354 STRUGGLE BETWEEN PLANT-COMMUNITIES sect, xvii 



Fire is one of the means by which man upsets natural conditions, and 

 is of direct service in the cause of agriculture in the tropics, where man 

 usually prepares land for cultivation by the felling and firing of forest.^ 

 So long as the soil is cultivated and this is often only for a few years 

 man has constantly to contend with wild plants, for instance with the 

 stool-shoots and suckers of old forest trees, if he is to protect cultivated 

 plants. Hardly has the soil been left to itself before it is invaded by wild 

 plants. The first to settle are numerous annual and other herbs, also 

 shrubs, which form a commonplace vegetation of weeds, whose seeds 

 and fruits blow hither from all sides or are brought by birds. Thus 

 arises a community which is gradually converted into weed-bushland ! 

 (a ' secondary formation '). But soon the forest-plants start to grow 

 afresh ; shoots sprout forth from stumps and roots, and perhaps from 

 seeds that lay hidden in the soil : and after a number of years forest 

 once more occupies the site. 



But there are cases in which a secondary formation does not revert 

 to the primary one. According to Pearson,^ the patanas in Ceylon have 

 been derived from savannah-woodland, but are now to be regarded as 

 permanent grasslands that cannot ever be changed back into forest 

 because the soil has been modified. 



According to Kihlman,^ forest fires prevent the spread of the common 

 spruce in certain parts of the northern forest-zone. The Scots pine has 

 expelled it from regions where it was formerly abundant. The farther 

 north a station lies the more serious is the effect of forest fires, because 

 the maturation of seeds becomes increasingly difficult. Between Kola 

 and Lake Imandra Kihlman discovered an elevation, three kilometres 

 in length, which had been devastated by fire several years earlier : here 

 the spruces, which were formerly dominant, had all perished, but there j 

 were isolated pine-trees that had survived the ordeal. With this excep 

 tion the soil was occupied by a J^oung, tolerably dense association of j 

 birches, among which one sought in vain for conifers. In this case it I 

 appears that aided by fire the birch will suppress the spruce because its 

 seeds ripen more readily. Hult ^ showed the very grave extent to which ; 

 forest-fires in Blekinge affect the competition among the different kinds 

 of vegetation. i 



Krassnoff,^ travelling in the inner valleys of the Altai mountains, i 

 observed forests that had been burned over a distance of ten or eleven 

 kilometres. Although a long time had elapsed since the fire raged, no 

 new forest had arisen, but in its place there was a waving sea of herbs, 

 3 feet in height, which, however, did not form a continuous covering over 

 the soil ; the herbs included Helleborus, Aconitum, Thalictrum, Ligu- 

 laria, Paeonia, and Pedicularis. Forest in this case seems to have been 1 

 suppressed by a community of an entirely different class. i 



Fires on heaths provide another example of the production of new! 

 soil. Often the first vegetation to appear differs from that occurring; 

 subsequently, but finally Calluna recovers the ground ; yet often Calluna ' 

 from the outset gives rise to an association, and the plants burnt emit 

 shoots from the ground, while countless seedlings sprout forth. Fires on<; 



' Warming, 1892. ^ H. H. W. Pearson, 1899; seep. 298. 



" Kihlman, 1890. * Hult, 1885. ' Krassnoff, 1888. 



