DECIPHERING LANDSCAPE MOSAICS OF NEOTROPICAL TREES - GIS AND SYSTEMATIC-SAMPLING PROVIDE NEW VIEWS OF TROPICAL RAIN-FOREST DIVERSITY

Authors
Citation
Da. Clark, DECIPHERING LANDSCAPE MOSAICS OF NEOTROPICAL TREES - GIS AND SYSTEMATIC-SAMPLING PROVIDE NEW VIEWS OF TROPICAL RAIN-FOREST DIVERSITY, Annals of the Missouri Botanical Garden, 85(1), 1998, pp. 18-33
Citations number
60
Categorie Soggetti
Plant Sciences
ISSN journal
00266493
Volume
85
Issue
1
Year of publication
1998
Pages
18 - 33
Database
ISI
SICI code
0026-6493(1998)85:1<18:DLMONT>2.0.ZU;2-L
Abstract
How are tree species within tropical rainforests distributed at the la ndscape scale? One research site, the La Selva Biological Station in C osta Rica, offers exceptional tools for addressing this question: a do cumented flora, soil and topographic maps. a reserve-wide grid, and a Geographical Information System (CIS). My colleagues and I have combin ed these tools with highly replicated systematic sampling over 600 ha of old growth to investigate patterns of forest composition within thi s lowland tropical wet forest. This approach has revealed Features of within-forest heterogencity that had remained ''invisible during exten sive fieldwork by mane researchers at La Selva. Examples are: a doubli ng in density of the guild of subcanopy and canopy palms between flat terrain and increasing steep topography; strong shifts in density of m any palm and tree species over La Selva's limited gradients of soils a nd topography: evidence of human harvesting of one palm species from o ld-growth Forest: and evidence suggesting indigenous human activity de ep within the reserve (the co-occurrence of a previously unrecognized zone of alluvial soil, buried charcoal, and an avocado tree). These st udies have also added 15 tree species to the known flora of this inten sively researched forest. Although La Selva's support For such landsca pe-scale studies is exceptional, even in remote tropical forests it is now possible to systematically sample and geo-reference information o n site variation and species distributions using newly available Globa l Positioning Systems. Findings can then be rous-referenced with curre nt and Future site data, using a GIS, Although such efforts, especiall y the development of a GIS, require considerable investments of time a nd expertise. the payoff can be a more robust understandings of the di stribution of tree diversity and species abundances over tropical rain forrst landscapes.