2gO VELD-BURNING EXPERIMENTS AT GROENKLOOF, PRETORIA. 



replaced by Andropogon amplcctens, A. schirensis, var. angusti- 

 folia and Trachypogon polymorphus. 



Plots A, B, C, and D. 



Though it was unfortunate that the control plot A was acci- 

 dentally burnt, yet it was interesting to note the same life-history 

 taking- place in the succession as was observed on both the plots 

 E^ and E,, but the order of succession was slightly different. The 

 various species, such as Gazania krcbsiana^ Lasiosiphon linifolius, 

 Gerbera piloselloides^ etc., which appeared in such definite succes- 

 sion on E appeared here more or less at the same time. The only 

 fact perhaps worthy of note was the large number of plants of 

 Oldcnlandia {Hedyotis amatymbica) , which was far more common 

 than on either E^ or Eg. After passing through the various 

 phases as noted in E^ and Eo, but only in a less marked degree, the 

 area is taken possession of by grasses. The two vernal aspect 

 societies produced by Vcrnonia laisiochda and Hypoxis rigidula 

 do not appear at all so markedly, and also the small shrubs, such 

 as Elephantorhisa Burchellii, Indigofera hedyantha, Vernoma 

 lasioclada, are not nearly so common. In the final stage of the 

 succession by far the most dominant grass was Trachypogon poly- 

 ntcrphus. The other dominant species were Andropogon amplec- 

 tens and A. schirensis, var. angustifolia. Panicwtn natalense 

 is very common; Andropogon nardus and A. hirtifloriis common. 

 Occasionally may be seen plants of Tricholccna setifera, and 

 more rarely Digitaria diagonalis. On this plot Anthistiria 

 imberhis is of very secondary importance, and in some localities 

 almost completely absent. These observations bear out the 

 statements by Bews that Anthistiria imberbis is not a stable 

 stage ; but, if left undisturbed, gradually gives way to the tall 

 species of Andropogon, a stage in the succession to bush. On 

 all the other plots which had been burnt Anthistiria imberbis 

 was at one period of the succession the dominant grass, but this 

 was not the case on A, which has not been burnt for at least two 

 years, and probably longer. 



Bews's observations that grass-burning has the effect of 

 increasing the abundance and extent of vernal aspect 'societies has 

 also been verified. As I pointed out when dealing with the 

 plots El and E._„ this is seen in Vcrnonia hasinclada and Hypoxis 

 rigidula, but whilst not definite on A, and the same applies to the 

 plots B, C, D, which were not burnt last year at all. 



On the plot D the dominant grasses are Trachypogon poly- 

 morphus and Anthistiria imberbis. The former, being a tall 

 grass, overtops the AntJiis^tiria, which is not seen unless one walks 

 through the plot. Panicnrn natalense is common, while Andro- 

 pogon nardus is only occasional. The grasses Digitaria diagon-alis, 

 Elionnrus argcnteus, Andropogon amplectcns are rare, while only 

 a few specimens of Panicnrn serratum were observed on old 

 ant-heaps. Vigna avgnstifolia is a very common shrub, but, 

 except in very rare cases, had finished flowering. Pcntanisia 



1 



