Careful observation shows that, quite apart from the periodic rhythm of the seasonal changes, no community is stable. The physical environment itself is slowly altering, partly as a consequence of climatic action, partly through the influence of the living covering. Commonly, these changes are too secular in character to be readily appreciated, and it is only where the alterations in the physical environment are unusually rapid that the existence of a definite life-history in the plant community can be easily observed. In general, the tendency is for communities characterized by extreme conditions, as of dryness or wetness, to approach the mean. The process is well illustrated by the forma tion of sand-dUnes, where the early stages consist of sandy accu mulations around plants of marram grass (Psamma) or other pio neer species ; these by their continued growth, as fresh layers of sand are added, not only stabilize the sand but by the decay of their older parts add organic material and so materially increase the water-retaining capacity of the soil. With the consequent amelio ration of the extreme desert conditions species less specialized than the first colonizers can become established, and so there is a cumulative increase in the stability and quality of the habitat. The original dune-units increase in size and coalesce, so that the "yellow dune" gives place to the "grey dune," carpeted with continuous vegetation, and ultimately comes to bear scrub or even forest. Similarly the first occupation of the submerged mud in some estuary by almost microscopic seaweeds is the vanguard of a sequence of plants which progressively raise the level of a salt-marsh till it becomes pasture often of high economic value. The zones of vegetation which can be seen bordering the margin of a shallow lake present a sequence in space from within out wards which corresponds closely with the sequence of changes in time which led to the present condition of trie outermost. An other familiar example is the passage of enclosed pasture to scrub and finally woodland. In such successions, as these se quences are termed, each phase is distinguished by the presence of certain species, and the number and size of these tends usually to increase as the succession proceeds. Frequently in the final stages, however, one or two species again come to predominate, and since these often determine the physiognomy of the commu nity they are frequently referred to as the dominants. Successions have been extensively studied by American ecologists and F. E. Clements has utilized this aspect as the basis for a classification of plant communities.
Except perhaps in the so-called "open" com munities such as those of deserts, where the individual plants are separated by bare soil, there is probably always direct competition between the constituent members. The conditions of the habitat
exert a selective action, stimulating the development of certain species and depressing the vigour of others. If protected from competition many species can grow in habitats where they do not occur naturally. Many aquatics can survive under terrestrial con ditions, and a large proportion of alpines succeed under cultiva tion in the lowlands. It is thus commonly not a question of whether or not a species can grow, but whether it can grow well, and competition often resolves itself into a capacity to obtain the requisite share of such essentials as light, air, water and mineral salts. In competition for light the naturally taller species is at an advantage; it can grow above and shade its competitor. Thus, wherever soil and climate can support forest vegetation, this tends to be the end phase or "climax" in the succession. So, too, trees which cast a deep shade are usually successful in competition with those having a light canopy. The birch, which by reason of its light wind-born fruits readily colonizes denuded woodland areas, eventually gives place to trees of heavier canopy such as the oak. Large size, however, is usually associated with delay in reproductive activity and the prolonged life-history renders the arboreal habit less successful than the herbaceous under the stress of conditions imposed by the presence of man. This is reflected in the steady diminution of the area occupied by natural forests.
Almost any feature in the organization or life-history of the individual which fits it for its particular habitat is at the same time a passive agent in its survival ; but amongst the features which may be regarded as of importance in aggressive competition may be mentioned especially the rate of vegetative spread, the annual output of seeds, the efficiency of the mechanism of disper sal, and the extent and character of the root and shoot systems.
Finally, mention should be made of the importance of the study of plant communities and their succes sions for the proper control and exploitation of forests and the maintenance of high yielding pastures. The naturally occurring species of an area serve as a valuable guide to agricultural prac tice, since their presence is in effect a product of the complex built up of soil and climate. A study of the natural relations of species is, moreover, a useful guide to their artificial juxtaposition in the practice of agriculture, forestry and horticulture (qq.v.).