PARENT SESSION
Session #18: Mechanisms underlying the maintenance of diversity and mode of regeneration in tropical forests: tests of alternative hypotheses. Organized by: W.P. Carson, C. Gehring, and T. Theimer.
Wednesday, August 8, 2001. Madison Ballroom B
Seedling regeneration over 35 years in Australian tropical and subtropical rain forests.
Connell, Joseph1, 1
ABSTRACT- We have carried out a series of standard censuses of the recruitment, growth, and mortality of mapped seedlings, saplings, and trees on precisely the same permanently marked transects at two rainforest plots in Queensland, Australia. They are in a tropical rainforest plot at 17degrees S lat., and a subtropical rainforest plot at 28 degrees S lat. We have monitored these long-term plots at intervals over 35 years, between 1965 and 2000. Over this time, seedling numbers varied over several orders of magnitude. On the tropical plot, of the 121 species that had at least 2 adults, the number of new seedlings mapped and tagged since 1965 has ranged from zero to 12,677. In one common species the seedling numbers per census ranged from 10 to 3900; 98% appeared in 5 peaks spaced at 4 to 11 year intervals. Many other common species showed a similar pattern of markedly episodic recruitment. Long-term trends in recruitment varied considerably. Over 35 year on our permanently-marked belt transects, seedling recruitment per yr of some species gradually increased, others decreased, others showed a reversing trend (rising followed by falling, or vice-versa), and still others showed no significant pattern. Long-term records were essential in revealing these patterns or lack of them. Even a 15-year study would have misclassified those with a reversing trend as having directional trends instead. The idea that biodiversity confers advantages on an ecological community has a long history, beginning with Darwin. To date, tests of the hypothesis with plants have used experiments with herbaceous vegetation in the laboratory or in field plots planted for this purpose. In contrast we have tested the hypothesis under natural conditions in our rainforest plots. Newly established seedlings usually occur in conspecific patches; thus their 'pattern diversity', i.e., degree of intermingling of different species, is low initially. However we have found that survival of a seedling or sapling is worse when its neighbors are conspecific than when they are a different species. Thus greater pattern diversity seems to confer advantages to seedlings in an Australian tropical and subtropical rainforest.
KEY WORDS: rain forest, diversity, seedling regeneration, Australia