In a world where the human population is expected to increase by 50% to 100% over the next 50 years, understanding the ecological, biological and management issues of forest fragmentation is one of the main challenges of conservation biology. Over the past few years, several studies have focused on defining and articulating issues related to forest fragmentation, among them, studies of deforestation in tropical and temperate ecosystems, losses of biodiversity, and changes in physical and ecological processes. NASA’s Earth Science Enterprise (ESE) through its Land Cover and Land Use Change (LCLUC) program has focused on some of the ecological, climatic, and social causes and impacts of landscape conversion. Within this multidisciplinary program, regional and local studies focus on using a combination of space observations, in situmeasurements, process studies and numerical modeling to address scientific issues regarding forcing factors, processes, and consequences of land-cover and land-use change. This program is also helping to develop the capability for repeated global inventories of land cover from space through such activities as the Global Observation of Forest Cover (GOFC) initiative under the auspices of the Committee on Earth Observation Satellites (CEOS).

Currently, at least two topics of interest to the conservation biology community, related to the activities of the LCLUC program, would benefit from increased attention by NASA’s ESE: 1) The impact of forest and landscape conversion on biodiversity, and 2) the means of managing and conserving landscapes for continued and sustainable biodiversity. The symposium sought to begin to address these topics with the intention of generating a set of recommendations for possible incorporation into the NASA ESE program. This summary discusses issues related to forest fragmentation and its impact on biodiversity, addresses relevant remote sensing tools and other baseline data for mapping and monitoring forest fragments, and highlights those activities that can be readily incorporated into the NASA ESE program.

Natural landscapes, in particular forests, have been under severe pressure as a result of human population increase and associated economic development. Human impacts on the biotic richness and ecological roles of forests have included: clearing, burning, logging, and the resulting fragmentation. In recent decades, the rate of deforestation has increased in tropical regions where species richness is highest. The immediate consequences of large-scale forest loss are corresponding declines in biodiversity, massive soil erosion, siltation of streams and destabilization of watersheds, loss of sustainable forest use, and threats to indigenous peoples (Laurance and Bierregaard, 1997).

At the symposium, William Laurance of the Smithsonian Institution described three main proximate causes of forest fragmentation in the Amazon region:

In addition, there appears to be an alarming synergism between land-use patterns and climate change scenarios. For example, drought conditions prompt large and frequent fires in Amazonia, which are unusual in a rainforest environment. Recent research by Nepstad et al. (1999) shows that surface fires burn large areas of standing forest that have been affected by logging. The thinning of forests by removing certain valuable trees can make the forest vulnerable to fires in dry seasons or during severe droughts that can result from climate variations (such as El Nino episodes). Regional-scale water balance models also indicate that fragmentation can cause significant impoverishment of forests and increase their vulnerability to future burning.

Descriptions of the Biological Dynamics of Forest Fragments Project in the Central Amazon by William Laurance and also by Claude Gascon of Conservation International's Center for Applied Biodiversity Science, highlighted key issues for the study of forest fragmentation and for relating it to biodiversity and conservation issues. The results from this project demonstrated that fragmentation has an impact on forest dynamics and ecological functions through: 1) increasing rates of tree mortality and damage in fragmented forests, 2) decline in biomass near forest edges and emission of CO2 and other greenhouse gases, and 3) changes in diversity of fauna and flora near forest edges. Forest edges are littered with dead, dying, and damaged trees. They succumb to drought stress, but remain standing until toppled by wind shear, which is enhanced at the forest edge by the denuded surrounding landscape. This strong wind vector makes the trees more vulnerable to the winds’ punishing effects. Often, twice as many trees are seen dying at the forest edge as opposed to within the forest interiors. An unanswered question is whether the edge is static or is increasing in depth. It can be hypothesized through observation that the fragments are shrinking and that disturbances may produce additional disturbances.

Fundamental changes in forest dynamics as a result of fragmentation can also have a number of implications for forest ecology. As a result of edge effect, vegetation structure changes and forest plots start losing biomass. In the first few years after fragmentation, there is a striking loss of biomass and an upsurge in tree mortality. Subsequently, the secondary vegetation recovers and a new equilibrium is reached between mortality and rejuvenation, albeit within a scrubbier forest that harbors less biomass. Simulations attempting to estimate the importance of this collapse in biomass suggest that Amazonian forest fragmentation is producing 3 to 15 million tons of carbon per year. Fragmentation of tropical forests on a global scale is modeled to produce between 20 to 150 million tons of carbon annually. When placed within the context of global warming, forest fragmentation impacts are not trivial.

This decrease in biomass also leads to dramatic changes in overall forest composition. Dense, tangled vines (liana) become more common near forest edges, where the vines compete with trees for nutrients, water, and light. Moreover, by growing up tree trunks, liana place dynamic physical stress on trees. While the increased growth of liana may compensate somewhat for the loss of biomass, it does not offset the impact of fragmentation. In fact, the compensation effect accounts for only 12% of the biomass lost as a result of tree mortality.

Changes in forest structure affect various taxa including amphibians and undercanopy birds. When forest edge spaces expand, certain species that naturally inhabit the forest edge gain an advantage over their competitors.

A mathematical model that is used to gauge the percentage of the fragment area that is affected by edge effects shows that it increases dramatically on plots below 400 hectares. Data also suggest that the impacts of edge effects are detectable up to 300 meters away from the edge. More sensitive analyses are needed to study the complete influence of edge effects on forest ecology. As fragmentation and forest-nonforest edges are currently increasing in tropical forests, the need for suitable remote sensing data in monitoring and assessing the impacts of fragmentation becomes clear. A review of current remote sensing capabilities suggests that sensors such as Landsat and synthetic aperture radars can be used to map and monitor changes in forest cover and to some degree the intensity of fragmentation. However, none of the existing systems is suitable for detecting changes associated with edge effects.

David Wilkie of Boston College and U.S. AID’s Central African Regional Program for the Environment (CARPE) emphasized the impact of logging on forest and habitat fragmentation. He described a series of threats to biodiversity that are currently occurring in the Congo basin. The Congo basin is the second largest tropical forest after Amazonia and it has a different history and different trajectories of disturbance. For example, the “fishbone” disturbance patterns often evident from images of South America are not apparent in the Congo Basin. Furthermore, our knowledge of forest fragmentation and the large-scale distributions of fragmentation in the region are limited. In 1999, for the first time, the area’s forest cover was mapped completely.

Logging threatens the greatest percentage of the forest area. However as most concessions are involved in highly-selective, old-growth forest mining (i.e., tree biomass accumulated over hundreds of years is being liquidated at a rate that far exceeds production rates), present impacts are primarily due to forest fragmentation and defaunation, and not land-cover conversion. The size class distribution and regeneration capacity of economically-valuable tree species, and the future of plants and animals dependent on these tree species, are certainly affected by selective logging that often removes over 95% of all commercially-valuable individuals of over 60 cm diameter at breast height.  However, the impact of logging-based forest fragmentation on woody biomass, carbon storage, plant and animal species composition and abundance, and animal behavior is not well understood.

In the Congo Basin, the rate of deforestation is low at present.  However, forests are being rapidly fragmented and degraded as a result of logging. Loss of woody biomass and changes in species composition and abundance within fragmented forest is a concern from both the global climate change and biodiversity conservation perspectives. The key factor driving forest fragmentation and degradation, and thus the key to near-term management of forests in the Congo Basin is logging.  Logging is particularly critical to protected area management because logging concessions bordering national parks provide essential habitat to wide-ranging species, and logging is the primary land use within corridors that connect isolated protected areas.

Poor nations within the Congo Basin are looking to their forests as sources of revenue to fuel economic expansion. At present, timber extraction generates more money per unit area of forest for national treasuries than other uses of forest, such as agriculture, Non Timber Forest Production (NTFP), and tourism (this is not to deny that NTFPs presently may generate more revenue for local communities than does timber). Consequently logging is and will continue to be an important component of national economic development plans.

Old-growth timber mining is sweeping across the region and there is probably less than two decades of unlogged forest remaining in most nations (Gabon and the Democratic Republic of the Congo may be exceptions).  During the mining phase access to forests is opened through the construction of logging roads and subsequently bushmeat flows out of the forest to urban markets. Forest fragmentation as a result of road construction increases the edge to surface area ratio of forest patches and may result in the massive biomass collapse seen in the Amazon. Once the old growth mining phase of logging is over we expect that the volume of timber that can be harvested from the forest per unit area will decrease (unless markets for new species appear) to sustainable levels (i.e., where harvesting equals production). If we assume that wood production is the primary land use within concession forests, then markets will determine the species that will be exploited and the size class that is harvested. Markets may also drive silvicultural practices to enrich the forest with valuable species thus effectively increasing the homogenization of the landscape, and the fragmentation and degradation of "natural" forests.

In order to understand and track the impacts of logging over time, the challenge will be to develop the tools and models needed to monitor forest fragmentation and forest patch degradation, particularly as the logging industry adapts to changing market conditions, and the progressive decline in old growth forest.  The challenge to those managing the impacts of logging will be to balance the economic utility gained from logging with the loss of biodiversity likely within forests utilized primarily for timber production.

Mapping and monitoring areas affected by logging represent a challenge for remote sensing techniques. Though the initial forest openings created by logging and timber extraction may be visible in satellite data, they are often misclassified due to their being covered over by regenerating vegetation. Better estimation, through remote sensing imagery, of the area and intensity of logging is important for understanding the impact of fragmentation and designing mechanisms for logging management and conservation.  Technological advances may one day make it possible to track illegal logging in protected areas around the world.

Claude Gascon emphasized the importance of educating NASA as to the needs of conservation biologists.  In order to manage landscapes effectively for conservation, conservationists need more details about landscape elements that are important to the dynamics of forest fragments.  A lot of traditional field conservation has taken the approach of defending landscapes through protected areas; however, there is now the growing realization that protected areas are vital, but left alone, they are not a viable long-term strategy. In fact, the empirical evidence suggests that much of what happens inside a protected area is dependent upon what is going on outside of the protected area.  Therefore, an emerging approach toconservation is large-scale conservation planning or corridor-based management as opposed to site-based.

It is known from research at Biological Dynamics of Forest Fragments Project sites that corridors can represent viable habitats and that certain species use these corridors for movement.  Linear forest remnants (corridors) of approximately 150 meters wide may be important in maintaining connectivity in a landscape for small mammals and litter frogs.  There may also be a correlation between corridor width and remaining species richness in a connected area.

The type of habitat that surrounds forest patches, as defined by the concept of matrix habitat, is now considered to be extremely important to the patches’ ability to maintain biodiversity.  Forest species are often found to use, to some degree, this surrounding matrix habitat, which can be pasture or low-growth forest. However, matrix habitats can also harbor invasive species.  One study found that 20% of observed species in an analyzed landscape were not found within it when it was continuous forest.  These species had presumably entered the area through patches of disturbed habitat or from adjacent ranches.  As more fragmentation occurs along a corridor, more new species invade.

The importance of the matrix habitat can be assessed in terms of which species can persist in the fragment.  Species abundance in the fragment is compared with its abundance in the continuous forest and both are then correlated to the species abundance in the matrix habitat.  If a high correlation is found, then, the species that are most able to use the continuous forest habitat are those species that are the most persistent in the matrix habitat.  For vertebrate groups, the matrix plays an important role in determining those that can best survive in the fragment.

During the meeting, the application of several remote sensing data sets to the study of forest fragments was discussed.  It was realized that the role of remote sensing goes farther than mapping land cover types.  Several spatial, temporal, and structural attributes of forests were identified as foci for remote sensing studies.

United Nations climate change initiatives arising from the meeting in Kyoto, Japan established economic incentives to sequester biomass. Effective implementation of these initiatives requires cost-effective and high-resolution remote sensors. Howard Schultz of the University of Massachusetts reported on the Environmental Monitoring Project, a collaboration between the University of Massachusetts and the Massachusetts Forest and Wildlife Management Service.  The project’s research goal is to apply computer science to forestry science and thereby improve methods for determining how much biomass a forest contains.  The scheme of data organization merges data from high-resolution video cameras with satellite imagery.  It was suggested that for detailed mapping of forest fragments and edge structure characteristics, digital videography, combined with laser altimetry, are the most suitable tools.  Recent data acquired by the University of Massachusetts over various forest types in Amazonia suggest that the data assembled can be used for classifying trees and providing estimates of tree height.  To estimate biomass, the system maps the forest as a simple model of poles and circles that symbolically represent heights and types of trees.  Crown diameters are handled by high-resolution video. This measure, when compared with field-derived measures of biomass demonstrated acceptable correlation.

The group also discussed the capability of high-resolution radar data to provide information about forest structure and biomass that can be used in fragmentation studies.  Donat Agosti and Sasan Saatchi presented results from the Atlantic Coastal forest of Brazil where radar data were effective in separating forest fragments from tree crops such as cocoa plantations.

Julie Robinson of the NASA Johnson Space Center also described how astronaut photographs taken from the Space Shuttle and other platforms can be used as remote sensing data.  Although the images are more variable than more widely-used satellite data, the 3-band (red, green, blue or infrared, red, green) images from hand-held cameras in orbit were useful in studying forest conversion and fragmentation at scales of 10 to 60 meter pixels. The length of record (since the 1960s) and public availability of images can make them an important ancillary data source.

The members of the breakout group recognized the importance of scale and the use of multiscale analysis for extrapolating our understanding of individual forest fragments to landscape and regional scale studies. This aspect of the work could directly link NASA's ESE capabilities to problems in the discipline of conservation biology. Remote sensing data, in conjunction with modeling and scaling exercises, should allow us to improve our ability to understand the impacts of forest fragmentation and develop conservation and policy plans.

The conservation biology community has immediate need for certain spatial and temporal remote sensing information.  These data needs present challenges to NASA and call for new areas of focus within its ESE program.

back to main

• Privacy Policy & Important Notices
• Contact Support
• Download Adobe Reader
• Accessibility

• NASA Official: Jon Ranson
• Curator:Leanne Kendig
• Last Updated: