Journal of Tropical Ecology (2002) 18:935–942. Copyright  2002 Cambridge University Press DOI:10.1017/S0266467402002614 Printed in the United Kingdom SHORT COMMUNICATION Two biases in estimating range sizes of Amazonian plant species KALLE RUOKOLAINEN*, HANNA TUOMISTO* 1 , JAANA VORMISTO* and NIGEL PITMAN† *Department of Biology, University of Turku, FIN-20014 Turku, Finland †Center for Tropical Conservation, Box 90381, Duke University, Durham, NC 27708, USA (Accepted 7th October 2001) KEY WORDS: Amazon basin, Arecaceae, ferns, Melastomataceae, palms, plant height, Pteridophyta, species identification, species range size, trees, tropical rain forest The Amazonian rain forests form one of the largest areas of continuous forest in the world, and they harbour a substantial part of the world’s biodiversity (World Conservation Monitoring Centre 1992). The extravagant species rich- ness of Amazonia makes collecting information on species distribution patterns there a formidable challenge. Relatively few studies have been able to provide data that cover both a large enough number of species and wide enough geo- graphical areas to allow relevant biogeographical and ecological generaliza- tions (such studies concerning plants include Balslev 1988, de Oliveira & Daly 1999, Pitman et al. 1999, Prance 1973, ter Steege et al. 2000, Terborgh & And- resen 1998, Tuomisto & Poulsen 1996). The scarcity of field observations is undoubtedly one of the reasons why very different ideas exist about the mech- anisms that control plant distributions and patterns of species richness in Amazonia (Balslev 1988, Condit 1996, Gentry 1988, Nelson et al. 1990, Pitman et al. 1999, Prance 1973, 1982; ter Steege et al. 2000, Tuomisto et al. 1995). Quantitative tree inventories, regional plant check-lists and taxonomic revi- sions are the main sources of information for documenting Amazonian plant distribution patterns, and luckily new information is continuously accumulating 1 Corresponding author. Email: hantuo@utu.fi 935