| The main objective of this dissertation is to explore the ecological relationships between lianas and trees, and the consequences of these relations for forest management. First, I describe the tree and liana communities of Oquiriquia, a tropical forest concession in Bolivia. Second, through a liana cutting experiment, I evaluate whether lianas reduce tree growth of their hosting trees by interfering with water availability during seasonal periods of water scarcity. Third, I assess whether lianas affect two aspects of tree regeneration, namely tree seedling survival and growth, and evaluate whether cutting lianas affects soil moisture and canopy openness. Finally, I quantify the cost of liana cutting as a silvicultural treatment, and evaluate its efficiency in terms of reducing liana density and number of liana infested trees.; The forest in Oquiriquia concession apparently has the highest liana density so far reported in the literature (ca. 2500 individuals {dollar}ge{dollar} 2 cm in diameter per ha). Frequency of liana infestation was also high, with 86% of trees carrying at least one liana. Liana species showed no association with particular tree species. Trees climbed by one liana had a higher probability of a second liana infection than expected based on the Poison distribution. In Oquiriquia, lianas that climbed using one of four different mechanisms climbed a similar number of trees. Larger diameter lianas climbed more trees than slender lianas. Plots with high liana density also had high palm density. Tree diameter was the most important variable associated with the number of lianas per tree, but the number of neighboring trees in the 10-30 cm dbh size class also affected this relationship.; Liana cutting around adult Senna multijuga trees resulted in increased water availability, and increased growth rates in stem girth over a period of one year. Liana cutting resulted in increased growth in height and total number of leaves produced in seedlings of Clarisia ilicifolia and Astronium fraxinifolium. Liana cutting did not affect gravimetric soil water content, and it increased canopy openness only after 26 months. Increased tree seedling growth in liana-cut plots, though statistically significant, was modest in comparison to increases in growth rates observed in seedlings growing in tree fall gaps. Tree seedling survival, and tree and liana seedling densities were not affected by liana cutting.; Liana cutting as a silvicultural practice proved effective in reducing liana density by 95%. After one year, 22% of liana stumps resprouted, most stumps producing two resprouts. Liana cutting also proved expensive, with a cost of 24 h per ha of forest. Thus, liana infestation should be prevented by usage of reduced impact logging practices, in which liana cutting could be included. (Abstract shortened by UMI.)... |
