464 INTRODUCTION TO PLANT GEOGRAPHY [CHAP. 



Primary xeroseres in the tropics may be observed for example on 

 recently emerged or volcano-devastated areas such as the East 

 Indian island of Krakatau, where after three years an associes con- 

 sisting of a lower layer of Blue-green Algae and an upper one of 

 Ferns and other vascular plants was found to clothe the surface of 

 the pumice and ash in some inland places. Eleven years later, in 

 1897, the interior supported a dense growth of Grasses, with isolated 

 shrubs and fairly numerous forbs as well as Ferns. There were, 

 however, very few lower cryptogams to be seen apart from ter- 

 restrial Algae. Nine years later still, the shore supported a belt of 

 well-developed maritime woodland complete with climbers, etc., 

 but the inland savanna persisted. Subsequently it developed into 

 a mixed woodland of fair luxuriance, and came to show every 

 indication of progressing ultimately to the local rain forest (apart 

 from some floristic depauperation and, probably, slowness of succes- 

 sion due to isolation and the consequent difficulties of colonization). 

 Thus the usual sequence, such as we have seen elsewhere, of domin- 

 ance by cryptogams and then by herbs and finally by trees, holds 

 true in this tropical xerosere, though it should be noted that 

 in rain-forest areas, such as this, there is a preponderance of 

 phanerophytes among the flowering plants — often from quite early 

 stages. 



As in the case of the hydrosere, it seems that in the tropical 

 xerosere the number of species goes on increasing to the end, whereas 

 in temperate regions the numbers of species in both hydrosere and 

 xerosere tend to rise to a maximum and then decline as the climax 

 is approached, the decline usually starting when the community 

 becomes closed. Other humid tropical regions appear to have 

 xeroseres of a generally similar nature to that observed on Krakatau, 

 even if the pioneers in some cases are forbs. Grasses, Sedges, or 

 even woody plants (especially in secondary successions). 



The sea-shore and littoral communities outside the normal forest, 

 already dealt with in the last section, are probably indicative of at 

 least potential successions. However, with the factors of the 

 environment as overwhelming as they often are in such situations, 

 it seems unlikely that the successions will become actual as long 

 as the shore-line continues as at present. Rather does it appear 

 that each zone is in equilibrium with its particular environment. 

 This, as we have seen, is apparenllv the case with many mangroves 

 as well as with the latest stages of tropical hydroseres. But where 

 accumulation can continue, as on some sandy shores, the communities 



