Master of Science (MS)


Biological Sciences

Document Type



Forest regeneration on abandoned land in the Brazilian Amazon depends first and foremost on prior land-use history. Abandoned clearcuts become dominated by Cecropia trees, but contain a rich mix of other arboreal genera and succession proceeds rapidly to a diverse forest. In contrast, abandoned land that has been burned repeatedly for pastures becomes dominated almost completely by Vismia trees and remains in monogeneric stands gaining few new genera over the same time interval. I tested the hypothesis that areas with repeated burns would have more re-sprouts, particularly Vismia re-sprouts. I also predicted that in monogeneric Vismia stands, seed dispersal would be limited to few bat-dispersed genera while there would be more diverse seed rain in Cecropia, particularly from bird dispersers. In order to test these hypotheses, stands with different land-use histories were evaluated in terms of tree, palm, shrub and liana abundance and diversity, as well as whether they germinated from seed or grew as re-sprouts from the root. I found that Vismia stems were one hundred percent re-sprouts in Vismia stands. In Cecropia stands, 93% of Vismia stems germinated from seed. There were no significant differences in the abundance of re-sprouts by Cecropia and all other species, regardless of stand type. To test how seed dispersal in Cecropia and Vismia stands differs, I collected seeds fallen into seed traps by bird and bat dispersers as well as from fecal samples collected principally from mist netted birds. I found that both bird and bat dispersers deposited a more diverse seed assemblage into Vismia stands than Cecropia stands. Overall, bats disperse more Vismia seeds while birds disperse more Miconia seeds, regardless of stand type. My results suggest that seed dispersal does not drive the differences in succession found in Cecropia and Vismia second growth. However, the capability of seeds to germinate and recruit in Vismia stands warrants further investigation. I suggest that the use of repeated fire should be limited as much as possible due to the long term effects on succession after site abandonment.



Document Availability at the Time of Submission

Release the entire work immediately for access worldwide.

Committee Chair

Williamson, G. Bruce