Hansen et al.
Edge effects across ecosystem types
Ecosystem Biomass as a Framework for Predicting Habitat Fragmentation Effects
Running Head: Edge effects across ecosystem types
Key Words: biomass, conservation principles, edge effects, forest interior species, habitat
fragmentation, microclimate
Word Count: 7150
Andrew James Hansen, Ecology Department, Montana State University, Bozeman, MT
59717-3460
Lisa Baril, Ecology Department, Montana State University, Bozeman, MT 59717-3460
Jennifer Watts, Land Resources and Environmental Sciences Department, Montana State
University, Bozeman, MT 59717-3460
Fletcher Kasmer, Montana State University, Bozeman, MT 59717-3460
Toni Ipolyi, Montana State University, Bozeman, MT 59717-3460
Ross Winton, Entomology Department, Montana State University, Bozeman, MT 597173460
Corresponding Author: Andrew Hansen, Ecology Department, 310 Lewis Hall, Montana
State University, Bozeman, MT 59717-3460. hansen@montana.edu
25 February 2008
1
Hansen et al.
Edge effects across ecosystem types
Abstract: Past studies of the consequences of human-induced edge effects on native
biodiversity have had mixed results leading to confusion on if and how to manage to avoid
negative edge effects in different forested ecosystem types. The purpose of this paper is to
evaluate if forest ecosystems differ predictably in response to edge effects based on forest
structure and density. The Biomass Accumulation Hypothesis (Hansen & Rotella 2000)
asserts that edge effects have the highest magnitude of influence in ecosystems that
accumulate high levels of standing aboveground vegetation biomass. We test two
predictions of this hypothesis: 1) The dense vegetation in mid to late seral habitats within
high-biomass ecosystems buffers microclimate, resulting in large differences in average
microclimate between disturbance patch edges and forest interiors; 2) In high biomass
ecosystems the sharp gradient in vegetation and microclimate results in finer habitat
partitioning by organisms and more forest interior specialists than is the case in lower
biomass ecosystems. These predictions were tested by statistical analysis of the results of
published studies in several of the major forest biomes. We found that magnitude of edge
influence of microclimate was significantly related to forest biomass for light intensity and
relative humidity. Trends for air temperature, soil temperature, vapor pressure deficit, and
soil moisture were as predicted, but not statistically significant. A model including all
microclimate samples and controlling for microclimate variable type was significant. The
percent of species that specialized on forest interiors was significantly related to biomass for
mammals and birds, and nearly significant for beetles. The results suggest that forest
fragmentation is most likely to negatively influence forest species in high biomass
ecosystems such as tropical and temperate rainforests, but may have little influence on forest
species in low-biomass ecosystems such as boreal or subalpine forests. These finding
2
Hansen et al.
Edge effects across ecosystem types
provides a basis for prioritizing the management of habitat fragmentation appropriately
across the world’s forest ecosystems.
3
Hansen et al.
Edge effects across ecosystem types
Introduction
The emerging discipline of conservation biology has increasingly converged on
principles dealing with how humans influence ecological systems and strategies for
mitigating negative effects (Grumbine 1994; Groom et al. 2006). Resource managers are
drawing on these strategies to retain ecosystem function and biodiversity in the face of
human expansion. A necessary step in the maturation of conservation biology is to
understand which conservation strategies are best employed in a particular ecosystem,
including those pertaining to human habitat fragmentation (Saunders et al. 1991).
The vast body of research on edge effects, a consequence of fragmentation, has
shown conflicting results (Paton 1994; Murcia 1995; Harper et al. 2005) leading to confusion
concerning management. A major review by Matlack & Litvaitis (1999) concluded, “The
negative repercussions of edge habitats have prompted changes in forestry practices,
especially regulation regarding the size, distribution, and configuration of clearcuts. Yet
there is considerable variability in edge response among forest sites and species, which
makes it difficult to formulate a consistent edge management policy. The most productive
approach is probably to consider the impact of edges on forest communities on a case-bycase basis.” (pg. 222).
In a more recent review, Fahrig (2003) concluded that there is strong evidence that
loss of habitat area negatively impacts biodiversity. However, change in spatial
configuration of habitat (e.g., patch size, edge density, patch isolation), independent of
habitat loss, “… has rather weak effects on biodiversity, which are as likely to be positive as
negative.” (pg. 508). She states with regards management, “The fact that effects of
4
Hansen et al.
Edge effects across ecosystem types
fragmentation per se [habitat configuration] are usually small and at least as likely to be
positive as negative suggests that conservation actions that attempt to minimize
fragmentation (for a given habitat amount) may often be ineffectual.” (pg., 508). A caution
is offered, however, that evidence suggests that negative edge effects on biodiversity are
much stronger in tropical than in temperate systems. But, “This prediction remains to be
tested.” (pg 508). Thus, substantial uncertainty remains on which ecosystem types are
sensitive to changes in spatial configuration of fragmented habitats and ecosystem-specific
guidelines for managers are generally not available.
Edge effects are the result of the interaction between two adjacent ecosystems that
are separated by an abrupt transition (Murcia 1995). Three categories of edge effects have
been identified: abiotic effects, direct biological effects, and indirect biological effects
(Harper et al. 2005). Abiotic effects include the alteration of microclimate and nutrients from
patch exterior to interior. Changes in species abundances are examples of a direct biological
effect. Edge conditions might lead to indirect biological effects seen in alterations of species
interactions, such as predation, herbivory, pollination, and seed dispersal.
5
Hansen et al.
Edge effects across ecosystem types
Introduction
Human expansion has led to fragmentation of many natural ecosystems (Saunders et
al. 1991). Habitat fragmentation results in both a reduction in total area of habitat and
change in spatial configuration of remaining habitat patches (Wilcove et al. 1986). Change
in habitat patch size and number lead to increases in abrupt edges between remaining habitat
patches and the expanding matrix as fragmentation proceeds. While some native species
thrive at habitat edges, others are adapted to habitat interiors and may become extinct in
landscapes where habitat is dominated by edge effects (Noss et al. 2006a).
The vast body of research on the consequences of human-induced habitat edge effects
on native biodiversity, however, has shown conflicting results (Paton 1994; Murcia 1995;
Harper et al. 2005) leading to confusion concerning management. A major review by
Matlack & Litvaitis (1999) concluded, “The negative repercussions of edge habitats have
prompted changes in forestry practices, especially regulation regarding the size, distribution,
and configuration of clearcuts. Yet there is considerable variability in edge response among
forest sites and species, which makes it difficult to formulate a consistent edge management
policy.” (pg. 222). More recently, Fahrig (2003) concluded that change in spatial
configuration of habitat (including edges), independent of habitat loss, “… has rather weak
effects on biodiversity, which are as likely to be positive as negative.” (pg. 508). She states
regarding management, “… conservation actions that attempt to minimize fragmentation (for
a given habitat amount) may often be ineffectual.” (pg., 508). A caution is offered, however,
that that negative edge effects may be much stronger in tropical than in temperate systems,
but, “This prediction remains to be tested.” (pg 508). Thus, substantial uncertainty remains
6
Hansen et al.
Edge effects across ecosystem types
on which ecosystem types are sensitive to edge effects and which management strategies are
most effective for maintaining native biodiversity in a given ecosystem.
The purpose of this paper is to evaluate if forest ecosystems differ predictably in
response to edge effects based on forest structure and density. The Biomass Accumulation
Hypothesis (Hansen & Rotella 2000) asserts that edge effects have the highest magnitude of
influence in ecosystems that accumulate high levels of standing aboveground vegetation
biomass. We examine how well the predictions of this hypothesis are supported by
published studies from a range of forest ecosystem types.
The abrupt edge transition comprising human-induced edges in forest systems is
typically in vegetation structure, from patches with lower density vegetation to forest stands
with higher vegetation density. This sharp gradient in vegetation structure across forest
edges is the primary mechanism driving abiotic effects such as changes in microclimate,
direct biological effects such as changes in plant and animal composition, and indirect
biological effects such as responses herbivory and predation (Matlack & Litvaitis 1999;
Kremester & Bunnell 1999; Noss et al. 2006a).
Dense vegetation buffers microclimate and disturbances. Consequently, forest patch
interiors often have less extreme microclimates and disturbance regimes than forest edges
because dense vegetation diminishes incoming solar radiation, wind, and many types of
disturbances (e.g., Chen et al.1999). These more moderate interior forest conditions often
support plant and animal species that are uniquely adapted to such conditions (Matlack &
Litvaitis 1999). Creating edges within or near forest interior habaitat can reduce this
vegetation buffer allowing more extreme conditions that may be intolerable for forest interior
species.
7
Hansen et al.
Edge effects across ecosystem types
While the gradient in vegetation density is a primary mechanism underlying edge
effects, many site-scale and landscape-scale factors may modify or override vegetation
density. These include the age of habitat edges, edge aspect, the combined effects of
multiple nearby edges, fragment size, the structure of the adjoining matrix vegetation,
influxes of animals or plant propagules from the matrix, extreme weather or disturbance
events, and land use in the surrounding landscape (Murcia 1995; Harrison & Bruna 1999;
Noss et al. 2006a; Laurance et al. 2007). Such local scale factors can produce striking
variability in edge effects within the same region (Laurance et al. 2007). These local factors
may cause difficulty in predicting the nature and magnitude of edge effects and may also
explain some of the conflicting results in the literature (Harper et al. 2005). The primary role
of vegetation structure in driving edge effects, however, may provide a basis for predicting
the sensitivity of forest ecosystem types to edge effects.
Vegetation height and density vary across the major forest ecosystems of the world as
a function of forest productivity and disturbance (Perry 1994). High levels of vegetation
structure result from high net primary productivity and infrequent loss of vegetation to
disturbance (Fig. 1). Net primary productivity is favored by warm temperatures, high solar
radiation, high moisture, and fertile soils (Running et al. 2004). Stand replacing disturbances
are often less frequent in highly productive forests (White & Jentsch 2001). This is because
fires are inhibited by high precipitation and humidity within the forest, deep fertile soils
inhibit windthrow, and healthy, productive trees are generally less susceptible to catastrophic
insect and disease outbreaks. Across the forests of the earth, vegetation structure, expressed
as aboveground biomass of mid to late seral stages, is highest in wet tropical and temperate
forests, intermediate in moist temperate forests, and lowest in dry or cold forests (Fig. 2).
8
Hansen et al.
Edge effects across ecosystem types
Merging concepts on drivers of edge effects and knowledge of global patterns of
forest structure, the Biomass Accumulation Hypothesis asserts that edge effects have the
highest magnitude of influence in ecosystems that accumulate high levels of standing
aboveground vegetation biomass (AGB). AGB accumulation refers to the total dry weight of
vegetation present within a forest in mid to late successional stages, typically quantified as
tons per hectare. High AGB forests are often characterized by high canopies, multiple
canopy layers, and high foliage height diversity.
The hypothesis predicts that in ecosystems with high AGB accumulation, sharp
gradients in microclimate, disturbance, and the resulting ecological processes will show high
contrast between edge and interior habitats (Fig 3). This steeper gradient in resources and
conditions may be more finely partitioned by plants and animals, leading to a greater
percentage of species specializing on forest interior, edge, or disturbance patch interior
conditions. In contrast, edges in ecosystems with low AGB are predicted to have less of a
buffering effect; microclimatic gradients and other processes will also be less pronounced. In
other words, environmental conditions will be relatively similar between disturbance patches
and the more open mid to late seral forests that characterize these systems. Hence, more
species are predicted to use forest without regard for edge or interior habitats.
The goal of this paper is to examine the degree to which previously published studies
support or refute the Biomass Accumulation Hypothesis. A key conservation concern with
regards to fragmentation is loss of species dependent upon interior habitats. Consequently
we choose as a meaningful measure of community response to fragmentation the proportion
of species in the community statistically associated with forest interiors because these are
species that are likely to be lost if fragmentation proceeds.
9
Hansen et al.
Edge effects across ecosystem types
The two predictions tested were:
1.
The dense vegetation in later seral habitats within high AGB ecosystems buffers
microclimate, resulting in large differences in average microclimate between
disturbance patches and later seral patches, and a sharp gradient from the patch
edge to the interior.
2.
This sharp gradient in vegetation and microclimate results in a finer habitat
partitioning by organisms from patch edge to interior, with a larger percentage of
the species significantly associated with habitat interiors than is the case in lower
AGB ecosystems.
This study is the first to evaluate the Biomass Accumulation Hypothesis. The notion that
edge effects differ predictably across the world’s forests as a function of their inherent
abilities to accumulate biomass, while simple and intuitive, has not been previously reported
in the peer review literature nor discussed in forest conservation and management texts. The
hypothesis may provide an framework for prioritizing conservation and management
strategies for habitat fragmentation across the major forest types of the globe.
Methods
Overview of Methods
For this study, we drew on published research on edge effects for forest ecosystem
types ranging from low to high in AGB. The goal was to summarize results on patterns of
microclimate and species distributions across abrupt forest edges. Thus, we focused on
studies that quantified response variables along transects placed perpendicular to forest
edges, expressly, newly created edges between nonforest or early successional forest and mid
10
Hansen et al.
Edge effects across ecosystem types
or late-seral forest. This allowed us to select studies that had very similar study designs and
allowed a direct test of the predictions.
We used standard literature search engines such as Web of Science to identify
published papers to consider for the study. Search terms such as “edge effects”, “ecotones”,
“aboveground biomass”, “microclimate”, and “species composition” were used. We also
identified candidate studies from the literature cited in published papers. Candidate papers
where then scrutinized relative to the criteria we specified for each of the predictions. The
data were analyzed by regressing the microclimate and interior species edge response
variables across edges on AGB of the forest.
Vegetation AGB
We estimated AGB for the edge study areas by drawing on 10 published studies of
AGB from similar forest types (Fig. 2). These biomass studies synthesized results from a
total of 59 ecosystems. We eliminated estimates from early seral forests, and non-native
forests, which left 53 estimates of AGB in mid to late seral forests for various ecosystems.
These studies used various methods to estimate AGB, including allometric, modeling, and
remote sensing approaches. The accuracy of the AGB data was not quantified. We grouped
the results of these 53 ecosystems into one of eight biome types and averaged AGB within
types to represent AGB levels typical of the biome.
Prediction 1: Microclimate
Criteria for inclusion of studies in this analysis were: 1) study designs compared edge
and/or matrix microclimate variables to interior microclimate variables; 2) the disturbance
patch adjacent to the forest had very low AGB and microclimate there approximated that on
ground uninfluenced by vegetation; 3) the mid and late-seral forests studied were in patches
11
Hansen et al.
Edge effects across ecosystem types
large enough to exhibit interior microclimate conditions; 4) the microclimate variables were
measured during the dry season in tropical evergreen forests or during the growing season in
tropical deciduous, temperate and boreal forests; and 4) the microclimate variables were
measured during mid-morning to mid-afternoon; 5) the results among studies were not
rendered incomparable due to edge orientation, season, or other factors. Thirteen studies met
these criteria and were included in the analysis (Table 2). These spanned the gradient in
biome types and AGB levels.
The microclimate variables reported differed among studies. The following variables
were studied frequently enough to include in our comparison: light levels, air temperature;
soil temperature; humidity; vapor pressure deficit; and soil moisture. All variables were
measured at intervals along transects from forest edge and/or matrix to forest interior using
either a hand-held sensor or a semi-permanent microclimate monitoring station situated near
the ground or a maximum of two meters in height. Measurements used were those taken
simultaneously or within two hours of each other along transects.
We extracted from each paper estimates for a given microclimate variable for forest
interior and forest edge. Interior values used were those that did not differ significantly as a
function of distance from edge. We assumed that this indicated that these locations were not
influenced by proximity to edges. Edge or matrix values were averages of those reported.
The distance from edge for interior conditions varied among studies (range: 30 m to 120 m).
Exact values were obtained from tables when available in the publications and estimations
from figures were used otherwise.
We used these data to estimate magnitude of edge influence (MEI). Following
Harper et al (2005), MEI was calculated as:
12
Hansen et al.
Edge effects across ecosystem types
(e-i)/(e+i)
e: value of the parameter at the edge
i: value of the parameter in the interior
MEI ranges between -1 and +1, with negative values indicating edge is lower than interior,
positive values indicating that edge is higher than interior, and a value of 0 indicating no
edge effect.
Prediction 2: Proportion Interior Species
Criteria for inclusion of papers in this analysis were: species abundances quantified
along forest edge to interior transects; forest edges were relatively recently created with
matrix AGB being low; sampling methods were comparable in sampling interval (spatial and
temporal); the results among studies were not rendered incomparable due to edge orientation,
season, or other factors; and statistical significance of species associations with forest
interiors were reported. Studies near urban or suburban development were excluded to
prevent complications due to species-human interactions. Most published studies we found
involved beetles, birds, and mammals, thus we restricted the analyses to these groups. A
total of 20 studies remained after the final selection (Table 3).
Species response to edge was quantified as the proportion of species found in forests
that were significantly more abundant in forest interior than at the forest edge. This measure
was obtained for each study by dividing the total number of species associated statistically
with interior sites by the total number of species included in the statistical analysis of forest
species. MEI based on abundances of individual species was not appropriate because the
13
Hansen et al.
Edge effects across ecosystem types
species present differed between studies. Moreover, measures of total abundance among all
species or species richness are not considered highly relevant to conservation (Noss 2006a).
Statistical Analyses
Statistical inference from research findings across studies is increasingly drawn using
meta-analysis (Gurevitch & Hedges 1999, Osenberg et al. 1999). Meta-analysis requires
information on sample size and sampling variation for each study. Few of the studies we
selected for analysis reported sampling variation. Most of these papers were done early in
the period of conservation biology when statistical reporting standards were less stringent
than at present. Among the 31 studies included in the analyses, only 5 reported sampling
variance numerically. We concluded that a rigorous meta-analysis is not possible with the
current studies.
These studies do, however, represent the best current opportunity to evaluate the
Biomass Accumulation Hypothesis. If these studies provide support for the hypothesis, new
well designed research is justified. This approach is consistent with Gurevitch and Hedges
(1999) who argue that, “given a body of literature, some information regarding the overall
findings is much better than no information, and that therefore it is desirable to develop
methods for data synthesis of poorly reported data (e.g., where no estimate of sampling
variance is published)” (pg 1146). They suggest unweighted standard parametric statistical
tests such as ordinary least-squares regression can sometimes be used and that the
assumption of homogeneity of variances of conventional statistics “does not always seriously
compromise the Type I error rates of these tests” (pg 1146). Sample variance is strongly
influenced by sample size. If the sample size is not biased relative to the magnitude of the
predictor variable, regression may provide reliable results. We examined the sample sizes
14
Hansen et al.
Edge effects across ecosystem types
for studies used in each of our analyses and found they were either random relative to
estimated AGB or were larger in lower AGB forests, which increases the likelihood of
significant relationships in low AGB forests and makes this analysis conservative for testing
the hypothesis of greater differences in high AGB forests. We conclude that use of
regression to relate response variables to forest AGB is justified for these analyses.
MEI for each microclimate variable was regressed on AGB. Because sample sizes
were relatively small, we additionally used an analysis of covariance linear model for
microclimate, where microclimate type was the classification variable, MEI was the response
variable, and AGB and microclimate variables were the predictors. The model allowed the
slope and intercept to differ among microclimate variables. The absolute value of MEI was
used in this analysis. If AGB was significant in the model, we considered this as evidence in
support of the AGB Accumulation Hypothesis. Because sample sizes for species responses
were higher, we analyzed each taxonomic group separately. The proportion of interior
species specialists was regressed against AGB with linear models. Results were considered
significant at the P<.05 level.
Results
MEI of light intensity increased significantly with AGB (n=13, F=8.65, p<0.01) (Fig
4). MEI for forests with AGB of less than 200 t/ha ranged from 0.5 to 0.9 and was generally
0.9 to1.0 for forests higher in AGB. MEI for humidity was significantly negatively related to
AGB (n=6, F=27.49, p<0.006). This indicated a more humid environment in forest interiors
than in edges.
15
Hansen et al.
Edge effects across ecosystem types
The remaining microclimate variables generally had the weakest MEI in low AGB
forests, however the univariate relationships were not statistically significant. VPD (n=5)
was similar to humidity showing moister conditions in forest interiors (Table 1). MEI for air
temperature data (n=7) generally increased with AGB, except for one observation in boreal
forest that showed a very high MEI. For soil moisture (n=3) and soil temperature (n=4)
there was no apparent relationship between MEI and AGB.
The overall model of microclimate MEI on AGB, controlling for microclimate
variable was significant (n=37, F=79.94, p<.0001). Within this model, AGB was possitively
related to the absolute value of MEI (F=251.02, p<0.0001).
Prediction 2: Interior Species
The percent of forest species specializing on forest interiors exhibited a positive
significant relationship with AGB for mammals (n=6, F=28.93, P<0.006) and birds (n=8,
F=6.57, P<0.043) (Fig. 5). The relationship for beetles was positive, but P-value slightly
exceeded the 0.05 cut-off for statistical significance (n=6, F=6.72, P<0.060). AGB explained
90%, 67%, and 65% of the variation in mammals, birds, and beetles respectively.
Among mammal studies, no forest interior species were found in three studies with
forest AGB below 140 t/ha. Percent interior species was 11-25 for mammals in forests of
312-412 t/ha. Percent interior bird species increased in forests with AGB of 40 t/ha to about
200 t/ha. Above this AGB level, the percent interior species remained near 30. The
percentage for beetles was 0 in forests below75 t/ha, approximately 18 in forests 78-423 t/ha,
and was 50 in one forest with a AGB of 450 t/ha.
Discussion
16
Hansen et al.
Edge effects across ecosystem types
The results were supportive of the two predictions of the Biomass Accumulation
Hypothesis. The microclimate analyses were limited by very small samples for four of the
five variables. Even so, there was a significant relationship between MEI and AGB for light
and humidity and the overall effect of AGB on MEI across microclimate variables was
significant. These results suggest that the greater amount of vegetation in higher AGB forest
ecosystems more fully buffers the microclimate in forest interiors and creates sharp gradients
from interior to edge. The magnitude of MEI for microclimate in tropical rainforests is
known to be sufficient to have pronounced ecological effects such as tree and plant mortality
(Laurance et al. 2002) and substantially elevated fire frequencies (Cochrane & Laurance
2002; Cochrane 2003).
The proportion of species significantly more abundant in forest interior was also
positively related to AGB for two of the three taxonomic groups we examined. This was
especially the case with mammals for which three studies with AGB below 200 t/ha had no
forest interior specialists. These results are consistent with the prediction of the Biomass
Accumulation Hypothesis that interior species were poorly represented in these low AGB
forests because forest interiors and edges did not differ much in microclimate, vegetation
structure, and other direct edge effects. In support, several studies found significant
correlations between microclimate, vegetation structure, and species abundances along edge
to interior gradients (Didham et al. 1998; Magura et al. 2001).
The trends in percent interior species with AGB were generally similar for the three
taxonomic groups. The results suggest that beetle species may partition the edge-to interior
gradient in lower AGB forests than mammals and birds. The results also suggest that the
proportion of interior bird specialists plateaus at forest AGB levels above about 200 t/ha. It
17
Hansen et al.
Edge effects across ecosystem types
is likely that differences in body size, home range size, and other aspects of scaling between
taxonomic groups influences the relationship between edge effects and AGB. More rigorous
study and larger sample sizes would be needed to test this hypothesis.
Scope and Limitations
We acknowledge several limitations to our treatment of this issue. We choose to limit
the analyses to microclimate, beetles, birds, and mammals in forested systems. Forested
ecosystems were studied because the hypothesis most obviously applies to forests.
Conceptually, the hypothesis should apply to grassland and shrubland systems, but less so
than for forests due to their greater AGB. Moreover, studies of edge effects in nonforest
vegetation types are fewer and estimates of AGB are more difficult to obtain.
The analysis of species specialization was restricted to beetles, birds, and mammals
because too few studies were available for other taxonomic groups. Also, we suspect that
edge effects for tree species composition are more difficult to quantify because of the
relatively slow response of their distribution to changing conditions such as edge creation.
Edge effects for amphibians may also be difficult to assess because these species require
various aquatic and terrestrial habitats at different life stages (Becker et al. 2007).
An assumption of our approach for species distributions across edges was that no forest
interior species had become locally extinct after habitats had become fragmentation but
before the studies were conducted. Local extinction in forest fragments has been
documented (Saunders et al.1991: Robinson 1999). The patch sizes at which extinction
occurs is known to vary with trophic level, home range size, and population abundance, with
top-level carnivores requiring the largest habitat sizes for persistence (Woodroffe & Ginsberg
1998). The species sampled in studies we used were relatively small in body size and
18
Hansen et al.
Edge effects across ecosystem types
sufficiently abundant to be detected at levels adequate for statistical analysis by the
multispecies sampling methods used in these studies. Thus, they likely had relatively small
home ranges. Moreover, all the studies were done in relatively large forest stands. We
conclude that it is unlikely that our results were influenced by previous local extinctions of
forest interior species.
The study considered only the magnitude of edge influence and not depth of edge
influence. Depth of edge influence (DEI) is the distance that edge effects penetrate the
habitat (Chen et al. 1999; Harper et al. 2005). This was done to make this initial analysis
more tractable. We speculate that DEI has a unimodal relationship with AGB, reaching a
maximum in forests with intermediate AGB. In forests higher in AGB forests, the dense
vegetation should inhibit depth of penetration of direct and indirect edge effects. In forest
with lower AGB, DEI should diminish as conditions in the forest interior are similar to those
in the matrix.
Within the response variables we chose for analysis, sample sizes were relatively small.
This was because of the moderate number of edge studies that have been published and
because we excluded many of the published papers because of incompatible study designs,
lack of rigorous statistical analysis, or because of confounding factors. To best draw
inference from the limited sample, we reported results for both the relationships between
AGB and MEI of individual microclimate variables and across all microclimate variables and
controlling for variable type.
Although the studies included in the analyses were selected based on the stated criteria,
they differed to some extent in site-level factors such as edge age and development,
orientation, and matrix type. We assume that these biases were random relative to the effects
19
Hansen et al.
Edge effects across ecosystem types
of AGB. Laurance et al. (2007) describes how such local factors can cause considerable
variation in edge effects within a given region. The fact that significant relationships were
found in our analyses despite the variability introduced by differences in the site conditions
of studies we drew upon, suggests that the effect of AGB on edge effects was strong enough
to be detected despite the variability due to confounding factors.
Finally, the studies did not provide adequate information to allow formal meta-analysis
(see Methods), we feel the regression methods used were justified for this analysis.
Despite these limitations, we feel that this analysis is an appropriate first test of the
hypotheses. Given the results of this analysis, the cost of conducting a replicated experiment
across biomes may now be justified.
Comparison with Current Knowledge
Previous efforts to develop general theory on the situations where edge effects are
most pronounced have focused largely on site-level factors (Donovan et al. 1997; Laurance et
al. 2002). These include: attributes of the habitat such as vertical structure and soil depth;
edge attributes such as age, orientation, and vegetation contrast; and matrix characteristics
such vegetation structure and composition and land use. The susceptibility of major
ecosystem types to edge effects has been little studied. A recent paper by Harper et al.
(2005) suggested that the traits of ecosystems that are especially prone to edge effects
include: climate (high mean air temperature and solar radiation, low cloud cover, frequent
extreme winds); disturbance (infrequent stand replacing disturbances); community structure
(many early-seral and invasive species); landscape patterns (low inherent habitat patchiness).
Additionally, Hansen & Urban (1992) suggested that the life history attributes of species in
the community vary among ecosystems and influence direct biotic edge effects.
20
Hansen et al.
Edge effects across ecosystem types
The Biomass Accumulation Hypothesis integrates several the ecosystem factors
described by Harper et al. (2005) and Hansen & Urban (1992). Forests with high AGB
accumulation tend to have warm temperatures, high solar radiation, periods of low clouds,
high primary productivity, infrequent stand replacing disturbance, and low natural patchiness
in forest/non-forest conditions (Brown & Lugo 1982). Moreover, species in high
productivity and high AGB systems tend to have small home ranges, low dispersal, and
habitat specialization. Hence, biomass accumulation appears to represent a syndrome of
ecosystem characteristics that cause magnitude of edge effects to be pronounced.
Implications
This study is the first to test the Biomass Accumulation Hypothesis. The results
support the hypothesis that biomass accumulation influences the magnitude of one
component of fragmentation, edge effects. Edge effects, in turn, are a component of patch
size effects (Turner et al. 2001; Noss et al. 2006b). Less area within a patch will display
habitat interior conditions when edge effects are pronounced. Consequently, we expect that
extinction rates of interior species in small patches will be higher in high biomass
ecosystems.
Our results suggest that habitat fragmentation has the strongest negative effects on
ecosystems with high productivity and biomass accumulation, such as wet tropical and wet
temperate forests. Thus, efforts to minimize or mitigate fragmentation are most appropriately
applied to such high biomass ecosystems. It is important that managers in high biomass
forests not be confused by the ambiguous results of edge studies globally and take seriously
the management of habitat configuration within their forests. In ecosystems with low
biomass accumulation such as boreal, dry, or cold forests, conservation strategies to manage
21
Hansen et al.
Edge effects across ecosystem types
edge and patch effects may be a lower priority. Further work should be directed towards
identifying thresholds in the relationship between forest biomass and edge effects that could
be used to more precisely guide the management of landscape configuration to local
biophysical conditions.
Overall, application of the Biomass Accumulation Hypothesis is an example of using
fundamental ecosystem properties as a basis for tailoring conservation to local ecosystem
conditions, a needed next step for conservation biology.
Acknowledgements
We thank Diane Debinski and Jiquan Chen for comments on earlier drafts of the
manuscript. Jim Robinson-Cox provided guidance on statistical analyses. We also thank the
authors of the numerous publications that made this study possible.
Literature Cited
Baker, J., K. French, R. J. Whelan. 2002. The edge effect and ecotonal species: Bird
communities across a natural edge in southeastern Australia. Ecology 83:3048–3059.
Becker, C.G., C.R. Fonseca, C.F.B. Haddad, R.F. Batista, P.I. Prado. 2007. Habitat Split
and the Global Decline of Amphibians. Science 318 (5857): 1775-1777.
Berry, L. 2001. Edge effects on the distribution and abundance of birds in a southern
Victorian forest. Wildlife Research 28:239-245.
Brand, L. A., and T. L. George. 2001. Response of passerine birds to forest edge in coast
redwood forest fragments. The Auk 118:678-686.
22
Hansen et al.
Edge effects across ecosystem types
Brown, S., and A. E. Lugo. 1982. The storage and production of organic matter in tropical
forests and their role in the global carbon cycle. Biotropica 14:161-187.
Brown, S., and A. E. Lugo. 1984. AGB of tropical forests: A new estimate based on forest
volumes. Science 223: 1290-1293.
Brown, S. L.R. Iverson, A. Prasad, and D. Liu. 1993. Geographical distributions of carbon
in biomass and soils in tropical Asian forests. Geocarto International 4:45-59.
Burke, D. M., and E. Nol. 1998. Edge and fragment size effects on the vegetation of
deciduous forests in ontario, Canada. Natural Areas Journal 18:45-53.
Cairns, M. A., I. Olmsted, J. Granados, and J. Argaez. 2003. Composition and aboveground
tree AGB of a dry semi-evergreen forest on Mexico’s Yucatan Peninsula. Forest
Ecology and Management 186:125–132.
Chazon, R. L., and N. Fetcher. 1984. Photosynthetic light environments in a lowland tropical
rain forest in Costa Rica. The Journal of Ecology 72:553-564.
Chen, J., J.F. Franklin, and T.A. Spies. 1993. Contrasting microclimates among clearcut,
edge and interior of old growth Douglas-fir forest. Agricultural and Forest
Meteorology 63:219-237.
Chen, J., S. C. Saunders, T. R. Crow, R. J. Naiman, K. D. Brosofske, G. D. Mroz, B. L.
Brookshire, and J. F. Franklin. 1999. Microclimate in forest ecosystem and landscape
biology. BioScience 49:288-297.
Cochrane, M.A. and W.F. Laurance. 2002. Fire as a large-scale edge effect in Amazonian
forests. Journal of Tropical Ecology 18:311-325.
Cochrane, M.A. 2003. Fire science for rainforests. Nature 421:913-919.
23
Hansen et al.
Edge effects across ecosystem types
Didham, R. K., P. M. Hammond, J. H. Lawton, P. Eggleton, and N. E. Stork. 1998. Beetle
species responses to tropical forest fragmentation. Ecological Monographs 68:295323.
Donovan, T., P. W. Jones, E. M. Annand, and F. R. I. Thompson. 1997. Variation in localscale edge effects: Mechanisms and lanscape context. Ecology 78:2064-2075.
Fahrig, L. 2003. Effects of habitat fragmentation on biodiversity. Annu. Rev. Ecol. Evol.
Syst. 2003. 34:487–515.
Gehlhausen, S. M., M. W. Schwartz, and C. K. Augspurger. 2000. Vegetation and
microclimatic edge effects in two mixed-mesophytic forest fragments. Plant Ecology
147:21-35.
Gerwing, J.J., and D. L. Farias. 2000. Integrating liana abundance and forest structure into
an estimate of total aboveground biomass for and eastern Amazonian forest. Journal
of Tropical Biology 16:327-335.
Ghuman, B. S., and R. Lal. 1987. Effects of partial clearing on microclimate in a humid
tropical forest. Agricultural and Forest Meteorology 40:17-29.
Groom, M. J., G. K. Meffe and C. R. Caroll, editors. 2006. Principles of Conservation
Biology. 3rd edition. Sinauer, Sunderland, MA.
Grumbine, R. E. 1994. What is ecosystem managment? Conservation Biology 8:27-38.
Gurevitch, J., and L.V. Hedges. 1999. Statistical issues in ecological meta-analyses.
Ecology 80:1142-1149.
Hansen, A. J. and D. L. Urban. 1992. Avian response to landscape pattern: The role of
species life histories. Landscape Ecology 7:163-180.
24
Hansen et al.
Edge effects across ecosystem types
Hansen, A. J. and J. J. Rotella. 2000. Bird responses to forest fragmentation. Pgs 202-221 in
R.L. Knight, F.W. Smith, S.W. Buskirk, W.H. Romme, and W.L. Baker, (eds.) Forest
Fragmentation in the Southern Rocky Mountains. University of Colorado Press,
Boulder, CO.
Hansson, L. 1994. Vertebrate distributions relative to clear-cut edges in a boreal forest
landscape. Landscape Ecology 9:105-115.
Harper, K. A., S. E. Macdonald, P. J. Burton, J. Chen, K. D. Brosofske, S. C. Saunders, E. S.
Euskirchen, D. Roberts, M. S. Jaiteh, and P. Esseen. 2005. Edge influence on forest
structure and composition in fragmented landscapes. Conservation Biology 19:768782.
Harrison, S., and E. Bruna. 1999. Habitat fragmentation and large-scale conservation: What
do we know for sure? Ecography 22:225-232.
Heliola, J., M. Koivula, and J. Niemela. 2001. Distribution of carabid beetles (coleoptera,
carabidae) across a boreal forest-clearcut ecotone. Conservation Biology 15:370-377.
Heske, E. J. 1995. Mammalian abundances on forest-farm edges versus forest interiors in
southern illinois: Is there an edge effect? Journal of Mammalogy 76:562-568.
Hill, C. 1995. Linear strips of rain forest vegetation as potential dispersal corridors for rain
forest insects. Conservation Biology 9:1559-1566.
Kapos, V. 1989. Effects of isolation on the water status of forest patches in the Brazilian
Amazon. Journal of Tropical Ecology 5:173-185.
King, D. I., C. R. Griffin, and R. M. DeGraaf. 1997. Effect of clearcut borders on distribution
and abundance of forest birds in northern New Hampshire. Wilson Bulletin 109:239245.
25
Hansen et al.
Edge effects across ecosystem types
Koivula, M., V. Hyyryläinen, and E. Soininen. 2004. Carabid beetles (coleoptera: Carabidae)
at forest-farmland edges in southern Finland. Journal of Insect Conservation 8:297309.
Kremsater L, Bunnell FL. 1999. Edge effects: theory, evidence and implications to
management of western North American forests. Pages 117–153 In Forest
Fragmentation: Wildlife and Management Implications, ed. J.A. Rochelle, L.A.
Lehmann, J. Wisniewski. Boston, MA. 301 pp.
Kroodsma, R. L. 1982. Edge effect on breeding forest birds along a power-line corridor. The
Journal of Applied Ecology 19:361-370.
Laurance, W. F. 1994. Rain-forest fragmentation and the structure of small mammal
communities in tropical Queensland. Biological Conservation 69:23-32.
Laurance, W. F., T. E. Lovejoy, H. L. Vasconcelos, E. M. Bruna, R. K. Didham, P. C.
Stouffer, C. Gascon, R. O. Bierregaad, S. G. Laurance, and E. Sampaio. 2002.
Ecosystem decay of Amazonian forest fragments: A 22-year investigation.
Conservation Biology 16:605-618.
Laurance, W.F., H.E.M. Nascimento, S.G. Laurance, A. Andrade, R.M. Ewers, K.E. Harms,
R.C.C. Luizao, J.E. Ribeiro. 2007. Habitat Fragmentation, Variable Edge Effects,
and the Landscape-Divergence Hypothesis. PLoS ONE 2(10): e1017, pgs 1-8.
Lefsky, M. A., W. B. Cohen, D. J. Harding, G. G. Parker, S. A. Acker, and S. T. Gower.
2002. Lidar remote sensing of above-ground AGB in three biomes. Global Ecology
& Biogeography 11:393–399.
26
Hansen et al.
Edge effects across ecosystem types
Lehman, S. M., A. Rajaonson, and S. Day. 2006. Lemur responses to edge effects in the
vohibola III classified forest, Madagascar. American Journal of Primatology 68:293299.
Luken, J. O., and N. Goessling. 1995. Seedling distribution and potential persistence of the
exotic shrub lonicera maackii in fragmented forests. American Midland Naturalist
133:124-130.
Magura, T., B. Tothmeresz, and T. Molnar. 2001. Forest edge and diversity: Carabids along
forest-grassland transects. Biodiversity and Conservation 10:287-300.
Matlack, G. R. 1993. Microenvironment variation within and among forest edge sites in the
eastern united states. Biological Conservation 66:185-194.
Matlack, G. and J. Litvaitis. 1999. Forest Edges. Pages 210-233 In: M.L Hunter, Jr. (ed.).
Maintaining biodiversity in forest ecosystems. Cambridge University Press,
Cambridge, U.K.
McDonald, D., and D. A. Norton. 1992. Light environments in temperate new zealand
podocarp rainforests. New Zealand Journal of Ecology 16:15-22.
Means, J. E., S. A. Acker, D. J. Harding, J. B. Blair, M. A. Lefsky,: W. B. Cohen, M. E.
Harmon, and W. A. McKee. 1999. Use of large-footprint scanning airborne Lidar to
estimate forest stand characteristics in the Western Cascades of Oregon. Remote
Sensing of the Environment. 67:298-308.
Messier, C., S. Parent, and Y. Bergeron. 1998. Effects of overstory and understory vegetation
on the understory light environment in mixed boreal forests. Journal of Vegetation
Structure 9:511-520.
27
Hansen et al.
Edge effects across ecosystem types
Murcia, C. 1995. Edge effects in fragmented forests: Implications for conservation. Tree
10:58-62.
Myneni, R. B., J. Dong, C. J. Tucker, R. K. Kaufmann, P. E. Kauppi, J. Liski, L. Zhou, V.
Alexeyev, and M. K. Hughes. 2001. A large carbon sink in the woody AGB of
Northern forests. PNAS 98:14784-14789.
Noss, R. F. 1991. Effects of edge and internal patchiness on avian habitat use in an oldgrowth Florida hammock. Natural Areas Journal 11:34-47.
Noss, R., B. Csuti, and M. J. Groom. 2006a. Habitat fragmentation. Pages 213-251 in M. J.
Groom, G. K. Meffe and C. R. Caroll, editors. Principles of Conservation Biology.
3rd edition. Sinauer, Sunderland, MA.
Noss, R. F., J. F. Franklin, W. L. Baker, T. Schoennagel, and P.B. Moyle. 2006b: Managing
fire-prone forests in the western United States. Frontiers in Ecology and the
Environment 4:481–487.
Olson, D. M., E. Dinerstein, E. D. Wikramanayake, et al. 2001. Terrestrial ecoregions of the
world: A new map of life on Earth. BioScience 51:933-938.
Osenberg, C. W., O. Sarnelle, S. D. Cooper, and R. D. Holt. 1999. Resolving ecological
questions through meta-analysis: goals, metrics, and models. Ecology 80:1105–1117.
Pardini, R. 2004. Effects of forest fragmentation on small mammals in an Atlantic forest
landscape. Biodiversity and Conservation 13:2567-2586.
Paton, W. C. 1994. The effect of edge of avian nest success: How strong is the evidence?
Conservation Biology 8:17-26.
Penner, M., K. Power, C. Muhairwe, R. Tellier, and Y. Wang. 1997. Canada’s Forest AGB
Resources: Deriving Estimates from Canada’s Forest Inventory. Pacific Forestry
28
Hansen et al.
Edge effects across ecosystem types
Centre, Canadian Forest Service, Victoria, British Columbia. Information Report BCX-370
Perry, D. 1994. Forest Ecosystems. Baltimore, Maryland. The Johns Hopkins University
Press. 649 pg
Reifsnyder, W. E., G. M. Furnival, and J. L. Horowitz. 1971/1972. Spatial and temporal
distribution of solar radiation beneath forest canopies. Agricultural Meteorology
9:21-37.
Restrepo, C., and N. Gomez. 1998. Responses of understory birds to anthropogenic edges in
a neotropical montane forest. Ecological Applications 8:170-183.
Robinson, W.D. 1999. Long-term changes in the avifauna of Barro Colorado Island,
Panama, a tropical forest isolate. Conservation Biology 13 (1):85–97.
Running, S. W., R. R. Nemani, F. A. Heinsch, M. S. Zhao, M. Reeves, and H. Hashimoto.
2004. A continuous satellite-derived measurement of global terrestrial primary
production. Bioscience 54: 547-560.
Saunders, D. A., R. J. Hobbs, and C. R. Margules. 1991. Biological consequences of
ecosystem fragmentation: A review. Conservation Biology 5:18-32.
Sekgororoane, G. B., and T. G. Dilworth. 1995. Relative abundance, richness, and diversity
of small mammals at induced forest edges. Canadian Journal of Zoology 73:14321437.
Strong, A. M., C. A. Dickert, and R. T. Bell. 2002. Ski trail effects on a beetle (coleoptera:
Carabidae, elateridae) community in Vermont. Journal of Insect Conservation 6:149159.
29
Hansen et al.
Edge effects across ecosystem types
Tate, K. R., D. J. Giltrap, J. J. Claydon, P. F. Newsome, A. E. Atkinson, M. D. Taylor, R.
Lee. 1997. Organic carbon stocks in New Zealand's terrestrial ecosystems. Journal
of The Royal Society of New Zealand 27:315-335.
Turner, M. G., R. H. Gardner, R. V. O’Neill. 2001. Landscape Ecology in Theory and
Practice: Pattern and Process. Springer Verlag, New York.
Turton, S. M, and H. J. Freiburger.1997. Edge and aspect effects on the microclimate of a
small tropical forest remanant on the Atherton Tablelands, northeastern Australia.
Pages 45-54 in W.L. Laurance and R.O. Bierrgaard,Jr., editors. Tropical forest
remnants: ecology, managment, and conservation of fragmented communities.
University of Chicago Press, Chicago, Illinois, USA.
Vaillancourt,D. A. 1995. Structural and microclimatic edge effects associated with clearcutting in a Rocky Mountain forest. Master's thesis, University of Wyoming,
Laramie, Wyoming.
White, J. D., N. C. Coops, N. A. Scott. 2000. Estimates of New Zealand forest and scrub
AGB from the 3-PG model. Ecological Modelling 131:175–190.
White, P. S., and A. Jentsch. 2001. The search for generality in studies of disturbance and
ecosystem dynamics. Progress in Botany 62:399-450.
Woodroffe, R., and J. R. Ginsberg. 1998. Edge effects and the extinction of populations
inside protected areas. Science 280:2126–2128
Young, A., and N. Mitchell. 1994. Microclimate and vegetation edge effects in a fragmented
podocarp-broadleaf forest in New Zealand. Biological Conservation 67:63-72.
30
Hansen et al.
Edge effects across ecosystem types
31
Hansen et al.
Edge effects across ecosystem types
Table 1. Studies of microclimate across forest edges included in the analyses. Biomes are
defined as: TrRF = tropical rainforest; TeDF = temperate deciduous forest; TeRF = temperate
rainforest; BDF = boreal deciduous forest; BCF = boreal coniferous forest; SAF = sub-alpine
forest. AGB refers to aboveground standing vegetation. MEI is magnitude of edge influence
as defined in Methods.
Biome
Study
Region
AGB
Variable
MEI
(t/ha)
Light
BDF
Burke & Nol 1998
Ontario, Canada
63
0.60
BDF
Messier et al. 1998
Quebec, Canada
63
0.78
BCF
Messier et al. 1998
Quebec, Canada
63
0.79
BDF/BCF Messier et al. 1998
Quebec, Canada
63
0.86
SAF
Vaillancourt 1995
Wyoming, USA
100
0.51
TeDF
Luken & Goessling
Kentucky, USA
104
0.88
1995
TeDF
Reifsnyder et al, 1971
Connecticut, USA
142
0.83
TeDF
Matlack 1993
Pennsylvania, USA 180
0.82
TrRF
McDonald & Norton
New Zealand
343
0.95
1992
TrRF
Young & Mitchell 1994
New Zealand
378
0.86
TrRF
Kapos 1989
Brazil
400
0.94
TrRF
Ghuman & Lal 1987
Nigeria
412
0.94
TrRF
Chazdon & Fetcher
Costa Rica
450
0.97
32
Hansen et al.
Edge effects across ecosystem types
1984
Air temperature
(°C)
TeDF
Magura et al. 2001
Hungary
78
0.10
TeDF
Gelhausen et al. 2000
Illinois, USA
100
0.03
TeDF
Matlack 1993
Pennsylvania, USA 180
0.05
TrRF
Turton & Frieburger
Australia
270
0.01
1997
TrRF
Young & Mitchell 1994
New Zealand
378
0.04
TrRF
Kapos 1989
Brazil
400
0.08
TeRF
Chen et al. 1993
Washington, USA
687
0.10
Humidity (%)
BDF
Burke & Nol 1998
Ontario
63
0.00
TeDF
Magura et al. 2001
Hungary
78
-0.07
TeDF
Gelhausen et al. 2000
Illinois, USA
100
-0.06
TeDF
Matlack 1993
Pennsylvania, USA 180
-0.04
TrRF
Ghuman & Lal 1987
Nigeria
412
-0.26
TeRF
Chen et al. 1993
Washington, USA
687
-0.29
VPD (mb)
TeDF
Matlack 1993
Pennsylvania, USA 180
0.17
Turton & Frieburger
Australia
270
0.50
New Zealand
378
0.15
1997
TrRF
Young & Mitchell 1994
33
Hansen et al.
TrRF
Kapos 1989
Edge effects across ecosystem types
Brazil
400
0.68
Soil Moisture (%)
BDF
Burke & Nol 1998
Ontario, Canada
63
-0.07
TeDF
Gelhausen et al. 2000
Illinois, USA
100
-0.22
TrRF
Kapos 1989
Brazil
400
-0.14
Soil Temperature
(°C)
BDF
Burke & Nol 1998
Ontario
63
0.07
TeDF
Magura et al. 2001
Hungary
78
0.05
TrRF
Turton & Frieburger
Australia
270
0.01
Washington, USA
687
0.04
1997
TeRF
Chen et al. 1993
34
Hansen et al.
Edge effects across ecosystem types
Table 2. Studies of animal species composition used in the analyses. AGB refers to
aboveground standing vegetation.
Biome
Study
Region
AGB
Taxonomic
Total
Percent
(t/ha)
Group
Species
Interior
Specialists
Boreal
Heliola et al.
Finland
40 beetle
34
0
Finland
75 beetle
12
0
130 beetle
11
18
78 beetle
17
17
Brazil
423 beetle
32
19
2001
Boreal
Koivula et al.
2004
Temperate
Strong et al.
Vermont,
Coniferous
2002
USA
Temperature Magura et al.
Hungary
Deciduous
2001
Wet
Didham et al.
Tropical
1998
Wet
Hill 1995
Australia
450 beetle
4
50
Boreal
Hansson 1994
Sweden
39 bird
16
0
Temperate
Baker 2001
Australia
100 bird
27
11
Intermediate Berry 2001
Victoria,
157 bird
13
0
Tropical
Australia
181 bird
27
15
Tropical
Deciduous
Temperate
Noss 1991
Florida,
35
Hansen et al.
Edge effects across ecosystem types
Deciduous
USA
Temperate
Kroosma et al.
Tennessee,
Deciduous
1982
USA
Temperate
King et al.
New
Deciduous
1997
Hampshire,
244 bird
26
35
266 bird
7
29
600 bird
14
29
USA
Temperate
Brand &
California,
Coniferous
George 2001
USA
Boreal
Hansson 1994
Sweden
39 mammal
9
0
Temperate
Sekgororoane
New
84 mammal
9
0
Coniferous
et al. 1995
Brunswick,
137 mammal
11
0
Canada
Temperate
Heske 1995
Illinois,
Deciduous
USA
Intermediate Pardini 2004
Brazil
314 mammal
9
11
Laurance 1994
Australia
400 mammal
4
25
Temperate
Lehman et al
Madagascar
412 mammal
6
17
Deciduous
2006
Wet
Restrepo 1998
Columbia
552 bird
25
32
Tropical
Wet
Tropical
Tropical
36
Hansen et al.
Edge effects across ecosystem types
Figure Legends
Figure 1. Conceptual model of interactions between ecosystem productivity and disturbance
regimes in influencing aboveground biomass accumulation.
Figure 2. Average aboveground biomass of late seral forest in seven forested ecoregions
(defined in Olson et al. 2001). Data derived from published biomass values (Brown et al.
1993; Penner et al 1997; Lefsky et al 2002; Myneni 2006; Brown & Lugo 1984; White et al
2000; Tate et al 1997; Gerwing et al 2000; Means et al. 1999; Cairns et al. 2003). Temperate
coniferous studies were from extremely high biomass systems (California redwoods, Pacific
Northwest rainforest) and is not likely representative of the broader ecoregion.
Figure 3. Conceptual model of edge effects in ecosystems with high and biomass
accumulation. Depicted are adjacent early and late seral forest patches. Aboveground
vegetation biomass (line 1) increases abruptly between early seral and late seral forest
patches, causing a strong gradient in microclimate (line 2) from forest edge to forest interior,
and allowing species (line 3) to specialize in either forest interior, edge, or early seral habitats
resulting in narrow distributions in abundance of guilds across the edge to interior gradient.
In lower biomass ecosystems, the gradient in biomass is less pronounced, consequently
microclimate is less different between early and late seral patches and species do not
specialize highly on interior or edge habitats.
37
Hansen et al.
Edge effects across ecosystem types
Figure 4. Magnitude of edge influence (MEI) for microclimate variables across AGB levels.
MEI ranges between -1 and +1, with negative values indicating edge is lower than interior,
positive values indicating that edge is higher than interior, and a value of 0 indicating no
edge effect. VPD is vapor pressure deficit. Data are from the studies in Table 1.
Figure 5. Percentage forest interior species for birds, beetles, and mammals across
aboveground biomass levels. Data are from studies in Table 2.
38
Edge effects across ecosystem types
Infrequent
Hansen et al.
Biomass
Frequent
Stand-replacing
Disturbance
Accumulation
Low
NPP
High
Solar radiation, Temperature, Precipitation, Soil Fertility
39
Hansen et al.
Edge effects across ecosystem types
40
Hansen et al.
Edge effects across ecosystem types
Magnitude
Low
High Biomass Ecosystem
2. Microclimate (e.g.,
wind speed)
1. Vegetation biomass
High
3. Species Guild
Abundances
Early Seral Patch
Edge
Later Seral Patch
Magnitude
Low
Low Biomass Ecosystem
2. Microclimate (e.g.,
wind speed)
1. Vegetation biomass
High
3. Species Guild
Abundances
Early Seral Patch
Edge
Later Seral Patch
41
Hansen et al.
Edge effects across ecosystem types
y= 0.0006 * AGB + 0.70
2
0.4
R =0.44
0.0
MEI
0.8
Light Intensity
0
100
200
300
400
500
AGB (t/ha)
0.0
Humidity
-0.3
-0.6
MEI
y= 0.0005 * AGB + 0.001
2
R = 0.87
0
100
200
300
400
500
600
700
AGB (t/ha)
42
Hansen et al.
Edge effects across ecosystem types
Percent of Species Specializing on Forest Interiors
60
Beetles
y = 0.0771x + 2.3499
R2 = 0.6457
50
Birds
y = 0.07x - 0.6979
R2 = 0.6782
40
30
Mammals
y = 0.0664x - 5.7805
R2 = 0.8958
20
10
0
-10
0
100
200
300
400
500
600
700
Biomass (t/ha)
43