conditions (continuous variation), and where neighbouring environmental conditions are different enough to result in sudden changes between vegetation associations (discontinuous variation) (Bunt and Williams 1981). Thus vegetational changes can be continuous, discontinuous, or a combination of both. This is why it is crucial to consider scale before attempting an analysis of mangrove forest and why, in the absence of such consideration, data from isolated transects, even from the same area, are so hard to interpret. Comparisons of transect data from different sites are useful in compiling inventories and in noting similarities, but are not a basis for a discussion of zonation which should be based instead on air photos of, and ground surveys over, parts of a number of forests.
Figure 2.18. Profile diagram through somewhat disturbed mangrove forest at Lapangga, Morowali National Park. Note: the general predominance of R. apiculata and discrete occurrence of the other species.
After Darnaedi and Budiman 1984
The tendency of mangrove forests to occur in distinct zones has been interpreted variously by different authors as a consequence of plant succession, geomorphology, physiological ecology, differential dispersal of propagules and seed predation. Each of these is considered next.
Figure 2.19. Intsia bijuga. Scale bar indicates 1 cm.
After Soewanda (n.d.)
Plant Succession. Plant succession is a classic ecological concept and is defined as being the progressive replacement of one plant community with another of more complex structure (p. 366). Much of the early work on mangrove forests focused on its supposed land-building role and it seemed clear from this that one species colonized an exposed bank of mud and, as conditions changed (such as an increase in the organic debris of the mud), so other species took over. For example the colonization of a new shallow or exposed substrate by Avicennia or Sonneratia trees, such as on the banks of the Rongkong River delta at the north of Bone Bay, produces a network of erect pneumatophores which have three indirect functions:
Figure 2.20. Succession in mangrove forests.
After Chapman 1970 in Walsh 1974
they protect the young trees and germinating seeds from wave damage;
• they entangle floating vegetation which decays and becomes incorporated in the soil; and
• they provide a habitat for burrowing crabs which help to aerate the soil (Chambers 1980).
These changes, and subsequent additional sedimentation, lead to a succession of species or communities of species over a period of time (fig. 2.20). It is clear, however, that the stages of succession are not always consistent and different local environmental conditions and man's impact on those have an influence (Steup 1941; Anon 1980a).
A total of 20 characteristics have been noted for secondary succession in tropical forests (Boudowski 1963). If the characteristics are examined in relation to the succession of mangrove forests, only seven of the characters apply, nine do not, and four are inconclusive. This suggests that attributing the apparent zonation to succession is not the whole story (Snedaker 1982b).
Geomorphological Change. As stated above, early workers on mangrove felt that it was mangroves that 'built land'. However, it is clear from observations in the huge deltas of the Ganges, Indus and Irrawaddy on the coast of the Indian sub-continent, that it is the process of sediment deposition that builds land. It is now generally agreed that mangroves do not have any influence on the initial development of the land forms. Mangroves may accelerate land extension but they do not cause it (Ding Hou 1951).
From a geomorphological perspective, it is the shape, topography and history of the coastal zone that determine the types and distributions of mangrove trees in the resulting habitats. The position of species relative to tidal levels (and thus soil type) is obviously important and the pattern of tidal inundation and drainage has been considered to be the major factor in mangrove zonation (Watson 1928). This idea has been developed to include variation in the salinity of the tidal water and the direction of its flow into and out of the forest (Lugo and Snedaker 1974). Although good correlations exist between salinity and zonation, they are not proofs of direct cause and effect.
Physiological Response to Soil-Water Salinity. If salinity is an actual cause of zonation in mangrove forests it needs to be shown that the plants actually respond through their physiology to salinity gradients and not to a factor such as oxygen levels which, under certain conditions, will fall with increasing salinity. It has already been mentioned that mangrove trees are able to live in freshwater (i.e., they are facultative, not obligate, halo-phytes), but each species probably has a definable optimum range of salinity for its growth. Indeed, it has been found that within each zone the characteristic species had apparently maximized its physiological efficiency and therefore had a higher metabolic rate than any invading species (Lugo et al. 1975). Thus invading species would be at a competitive disadvantage due to their lower metabolic efficiency in that habitat. Similarly, it has been found that each species of the mangrove forest grows best under slightly different conditions such as the amount of water in the mud, the salinity, and the ability of the plant to tolerate shade. This means that the various species are not mingled together in a haphazard way but occur in fairly distinct zones.
Salinity can vary considerably between high and low tides and between seasons (p. 109), and thereby presents a confusing picture to a scientist conducting a short-term study. Thus, to identify the salinity levels to which the different species are optimally adapted requires long-term and detailed measurements to determine long-term averages and ranges of salinity. It is often the case that a species' ecological limits are defined by relatively rare events such as occasional extremely dry years (p. 22) or, in the case of mangroves, high salinity levels in the dry season when there is little freshwater input to the system.
Differential Dispersal of Propagules. The suggestion that differential dispersal of propagules (fruit, etc.) influences zonation of mangroves rests on the idea that the principal propagule characteristics (e.g., size, weight, shape, buoyancy, viability, numbers, means of release and dispersal, and location of source areas) result in differential tidal sorting and therefore deposition. There are as yet few data to support this hypothesis (Rabi-nowitz 1978; Snedaker 1982b).
Seed Predation. Experiments conducted in Australia have shown that about 75% of the propagules of five species of mangrove trees were consumed by predators, primarily grapsid crabs, and there were significant differences among species and between forest types. Rhizophora stylosa was preyed upon the least and Avicennia marina the most. As might have been expected, predation was generally highest where a particular tree did not have neighbouring trees of the same species. Predation rates seem to be associated with chemical composition because A. marine, selected by predators over all the other species, had the highest concentrations of protein and sugars and the lowest concentrations of fibre and tannins. The propagules with the highest tannin content appeared to be preyed upon only by a single, specialist crabs (Smith 1986). These differences will influence the pattern of seedling establishment in a mangrove forest, such that establishment is most likely to occur in areas where seed predators are least likely to occur. In the case of A. marina this may mean establishment in the areas most frequently inundated.
Geomorphology, physiology and seed predation thus appear to be the most relevant forest. The impact of human activities, so ubiquitous in coastal regions, plays a major role in modifying species composition and physical conditions, however, and so should be considered first in any
