PLANT SUCCESSION'. 



239 



•effects of grass fires. The eastern slopes of the valleys under con- 

 sideration in this paper are considerably moister than the western, 

 probably owing to the dip of the underlying rocks, and conse- 

 quently migration on to these slopes has been possible. In f he 

 case of the "central valley" the lcng tributary depressions are not 

 only moister, but serve also to pi^vide «ome shelter against grass 

 fires. 



Further evidence of this eastward migration is afforded by the 

 fact that the largest and thickest clumps occur nearest to the 

 valley, the smallest ones furthest east. This does not, of course, 

 hold absolutely, since each clump may act as a fresh base of 

 •colonisation, and one often finds small clumps starting close to 

 much older ones. 



VI. The Formation of Clumps. 



Outside the valleys the chief pioneer in the invasion of the 

 grassland is Gussonia spicata. Unfortunately owing to continual 

 grazing and repeated grass burning not many young clumps are 

 to be found, but in those which occur Cussonia is the most frequent 

 pioneer. There is some evidence, principally from one or two of 

 the older clumps to show that occasionally Clerodendron glabrum 

 and Combretum kraussii may act as pioneers. In most of the 

 mature clumps Cnnnonia sjjicata is found in the centre, and is 

 unmistakably the oldest tree present, though its place is sometimes 

 taken by one of the other two trees just mentioned. 



The earliest stage of the succession, noticed by the present 

 writer, was a young Cussonia, probably not more than two years 

 old, growing right out in the open veld at a considerable distance 

 from the valley. Alongside it was a young Finis capensis, but, 

 since this does not occur in any of the mature clumps in the open, 

 it will probably be killed out eventually. There seems no reason 

 for supposing that the seeds of the pioneers only are successful in 

 germinating in the open veld, and many of the seedlings may even 

 survive for a year or two, though they are eventually killed, 

 probably owing to the intense illumination to which they are sub- 

 jected. The pioneers differ in their ability to withstand this and 

 to grow and thrive in their exposed situations in the open veld. 



Once the pioneer is firmly established other species soon follow 

 and grow up in its shade. Amongst the earliest arrivals is Ehretia 

 hottentotica. Where Cussonia is the pioneer it is soon followed 

 by Combretum and Clerodendron. Later arrivals include Zizyphus 

 mucronata (always found at the edge of the clump), Fagara 

 capensis, Clausena inaequalis, Euclea sp. (probably ovata), 

 Dombeya rotundifolia, and a few others. Shrubs such as 7 J Iecfronia 

 spinosa appear quite early in the succession, and climbers very soon 

 grow up over the clump. 



In most of the clumps at the lower levels the succession pro- 

 ceeds no further, and the result is the formation of a small group 

 of trees usually with a clear space in the centre. (See PI. Ill, 

 Fig. 3.) In some of the clumps higher up, however, a further 



