We did not inherit the earth from our ancestors - we borrow it from our children
In the end we will conserve only what we love; We love only what we understand;
We understand only what we are taught.
What is Forest Fragmentation?
Simply put "Forest Fragmentation" is what happens when large contiguous patches of forests are fragmented, or split up, into several smaller patches. These remaining patches are separated by what is defined here as the "matrix" which is just anything other than mature forest and may inlcude clear cuts, development or young plantation forests. To conserve biological diversity (biodiversity) it is believed that these patches must have connectivity between them, based on the principles of landscape ecology.
What is Conservation Biology and Landscape Ecology?
Conservation biology can be defined as: "an emerging discipline that deals with the basic issue of eroding biological diversity...it derives its theoretical basis from the pure sciences, such as population genetics, demography, biogeography and community ecology. However, it uses the resulting principles to address applied problems in conservation, such as the loss of genetic diversity, loss of species diversity and loss of diversity of ecosystems." (Temple et al., 1988)
The discipline of landscape ecology can be considered to be either a subdivision or a complement to conservation biology. The basis of landscape ecology is a consideration of the composition and relative availability of habitat fragments, their spatial arrangement or geometry, and the way in which they are connected and used by species (Probst and Weinrich, 1994). The first step of any landscape ecological study consists of abstracting the landscape into patches, corridors, and the surrounding matrix (see the glossary for a detailed definition of these terms). The development of landscape ecological theory in the 1970s and 1980s drew upon studies showing that many animals use naturally occurring corridors between habitat patches when traveling through a human-dominated landscape matrix (Wegner and Merriam, 1985), and that corridors can enhance the persistence of populations (Fahrig and Merriam, 1985). These studies, and many more like them, demonstrated the importance of connectivity between natural elements in a landscape for determining the persistence and success rates of species in a fragmented landscape. Many similar studies analyze the dynamics of "metapopulations", which involves the study of recolonization and extinction events of local subpopulations within a set of partially isolated populations belonging to the same species (Hunter, 1996).
The disciplines of landscape ecology and conservation biology are still quite young, and the development and refinement of their theories and principles have also at times been rife with controversy since their conception (Jarvinen, 1982; Soule and Simberloff, 1986; Willis, 1984). However, several key prescriptions concerning landscape patterns for the purpose of biodiversity conservation are now widely accepted by scientists and resource managers alike (Noss and Cooperrider, 1994). Some controversy still inevitably persists today, and it is unlikely that such a discipline will ever come up with absolute, proven principles, simply because of the sheer complexity of ecological systems, and our inability to establish proper controls for rigorous scientific testing. However, when applied properly, these accepted principles presented below hold the promise of maintaining biodiversity at least more adequately than any other existing landscape management system. These commonly accepted basic ecological principles for landscape management and/or reserve design at the regional level have been adapted from Diamond (1975), and Wilcove and Murphy (1991), and can be summarized as follows in the sections below.
Representativeness
Existing natural patches should encompass the whole range of ecosystem types occurring in a given area. When looking at the fragmentation of ecosystems, an adequate representation of remaining fragments among ecosystem types would have less of a resulting impact on ecosystem structure and function. This would also have a lower impact on the species within an ecosystem if the remaining fragments are adequately distributed among the ecosystem types of the region, as it would allow for a more diverse range of ecosystems and habitats to be represented, which, in turn, would allow for a greater range of species to be represented. Furthermore these remaining fragments would provide source populations of seeds, larvae, propagules and emigrants that would be crucial for the development of diversity in the surrounding regenerating forests.
Patch Size
Patches of habitat that are larger are superior for achieving conservation goals than those that are smaller for several reasons. First of all, large areas will likely contain a larger amount of environmental heterogeneity. Thus, larger patches of habitat should offer more available niche spaces, and consequently will contain a higher number of species. Second, some species will likely be absent from small reserves even when an appropriate niche is available, either because they require large home ranges (ie: carnivores), or because they occur at low densities and are therefore unlikely to be in a small reserve simply due to chance effects (ie: rare plants), or because they are a small species which behaviorally prefers to select habitat patches that are large enough to support other members of their species (Stamps, 1991; Robbins et al., 1989; Hunter, 1996).
Large patches also ensure proper functioning of the ecosystem patch in question, since small patches tend to be more susceptible to external disturbances due to a high edge/interior ratio (see section 5.5.3). As a consequence, larger patches provide more habitats for forest interior species that cannot survive at forest edges, and forest interior species will thus survive and thrive better in larger patches. This is part of the reason behind the view that larger patch sizes will reduce the risk of extinction of individual species (Diamond, 1975). One other reason for this is that larger patches tend to harbour larger population sizes, which are less likely to go extinct (Soule, 1986). The smaller the fragment size, the more likely extinctions will occur in the remaining fragments since reduced population sizes make species vulnerable.
Finally, larger patch sizes have a lower susceptibility to catastrophic events, since most catastrophes tend to be of a local scale and thus cannot affect an entire large patch (Hunter, 1996). As an empirical example of these concepts, studies done on the effects of fragmentation on Amazonian dung beetles found that the negative impacts on these populations increased with decreasing fragment size (Klein, 1989). Furthermore, forest interior species, such as many of the beetles that were studied, were found to be highly prone to extinction in small fragments (Klein, 1989). Similar studies on other taxa have also shown the same results. For example, forest interior bird species are only found in the largest fragments (Newmark, 1991). This shows that preservation of areas large enough to preserve forest interior species is an important conservation goal.
Fragments of forest, providing that they are large enough so that extinction is not a serious concern, may be able to provide a temporary refuge for species long enough until the surrounding matrix grows back and is able to support those species (Kellman, 1996). This could be an important consideration providing that the surrounding matrix is allowed to grow back to conditions that support sensitive species, such as those that are old-growth dependent.
Edge/Interior (E/I) Ratio
Forest edges are characterized by biophysical conditions that are different from the forest ecosystem in the interior of a patch (Chen et al, 1992; Hobbs and Humphries, 1995). From a conservation perspective, it is a very important fact that forest edges are relatively rare in undisturbed nature, but are very common in today's modified landscapes (Franklin and Forman, 1987; Noss and Cooperrider, 1994). This has resulted in serious ecological changes that may have significant impacts upon forest interior species. It is commonly accepted that most threatened species tend to be those that are disfavoured by the creation of edge habitat (Franklin and Forman, 1987; Hunter, 1996; Noss and Cooperrider, 1994). Therefore, in terms of conserving biodiversity, forest patches with a higher percentage of edge habitat are undesirable.
For the most part of this century, land managers and wildlife managers considered edges as beneficial to wildlife because species diversity generally increases near habitat edges (Yahner, 1988). However, edges can have negative consequences for wildlife by modifying distribution and dispersal and by increasing the incidence of nest predation and parasitism (Yahner, 1988), as well as being detrimental to species requiring interior habitat (Harris, 1988). The relationship between edge effect or edge width and wildlife communities has been poorly studied and poorly documented (Forman and Godron, 1986). However, it has been determined that edge effects can be characterized well as a function of edge length, as opposed to edge width (Yahner, 1988). The magnitude of the edge effect is inverse to the habitat quality of the adjacent patches, so the dysfunction of edges as ecological traps (ie: increased parasitism and predation) is probably accentuated in secondary habitats, and by practices such as plantation forestry (Harris, 1988) which may produce a habitat of poorer quality than what was there previously.
Given that true edges are rare in nature a low E/I ratio is most desirable. Therefore, a single large reserve (above left) is better than a group of smaller ones of the same total area (above middle), because of the large E/I ratio in the latter case. Furthermore, a round reserve of the same size is also better than a long, thin reserve because of the large E/I ratio in the latter case (above right). In addition, long, convoluted boundaries will have a larger total edge effect than will straight and simple boundaries (Franklin and Forman, 1987).
Previous studies have shown that edge effects on the microclimate can extend between 5 and 250 m, depending on the type of forest and azimuth (Burke and Nol, 1998, Fraver 1994, Chen et al 1992, Laurance 1991a). Furthermore, edge effects on wildlife can extend further into the patch then the actual changed physical microclimatic parameters. For example, in the eastern U.S., physical edge effects have been documented at about 10-20 m (Ranney et al, 1981). However, population declines of songbirds in the Eastern U.S. in relation to forest fragmentation concluded that the edge-related increase in predation may extend from 300-600 m inside the forest (Wilcove, 1985).
At forest edges, microclimatic variables include light intensity and duration, relative humidity, air temperature, and soil characteristics such as pH, organic carbon levels, total nitrogen, available phosphorous, and soil moisture and temperature (Oosterhorn and Kappelle, 2000). These variables differ strongly over short distances (Laurance, 1991b). Studies suggest that edge effects can penetrate some old growth forests up to 180m, and on hot, windy days, this reduced humidity and increased temperature can penetrate up to 240m into the forest interior (Laurance, 1991a).
It is important for our purposes not to get caught up by actual measures of diversity (e.g: species richness) when referring to forest edges. Actual measured species richness tends to increase as we travel from the forest interior towards the edge (Oosterhorn and Kapelle, 2000; Kunin, 1998). This is mainly due to "spatial mass" effects, in which forest plant species more typical of the interior forest share their space with the pioneer and weedy introduced species at the edge who take advantage of the changed biophysical conditions (Oosterhorn and Kappelle, 2000). It is typical in fragmented systems for forest edge species to be favoured in terms of survival due to a higher amount of forest edge. One serious problem therefore is the proliferation of edge species at the expense of forest interior species. This has been experimentally shown not only for plants (Williams-Linera et al., 1998; Kunin, 1998; Hobbs and Humphries, 1995; Laurance 1991b), but also for beetles (Baldi and Kisbenedek, 1994), for birds in several regions (Buford and Capen, 1999; Germaine et al, 1997), and for mammals (Laurance, 19991). From the perspective of a forest interior bird, a small forest patch may be entirely edge and will therefore be uninhabitable. Another problem is that while certain birds may be attracted to these edge environments for nesting sites, their ability to fledge their young may be drastically reduced due to increased predation and parasitism (Noss and Cooperrider, 1994).
Dispersal Between Patches
Until recently, most species lived in well connected landscapes (Gilpin and Soule 1986; Meffe and Carroll 1997). This served to enable different types of movement of many animals at many scales. These include: the daily movements of animals over their home ranges, the annual migratory routes of animals, the immigration or emigration routes between subpopulations or populations, and, finally range shifts in response to climate change (Hunter, 1996). Thus, an important component of fragmentation studies must consider how changes to the landscape have impacted upon the ability of plants and animals to maintain these movement patterns in the altered landscape.
Dispersal ecology will vary greatly among species; unfortunately, there is a general lack of studies on dispersal of animals because it is a difficult phenomenon to study (Gaines and Bertness 1993). However, we can say that the natural ecological characteristic of regional and local connectivity has been altered by the formation of a patch-matrix-corridor system. Several experimental studies have demonstrated the important role of connectivity in maintaining population persistence. One study showed this for Drosophila in a lab (Forney and Gilpin 1989), another showed it for small mammals (Merriam and Lenoue, 1990), and a third demonstrated it for salamanders (Rosenberg 1994). In any case, in the absence of more complete information, most ecologists would agree that it is safe to assume that a connected landscape is preferable to a fragmented landscape (Beier and Noss, 1998).
In fragmented landscapes, there are three main ways in which some degree of connectivity is maintained. The first is by corridors of forest habitat between patches (or, alternatively, and even better, by long strips of intact natural habitat). The second is by "stepping stones" of forest patches throughout the landscape. The third is simply through the matrix itself.
1. Dispersal Routes Through Corridors
Blocks of habitat inter-connected by corridors are better than isolated blocks. Although direct and extensive experimental scientific evidence in support of corridors as a conservation tool has been weak, many studies have provided evidence that corridors are valuable conservation tools for many different species, whereas direct evidence to the contrary is absent (Beier and Noss, 1998). Connectivity allows for easier dispersal of species. This would be particularly important for nomadic species or species with large home ranges, whose needs cannot be met within the boundaries of a patch or protected area. Connectivity also aids birds and other animals in dispersing from their birth location to their adult home ranges and breeding sites (Newmark, 1991). Dispersal across a region also provides demographic and genetic exchanges between populations or between metapopulations, which is important in avoiding genetic effects such as inbreeding depression, as well as in maintaining the potential for the reestablishment of species in areas from which the species has been extirpated or severely reduced in numbers (Jeo et al, 2000). Finally, enhanced dispersal capability would also be very important for species that must cross distances between patches in order to shift their ranges in response to climate change (He et al 1999)
Many studies predict that immigration rates will be reduced in our current fragmented systems due to the isolation of the species contained within the fragments. Island-biogeography theory predicts that when species are isolated on newly formed islands their subsequent isolation leads to extinction over time (MacArthur and Wilson, 1967). This theory has been applied to the concept of habitat "islands" in our current fragmented systems. However, many feel that this theory is a gross oversimplification that masks fundamental differences in the surrounding matrix, in site history and age, as well as differences in colonization and extinction rates (Klein, 1989 and Kellman, 1996). This is because habitat islands are not strictly comparable to true islands. The surrounding sea of vegetation is dynamic, changing with natural succession and pasture maintenance, and not completely inhospitable to many of the organisms in the habitat island (Bierregard, 1992).
Some believe that for some species, extinction rates might fall after isolation, due to protection from pathogens and predators, and reduced competition from new immigrants. They also point to the fact that species with animal seed dispersers able to survive in fragmented systems will be able to disperse well in these new systems (Kellman, 1996). However, though some notable exceptions such as those mentioned above might occur, in terms of general prescriptions for landscapes, many of the recommendations for managing forests based on island biogeography as a rough guide have proven, through experience, to be valid (Noss and Cooperrider, 1994).
In general, it has been found that forest fragments have fewer species, sparser populations, and in some species, such as beetles, their sizes are even reduced as compared to their sizes in contiguous forests. Studies done on dung beetles in the Amazon showed that these fragmentation effects were observable even though the patches were isolated by a relatively short distance (less than 350m) and for a short ecological time (2-6 years), showing that the effects of fragmentation on populations can be very rapid. However, beetles may be a more sensitive species to fragmentation since those studied rarely entered clearings (Klein, 1989). In addition, we tend to see some convergent patterns in terms of species that are most at risk in poorly connected landscapes: they tend to be species found more commonly in forest interiors and they often tend to be of conservation importance (ie: threatened), or at least less common than other species less affected.
Finally it is important to remember that a single corridor can be problematic due to increased predation so that more corridors are, in general, preferable. Another problem is that narrow, edge-dominated corridors may do more harm than good if favour is given to weedy generalists and encourages the invasion of reserves by these opportunistic species (Noss and Cooperrider, 1994).
2. Dispersal Between Patches
Organisms will often disperse through the matrix in the absence of corridors, as well. Indeed, immigration is expected to be much more common for a given system of forest patches than for a system of islands, which has led to some of the major criticisms of island biogeographic theory (Berry, 1992).
When we are talking about species using "stepping stones" for dispersal, blocks of habitat close together are superior to blocks that are far apart. Patches that are closer together allow more movement of species as species are able to essentially "hop" from one island to another. Patches of habitat clustered compactly are better than long, linear patches because they offer more routes for dispersal of biota. One significant issue is that a single corridor or dispersal route can be very problematic for biodiversity conservation since it allows predators easy access to prey as they move through the corridors which lack significant space and cover for the prey to hide. As the number of corridors increases, this problem is reduced.
3. Dispersal Through the Matrix
In addition, connectivity should also be measured by the degree of ease with which species can move in between patches through the matrix. For some species that enter or even live in clearings, the matrix is not an issue. However, as described in the sections above, the hospitability of the matrix is a serious concern, particularly for forest interior species. For example, a study of forest bird species showed that only 23% of the forest interior species were capable of moving between fragments, as compared to 100% of the forest edge bird species (Newmark, 1991).
Studies done on Australian rain forest mammals also showed that the tolerance of modified habitats (the matrix) was the single predictor of extinction proneness in fragments, once the effect of relative abundance was removed. Species that were able to move through, or even exploit, these modified habitats remained stable or increased in numbers in fragmented habitats, while those that avoided the matrix often disappeared from the fragments (Laurance, 1991). Thus, the ability of a species to survive in fragments depends on its ability to move through the matrix.
This is an important consideration since a large portion of the matrix in tropical countries is highly modified by slash and burn agriculture and conversion to pasture. The greatest loss of species is encountered when pasture land is created from an area that was once forest (Fujisaka, 1998). In addition, the soil often becomes compacted and degraded, which makes future regeneration more difficult (Fujisaka, 1998). In the temperate rainforests of the BC coast, the matrix often consists of tree plantations that have proven to be biologically impoverished as compared to old growth and naturally regenerated forests (Noss and Cooperrider, 1994). Given these facts and that many species are unable to move through these kinds of land use areas, patch connectivity and corridors will be crucial requirements in the long term conservation of biological diversity. .
