Wetlands Ecology and Management 11: 233–242, 2003. © 2003 Kluwer Academic Publishers. Printed in the Netherlands. 233 Annual growth rings and long-term growth patterns of mangrove trees from the Bragança peninsula, North Brazil M. Menezes 1 , U. Berger ∗ & M. Worbes 2 1 Universidade Federal do Par´ a, Campus de Bragança, Alameda Leandro Ribeiro s/n, 68600-000 Bragança (PA), Brazil; 2 Institute of Forest Botany, University of Göttingen, Büsgenweg 2, 37077 Göttingen, Germany; ∗ Author for correspondence: Center for Tropical Marine Ecology, Fahrenheitstr. 6, D-28359 Bremen, Germany; E-mail: uta.berger@zmt.uni-bremen.de Received 15 July 2000; accepted in revised form 4 April 2002 Key words: dendrochronology, mangrove forest dynamics, Rhizophora mangle Abstract Ring analysis was carried out on 39 Rhizophora mangle trees from two saline and one brackish forest sites on a peninsula in north Brazil. All trees showed constant growth over their entire life span, however the distinctiveness of growth rings was greater in trees from the saline areas than in trees from the brackish site. The mean radial increments form a pattern of three distinct groups (‘fast’, ‘medium’, and ‘slow’ growth). Although all three growth groups contained trees from each study site, trees from the brackish and the frequently inundated saline area presented a significantly higher growth rate (3.3 mm y −1 ) than the trees growing in a saline, seldom inundated area. Radiocarbon analysis showed that R. mangle forms annual rings in the region. The mean age of each forest, derived from the mean stem radius, the growth rates, and the tree affiliation to each growth group, explains forest’s density and basal area. For the trees from a saline area belonging to the medium growth group (mean increment 2.5 mm y −1 ), the cambial growth correlates significantly with the precipitation in the transition months between the dry and the rainy season. The slowest growing trees (1.2 mm y −1 ) showed a close relationship between the ring width and the number of months with rainfall < 50 mm. Based on these results, we propose that the local abiotic factors influence the individual growth rates but their effect on the forest structure is modified by biotic factors, such as neighborhood competition. Introduction Mangrove trees of the same species can show a high plasticity in growth, e.g. height, stem diameter, and tree architecture, depending on specific environ- mental factors such as nutrients, salinity or inundation frequency (Tomlinson, 1986; Suarez et al., 1998). A detailed analysis of mangrove forest dynamics is therefore complex, and is still not covered by a gen- eral theory (Fromard et al., 1998). Several researchers have tried to investigate the development of mangrove forests by means of simulation models (Chen and Twilley, 1998; Berger and Hildenbrandt, 2000). How- ever, their success has been limited since key data of demographic processes are still unavailable, e.g. site- and species-specific growth rates, or medium age of individuals and stands. It has been generally assumed that datable growth rings do not exist in mangrove trees (Ash, 1983; Tomlinson, 1986; Gill, 1971). This assumption is partially linked to the general assump- tion that annual growth periodicity does not occur in the tropics (Lamprecht, 1959; Hallé et al., 1978; Uhl, 1980). Nevertheless, a number of investigations have shown that tropical trees exposed to distinct rainy/dry seasons or to extended periods of flooding form annual rings (Coster, 1927, 1928; Mariaux, 1967; Worbes, 1985). In these cases, ring analysis enables the de- termination of tree age (Jacoby, 1989; Worbes, 1989; Worbes and Junk, 1989; Worbes and Junk, 1999) and provides the basis for an appraisal of the age of forest stands. This is important for identifying and understanding successional sequences (Worbes et al.,