Vegetation patterns in the semi-desert plains of British Somaliland.

During A geological and water supply investigation carried out in 1947-48, I had occasion to study a large number of air photographs of two areas of semi-desert, the Haud, and the Sawl Haud 2 plains of British Somaliland. Many show ground devoid of appreciable interest, scanty or thick bush growing haphazard and without apparent design, though roads and tracks occasionally show lines of denser vegetation, due no doubt to the delayed run-off in ruts. Other photographs show remarkable rhythmic patterns; "vegetation arcs" as I have called them, at first sight reminiscent of the ripple marks produced by blown sand, but on a large scale; and more rarely "water lanes," one striking multiple example of which resembles a gigantic ploughed field. In some places the patterns are simple, but elsewhere interference effects of the two patterns are to be seen. Detailed investigation, both of the photographs and in the field, showed that the patterns are not in fact made by sand but by the rhythmical arrangement of concentrations of vegetation?trees, shrubs, herbs and thick grass? which photograph black against the light background of bare desert surface, usually of red sandy soil, silty sand, fine gravel or dried mud. A third type of pattern, "termitaria peppering," of lesser interest and quite different origin, was found to be due to the abundant growth of termite mounds, which show over limestone outcrops on the photographs. This pattern again is made visible by conditioned growth of the vegetation. The study of the first two patterns, of which I have been able to find no previous recognition, is not conveniently to be pigeon-holed under any one accepted branch of knowledge, and the phenomena are thus awkward to classify. They are manifestly within the province of botany and ecology; the essential background concerns geomorphology and meteorology; the causes, as I believe, must be investigated by physics and mathematics; and the whole matter must be studied on air photographs, since on the ground it proved difficult to recognize the patterns at all. While the superficial deposits are of importance, the underlying solid geology seems to have no particular signifi? cance except in its influence in moulding the pre-requisite geomorphology. The synthesis of these factors is perhaps a problem for geographers. The subject proved to be of more than academic interest when it was found that the vegetation arcs were of considerable help in the water supply investi? gation. They indicate precisely the direction of the very low slopes; and