352 W. H. PEARS ALL 



The average level of the latter will tend to remain below that representing 

 the average amount of food it is possible to take without damaging the food 

 population. The fraction representing the conversion ratio will thus always 

 be less than one-tenth say (if this is the best possible) and the fraction will be 

 smallest where the habitat is in other respects most severe. 



There are, of course, other possible explanations of variations in the 

 conversion ratio. It seems probable that well-adapted animals, well adapted 

 to a particular habitat, would give a larger conversion ratio than unadapted 

 ones. This at least seems to be the implication of Harthoorn's data for buffalo, 

 that they flourish on diets on which domestic animals cannot maintain 

 themselves. This also would explain how it may be that the plains game 

 give as high a stocking intensity in a less favourable habitat as Masai cattle 

 do in a more favourable one. 



But in the latter case we have already seen what may be still a third 

 explanation of a low conversion ratio — the unsuitable and unpalatable 

 character of much of the fire-climax vegetation which has resulted from bad 

 land use. This means that only a fraction of the vegetation is used as food. 



We are thus brought back to what many distinguished conservationists 

 have emphasized as the first necessity of wild-life management — the 

 conservation of the habitat. In semi-arid Africa this is, I believe, first and 

 foremost — the maintenance of a natural vegetation cover or else its replace- 

 ment by one which possesses similar quahties in respect of light absorption, 

 protein synthesis and soil protection — throwing useful shade, maintaining 

 soil structure and the nitrogen cycle, and preventing soil erosion. 



Secondly, if we are to use animals for protein harvesting, we shall need 

 suitably adapted species, even if they are wild, in order to get a high yield. 

 The criterion of suitability is adaptability in that particular habitat and not 

 herdability or any of the other traits characteristic of many domesticated 

 breeds. 



In conclusion, I think the greatest need is for the fullest possible study of 

 the eco-systems in these somewhat arid habitats. The ecology of 'bush' and 

 the neighbouring seasonal grasslands is practically unknown in spite of its 

 quite reasonable natural production. No satisfactory form of land use for it 

 by man has been devised. A study of the natural biological potential would 

 undoubtedly be of the greatest value both from scientific and economic 

 points of view. Here is the real reason for national parks and for a trained 

 biological service to maintain them and study them. 



REFERENCES 



Bere, R. M. (1959). Queen Elizabeth National Park, Uganda: The hippopotamus problem and 



experiment. Oryx^ 5, 116. 

 Deevey, E. S. (1956), The human crop. Sci. Amer., 182, 58-60. 



