Tropical Forest Diversity and Dynamism: Findings from a Large-Scale Plot Network

Tropical Forest Diversity and Dynamism: Findings from a Large-Scale Plot Network


View All Available Formats & Editions
Usually ships within 6 days

Product Details

ISBN-13: 9780226493466
Publisher: University of Chicago Press
Publication date: 10/28/2004
Edition description: 1
Pages: 688
Product dimensions: 6.00(w) x 9.00(h) x 1.40(d)

About the Author

Elizabeth Losos is director of the Center for Tropical Forest Science at the Smithsonian Tropical Research Institute in Washington, DC. Egbert Giles Leigh Jr. is a biologist with the Smithsonian Tropical Research Institute in Balboa, Panama.

Read an Excerpt

Tropical Forest Diversity and Dynamism
Findings from a Large-Scale Plot Network

The University of Chicago Press
Copyright © 2004 The University of Chicago
All right reserved.

ISBN: 978-0-226-49346-6

Chapter One
The Growth of a Tree Plot Network Elizabeth C. Losos and Egbert G. Leigh, Jr.

On the Malay Peninsula, mature tropical forests reach 60 and even 70min height and shelter a dizzying array of mammals, birds, reptiles, and insects. These southeast Asian forests are also host to some of the most profitable timber stands in the world. Throughout the last century and up to the present, foresters have grappled with the dilemma of extracting timber without diminishing the productivity of these lush forests. This pursuit has drawn attention to several fundamental questions: What factors allowed these diverse, productive tropical forests to arise in the first place? How does human intervention-such as logging or conversion to agriculture-upset a forest's balance? Do the natural mechanisms by which tropical forests maintain their productivity and diversity offer us lessons on how to conserve or manage tropical forests more effectively?

On the other side of the world, researchers have been struggling with similar questions in neotropical forests. On Barro Colorado Island (BCI), Panama, biologists working at the Smithsonian Tropical Research Institute have long been seeking to understand what factors maintain the diversity and productivity of flora and fauna in that Central American forest. How did diversity arise in that tropical forest? Has specialization to the island's physical features allowed the coexistence of these diverse plants and animals? What role have history and chance played in that forest's development?

In the last three decades, real strides have been made toward addressing such questions, with Barro Colorado Island playing a key role in this advance. To discern more clearly what processes maintain the diversity of the island's trees, Stephen Hubbell and Robin Foster initiated a study of a 50-ha permanent plot within the island's mature forest. In this "Forest Dynamics Plot," all free-standing, woody trees and large shrubs over 1 cm diameter at breast height (dbh) were tagged, measured, mapped, identified to species, and repeatedly censused. Two decades, almost 350,000 tagged trees, and hundreds of publications later, findings from this 50-ha Forest Dynamics Plot, together with other research on BCI, have provided perhaps the best foundation for understanding the structure, dynamics, diversity, species distribution, and interactions of plants, animals, and fungi in the lowland tropics. In chapter 2, Hubbell describes how BCI's 50-ha plot has transformed ecologists' thinking on topics such as the stability of species composition within tropical forests, the influence of adult abundance and distribution on conspecific sapling recruitment, and how quickly tree population sizes change in response to the changing climate.

How well do findings from BCI's 50-ha plot apply to tropical forests elsewhere in the world? Certainly, this Panamanian forest has some unique features: The island's biota reflects the interchange of plants and animals over a land bridge that connected the Americas 3 million years ago; no such event influenced the biotas of the other tropical areas in Africa or southeast Asia. Twelve thousand years ago, the isthmus lost its megafauna. More recently, during the construction of the Panama Canal in 1910, Barro Colorado was severed from the adjacent mainland and transformed into a 1600-ha island. Climatically, BCI has a formidable dry season; its median rainfall for the year's first quarter is only 100 mm-far less rain than everwet tropical forests experience. Barro Colorado Island also tends to be far less diverse than everwet forests. On BCI, the trees over 10 cm dbh in a 25-ha subplot of the Forest Dynamics Plot account for 210 species, whereas the 25-ha subplots of everwet forest at Yasuni in Amazonian Ecuador and Lambir in Sarawak, Malaysia, respectively, contain 822 and 920 species of tree over 10 cm dbh. What causes these differences and how much do these differences matter when researchers analyze the factors attributable to tropical diversity?

To find out, in 1986 Peter S. Ashton and Salleh Mohammed Nor convinced the Forest Research Institute of Malaysia, Harvard University, and the Smithsonian Tropical Research Institute to establish a comparable 50-ha Forest Dynamics Plot in the Pasoh Forest Reserve, peninsular Malaysia. To create comparable datasets, the trees in the Pasoh plot were tagged, mapped, measured, and recensused using an identical protocol to that developed by Hubbell and Foster for the BCI plot. While both plots lie on relatively flat terrain, Pasoh has only 1 month per year that averages less than 100 mm rainfall, compared to BCI's 3; Pasoh's 50-ha plot contains 678 species of tree over 10 cm dbh, compared to BCI's 226. Do the same processes maintain tree diversity on both plots? Comparisons of the two Forest Dynamics Plots revealed that Pasoh has a far smaller number and far fewer species of pioneer trees specialized for colonizing large treefall gaps than does BCI. Thus specialization to treefall gaps of different sizes-a popular explanation for the diversity of tropical trees-could not explain why Pasoh's tree diversity is somuch higher than BCI's. Increasing evidence from Pasoh indicates that sapling recruitment is diminished near conspecific adults, though nomore so, apparently, than on BCI. And so the pursuit to explain why Pasoh's tree diversity is higher than BCI's continues.

The third 50-ha Forest Dynamics Plot provided a drier contrast to BCI. In 1987, R. Sukumar of the Indian Institute of Science established a large plot at Mudumalai Wildlife Sanctuary in southern India. The driving force behind this plot was the newly created Nilgiri Biosphere Reserve and World Heritage Site-of which Mudumalai was a part-and the need for rigorous information on forest regeneration to develop appropriate management plans for the reserve. Mudumalai averages only about 1200 mm of rain per year and its dry season lasts 4 months, while Barro Colorado Island averages 2600 mm per year and has a 3-month dry season. Mudumalai's 50-ha plot has only 63 species of tree over 10cm dbh, compared to BCI's 226. Past logging allowed grass to take over Mudumalai's understory, providing fuel for frequent anthropogenic fires. Recent evidence has shown that many species recruit only when several years have passed without fire. BCI, by contrast, has suffered no fires in at least the past 80 years. Moreover, Mudumalai has elephants and wild cattle, which are far larger and more common than BCI's biggest herbivore, the tapir. Animal seed dispersers and seed predators play a less active role in Mudumalai, where most of the common species do not produce fleshy fruits attractive to frugivorous mammals or birds. Given these differences, can the processes shaping forest structure, dynamics, and diversity at Mudumalai bear any resemblance to those at work on BCI? Tree species composition is less stable and tree density is lower at Mudumalai than BCI, in part because of heavy and selective browsing by elephants and other mammalian herbivores at Mudumalai. An even more striking difference involves patterns of tree mortality. Barro Colorado Island resembles many, if not most, moist and wet tropical forests in that tree mortality is the same in nearly all size classes between 7 and 70 cm in trunk diameter. In the dry deciduous forest of Mudumalai, however, mortality is higher for smaller trees: a tree must attain a trunk diameter of 20 cm or more to be reasonably safe from fire and elephants. Moreover, unlike BCI, Mudumalai offers little evidence to link habitat specificity or density dependence to the regulation of its tree populations.

Both the similarities and the differences among these first three plots aroused widespread interest in establishing large-scale Forest Dynamics Plots in other tropical forests. As a result, in the late 1980s and early 1990s, plots were established in Sarawak, Thailand, Puerto Rico, Sri Lanka, Democratic Republic of Congo, Cameroon, Ecuador, Colombia, Philippines, Taiwan, and Singapore. In 1992, under the guidance of its director, Ira Rubinoff, the Smithsonian Tropical Research Institute (STRI) took the lead in establishing the Center for Tropical Forest Science (CTFS) to coordinate the rapidly growing network of Forest Dynamics Plots and widely disseminate plot findings. Since that time, CTFS has represented a voluntary global consortium of forestry agencies, research institutions, universities, and nongovernmental organizations, each involved in one or more Forest Dynamics Plot. The mission of the center is to promote and coordinate long-term research in the natural and social sciences based on standardized data from Forest Dynamics Plots and to translate these findings into information relevant to tropical forest conservation, management, and natural resource policy. In pursuit of these goals, over the last decade CTFS has developed a standardized forest censusing protocol; provided assistance in field training, data management, and data analysis to Forest Dynamics Plot programs; promoted communication within and outside the network; and catalyzed conservation and management applications of the plot research. By 2003, CTFS included more than three dozen research institutions and hundreds of scientists at 16 Forest Dynamics Plots in tropical forests spanning the globe. The network is nowmonitoring nearly 3 million trees of about 6000 species.

Due to the intensive sampling protocol of Forest Dynamics Plots, each individual plot supports detailed investigations that address the maintenance of tree diversity within its local setting. Researchers at each plot can examine, for example, whether tree composition is due to the specialization of individual tree species to habitats or light gaps, whether it is related to the concentrated effect of pests on seeds and saplings near their parents, or whether it is more attributable to the site's biogeographic history. Many plots are well suited for asking additional questions particular to their own forests: The Forest Dynamics Plots in Puerto Rico, Taiwan, and the Philippines are good places to investigate the effect of hurricanes or typhoons on the structure, dynamics, and diversity of these forests. The plots in the Congo's Ituri Forest were chosen to determine what factors allow a single tree species to dominate the canopy and why this happens in some areas but not others. Elsewhere in Africa, Cameroon's Korup National Parkis distinctive for its nutrient poor white sandy soils. Thailand's Huai Kha Khaeng and India's Mudumalai plots are well placed to investigate the influence of large herbivores and fire on tropical forests. With its high endemism, Sri Lanka's Sinharaja provides an opportunity to evaluate the dynamics of an island flora. The plot in Sarawak's Lambir Hills National Park is ideal for assessing the effect of heterogeneity in soil and topography on the distributions of different tree species. Montane forests can be explored in Thailand's Doi Inthanon and Colombia's La Planada.

Together, these plots can do even more. The network of Forest Dynamics Plots provides opportunities for global comparisons and the synthesis of research that could not be accomplished through any individual plot. For example, what climatic feature best predicts tree diversity in tropical forests? One consistent pattern that arises throughout tropical forests, whether in America, Africa, or southeast Asia, is that diversity is greater in more aseasonal climates. Yet, despite their rainy everwet conditions, tree diversity is much lower in islands such as Puerto Rico or depauperate areas such as the Deccan plate of India and Sri Lanka. What conditions must be satisfied for a tropical forest to attain the diversity appropriate to its climate? Similarly, how does soil quality influence the ways trees die, the role of pioneer species in forest dynamics, the forest's turnover rate, and the diversity of tree species? What factors influence species distributions? To facilitate global analyses of the features thought to most likely influence species composition and diversity, CTFS has selected a core set of Forest Dynamics Plots that represent a range of climatic conditions, soil topography and geology, and natural disturbance regimes. The introduction to part 2 describes the rationale behind the plot selection in greater detail.

This book represents the first attempt to assemble comparable data from the individual Forest Dynamics Plots. After reviewing the two-decade history and findings of the "mother plot" on Barro Colorado Island (chap. 2), this volume presents what we now know about the entire network of plots. It surveys basic features of tropical forests within the plots-forest structure, diversity, species accumulation, floristics-relating them to global variation in climate, biogeographic history, natural disturbances, and soil quality and topography (part 2). The volume then turns toward more detailed analyses of individual plots, illustrating the depth and range of information generated by Forest Dynamics Plots on tree diversity, seed dispersal limitation, canopy disturbance, fire response, canopy monodominance, pollination-mediated population dynamics, and forest fragmentation (parts 3-6). With the exception of a handful of chapters, most of these studies focus on an individual plot. All, however, present methods, analyses, and findings that can be readily transferred to other Forest Dynamics Plots and sites around the world. Finally, this book documents standardized qualitative and quantitative baseline information for each plot (part 7). These data, for usewithin this volume and beyond, provide the framework for assessing and comparing how biotic, abiotic, and stochastic factors affect the dynamics and diversity of tropical forests worldwide.

Chapter Two
Two Decades of Research on the BCI Forest Dynamics Plot

Where We Have Been and Where We Are Going

Stephen P. Hubbell


In community ecology research, there is perhaps no greater challenge than explaining the origin and maintenance of biodiversity in tropical rainforests, which are arguably the most species-rich ecosystems in the world. In particular, the tree component in the richest of these forests can be remarkably diverse, sometimes exceeding 1100 tree species in a single 25- or 50-ha plot, such as the 25-ha Yasuní Forest Dynamics Plot in Amazonian Ecuador or the 52-ha Lambir Forest Dynamics Plot in Sarawak, Malaysia. To put these plots' species richness in a global perspective, 1100 species is approximately double the entire tree flora native to continental North America north of Mexico, an area 40 million times larger. In this chapter, I summarize a few of the most salient results from23 years of research on these questions in one particular tropical forest, Barro Colorado Island (BCI), Panama. Then, based on this experience, I offer a few suggestions for future research directions for the BCI Forest Dynamics Plot, as well as more generally for the whole set of Forest Dynamics Plots coordinated by the Smithsonian Tropical Research Institute's Center for Tropical Forest Science (CTFS).

Robin Foster and I began the first 50-ha Forest Dynamics Plot of the CTFS network on BCI in 1980 (Hubbell and Foster 1983). At that time tropical forest ecology was still using methods designed for the temperate zone that we considered inadequate for studying community-level questions about tropical forests. Tropical moist forest tree communities are extraordinary not only in species richness but also in the extreme rarity of many of their tree species. In 1980, the standard tropical forest study plot was 1 ha or smaller. We assumed that we would need a larger plot if we wanted to obtain demographic information at the individual species level rather than pooling together, as is necessary when using small plots. The final choice of 50 ha was an estimate, however, because the 50-ha BCI plot was the first of its kind, and of course we did not know BCI tree species abundances in advance. Retrospectively, we now know that although a single hectare contains about half the species found in 50 ha, most are too rare for species-level analysis. For example, suppose we take 100 individuals as a reasonable minimum number of trees for statistical analysis of survival and growth. Then 10 or fewer species meet this abundance criterion in a single hectare on BCI, even including small trees down to 1 cm diameter at breast height (dbh). Approximately half of the species in the 50-ha BCI Forest Dynamics Plot, however, meet this abundance criterion. In fact, this is true for five of the six Forest Dynamics Plots that are 50 ha or more in size.


Excerpted from Tropical Forest Diversity and Dynamism Copyright © 2004 by The University of Chicago. Excerpted by permission.
All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.

Table of Contents

Part 1: Introduction
Part 2: The Whole Is Greater Than the Sum of the Plots
Part 3: Habitat Specialization and Species Rarity in Forest Dynamics Plots
Part 4: Local Variation in Canopy Disturbance and Soil Structure
Part 5: The Diversity of Tropical Trees: Background
Part 6: The Diversity of Tropical Trees: The Role of Pest Pressure
Part 7: Forest Dynamics Plots
List of Contributors

Customer Reviews

Most Helpful Customer Reviews

See All Customer Reviews