362 STRUGGLE BETWEEN PLANT-COMMUNITIES sect, xvii 



settlers. These changes may be of very divers kinds, and so slow as to be 

 almost imperceptible. What factors play the most important part in the 

 development of vegetation in such cases it is often extremely difficult 

 to decide ; it is usually not a single factor, but a whole series of factors 

 which interpose and co-operate. 



Change in depth of water-table and in the amount of water in soil 

 represent one factor, change in the chemical nature of soil is another. It 

 has already been mentioned ^ that in Russia steppe and forest compete 

 with each other; if Tanfiljew^ be correct, it is the slow but constant 

 elutriation of the soil that gives victory to forest. 



In Central Europe steppes existed at one time, namely, after the Tundra- 

 period that succeeded the Glacial Epoch. ^ These steppes subsequently 

 became forests, and probably differed from modern steppe-like communi- 

 ties in the amount of salt present in the soil. The causes for this change 

 of vegetation are not known, but must be sought in chmatic or physico- 

 geographical changes or influences. In more recent times the forests over 

 wide areas have had to give way to arable land. 



In forest itself a change in the constituent species has taken place, and 

 is still proceeding. Steenstrup's * investigations of moors taught us that 

 in Denmark one kind of vegetation succeeded another.^ Nathorst * 

 amplified these investigations by discovering arctic tundra-vegetation 

 underlying the moors. While Vaupell '^ has elucidated the latest phases- 

 of the struggle, namely, that between oak and beech. Here, too, we may 

 allude to P. E. Miiller's ^ researches deahng with the struggle between 

 forest and dwarf-shrub heath. 



What causes are responsible for the successive changes of vegetation 

 in the course of thousands of years it is difficult to decide.^ An important 

 part has perhaps been played by changes of climate, which on the whole 

 has become milder. It is not probable that any vast, secular rotation of 

 crops takes place, or that one species renders the soil more suited to 

 a successor and less so for itself, as do certain low organisms. According 

 to Whitford,^" in Michigan soil is improved by coniferous forest. When 

 a considerable amount has accumulated dicotylous forest establishes 

 itself. Conversely, by the removal of leaf-mould the soil may be im- 

 poverished with a consequence that dicotylous forest is vanquished by 

 coniferous forest. Yet, under certain conditions, as there is a progressive 

 improvement of the soil by an accumulation of humus, more exacting 

 species will be favoured at the expense of the more accommodating species 

 which were the first to occur : this holds true only if the soil be not 

 robbed of the plant-remains (in the shape of crops of any kind, hay or 

 wood) which contain its chief nutritive ingredients. Among forest-trees, 

 oak and beech are exacting, whereas birch and Scots pine are accommo- 

 dating. There can be no doubt that an important part has also been 

 played by the varying relations between trees and hght.^ In the compe- 

 tition between oak and beech in Denmark, human action (felling of trees, 

 and drainage) has affected the result in favour of the beech, so that the oak 

 has been able to maintain itself only on moister spots and poorer parts of 



^ See p. 282. " Tanfiljew, 1894, 1905. ^ Nehring, 1890. 



* Steenstrup, 1841. ^ See p. 359. " Nathorst, 1870. 

 ' Vaupell, 1857, 1863. * P. E. Miiller, 1878, 1884 (1887 a). 



* See also Adams, 1905. ^^ Whitford, 1901. " Seep. 18. 



