Search

Dendrochronology is developing outside temperate and boreal regions. Over the past decade substantial progress has been made in Mediterranean and wet tropical regions. However, research in dry tropical regions, notably those of sub-Saharan Africa, has remained fragmentary. Here, we try to identify the unique challenges and opportunities of dendrochronology in the dry tropics. First, we briefly review the status of dendrochronology outside temperate and boreal regions with an emphasis on sub- Saharan Africa. Subsequently, we focus upon one of those areas where dendrochronology in the dry tropics is at the forefront of scientific advance: Ethiopia. A detailed review of tree ring studies in the lowlands and highlands highlights the complexity of ring formation and made us identify four major types of growth ring expression: anatomically notdistinct rings, multiple rings per year, annual rings and multiple missing rings. This complex tree growth behaviour is associated with large-scale variations in precipitation regime (unimodal to multimodal) and relatively small-scale variations in tree sensitivity to water availability. Literature results are used to develop a scheme that can be used to predict differences in growth ring formation along gradients in these two factors. Because of the exceptional growth sensitivity of and the importance of local site conditions (topography, biological factors, etc.) for most trees sampled, those growing at the limits of their ecological amplitude are prone to possess multiple rings per year or multiple missing rings. In such circumstances, site selection should not always take place at the limits of the ecological amplitude of a species, but may sometimes have to be diverted to more mesic environments. Successful studies are now appearing, such as those reporting correlations between tree ring chronologies and Blue Nile river flows.

Most semiarid regions are facing an increasing scarcity of woody vegetation due mainly to anthropogenic deforestation aggravated by climate changes. However, there is insufficient information to reconstruct past changes in climate and to evaluate the implications of future climate changes on the vegetation. Tree-ring analysis is a powerful tool for studying tree age, population dynamics, growth behavior, and climate-growth relationships among tropical tree species and for gaining information about the environmental forces driving growth change as well as for developing proxies for climate reconstruction. Wood anatomical and dendrochronological methods were used on Juniperus procera trees from two Ethiopian highland forests to check (i) whether tree-ring series of juniper are cross-datable and hence suitable for building tree-ring chronologies, and if so, (ii) which climate factors mainly drive wood formation in juniper from this region. Visible growth layers of the juniper wood were shown to be annual rings. Tree-ring sequences could be cross-dated between trees growing at the same site and between trees growing at sites 350 km apart. Evidence was found that annual growth of junipers is mainly controlled by one climatic factor, precipitation. This strong precipitation influence proves the potential of African juniper chronologies for accurate climate reconstruction and points out the relevance of building a network of juniper chronologies across East Africa.

Nature and periodicity of growth rings were investigated in Sonneratia apetala and Heritiera fomes, two Bangladeshi mangrove species. From both species we collected three stem discs in the natural forest reserve of the Sundarbans. In addition, three discs were sampled from plantation- grown S. apetala trees of known age. Sanded stem discs revealed distinct growth rings but no periodic fluctuations in vessel variables (vessel density, vessel diameter, vessel grouping), which were measured at high resolution along a transect from pith to bark. The number of growth rings in plantation-grown S. apetala trees corresponded with the documented tree age, hence strongly suggesting the growth rings to be annual. Within species, the annual nature of the rings was further supported by a good match between the tree-ring series. The similar mean curves of S. apetala and H. fomes, growing at the same site in the Sundarbans, pointed to the presence of an external factor influencing their growth. A combination of precipitation and temperature was suggested influencing substrate salinity and phenological events. It became evident that tree-ring research in combination with the analysis of vessel patterns is a valuable tool to further investigate the complex interactions between tree growth and site ecology in mangrove forests.