only effect of the maximum soil moisture level on the computation of S and D is that these values will vary inversely with a change in the maximum storage selected, i.e., an increase in the maxi- mum soil moisture level will result in a decrease in the amounts of S and D, but the difference between S and D will remain the same.1 Moisture Need The distribution of potential evapotranspiration (moisture need) in Northeast Africa (Map 6) displays a close relationship to topography. Regions of higher ele- vation generally have lower mean temperatures and thus lIn the original formula, 1m = lOO(S-.6D)/PE, the maximum soil moisture storage would obviously affect the index value. Since S and D vary with each different Inaximum storage level, the 0.6 factor would, of course, change the resultant value of the moisture index. 19 lower potential evapotranspiration values. The lowest values are found in the Ethiopian highlands. Additional examples of the effect of altitude on potential evapo- transpiration occur in the highland parts of the northern Somali Republic and the mountainous areas at Moyale and Marsabit in North Central Keyna. Conversely, the lower elevations generally show a marked increase in moisture need. The upper valleys of the Shebeli and Juba river systems are striking examples of this phenomenon.1 An increase in moisture need is also obvious in the rift valley as altitudes decrease. Notable exceptions to the elevation—potential evapotranspiration relationship occur along the Indian Ocean coast. There are two phenomena along the coast that produce cooler mean temperatures than expected. In the lower valleys of the Juba and Shebeli, the surface area of these rivers is expanded and actual evaporation increases immensely. The large marshy areas in the lower river valleys also add great amounts of moisture to the air through transpiration. Evapotranspiration (1) reduces surrounding air temperatures by the use of specific heat, (2) increases relative humidity, (3) increases cloudiness and fog, and (4) results in reduced moisture need. The 1The highest station value of potential evapo— transpiration is at Lugh Ferrandi (76) in the Juba river valley. The PE value here is an incredible 2067mm. v... I 5. I. " 't Ho- I, l ! .-.' (J "N. t. '\._ ‘t .\ —5. \ I ‘ “.-.A / - \ “ w . / ‘e 1"" ,2 . ~ a} ................... .................. \// , ' 7’//// ., .//- e... a... ----- --n. ..... .... ----- s... ..... //’% ///.//‘. [III/II, A //////’. -’///////{\\., .\ /‘ c‘ \\\ \\\\\\ \ \\\\/ w W // \\\\\\ \ \ 45- 1 Moisture Need Annual Potential Evapotranspiration \\\< \\ C/ 175.0 c... I425 I l4.o 95.5 57.0 / N UD§I§I SSSSS 6?. “EC '7' l i- ...‘y i .4 ., ”ms”- .1— Eg_‘ .4 21 second influence along the coast is the cold ocean current and its cooling effect on coastal temperatures. The most distinct example is at Guardafui on the tip of the Horn where potential evapotranspiration is comparable to inland stations of over 2000 feet above sea level. Moisture Surplus When precipitation exceeds evapotranspiration, the excess moisture is first used to bring the soil moisture storage to its maximum level. After this level is reached, the remaining excess water is considered moisture surplus (Map 7). Because of the low amounts of precipitation throughout the region, soil moisture rarely reaches its maximum in the winter months, even at the wettest of stations (Carter, 1954). In fact, large moisture deficits often occur in locales where, during the summer months, moisture surpluses are recorded.1 Moisture surplus in Northeast Africa is found pri— marily in the higher elevations of Ethiopia. In the plateau regions the distributional pattern of rainfall is an indicator of the location of areas where the soil moisture level is exceeded. Such areas are present in the Southern plateau, but not in the Northern plateau and Ogaden. Only in highland parts of the Northern Somali 1As an example, Bishoftu (27) has a relatively large surplus of 216mm and an even larger deficit of 307mm. 22 (0‘4 *4? MAP 7 Moisture Surplus w n um I" 4- 00-1 Ansel Islstws Quotes 0000 ‘ I . a 7/ Am // -O' Besed on map in Catt. IOBQ. Ieetlsel 6}“ use ’7: 4" I 23 Republic are surpluses evident. The only other significant area where the amount of precipitation surpasses the water holding capacity of the soil is along the coast of Kenya. This coastal region is decidedly unique, for it is the only area of moisture surplus in Northeast Africa that is not related to high altitude. Moisture Deficit Like other climatic parameters in Northeast Africa, moisture deficit (Map 8) is to a large degree a reflection of the topography. Low highland values and high coastal values form the normal distributional pattern, but there are a few notable exceptions. At Guardafui, the upwelling of the cold sea water tends to reduce the moisture need and, in turn, the mois- ture deficit. A second major exception to the high coastal values is in the Lamu area of Kenya. The high amount of rainfall here greatly reduces the difference between potential and actual evapotranspiration. North of Lamu, towards the interior, rainfall de— creases rapidly, and the moisture deficit exceeds 1500mm. The rapid evaporation of water from the rivers that flow through this interior area is the cause of many rivers nevEer reaching the sea. However, the river water supplied by eevaporation to the water balance results in higher acthal evapotranspiration and lower deficit values near the rivers. 24 ....... nnnnnn ssssss eeeeeee ssssss 'x \ . ‘‘‘‘‘‘ Moisture Deficit nnual Moisture Deficit = ISO cm l20 "‘-. m ‘x ‘ so "\ u\\\\ i 60 "x. 30 2‘. I g Io Scale Based on map in Carter, l954. ° 5° '00 20° 30° NOUNCOI miles 35' “0° 45' 50°. HEC 1 1 1 1 25 The moisture deficit approaches zero only in the highlands. At all stations used in this study there was 1 a deficit during at least one month of the year. Moisture Index The map of the moisture index (Map 9) shows that a vast part of Northeast Africa is within an arid climate. The result of the revision of the Thornthwaite classifi- cation (Carter and Mather, 1966) reveals that the arid climates occupy an even more extensive area than under the original classification. In addition to the large arid region surrounding the Ethiopian highlands, there is one small area of arid climate located in the Lake Rudolf region of Southern EthiOpia and Northern Kenya. The presence of the arid climate means that Lake Rudolf is dependent upon massive inputs of water from distant sources for its very existence. The semiarid climate is found primarily encircling the Ethiopian highlands. Elevations of the semiarid cli- Inate boundary range from about 1500 feet in the south and Inorth to about 4000 feet in the east. The low values of ervapotranspiration in the highlands of the Northern Somali Republic result in'isolated areas of semiarid climate. ”PM": patches of semiaridity are found in the lower valleys <>f5 the Shebeli and Juba rivers where evapotranspiration The range of moisture deficit in the study area Was from 15mm at Cencia (30) to 1857m at Eil (8) . 0. Based on area In Carter. ”64. 26 e- Moisturo Index Index Climatic type \/ IOO - ”mm \;,-._- Hamid Moist when“ Dry sum“ mum -..'. D Arid '33.! Deals 0 so too 100 300 1::— ttattoet “so 5'4 27 values are reduced by the effect of river evaporation. The relatively high rainfall amounts in the Southern Somali Republic and Kenya also help to produce a semiarid climate. The dry subhumid climate is generally confined to the Ethiopian highlands. Minor areas also occur along the Southern Kenya coast where rainfall is more plentiful. The rift valley in Ethiopia and the higher mountains in Kenya also have some areas of the dry subhumid climate. Elevation—Moisture Index Relationship Throughout this study, the importance of elevation in relationship to climatic factors has been stressed. The effect of height upon the moisture index is seen in Figure 2. Arid climates are found from sea level up to 4500 feet. The semiarid areas range from about 400 to 5600 feet, and the dry subhumid climates are found be— tween 2500 and 7000 feet. Humid climates occur only above 5000 feet.1 Most of the points on the graph above the regres- sion line represent stations on the drier, eastern side of the Ethiopian highlands. Below the regression line, the points indicate stations which are primarily on the ‘windward western side of the highlands. Some are coastal ‘ 1The only station in Appendix I that is not on 'this graph is Lamu, which is on the Kenya coast. At this location, elevation is decidedly secondary to other cli- Inatic variables. 28 2 Figure _.-_.l 6 .0: .0 3.3.3:... .m . 5:265 I00 20 Index, ‘66.6 '33.?» ‘I00 'm Moisture 29 sites where elvation plays a smaller role in influencing the climate. Examples where altitude seems to have little effect upon climate serve to caution against strict ad- herence to elevation as a guide to climatic patterns. Cases which illustrate this fact are in the lower Juba and Shebeli valleys, where the Baidoa plain has a moister climate than the surrounding areas of higher altitude. An understanding of the circulation pattern and local factors affecting climate is thus a necessity in explain- ing exceptions in the relationship between climate and elevation. However, with these qualifications in mind, the altitude—climate correlation in Northeast Africa is quite strong, and the topographic map is perhaps the most valuable tool in determining unknown values of the regional water balance. III. LAND USE PATTERNS, NON-CLIMATIC FACTORS, AND THE REGIONAL WATER BALANCE Bases for Land Use Data and Definitions The information on land use activities in North- east Africa is scarce and existing land use maps are drawn at relatively small scales. Map 10 is a compilation of several descriptions and various land use maps. The dichotomous choice of non-irrigated agriculture and herding as the main economic pursuits in Northeast Africa greatly simplifies the land use pattern of the region. In addition to the obvious problem of generali— zation, there are also difficulties which are derived from trying to define herding and non-irrigated agricultural lands. Herding and agriculture are sometimes practiced in the same area during the year. Climatic variability may annually affect the agricultural land use pattern, and livestock numbers can fluctuate widely from year to year depending on rainfall and the availability of water. How— ever, the long-term land use patterns which were compiled are on such a scale that annual variations are usually not ‘ 1The sources for land use data include Mariam (1962), Fullard (ed., 1962), Huffnagel (1961), Kaplan (1969), Hunt (1951), and FAQ (1968). 30 31 significant. Since the climatic parameters are based on averages, comparing climate to a particular annual land use pattern would not serve the purposes of this research. The selection of "non-irrigated agriculture" as a land use type, rather than any form of agriculture, was based on the fact that irrigation water, generally, is not water produced by local climatic conditions. In the case of Northeast Africa, irrigation water is often a product of precipitation which has fallen hundreds of miles away. Irrigation plays a large role in producing the small amount of agriculture in the Somali Republic. Subsistence, as well as plantation agriculture, is prac— ticed in the riverine areas where the presence of the tsetse fly makes it difficult to raise livestock (Karp, 1960). The Somali government has utilized United Nations assistance to further develop the country's irrigated agriculture between the Shebeli and Juba rivers (Kaplan, 1969). However, farming in this area is almost totally dependent upon runoff from outside the Somali Republic. At present, there is no international agreement on the use of the two large rivers originating in Ethiopia (Mariam, 1964). Any utilization of the headwaters of these two rivers by Ethiopia would severely curtail development, or even destroy, the present land use pat— tern in the inter—river area. 32 Another major area of irrigated agriculture is located in the lower Awash river valley (Huffnagel, 1961). Agricultural enterprises in this region almost totally rely upon water produced by precipitation in the Ethiopian highlands. Thus, for the purposes of this study, non—irrigated agricultural land is considered land outside the lower Shebeli, Juba and Awash river systems. Herding is assumed to take place wherever agriculture is not present, although some places may not even be suitable for nomadic activity. Non—climatic Influences on the Land Use Pattern The Impact of Culture The most prominent cultural division in Northeast Africa is between the Islamic nomads, called Somalis, and the other peoples (primarily Galla and Afar) of the region. The cultural influences on land use in the Somali—inhabited areas are readily apparent when contrasted with most non- Somali lands. For almost all Somalis, the common denominator in their lives is the raising of livestock. Wealth and status are reckoned in stock, and most cultivators have some animals (Lewis, 1955). Agriculture has always been an activity essentially subsidiary to herding within the Somali perceptual scheme (Karp, 1960). Agricultural work ‘was left to the slaves or dependents of the Somali, and 33 land cultivation has always been performed by low-class citizens. From the perspective of the tradition-oriented Somali, Somali lineages that have taken up agriculture have lost prestige, even if they have gained economically (Kaplan, 1969). The image of the nomadic herdsman and warrior has been held in high esteem by the Somalis. Such reverence has led to constant conflict between the Somalis and neighboring peoples. After their arrival in Northeast Africa, from Arabia between the seventh and thirteenth centuries (Lewis, 1955), the Somalis continued their per— sistent expansion southward from the Northern Somali coast. The Somali warriors played an important role in displacing the Galla peoples to the west and south. The Somali expansionism resulted in the migration of many Galla tribes into regions which were moister than those which they had inhabited. Most of the Galla tribes adOpted agriculture as a primary occupation in these wetter areas, while the Somalis retained their proud heritage of nomadic herding. Considering the present distribution of Somali tribes, the Somalis seem to have preferred herding their livestock in the dry habitat to which they were accustomed. Although most of the Galla tribes practice agri— culture, the Boran Galla of Southern Ethiopia and Kenya display cultural characteristics similar to Somalis. With the exception of a few semi-permanent farmers on Marsabit 34 Mountain and near Moyale, the Boran do not cultivate (Baxter, 1954). Even in those places where the soil and climate are comparable to other agriculturally developed areas, the Boran, like the Somali, shows a definite prefer- ence for nomadic herding. The northern extent of the Boran almost exactly coincides with the division between herding and agricultural land (Map 10). The territory is bounded by the Somalis on the east, and the western bound- ary of the study area approximates the extent westward of the Boran. Recently, changes have occurred in the traditional Somali disdain for agriculture. The ethnic distinction that once prevailed between cultivators and herders seems to be disappearing to some extent (Kaplan, 1969). Perhaps the best examples are the agriculturalists in the areas of Hargeisa and Borama in the Northern Somali Republic. The clan-families of these Somalis were formerly of a nomadic tradition (Kaplan, 1969), but they have now chosen to practice dry land farming. Nevertheless, the majority of the Somalis still prefer nomadic herding as a way of life. The cultural milieu undoubtedly influences the Somalis' perception of their arid environment, and any change in land use, even in the wetter areas, will require corresponding changes in social values and customs. 35 Political Factors in Somali Areas The rise of nationalism in the developing nations has had its effect on the land use pattern of Northeast Africa. Following World War II, nationalist fervor led Ethiopia to claim vast areas of land which had formerly been the domain of nomadic Somali herders. Ethiopia's claims had historic bases, but the government of H.I.M. Haile Selassie had rarely administered previously in the Somali—inhabited areas (Touval, 1963). By the time the Somali Republic received its independence in 1960, skir- mishes had occurred to protect the rights of the Somalis in Ethiopian territory. The threat of war caused the authorities to restrict movement across the international boundaries. Such restriction caused hardships to the Somali nomads who annually crossed and recrossed the border in search of pasture and water. Many of the no— mads decided to forsake their transhumant life and wait for better times. They remained on one side of the border or the other with their flocks and, in the more favorable climatic areas, many began raising crops to supplement their basic milk diet. The easing of tensions in late 1967 was not followed by a mass return by the Somali cultivators to their former nomadic way of life. The political problems of Northeast Africa have thus furthered the decline in the pastoral nomadism and increased the use of cultivation to provide food for the Somalis. 36 An impetus to the rise of agriculture has also been the Somali government's development program (Kaplan, 1969). Somali requests for United Nations assistance have resulted in several worthwhile recommendations. Most of the United Nations effort has been concentrated on develOp— ing agriculture along the Juba and Shebeli rivers in con- junction with flood control and irrigation systems. The Somali government has reacted favorably to most agricul- tural development schemes. A statement in a United Nations report on Somalia (FAO, 1968) describes the Somali attitude: The need for government participation in a program for raising production of traditional agriculture has been recognized, and the Ministry of Agriculture has announced plans to form an extension service, to introduce land tenure legislation, to carry out re- search, to make credit available to cultivators, and to improve marketing. Such political pressures to develop agriculture should result in better utilization of available land. The under—utilization of land has been a serious problem in some parts of the country (Karp, 1960), and govern- mental assistance can help make agriculture more eco- nomically attractive to the nomadic Somali. The chang- ing land use patterns in the valleys of the Juba and Shebeli probably are evidence that development programs are having their effect (FAO, 1968). If the Somali Republic is successful in its effort to stimulate agricultural activity, it is likely that Somalis outside the country's borders will reassess 37 the value of agriculture in their traditional economic structure. Agricultural techniques which proved profit— able would be adapted to new areas, and the land use pattern of Northeast Africa could change drastically. Naturally, much remains to be done, and the development schemes along the two main rivers of the country are only a beginning to the development work needed to alter the present land use scheme of Northeast Africa. Spatial Interrelationships between Land Use Patterns, Ethnic Areas, and the Regional Water Balance in Northeast Africa The pattern of land use in Northeast Africa is obviously a reflection of the decisions of the inhabitants of the region. Certain factors are considered before peOple decide upon the activities which will occur in a given place. In the case of the area under study, the primary consideration is the availability of water. Super- imposed upon the framework of the regional water balance are the cultural and political factors which influence decision—making processes and thus the spatial arrange- ment of land use activities. Analysis of the interrelationships of land use, the ethnic areas, and the regional water balance in Northeast Africa reveals several discernible patterns. From these patterns, a territory may be described which fulfills the aims of this research, namely, a critical climatic zone where lands on either side tend to favor 38 the development of either agriculture or herding activities. Referring to Map 10, the first of these spatial interrelationships is that agriculture rarely occurs in areas with a moisture index below —66.6. Both the Somalis and the non—Somalis of Northeast Africa seem to perceive the semiarid/arid boundary as the limit below which agri- culture is impractical. The only exceptions are found in the irrigated regions along the Shebeli, Juba and Awash rivers, where agriculture is not based on locally produced water. Excluding the irrigated lands, little agriculture is found in Somali-inhabited parts of Northeast Africa. However, the Somali agriculture which does occur is ex— tremely important. Two isolated areas of dry land farm- ing, where the water balance is more favorable than the surrounding region, are found in (l) the Baidoa Plain between Iscia Baidoa and Bur Acaba, and (2) the area south- west of Mahaddei Ven. The more favorable water availa— bility, the government support of agricultural develop— ment, and the closer Somali ties to land combine to fur— ther the development of agriculture in these areas. Other significant Somali agricultural regions are found around Diredawa and Jijiga in Ethiopia, and between Borama and Hargeisa in the Northern Somali Republic. The factors which have given rise to cultivation in these areas are 39 35' 45° 45' 5‘0- 20% ~20- MAP IO . ,‘3": Land Use, Critical Climatic 5' Q, Zone, and Somali Regions .m- 5' of Northeast Africa "‘ i ‘ I I0'-i -s- -0' i-O’ tribes Monument extent at Somali Critical Climatic Zens -Moisture Index between a as -se.s Land Use Type [I] tlon-imgeted egrtosItsre C: Herding O SOIOO 200 I:— tleetloet mites rd 50° 4 46' J HEO 'TI 40 (1) changes in cultural attitudes, (2) the restrictions on movement across the political boundary, and (3) the relatively high moisture availability. One of the reasons for the dearth of agriculture in Somali—inhabited regions is undoubtedly the high degree of aridity. In fact, most Somalis reside in parts of Northeast Africa with a moisture index below —33.3 (the dry subhumid/semiarid boundary). Notable exceptions occur near Diredawa, Jijiga, and Borama--the same areas where agriculture is also found. A region where the moisture index is well above -33.3, but where the Somalis practice little cultivation, is near Daloh in the north central Somali Republic. Here, the Somalis seemingly prefer to utilize their lands for herding activities rather than attempt to raise crOps. In contrast to the lack of agriculture in Somali regions, non—Somali lands are predominantly used for agri- culture or agriculture and livestock~raising. A major case which does not fit this pattern is in Southern Ethiopia and Kenya where the tribal boundary of the Boran Galla almost coincides with the boundary between agricultural lands and herding areas. The Boran Galla seem to have an even greater disdain for agriculture than the Somali, for some of their herding areas are found in regions with a positive moisture index. The second major exception of non-Somali herding areas in Northeast Africa 41 is found in the northern part of the study area. This region extends from the western side of the rift valley to the Red Sea and is inhabited by the Afar (Danakil). Parts of this region are so devoid of water than even nomadic herding is a difficult occupation. Other non—Somali lands which are not agricultural are located adjacent to Somali— inhabited regions with a low moisture index. The non- Somali areas of Northeast Africa have a wide range of moisture index values, so it is difficult to draw any conclusions about interrelationships between these two spatial features. The critical climatic zone sought in this paper becomes apparent when all three of the variables discussed are correlated. The earlier observation that little agri- culture is found below a moisture index of -66.6 means that both Somalis and non-Somalis prefer not to attempt cultivation below that level of water availability. The arid-semiarid boundary thus forms a good choice for the lower limit of the critical climatic zone. However, the upper limit is not so easily discerned. With the ex- ception of the Boran Galla mentioned previously, non— Somali peoples always practice agriculture at a moisture index of -33.3 or above. On the other hand, the Somalis do not always use land for agriculture in areas where the moisture index approaches a positive value. There— fore, the limit of aridity as established by Thornthwaite 42 (Im=0) is perhaps the best selection for the upper boundary of the critical zone. Within this zone, the interrelationships between water availability and non- climatic factors combine to influence the inhabitants' perception of their environment and the resultant pattern of land use. IV. CONCLUSION Certain aspects of the Thornthwaite system are open to criticism (see, e.g., Chang, 1968); however, there is adequate evidence derived from this study that the water balance method is a useful vehicle for evaluating the effect of climate in a regional setting. The corres— pondence between areas of a certain moisture index and the location of specified economic activities provides an understanding of the relationship between climate and land use. The arid regions of the world pose special prob- lems, such as climatic variability, which are not con- sidered within the scheme of the climatic classification devised by Thornthwaite. This fact was recognized (Thorn— thwaite, 1956) and methods were proposed to assess the annual variability of precipitation using water balance information. Unfortunately, the data are not yet avail— able for obtaining an accurate assessment of the relia- bility of precipitation in Northeast Africa. However, in this study it has been assumed that the long term land use pattern is largely the result of climatic norms. 43 44 Thus, the factor of climatic variability should not affect the conclusions derived from this research nor the rele— vance of the Thornthwaite water balance method for deter— mining water availabilities. Perhaps the greatest usefulness of the water balance method of Thornthwaite is found in its basic reliance on only values of temperature and precipitation. Since these parameters are the most commonly recorded climatic characteristics, the system clearly has advantages over schemes requiring a multitude of input data. Even though the use of the water balance method may result in an accurate evaluation of climate, its util— ity in analyzing land use patterns is limited by the in- herent faults of viewing areal phenomena as products of single factors. The relationship between climate and land use is often modified by several influences upon the minds of the individuals who make the decisions concerning land use activities. In Northeast Africa, the close relationship be- tween climate and land use is affected by nonclimatic influences. The division between herding and non-irrigated agricultural activities has as its primary basis the de— gree of aridity as determined by the Thornthwaite method. There may be defined a critical climatic zone in which the division between land use activities are viewed within a framework of differing moisture index values. Anomalies 45 in the general correlation between the patterns of climate and land use usually can be explained through an investi- gation of the political and cultural factors operative in the region. The consideration of non—climatic influences will naturally provide a more thorough understanding of the regional interrelationships and the resultant land use pattern. BI BL IOGRAPHY BIBLIOGRAPHY Baxter, P. T. W. 1954. "The Social Organization of the Galla of Northern Kenya." Lincoln College, Oxford, Ph.D. Thesis. Brooks, C. E. P. 1960. "The Meteorology of British Somaliland." Quarterly Journal of the Royal Meteorological Society, v. 46, No. 196, October. Carter, D. B. 1954. "Climates of Africa and India According to Thornthwaite's 1948 Classification." Publications in Climatology, v. VII, No. 4, Johns Hopkins University. and Mather, J. R. 1966. "Climatic Classifi- cation for Environmental Biology." Publications in Climatology: v. XIX, No. 4, Elmer, New Jersey. Chang, J. 1959. "An Evaluation of the 1948 Thornthwaite Classification." Annals, Association of American Geographers, v. 49, pp. 24-30. . 1968. Climate and Agriculture. Chicago. de Martonne, E. 1927. "Regions of Interior Basin Drainage." Geographical Review, v. XVII, pp. 397-414. Flohn, H. 1964. "Uber die Ursachen der Ariditat Nordost— Afrikas." Wfirzburger G. Arb., v. 12, pp. 25-41. Food and Agriculture Organization (FAO) of the United Nations. 1968. Somalia: Agricultural and Water Surveys. Final Report, United Nations. ' Fullard, H., ed. 1967. Philips' Modern College Atlas " for Africa. Eighth Edition, London. Garnier, B. J. 1956. "A Method of Computing Potential Evapotranspiration in West Africa." Bulletin de l'Institut Francais d'Afrique Noire, v. 18, Dakar, pp. 665-676. 46 47 Glover, P. E. 1950. "Rain-water Penetration in British Somaliland Soils." East African Agricultural Journal, v. 16, pp. 26-33. Huffnagel, H. D. 1961. Agriculture in Ethiopia. Food and Agriculture Organization of the United Nations, Rome. Hunt, J. A. 1951. A General Survey of the Somaliland Protectorate, 1944-50. London. Kaplan, I. et a1. 1969. Area Handbook for Somalia. Washington, D.C. Karp, M. 1960. The Economics of Trusteeship in Somalia. Boston University Press. Kendrew, W. G. 1961. The Climates of the Continents. Oxford University Press. Koeppen, W. 1931. Grundriss der Klimakunde. Berlin. Lang, R. 1920. Verwitterung und Bodenbildung als Eihfuhrung in die Bodenkunde. Stuttgart. Lewis, I. M. 1955. Peoples of the Horn of Africa. London. Mariam, M. W. 1962. Preliminary Atlas of Ethiopia. University College, Addis Ababa. . 1964. The Background of the Ethio—Somalia Boundary_Dispute. Addis Ababa. Penck, A. 1910. "Versuch einer Klimaklassifikation auf Physiogeographischer Grundlage." Sitzungsberichte der Konigliphen Preussichen Akademie der Wissen- schaften, Berlin, pp. 236-246. Sibbons, J. L. H. 1962. "A Contribution to the Study of Potential Evapotranspiration." Geografiska Annaler, v. 44, pp. 279-292. Tato, K. 1964. "Rainfall in Ethiopia." Ethippian Geo- graphical Journal, v. 2, No. 2, December, pp. 28—36. Thompson, A. B. 1943. "The Water Supply of British Somaliland." Geographical Journal, v. CI, pp. 154-160. Thornthwaite, C. W. 1948. "An Approach Toward a Rational Classification of Climate." Geographical Review, v. 38, pp. 55-94. 48 Thornthwaite, C. W. 1956. "Climatology in Arid Zone Research." The Future of Arid Lands, American Academy for the Advancement of Science, pp. 67-84. and Mather, J. R. 1954. "Estimating Soil Tractionability from Climatic Data, Final Report." Publications in Climatolggy, v. VII, No. 3, Centerton, New Jersey. . 1957. "Instructions and Tables for Computing Potential Evapotranspiration and the Water Balance." Publicatipns in Climatology, v. X, No. 3, Center- ton, New Jersey. . 1955. "The Water Balance." Publications in Climatology, v. VIII, No. l, Centerton, New Jersey. Thornthwaite Associates. Laboratory of Climatology. 1962. "Average Climatic Water Balance Data of the Conti- nents, Part 1, Africa." Publipations in Climatology, v. XV, No. 2, Centerton, New Jersey. Touval, S. 1963. Somali Nationalism. Harvard University Press. Trewartha, G. R. 1960. The Earph's Problem Climates. University of Wisconsin Press, Madison. Walton, K. 1969. The Arid Zones. Chicago. ‘xi \V' Wernstedt, F. L. 1959. World Climatic Data, Africa, v. 1, Pennsylvania State University. APPENDIX 49 m .m m m o ummoomm moa.oo Zao.m o.mo- o omoa mmo omm mama m omega .ma o .m m oa moaa mmo.~v Zoo.aa o.omi o oama ooa ooa ooma om mammmmm .oa m .m a o o amo mov.ao zma.aa o.mm- o omo oao oao omma a muoao .mm N onwmd o .4 o oo m om mom.oo mma.~ «.mm- oa oao mam mam mama a mama .mm o .m m m m on mam.~o mm~.o m.mmi o omoa mmm mmm mooa m smegmam .am o .m m o o om mom.~o moa.o o.ooi o mmoa mom «on omaa a onsoao .om m .m m m m ma mom.~o moo.o o.ooi o mama ooo ooo omoa m mucoa .oa o .a m m m Nmm moo.~o zm~.o m.von o moma moo mmo ooma m smegma .ma m .m m o o mo moo.~o Zom.o m.mou o mmma an no mmma a muocmmmmam .oa m .4 m o o mm moo.ov zoo.a a.mou o mmaa aam aam omma a m>mum .ma o .m m o o mam m~v.om 2ov.a m.mml o maoa mm mm mmoa a mamcuo .ma m .m m o o mm mam.mo z~o.m a.mo- o mama moo moo amma a oaomaomooz .oa m .a m o o Nmm moo.mo Zoa.~ o.mmi o moaa mmm mmm mooa a aommm .ma o .a m m m amm www.mo z-.~ o.omi o mama omm omm ooma m mmamm .ma o .m o o o mom mom.mo zoo.~ v.mmi o omaa mom mom Nona a mono oamomaaa> .aa o .m m o o moo mo~.mo zmo.~ o.mmi o mama moo moo coma a cm> amoomnmz .oa o .m m m m mo mam.mo zm~.m a.mm- o nova mma mma amma m mango .m m .m m o o oaa moo.mo 2mm.o m.omi o omma o o amma a aam .m o .m m o o who moa.am zoo.aa m.omi o Nova a a mooa a aommoumoo .o m .m m m m mo mmo.om zmm.aa m.mmu o omoa oa ma mmma m maoam .o m .m m o o mm moa.mo Zea.aa o.mo- o Nmoa N N omoa m aammmo amoomm .m o .a m a oa mm mao.mv 2m~.oa m.mmi o omna mm mm mama mm mumoumm .v o .4 m o o mm mom.mv za~.aa o.mo- o mama so so oooa m oaamm .m o .a m mm ma mm maa.mo 2mm.aa m.mmi o mmoa mma mma oama m apoooaao .m o .m m o o mm mmv.~o Zao.ma m.moi o voma so on aoma a gamma .a a zOHomm age m .H. boom coaumooa xoch EE Be as E: E8 mama—om vaumawau mums» command/mam . m 0 mm m mm a xaozmmma 50 p .4 m b omvm mvo.mv zvm.m m.Nwl o mmm mom mom mmaa mm mcavao .vm v .4 m b b omvm mam.mv Zam.m w.bwl o omoa oaa ova omaa m omusm .mm c .4 m m 00mm mmm.m¢ zmm.m m.le o Nona mm mm amma am mopMHmo .Nm 0 .4 m h amom mmN.bv zmo.m mumml o hmva aoa aoa mmma xm dedm .am U .4 m h mmom www.mv Zmo.m o.mml o moma om om mmma xm Smame Acaammv .0m 0 .4 m m o omvm Mmo.mv sz.m m.oml o NMNa mma mNa amma m ovumw .mv v .4 m m m Nmaa mva.om zva.oa 0.0m! o mmma Mb mu mama m canaumfiom .mv U .4 m m ovam www.mv 2ma.oa b.0ml o hmaa hNa hma moma mm cmudm .hv O .4 m 5 04mm mma.®v 2mm.m m.mml o Nbaa mma oma mema mm aam3m4 am .ov U .4 m b momm mba.wv zwm.m a.aml 0 Ohm mmm mmm mmaa mm amam .mv n .4 m m m ooma mmo.mv zoa.oa m.mml o Nmma mmm mmm omma m Gonam £0mm4 when: .vv 4 ZOHomm 3.mmNU m ombo Mma.h¢ zbv.oa h.N I vaa hma man mmm mmm am noamo .mv w.vm Q m m @Nbv mma.mv zmm.m N.le o Nam bmv bmv mmoa mm fixamnm .Nv U.vm a h h oomv mmo.¢v Zam.m m.m@l o rho man man mooa m Mmammumm .av U.vm Q h omhv mum.mv sz.m o.mml o mam mvv mvw mmoa 3m manna maanwu .ov p.vm m h ovmv mwo.mv zNa.oa h.m®l o mvb vmm vmm mooa 3m c3mm .mm U. 4 m m ammooov maa.vv zwa.m m.ohl o aam vmm vmm mmaa mm wum34 .mm U.Nmau m m ummoomm mom.mv ZNN.m m.oml 0 0mm mum mbm amm m moanab .hm ©.mmNU ® ® mmam mbO.Nv Zma.m v.0 + mm am aam mmm mom 4 umnmm .mm 0. 4 a 9 ® ommm mam.av zhm.m «.mol 0 mmoa mm@ mm@ mfloa 4 m3mvouao .mm @.Nm Q m umoommm mbv.ov 2mm.@ o.mml o mmm com com mmm mm Oammmz .vm H.amNU m m ummooam mom.mm Zao.h m.m + boa mv mwm Nmm one Q mnow .mm p.mmao v ummooom mom.mm zom.b m.mal o ova man man mam om mummsx .Nm H.mmmo m ummoooo mo~.mm 2mm.m m.ma+ oma mm mum mmoa mam mm maumaom .am H.Nmmm v ummoomb mov.hm zba.m b.Nh+ mom ma mMB Noma vmh pm Maucmo .om m Zanmm M mmmm w v uwmooon ubv.hm zom.h o.mm+ om? mm nor mmaa mmh m MGaMmom .mm 6 mm a «N ammommm www.mm me.h h.mml o mmm mmm mmm 0mm om Sada afimw4 .mm N3 Nmao ® 9 ummoomm mao.mm va.m N.oal mam hon mmm 0mm 0mm 4 suwonmam .hN H wmmm hm hm Nmom mmv.mm zNo.m o.mw+ 40m mm Mbw hmma NMh m mnnmn4 mav©4 .om 51 o .4 4 ma m 4om mmm.mm mma.o 4.mmi o mmma mom mOm 4mma 4 mmmaamo .ma o .4 m m m omm mma.a4 zaa.a a.mm- o mama mmm mma mama m mumoamm .aa o .4 m m m amm mom.a4 zm4.m m.ami o aaaa mmm mmm mmoa m anomauma gooa .ma 4 .4 m o o ovm mmo.44 2mo.a 4.omi o maaa omm omm mmma 4 mnoo4 Hum .ma 4 .4 m m m mmo mam.m4 zao.m m.omu o mmma omm mmm aoma m auaom oasm .4a m .4 m m m 4mm maa.m4 zaoao. 4.mm- o amma mma mma oama 4 coo umamm .ma o .4 m m m mam mom.m4 za4.4 a.4oi o 4mma 4oa 4oa mmma 4 “am am .aa 4 .4 m m m 4mm mma.ao zoo.m m.4ou o «oaa 4m 4m mmma m camoamo .aa m zeaomm o .4 o, o o 4mma mao.m4 Zao.m a.mmi o mam mmm mmm mmma 4 mooamm maoma .oa o .4 o m m omoa mo4.m4 zoo.m, a.moi o maoa mom mom omma m mxoom mooamm .mm o .4 m m m mama mom.m4 zmo.4 m.aai o moma mam mam mmma m “ammo .mm o .4 m a a aooa mam.a4 zam.m a.mm- o mama aoa moa aooa 4 «noose: .mm m. 4 4 ma oa mom moo.o4 2m4.a m.mmi o oama maa maa mmaa 4 names .mm o.mmao om m ammo mom.am Zoa.a a.aai o aaa mom mom mam 4 uaammamz .mm 4.4m a ma m momm mmo.mm zam.m o.om- o mam mmm mmm oaoa 4 mamaoz .mm o.4m a o o mooa mom.mm zoa.m m.m4- o omo mmo mmo mom 4 aaammmz .mm m ZOHGMm o .4 m o o ammomaa moa.44 zam.m m.mo- o m4ma ma ma 4mma 4 aosmauoo .am o .4 m m oaaa moa.m4 zam.a o.mmi o mmma maa maa 4mma am gonna .am o .4 m m ommooaa mm4.m4 2am.a a.om- o maoa aaa aaa mmma am 0504 .om m .4 m m mava mma.m4 zma.m o.aou o amva aaa aaa mmma am amaam .mm o .4 m a a mama maa.a4 zma.m a.om- o amaa aaa aaa mama m @044 mma .mm m .4 m a mama mma.m4 Zam.m m.om- o 4aaa moa mma omma am onmca4 .mm m .4 m m oaaa mmm.m4 zom.m o.mmi o 4aaa mma mma aoma am nonmmno.ow .mm o .4 m a a mmmm moo.m4 zoo.m 4.mm- o maaa mma mma aoma m ansmso .mm Uwscaucoolt.a xHszmm4 52 conam nomm4 oncmz x mnoo a omuom m umamm o sumocmam o opumw a andpso n mmaomumm : mmanab m acammom o oano 3 .coaumuammcmuuomm>o awaucouom oumEaumo ou pouaaaus mums momma ousumaomsou omonz xapcomm4 on» Casua3 chaumum aonuo ucomoumou a nmsounu n muouuoa one .mm4i4m4 .mm .oaoa .amnouoo .moa .02 .mo .ao> .Numaoom amoamanuouoz ammom ogu mo amcu50h maaouumoa =.pcmaaamsom nmauaum mo umanuoouoz one: .mxooam .m .m .0 m .oousom moon coaumuamaooam Eoum muommap pa ma mump ousumuomaou any mo ooHsOm on» mumoapca muouuoa ammo um3oq .mmma .moamam>aco oumum .4amm .a .ao> .moaaa4 .mumo oaumsaao mayo: .uomumcams .a xoaumomum m l .33 ..o.z .couuouaoo .m .02 .>x .ao> .meaoumEaaU ca mc0aumoaansm :.moaum4 .a pawn .mucocaucoo onu mo madmaom Houmz oaUMEaaU ommuo>4= 4 “mooadom .co>ao uoz u a “oudumummEoB u a auaoamoo madumaoz u a «ceauouammcmauomm>m amsuo4 u m4 “moamasm oasumaoz u m moumfiaumm n umo «coaumuamaooam u m “coaumuammcmuuomm>m amaucouom n mm 53 m.mm am>o moamusm uoEEdm manna 3 m.mm uo>o moamaom nouca3 momma m m.mmin.ma msamuom “meson oumuoUOE 3 m.MMIb.©H WSHQHDm HOHCH3 QHMHOGOE m n.maio moamasm oasumaoe aamEm 0 mm i n m ooa i a am.o.aoo xmpca muapaESm moumsaau who om uo>o uaoamop nouca3 manna m3 ow uo>o uaoawop aoEEdm omnma mm omioa uaoamop Houca3 oumuopoe 3 omioa uaoamoo HoEESm oumaopoe m oaio uaoamop wadumaos aamEm a a mmmma n a aao.m.4wli xopca muapau4 moumfiaau umaoz Hmuuma w a mc0mmo u.a m unane umoam .m mva mucosa .o mmN amEuonuouoaz .au bm¢ aaauonuoaoaz .mu onm amEuonuomoz .am man aoEHoSUOmoz .mm mmm ameuocuOmoz .mm hmm amaaozuomoz .vm EEovaa amEuonummoz .4 ammo mommauaamm mpasaao umuuma anemone pcoomm oaa4 m 5.00: caudanm o m.mmr oassznsm a man 0 O vafidnndm N umaoz o ON a vafinm m 0v N UaEdm m 00 m vafifim m ow mason 4m ooa oassramm 4 xoch oumaaau Manama mHDumHOE . UMHHW ZOHB4UHmHmm4AU m.mBH43mezm0mB OB UZHQmOUU4 mmmwfi UHB4ZHAU