Loss of soil fertility following removal of the primary forest may not be a serious threat to the integrity of some tropical ecosystems. Evidence includes the absence, or at least scarcity, of documented cases of permanent deflection in the absence of continued human disturbance and the fact that effects attributed to nutrient loss may be due to competitive interactions. Likewise, the universally recognized rehabilitation of soils during succession would seem to argue that the vegetation is more or less independent of soil nutrients over a wide range of availability. THE QUESTION OF SOIL FERTILITY has always been important in considerations of the effect of forest disturbance in the tropics. Originally, it concerned the problems of declining crop yields in agricultural systems involving shifting cultivation, in that loss of fertility was considered to be one of several possible causes. Among the others were increased competition from weeds and increased damage by insect pests. The shifting cultivation problem, coupled with failure of attempts at sustained agriculture, led to the widespread belief that tropical soils were inherently poor, and to the paradox of apparently lush vegetation growth on poor soils. The resolution of the paradox seemed to involve several characteristics of tropical ecosystems including storage of a large proportion of the nutrients in biomass and nutrient cycles (Richards 1964). Assuming, then, that perpetuation of these lush tropical forests involves nutrient storage in biomass and nutrient cycles, it becomes important to know about the effect of forest disturbance. It might be that removal of the forest is in a sense like pulling the plug on the ecosystem, resulting in rapid loss of the nutrient capital, and consequently very slow recovery. In this context, the soil fertility problem involves possible slowing or deflecting of succession. This is the problem that I wish to address. Most of my discussion will involve tropical wet forest, the tropical rainforest of Richards ( 1964). But it should be remembered that these forests occupy only about 30 percent of the land area between the tropics of Cancer and Capricorn (Sanchez 1976). Actually, the work of Nye and Greenland (1960) on semideciduous forests of Ghana seems to indicate that these areas share the problems of declining crop yields, and the characteristics of weathered soils, high nutrient storage in biomass, and tight cycles, so perhaps it is legitimate to include them. These forests constitute an additional 15 percent of the land area. This total land area occupied by these forest types is not very different from the 51 percent of the tropics occupied by highly weathered, leached soils, previously called 'latosols,' which include oxisols, ultisols, and alfisols (Sanchez 1976). It is easy to agree that secondary succession involves almost any vegetation regrowth following disturbance. However, as the nature and scale of the disturbance change, so does the succession. And succession in the tropics is so poorly known in general that it may be difficult to ascribe causes to any differences we observe, i.e. it becomes nearly impossible to isolate the effects of fertility changes from effects of alteration of seed sources or suppression of woody sprouts, or even from effects of differences in site physical characteristics. The situation becomes even more complicated by questions about the driving forces of succession, and the possible role of early successional species in altering site properties to allow invasion of species characteristic of late stages. However, I wish to sidestep these issues for the moment and concentrate on the question in its simplest form: can it be demonstrated that the successional sequence is altered or productivity of successional vegetation is lowered by loss of soil fertility following forest disturbance? The evidence will be primarily circumstantial. To anticipate my conclusions, I would argue that alteration of successional recovery due to soil changes is unlikely without major erosion of topsoil. On the other hand, biological processes, particularly competitive interactions and human influences upon seed sources, are very likely to slow succession. SOURCES OF INFORMATION Several kinds of information might be used as evidence from which to make inferences about the question of soil fertility and succession. The first 8 TROPICAL SUCCESSION 8-15 1980 This content downloaded from 157.55.39.49 on Tue, 09 Aug 2016 06:14:16 UTC All use subject to http://about.jstor.org/terms involves the nature and properties of tropical soils in general. Soil scientists have in recent years accumulated considerable information concerning the behavior of tropical soils in a wide variety of circumstances (Moss 1968, Nat. Acad. Sci. 1972, Sanchez 1976, Young 1976). Of particular interest, studies of the forms of nutrients in the soil and their susceptibility to loss by leaching might provide insights into the severity of soil degradation as well as the possibilities for rehabilitation. The general relationship between vegetation and soils also bears on the question: if different soils will support different types of climax vegetation, then it is to be expected that succession will also vary with soil types. Documentation of the differences may help in distinguishing the effects of soil fertility loss on vegetation regrowth from more general effects of soils on vegetation. Another aspect of the relationship between soils and vegetation is the question of alteration of this relationship by man. Is there evidence of permanent deflection of succession by human-caused soil degradation? This evidence may take one of two forms. First, evidence from vegetation itself: do changes in species composition or community structure seem to be related to agricultural activity in the distant past, or over very long periods of human occupancy? Second, is there direct evidence of declining soil fertility? Studies have shown that fertility does decline following forest removal. But the evidence of short-term changes in soil fertility is usually accompanied by evidence that successional vegetation is remarkably good at regenerating soil fertility.
No takes yet. Share an insight, caveat, or question.
P. A. Harcombe (1980) studied this question.