Poster Session: Modeling Species Patterns

Rainforest landscape dynamics: Treefall gaps and Passionvines. Plowes, Robert*,1, 1 Section of Integrative Biology, Austin, TX

ABSTRACT- Rainforest treefall gaps provide dynamic landscapes for light-demanding plant species. I study Passiflora vines to investigate spatial and temporal dynamics of treefall gaps and consequences for these vines. The model explores dynamics of organisms living in ephemeral habitat patches, applicable to a wider range of systems. Treefall gaps provide short-lived habitat patches that regrow in 3 to 5 years, while vines have potentially longer life spans than their habitat patches. To cope with this constraint, the 9 local Passiflora species exhibit various strategies of dispersal, reproduction and dormancy. This unusual dynamic is explored using spatially explicit models of both treefall gaps and Passiflora demography. A survey of 500 treefalls in 300 ha provides parameters for landscape models. Treefall densities were measured by field surveys, airphoto and landsat image interpretation. Subpixel analysis of Landsat images allows interpretation of older images predating the study, covering a range of disturbances. New gaps are generated at 0.4 gaps/ha/year, while vine occupancy of the commonest species is 0.5 plants/ha. An initial constraint on patch suitability is canopy-opening size, since germination requires 4 hours of direct sunlight. This restricts the size, orientation and age of suitable treefall gaps. Regrowth inhibits further germination after 2 to 4 years. The vine status was recorded annually, and used to estimate transition parameters for a markov model. I also estimated population growth rates () using non-spatial population projection matrices. Preliminary results show that population growth rates derived from spatially explicit models differ considerably from non-spatial estimates.

KEY WORDS: treefall, rainforest, demography, spatially explicit, Passiflora


Online publishing provided by
Allen Press, Inc. | 810 E. 10th St. | Lawrence, Kansas 66044 USA
e-mail abserv@allenpress.com | Web www.allenpress.com
All material is copyright © 2004 rama