Download Edge effects and their influence on lemur density and distribution in

AMERICAN JOURNAL OF PHYSICAL ANTHROPOLOGY 129:232–241 (2006)
Edge Effects and Their Influence on Lemur Density
and Distribution in Southeast Madagascar
Shawn M. Lehman,1* Andry Rajaonson,2 and Sabine Day2
1
2
Department of Anthropology, University of Toronto, Toronto, Ontario M5S 3G3, Canada
Department of Paleontology, University of Antananarivo, Antananarivo, Madagascar
KEY WORDS
forest fragmentation; diet; food quality; predation; protected areas
ABSTRACT
Edge effects are caused by the penetration of abiotic and biotic conditions from the matrix into
forest interiors. Although edge effects influence the biogeography of many tropical organisms, they have not
been studied directly in primates. Edge effects are particularly relevant to lemurs due to the loss of 80–90% of
forests in Madagascar. In this study, data are presented
on how biotic edge effects influenced the distribution and
density of lemurs in the Vohibola III Classified Forest in
southeastern Madagascar. In total, 415 lemur surveys
were conducted during June–October 2003 and May–
September 2004 along six 1,250-m transects that ran
perpendicular to the forest edge. Data were also collected
on lemur food trees along the six transects (density,
height, diameter at breast height, area, volume, and distance to forest edge). Four nocturnal species (Avahi
laniger, Cheirogaleus major, Lepilemur microdon, and
Microcebus rufus) and four diurnal species (Eulemur
rubriventer, Eulemur fulvus rufus, Hapalemur grisesus
griseus, and Propithecus diadema edwardsi) were
sighted during surveys. Regression analyses of lemur
densities as a function of distance to forest edge provided
edge tolerances for A. laniger (edge-tolerant), M. rufus
(edge-tolerant), E. rubriventer (edge-tolerant or omnipresent), and H. g. griseus (omnipresent). The density and
distribution of M. rufus and their foods trees were correlated. Edge-related variations in food quality and predation pressures may also be influencing lemurs in Vohibola III. Tolerance for edge effects may explain, in part,
how lemurs have survived extreme habitat loss and forest fragmentation in southeastern Madagascar. Am J
Phys Anthropol 129:232–241, 2006. V 2005 Wiley-Liss, Inc.
Forest habitats are becoming increasingly fragmented
in most tropical regions of the world (Laurance, 1999).
One of the most significant consequences of forest fragmentation is an increase in amount of habitat edge (Lovejoy et al., 1986; Laurance and Yensen, 1991; Chen et al.,
1992). Edges are dynamic zones characterized by the penetration, to varying depths and intensities, of conditions
from the surrounding environment (matrix) into the forest
interior (Malcolm, 1994). Hypothetically, if edge effects
penetrate 300 m into a 100-ha square-shaped forest fragment, then approximately only 16% of the total forest
amount will be unaffected by edge effects. For example,
Curran et al. (1999) found that edge effects reduced the
abundance of Dipterocarp seedlings up to 5 km from the
edge into forests in Borneo. These edge-related changes in
forest dynamics may have deleterious effects on resident
primate populations by reducing the distribution and
abundance of food trees near the forest edge. Changes in
species interactions (e.g., herbivory, frugivory) as a consequence of edge effects were defined as indirect biological
effects (Murcia, 1995). Indirect biological effects represent
a significant but poorly studied aspect of the ecological
consequence of edge effects for primates in tropical forests
(Norconk and Grafton, 2003). Although some researchers
invoked edge effects as a significant determinant of primate distributions (Mbora and Meikle, 2004; Tweheyo
et al., 2004), the associated data were based on relative
amounts or presumed limits of forest edges. Therefore, we
have few data on how the distribution and density of primates are affected directly by edge effects.
Edge effects may be particularly relevant to studies of
forests and primates in Madagascar. Lemurs are one of
the most threatened primate taxa in the world due to the
loss of 80–90% of forest habitats in Madagascar (Green
and Sussman, 1990; Du Puy and Moat, 1998). The remaining forest is highly fragmented and, therefore, may be
prone to extreme edge effects. However, there are few data
on how edge effects influence lemur biogeography in
Madagascar. Lidicker (1999) constructed an ecological
model of how species respond indirectly to variables, such
as changing resource conditions, associated with edge
effects. This model predicts that the response of a target
species to edge effects can be measured as matrix or ecotonal effects. A matrix effect occurs when the target species
responds directly to some aspect of the edge. Thus, lemur
species that have their highest densities near the edge
were defined as edge-tolerant. Conversely, lemur species
that avoid forest edges were defined as edge-intolerant.
Conversely, ecotonal effects occur when the organism
shows little or no response to the edge. In this case, a
lemur species can be referred to as being omnipresent
because it will range between the edge and forest interior
C 2005
V
WILEY-LISS, INC.
C
Grant sponsor: NSERC; Grant sponsor: Connaught Foundation;
Grant sponsor: University of Toronto.
*Correspondence to: Dr. Shawn M. Lehman, Department of
Anthropology, University of Toronto, 100 St. George St., Toronto,
Ontario M5S 3G3, Canada. E-mail: [email protected]
Received 17 March 2004; accepted 17 November 2004.
DOI 10.1002/ajpa.20241
Published online 1 December 2005 in Wiley InterScience
(www.interscience.wiley.com).
233
LEMUR RESPONSES TO EDGE EFFECTS
in a random pattern (Lehtinen et al., 2003). These definitions follow the classic multidimensional niche model
described by Hutchinson (1957). The multidimensional
niche concept is a theoretical explanation of how different
environmental factors limit abundance and distribution.
Because each species has a range of tolerances along every
niche axis, a species can only occur in those areas where
niche axes are within ranges of tolerance. Moreover, indirect edge effects should be particularly strong during the
June–September period of food scarcity in southeastern
Madagascar.
Using the model by Lidicker (1999) and data on lemur
feeding ecology, predictions were generated on how lemurs
may respond to forest edges in the Vohibola III Special
Reserve in southeastern Madagascar. There are in total
10 lemur species in Vohibola III: one nocturnal folivore
(Avahi laniger), three nocturnal omnivores (Cheirogaleus
major, Daubentonia madagascariensis, and Microcebus
rufus), one nocturnal folivore/frugivore (Lepilemur microdon), two cathemeral frugivores (Eulemur fulvus rufus
and Eulemur rubriventer), one diurnal folivore (Hapalemur griseus griseus), and one diurnal folivore/frugivore (Propithecus diadema edwardsi). If edge effects negatively influence the distribution and density of lemur food trees, as
was observed in other tropical forests (Laurance et al.,
1997; Norconk and Grafton, 2003), then these patch
dynamics may be of particular consequence for frugivores
and other lemurs in which fruit is an important component of their diet. Fruiting trees tend to occur at low densities and produce few fruit crops in Madagascar (Ganzhorn, 1995a). Moreover, fruit crops tend to be lost due to
increased wind turbulence near the forest edge, particularly during the annual January–March cyclone season.
For example, cyclonic winds resulted in the total devastation of most fruit patches in the Manombo Special Reserve
(Ratsimbazafy, 2002). Therefore, the frugivores (E. f. rufus
and E. rubriventer) were predicted to be edge-intolerant.
Bamboo specialists (H. g. griseus) and folivores (A.
laniger) may not face the same challenges of resource
acquisition because of the relatively high abundance and
availability of leaves and bamboo compared to fruits in
humid forests (Overdorff, 1993; Tan, 2000; Powzyk, 2003).
H. g. griseus and A. laniger were predicted to be omnipresent in forest fragments. Omnivores are often the organisms least affected by edge effects (Malcolm, 2001). Thus,
M. rufus, C. major, and D. madagascariensis should be
omnipresent in forest fragments. Insectivores may prefer
forest edges because of abundant insect prey in this microhabitat (Passamani and Rylands, 2000; Spironello, 2001).
Based on these assumptions, omnivores with a diet composed of insects (M. rufus) should range nearer to forest
edges than omnivores that rarely eat insects (C. major
and D. madagascariensis). Changes in the availability of
large fruit trees may influence ranging patterns in folivore/frugivores (P. d. edwardsi and L. microdon). Therefore, P. d. edwardsi and L. microdon should tend to range
at relatively intermediate distances from the edge compared to folivores (A. laniger and H. g. griseus) and frugivores (E. f. rufus and E. rubriventer).
In this paper, data are presented on how biotic edge
effects may influence the density and distribution of
lemurs in the Vohibola III Special Reserve in southeastern
Madagascar. Specifically, we sought to answer the following questions: 1) do the distribution and density of lemurs
and lemur food trees vary as a function of proximity to forest edges, and 2) are there ecological correlates between
the distributions of lemurs and lemur food trees?
Fig. 1. Location of study site in Vohibola III Classified Forest. Triangle indicates location of Camp Mangiatsika.
METHODS
The data presented here were collected from June 1–
October 29, 2003 and May 28–September 26, 2004 at
Camp Mangiatsika in the Vohibola III Classified Forest in
southeastern Madagascar. These data were used specifically to avoid conflating seasonal variations in ranging
patterns with edge effects (Fortin et al., 1996). Moreover,
this time period is associated with a period of food
resource scarcity for lemurs in southeastern Madagascar
(Overdorff, 1993). Thus, edge effects should be particularly relevant to the ranging patterns and feeding ecology
of lemurs during June–October. Vohibola III is a 2,034-ha
forest fragment located at 208 430 South and 478 250 East,
200 km southeast of the capital city of Antananarivo
(Fig. 1). Vohibola III is at the southern end of the Fandriana-Marolambo forest corridor (Lehman, 2000). Camp
Mangatsiaka is located at 208 410 32@ South, 478 260 15@
East (1,180-m altitude) in the central section of Vohibola
III. Rainfall amounts average 2,650 mm per year, and the
heaviest rains tend to come during the October–March
warm, wet season in southeastern Madagascar (Wright,
1999). Average annual temperature is 218C, with annual
lows (48C) occurring between June–September (Overdorff,
1993).
Vohibola III is located in the midaltitude humid forest
region of southeastern Madagascar (Nicoll and Langrand,
1989). Midaltitude humid forest tends to be composed of
endemic species of Tambourissa (Monimiaceae), Ephip-
234
S.M. LEHMAN ET AL.
piandra (Monimiaceae), Ocotea (Lauraceae), Breonia
(Rubiaceae), Oncostemum (Myrsinaceae), and Cyathea
(Cyatheaceae). The shrub and herb layers include various
species of Compositae, Rubiaceae, and Myrsinaceae.
There is also a high diversity of Pandanus species (Pandanaceae), bamboos (Poaceae), and epiphytic plants (Nicoll
and Langrand, 1989; Lowry et al., 1997). The canopy is
continuous and low (ca. 10 m in height), and the tallest
trees are 25 m in height.
The matrix is composed entirely of intensive cultivation
in the area surrounding Vohibola III. Cultivation involves
rice paddies and agricultural crops such as sugar cane
(Saccharum officinarum, Poaceae) and tobacco (Nicotiana
tabacum, Solanaceae). Most cultivation involves slashand-burn agriculture, known locally as tavy, in which
native and secondary forests are cleared and burned. Various crops, mostly dry land rice and sugar cane, are
planted for approximately 3–5 years and then abandoned
for approximately 15 years. Colonizing species, including
woody plants such as Harongana madagascariensis (Clusiaceae), form a secondary thicket in abandoned cultivated
areas. The tavy cycle is repeated until all vegetation is
reduced to an impoverished secondary grassland. There is
limited farming that involves tiered rice fields, and thus
no burning of local vegetation, but this type of agriculture
is not located near Vohibola III.
In total, six 1,250-m transects were set up for lemur and
botanical surveys in Vohibola III. Following Chen et al.
(1992) and Malcolm (1994), each of the six transects ran
perpendicular from the forest edge into the forest interior.
The first tree trunk encountered on each transect was
used as the edge point for a transect. Numbered flagging
tape was used to mark 10-m increments from the forest
edge (0-m mark) into the forest interior (1,250 m) for each
transect. Botanical surveys were conducted along both
sides of each transect to a depth of 1 m, for a total area
sampled of 1.5 ha. For all trees over 5 cm diameter at
breast height (dbh), data were collected on height (m), dbh
(cm), local name, and distance to forest edge (m). These
data were used to determine the area (m2) and volume
(m3) of each tree. Voucher specimens were collected for
trees identified by local name with the assistance of local
guides. Specimens were deposited for scientific identification by botanists at Parc Tsimbazaza in Antananarivo.
Transects were walked slowly (0.5–1.0 km/hr) during
the times of day best suited for locating lemurs (0700–
1100 hr and 1400–1700 hr). Surveys for nocturnal lemurs
were conducted from 1900–2230 hr along four of the
transects (I, II, III, and V). Nocturnal surveys of transects
IV and VI were not conducted due to the steep terrain (ca.
808 slopes) and heavy rainfall. Starting points for all surveys were rotated between the forest edge and the 1,250m mark to ensure that data were not biased. The following
data were collected whenever lemurs was seen: date, time,
transect number, participants, distance along trail from
first animal seen/middle of group, species/subspecies,
group composition and size, sighting distance from trail at
908, height (m) of first animal seen, group spread, and
method of detection. Species and subspecies characteristics described in Mittermeier et al. (1994) and Garbut
(1999) were used for field identification. No animals were
captured.
Density (number of individuals per km2) was estimated
only for species that had at least 80 total sightings of individuals. These values were chosen to ensure that small
sample sizes did not bias the density analyses. Lemur densities were obtained by dividing the number of individuals
surveyed by the total survey area. Densities were determined for 100-m increments (i.e., 0–100 m, 101–200 m,
etc.) from the forest edge into the interior. Species-specific
sighting widths for each 100-m increment were estimated
using the perpendicular distance (m) from the group to
the transect and the histogram inspection technique, with
a 50% criterion for falloff distance (Whitesides et al.,
1988). This method was found to provide accurate density
estimates for lemurs in southeastern Madagascar (Johnson and Overdorff, 1999).
Tree species that form an important component (>10%
of monthly feeding scores) of the May–October diet of
lemurs were identified using data from conspecifics at
Ranomafana National Park (Overdorff, 1993; Atsalis,
1999; Tan, 2000; Faulkner, 2005). Ranomafana is in the
same biogeographic zone, is similar in plant composition,
and contains the same lemur species as Vohibola III (Lehman, 2000). Botanical data on lemur food trees were also
used to estimate densities (number of trees per hectare)
for 100-m increments. The density method used for
lemurs was replicated for trees, except that the perpendicular distance was fixed at 2 m for botanical data. Each
density estimate used only those trees specific to the diet
of each lemur species. The following mean botanical variables were also used: height (m), dbh (m), area (m2), and
volume (m3).
Chi-square (v2) tests were used to determine if there
were significant differences in survey efforts between
transects for diurnal surveys and nocturnal surveys.
Mann-Whitney U-tests were used to determine if there
were significant: 1) between-year variations in edge
effects for each lemur species, 2) differences in sighting
distances from the forest edge for M. rufus vs. omnivores,
and 3) differences in sighting distances from the forest
edge for folivore/frugivores vs. folivores and frugivores.
Kruskal-Wallis tests (H) were used to determine if there
were significant between-month and between-transect
variations in proximity to forest edge for each lemur species. Spearman rank correlations (rs) were used to determine if lemur sightings were an artifact of survey effort
and if body size correlated with mean proximity to forest
edge. Body mass was included in analyses as an additional
measure of resource requirements (Smith and Jungers,
1997). Linear and polynomial (quadratic and cubic)
regression models were used to determine how lemur
densities and species-specific food tree characteristics
(dependent variables) varied as a function of distance
from forest edge (independent variable). Polynomial
regression analyses were used because there is no reason
to assume that edge effects and response variables vary
monotonically (Murcia, 1995). If more than one model
returned a statistically significant result, then a specific
model was chosen when it explained the greatest amount
of variation in the dependent variable(s). Spearman rank
correlations were used to determine correlates between
lemur densities and species-specific food trees (density,
mean height, mean dbh, mean area, and mean volume) as
a function of distance from the forest edge. Statistical
analyses were conducted using SPSS 11.5. The alpha level
was set at 0.05 for all analyses.
RESULTS
In total, 415 diurnal (N ¼ 321) and nocturnal (N ¼ 94)
lemur surveys were conducted in Vohibola III (Table 1).
There were no significant differences in the distribution of
diurnal surveys across the six transects (v2 ¼ 3.26, df ¼ 5,
235
LEMUR RESPONSES TO EDGE EFFECTS
TABLE 1. Frequency distribution of lemur surveys conducted
along six transects in Vohibola III
Lemur survey frequency
Transect
I
II
III
IV
V
VI
Total
Diurnal
Nocturnal
Total
62
55
47
50
64
43
321
24
23
23
0
24
0
94
86
78
70
50
88
43
415
P ¼ 0.19). The frequency distribution of nocturnal surveys
also did not differ across the four transects (v2 ¼ 0.02, df ¼
3, P ¼ 0.99).
In total, 589 individuals representing four nocturnal
species (A. laniger, C. major, L microdon, and M. rufus)
and four diurnal species (E. rubriventer, E. f. rufus, H. g.
griseus, and P. d. edwardsi) were sighted during surveys
(Table 2). Bite marks on tree branches and trunks indicated the possible presence of D. madagascariensis in
Vohibola III; however, no sightings were made of this
unique species. No sightings were made of V. v. variegata.
There was no correlation between number of lemur sightings and survey effort (rs ¼ 0.551, N ¼ 6, P ¼ 0.25),
indicating that statistical results are not an artifact of
variations in survey effort. There were no annual or
monthly variations in proximity to forest edge for any of
the lemur species in Vohibola III. Moreover, there were
no significant between-transect variations in edge proximity for any of the lemur species seen in Vohibola III.
Because the data presented here are not an artifact of
temporal or spatial variations in ranging patterns, species-specific lemur data were pooled across years and
transects.
Minimum sample size requirements for computing density estimates were met only for E. rubriventer, H. g. griseus, A. laniger, and M. rufus (Table 3). There were no significant relationships between density and proximity to
forest edge in E. rubriventer (Fig. 2 and Table 4). However,
removal of the low density data point at 100 m for E.
rubriventer resulted in a highly significant cubic relationship between the distribution of this species and proximity
to the forest edge (R ¼ 0.898, ANOVA F0.006 [3, 7] ¼ 9.72).
Therefore, E. rubriventer was classified as omnipresent
or edge-tolerant. There was no significant relationship
between density and proximity to forest edge in H. g. griseus. H. g. griseus was classified as omnipresent. A cubic
regression model revealed that proximity to the forest
edge was a major determinant of density for A. laniger (R
¼ 0.860, ANOVA F0.01 [3, 8] ¼ 7.59), explaining 74.0% of the
variation in distribution and density of this species. A.
laniger was classified as edge-tolerant. There was a linear,
negative relationship between proximity to forest edge
and density for M. rufus (R ¼ 0.641, ANOVA F0.02 [1, 10]
¼ 6.98). M. rufus was classified as edge-tolerant because its
highest densities occurred at the forest edge.
M. rufus ranged significantly closer to forest edges than
C. major (U ¼ 104.0, z ¼ 2.45, P ¼ 0.014). For folivores,
there were no significant differences in sighting distances
between A. laniger and H. g. griseus (U ¼ 1,258.0, z
¼ 1.589, P ¼ 0.112). For frugivores, average proximity to
the forest edge was 530.0 6 413.5 m in E. f. rufus (N ¼ 5)
and 590.2 6 405.3 m in E. rubriventer (N ¼ 91). These values did not differ significantly (U ¼ 205.0, z ¼ 0.371, P
¼ 0.711). For the frugivore/folivores, proximity to forest
edge did not differ between P. d. edwardsi and L. microdon (U ¼ 101.0, z ¼ 1.919, P ¼ 0.057). Thus, data were
pooled for species that exhibit similar dietary patterns to
produce data sets specific to folivores, frugivores, and folivore/frugivores. There were no significant differences in
sighting distances for folivore/frugivores compared to frugivores (U ¼ 1,874.5, z ¼ 0.633, P ¼ 0.527) or folivores
(U ¼ 1,585.5, z ¼ 0.728, P ¼ 0.46). Mean distance
to forest edge was not correlated with average body size
for either males (rs ¼ 0.455, N ¼ 7, P ¼ 0.25) or females
(rs ¼ 0.476, N ¼ 7, P ¼ 0.23).
Of 7,546 trees measured along the six transects, 10.7%
(N ¼ 813) were identified as being exploited as food
resources by A. laniger, M. rufus, E. rubriventer, and H. g.
griseus during the June–October time period (Tables 5
and 6). Proximity to forest edge was a significant linear
determinant of the density of A. laniger food trees (R ¼
0.590, ANOVA F0.044 [1, 10] ¼ 5.34), explaining 34.8% of the
variation in tree density. For food trees used by M. rufus,
distance to forest edge was significantly correlated with
dbh (R ¼ 0.598, ANOVA F0.040 [1, 10] ¼ 5.59), height (R ¼
0.778, ANOVA F0.049 [3, 8] ¼ 4.11), area (R ¼ 0.787, ANOVA
F0.043 [3, 8] ¼ 4.36), and volume (R ¼ 0.665, ANOVA F0.018 [1, 10]
¼ 7.91). The area (R ¼ 0.799, ANOVA F0.035 [3, 8] ¼ 4.71)
and volume (R ¼ 0.809, ANOVA F0.029 [3, 8] ¼ 5.08) of food
trees used by E. rubriventer varied as cubic functions of
distance from forest edge. Food trees exploited by H. g.
griseus varied as a function of distance from forest edge
for density (R ¼ 0.683, ANOVA F0.014 [1, 10] ¼ 8.76), area
(R ¼ 0.815, ANOVA F0.026 [3, 8] ¼ 5.30), and volume (R
¼ 0.828, ANOVA F0.021 [3, 8] ¼ 5.80).
The density of M. rufus was positively correlated with
the density of trees used as food resources by this lemur
species (rs ¼ 0.589, df ¼ 12, P ¼ 0.04; Table 7). There were
no significant correlations between the distribution and
density of A. laniger, E. rubriventer, or H. g. griseus and
any characteristics of species-specific food trees. Removal
of the low density outlier for E. rubriventer did not result
in any significant correlations with food trees exploited by
this species.
DISCUSSION
The first prediction was that E. rubriventer, the only
frugivore for which density estimates could be computed,
would be edge-intolerant due to the loss of fruit trees near
forest edges. However, E. rubriventer ranged widely (5–
1,250 m from the edge) throughout Vohibola III, and
regression analyses revealed that this species is either
edge-tolerant or omnipresent. Although the area and volume of food trees exploited by E. rubriventer showed
marked edge effects, there were no botanical correlates
with the distribution and density of this lemur species.
Lack of covariation between E. rubriventer and its food
trees may be because the survey data were collected during the period of relatively low fruit feeding and relatively
high exploitation of leaves for conspecifics at nearby Ranomafana National Park. For example, Overdorff (1993)
found that overall feeding time and percentage of feeding
time on fruits were lowest from August to mid-October.
Conversely, the percentage of feeding time E. rubriventer
spent feeding on leaves was highest during July to midSeptember. Thus, dietary patterns of E. rubriventer were
more like those of a folivore/frugivore during the time
period of this study, in which case this lemur should have
been predicted to be omnipresent in Vohibola III. An interesting question that cannot answered at this time is if E.
236
S.M. LEHMAN ET AL.
TABLE 2. Descriptive and comparative statistics on proximity to forest edge for eight lemur species in Vohibola III1
Number of sightings
Species
Individuals
A. laniger
C. major
E. f. rufus
E. rubriventer
H. g. griseus
L. microdon
M. rufus
P. d. edwardsi
85
6
16
212
109
19
96
46
Total
589
1
2
3
Groups
Distance from edge (m)
Mean
SD
Range
17
573.9
888.3
530.0
590.2
684.3
677.5
437.7
396.4
432.0
418.1
413.5
405.3
419.3
427.2
287.8
337.3
0–1,250
90–1,250
140–1,080
5–1,250
15–1,250
80–1,234
0–1,230
90–1,250
160
568.3
396.8
0–1,250
5
91
47
Temporal and spatial variations
Month2
3.39
2.14
3.80
4.59
6.94
2.51
2.84
5.24
Year3
Transect2
1.15 (0.24)
1.46 (0.33)
NA
0.475 (0.63)
0.427 (0.66)
0.399 (0.74)
0.507 (0.61)
1.64 (.013)
(0.33)
(0.14)
(0.28)
(0.33)
(0.13)
(0.47)
(0.58)
(0.26)
1.40
1.42
3.80
5.01
4.07
5.64
6.78
1.83
(0.30)
(0.69)
(0.28)
(0.41)
(0.53)
(0.06)
(0.06)
(0.76)
Statistics for group-living primates are based on sighting distance for group rather than individuals.
Kruskal-Wallis H (P-value).
Mann-Whitney z-score (P-value); E. f. rufus surveyed only in 2004.
TABLE 3. Density estimates (number of individuals/km2) as function of proximity to forest edge for four species
of lemurs in Vohibola III
E. rubriventer
H. g. griseus
A. laniger
M. rufus
Distance (m)
Mean
SD
Mean
SD
Mean
SD
Mean
SD
0–100
101–200
201–300
301–400
401–500
501–600
601–700
701–800
801–900
901–1,000
1,001–1,100
1,101–1,250
11.1
75.0
46.9
20.8
16.2
15.3
31.3
6.3
8.3
27.8
16.7
37.1
6.6
15.7
17.2
7.4
2.9
16.5
19.0
5.0
4.7
26.5
25.7
36.0
8.3
10.7
17.9
11.9
17.9
7.1
21.4
2.4
19
9.5
8.9
21.2
1.7
2.7
15.2
7.1
18.4
15.1
9.4
5.7
16.8
18.9
15.4
13.8
67.5
10.0
0.0
7.5
7.5
12.5
7.5
10.0
12.5
25.0
25.0
15.0
18.1
3.9
0.0
5.4
5.3
6.3
4.4
5.7
2.5
8.6
7.8
3.9
40.6
21.9
46.9
12.5
59.4
43.8
9.4
12.5
12.5
12.5
6.3
4.2
6.5
20.5
15.2
14.3
15.3
11.4
13.7
5.6
5.8
2.4
0.0
10.6
Total
26.7
10.2
13.4
12.2
16.6
6.6
rubriventer will be edge-intolerant during the period of
maximum fruit exploitation (ca. November–early June)?
Data are being collected during this time period, and this
question will be investigated in future studies.
The second prediction was that H. g. griseus (bamboo
specialist) and A. laniger (folivore) would be omnipresent
in Vohibola III. H. g. griseus was found to be omnipresent,
and there were clear negative edge effects associated with
the area and volume of food trees eaten by this lemur species. Despite this evidence of edge effects for food trees,
none of the botanical variables covaried with the distribution and density of H. g. griseus. Lack of correlations
between food trees and the distribution and density of H.
g. griseus is not unexpected, given that Tan (1999) found
that 72% of the annual diet of this lemur is comprised of
giant bamboo (Cathariostachys madagascariensis). Bamboo was not included in botanical surveys in Vohibola III.
A. laniger was clearly tolerant of forest edges, and thus
refuted the a priori prediction of its being omnipresent.
For example, three separate sightings were made of individuals resting in the canopies of trees that overhung the
matrix. However, none of the botanical measures correlated with the distribution and density of A. laniger. Clinal
variations in food quality rather than abundance represent a possible covariate to the distribution and abundance of A. laniger. Ganzhorn (1995) documented that
low-intensity logging increased light levels in western dry
forests, which resulted in higher protein concentrations in
leaves. Elevated light levels were documented near forest
edges in Vohibola III (Lehman, unpublished findings).
24
5.8
Thus, the quality of leaves may be highest near forest
edges in Vohibola III. These edge-related variations in food
quality are particularly relevant to A. laniger. A. laniger is
a small-bodied (600–1,300 g) folivore with a simple monogastric stomach. This lemur species lacks two of the key
morphological adaptations associated with folivory: large
body size and a complex sacculated stomach (Faulkner,
2005). If edges do contain higher-quality food sources for
folivores, then A. laniger should be expected to be edge-tolerant. Future studies will seek to test this hypothesis by
comparing leaf chemistry at differing proximities to forest
edges in Vohibola III.
M. rufus was predicted to be omnipresent and to range
nearer the forest edge than C. major. However, M rufus
was classified as edge-tolerant. Moreover, the density and
distribution of this lemur species were positively correlated with the density and distribution of its food trees in
Vohibola III. Tolerance for edge effects may also be due to
the abundance of insect prey near the forest edge (Corbin
and Schmid, 1995), although ecological patterns of insect
abundance have not been studied directly in southeastern
Madagascar. Previous surveys noted an abundance of M.
rufus near forest edges and secondary forests at sites in
the main forest corridor 35 km north of Vohibola III (Lehman and Ratsimbazafy, 2000). Furthermore, M. rufus
tended to preferentially consume arthropods during the
period of this study (Atsalis, 1998). Malcolm (1997) noted
similar patterns in the distribution of insectivorous mammals in Brazil. Specifically, he documented higher arthropod populations near forest edges, which explained why
237
LEMUR RESPONSES TO EDGE EFFECTS
Fig. 2. Density variations and polynomial regression coefficients as function of distance from forest edge for four lemur species
in Vohibola III. Arrow in plot for E. rubriventer indicates outlier value removed for subsequent analyses.
TABLE 4. Linear and polynomial regression models of relationship between lemur density and proximity to forest edge,
with statistically significant models in bold
Species
E. rubriventer
E. rubriventer
(outlier removed)
H. g. griseus
A. laniger
M. rufus
Model
Model R
Model R2
df
F
P
b0
b1
Linear
Quadratic
Cubic
Linear
Quadratic
Cubic
Linear
Quadratic
Cubic
Linear
Quadratic
Cubic
Linear
Quadratic
Cubic
0.276
0.437
0.565
0.456
0.872
0.898
0.155
0.164
0.477
0.170
0.600
0.860
0.641
0.657
0.680
0.076
0.191
0.319
0.208
0.761
0.807
0.024
0.027
0.228
0.029
0.360
0.740
0.411
0.431
0.463
1, 10
2, 9
3, 8
2, 9
3, 8
3, 7
1, 10
2, 9
3, 8
1, 10
2, 9
3, 8
10
2, 9
3, 8
0.82
1.06
1.25
2.36
12.71
9.73
0.25
0.13
0.79
0.30
2.53
7.59
6.98
3.41
2.30
0.386
0.385
0.354
0.159
0.003
0.007
0.629
0.883
0.534
0.596
0.134
0.010
0.025
0.079
0.154
35.69
53.45
25.16
46.34
106.01
141.08
11.29
12.21
9.88
21.98
48.85
92.32
45.28
37.90
24.36
4.70
8.90
12.28
2.68
23.65
43.86
2.66
1.25
8.32
8.10
1.20
44.61
3.34
3.10
9.83
b2
b3
0.560
3.000
0.190
1.000
4.386
0.146
0.029
0.147
0.0742
0.850
6.667
0.000
0.200
2.000
0.089
b refers to beta coefficients.
insectivorous mammals were unaffected by edge effects.
This hypothesis is supported by the fact that M. rufus
ranged significantly closer to the forest edge than C.
major, which do not rely heavily on insects. Ultimately,
detailed studies must be conducted on how edge effects
influence the abundance and availability of insects eaten
by lemurs in Vohibola III.
The final prediction tested in this paper was that folivores/frugivores (P. d. edwardsi and L. microdon) should
tend to range at relatively intermediate distances from
the edge compared to folivores (A. laniger and H. g. griseus) and frugivores (E. f. rufus and E. rubriventer). The
data presented here do not support this prediction. The
reasons that this prediction were not supported are diffi-
cult to determine, given the small sample sizes for P. d.
edwardsi, L. microdon, and E. f. rufus. However, each of
these three species tended to range widely throughout the
forest (80–1,250 m from the forest edge). Although the
species-specific mean distances from the forest edge
ranged from a minimum of 396.4 m for P. d. edwardsi to a
maximum of 677.5 m for L. microdon, the associated
standard deviations were very large (337.3 m and 427.2
m, respectively). These descriptive statistics tend to indicate that P. d. edwardsi, L. microdon, and E. f. rufus are
more likely to be omnipresent than edge-intolerant. Further surveys should increase sample sizes to the point
where more detailed comparative analyses can be conducted.
238
S.M. LEHMAN ET AL.
TABLE 5. Scientific and local names of plants eaten by four lemur species during June–October1
Scientific name2
Family
Local name
E. rubriventer2
H. g. griseus
A. laniger
M. rufus
Total
Anthocleista madagascariensis
Aphloia theiformis
Canthium sp.
Canthium sp.
Dombeya laurifolia
Erythroxylum nitidulum
Eugenia sp.
Ficus soroceoides
Gambeya madagascarensis
Gaertnera sp.
Grewia humblotii
Harungana modagascariensis
Memecylon aff. delphinense
Musaenda sp.
Nuxia pachyphylla
Oconstemon leprosum
Ocotea cymosa
Ocotea laevis
Psychotria polyphylla
Syzygium phyllyreifolium
Loganiaceae
Aphloiaceae
Rubiaceae
Rubiaceae
Sterculiaceae
Erythroxylaceae
Myrtaceae
Moraceae
Sapotaceae
Rubiaceae
Tiliaceae
Clusiaceae
Melastomataceae
Rubiaceae
Loganiaceae
Myrsinaceae
Lauraceae
Lauraceae
Rubiaceae
Myrtaceae
Variahy
Ravimboafotsy
Fantsikahitra
Ravimboanjo
Hafibalo
Ravimbolo
Ratrifotsy
Ravosa
Famakilela
Taolagnana
Hafipotsy
Harongana
Tomenjy
Fatora
Lambinana
Hazotohoka
Varongifinga
Varongy
Voafotsiala
Rotra
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
1
1
1
1
1
1
1
2
1
2
1
2
1
1
1
2
1
1
1
1
5
4
8
7
24
Total
1
2
X, plant species eaten by that lemur species; , not eaten by that lemur species.
E. rubriventer from Overdorff, 1993; H. g. grisesus from Tan, 2000; A. lanifer from Faulkner, 2005; M. rufus from Atsalis, 1999.
TABLE 6. Linear and polynomial regression models of relationship between characteristics of species-specific lemur food trees and
proximity to forest edge1
Species
Variables
Model
Model R
Model R2
E. rubriventer
Mean area
Mean volume
Density
Mean area
Mean volume
Density
Mean dbh
Mean height
Mean area
Mean area
Mean area
Mean volume
Cubic
Cubic
Linear
Cubic
Cubic
Linear
Linear
Cubic
Linear
Quadratic
Cubic
Linear
0.799
0.809
0.683
0.815
0.828
0.590
0.598
0.778
0.763
0.771
0.787
0.665
0.638
0.655
0.467
0.665
0.685
0.348
0.358
0.606
0.582
0.595
0.620
0.442
H. g. griseus
A. laniger
M. rufus
F
P
b0
b1
b2
b3
4.71
5.08
8.76
5.30
5.80
5.34
5.59
4.11
13.91
6.60
4.36
7.91
0.035
0.029
0.014
0.026
0.021
0.044
0.040
0.049
0.004
0.017
0.043
0.018
0.59
10.04
35.70
0.24
3.45
159.15
31.72
10.47
0.12
0.13
0.15
1.40
0.003
0.520
0.430
0.009
0.015
6.110
1.220
2.100
0.008
0.011
0.000
0.001
4.706
9.036
0.209
0.423
0.0001
0.0022
0.0006
0.0092
3.200
0.142
0.000
0.000
0.000
df
3,
3,
1,
3,
3,
1,
1,
3,
1,
2,
3,
1,
8
8
10
8
8
10
10
8
10
9
8
10
1
Only statistically significant relationships are presented.
b refers to beta coefficients.
TABLE 7. Spearman rank correlations between lemur densities and characteristics of species-specific food trees,
both measured as function of distance from forest edge
Species
E. rubriventer
E. rubriventer1
H. g. griseus
A. laniger
M. rufus
1
Density
0.403
0.296
0.037
0.218
0.589
(0.19)
(0.37)
(0.98)
(0.41)
(0.04)
Mean dbh
0.342
0.300
0.408
0.304
0.377
(0.27)
(0.37)
(0.18)
(0.33)
(0.22)
Mean height
0.112
0.086
0.122
0.474
0.386
(0.72)
(0.79)
(0.70)
(0.11)
(0.21)
Mean area
0.336
0.291
0.049
0.320
0.557
(0.28)
(0.38)
(0.87)
(0.30)
(0.06)
Mean volume
0.307
0.281
0.171
0.361
0.520
(0.33)
(0.40)
(0.593)
(0.24)
(0.08)
100-m outlier removed.
Lemur body size was not correlated with proximity to
forest edge in our study. Body size is associated allometrically with metabolic rate, energetic demands, and physical
performance in animals (Schmidt-Nielsen, 1997). In terms
of energetic demands and body size, large animals tend to
require larger home ranges and feeding areas than small
animals (Calder, 1984). Thus, it is hypothesized that
larger-sized animals are more at risk of extinction, and it
could be argued that smaller-bodied primates would be
best suited to exploit habitat edges (Gaston and Blackburn, 1996). However, there is little support for relation-
ships between body size and edge responses, forest fragmentation, and habitat loss (reviewed in Davies et al.,
2000). Ultimately, the lack of relationship between body
size and edge effects may be a consequence of the many
ecological and phylogenetic covariates to body size in primates (Gittleman and Purvis, 1998).
Understanding generalized lemur responses to edge
effects may explain how they have survived dramatic habitat loss and forest fragmentation in Madagascar. Because
lemurs are one of the world’s top conservation priorities,
there is considerable research being focused on why cer-
LEMUR RESPONSES TO EDGE EFFECTS
tain species are rare, and whether or not we can predict
which species are most likely to go extinct (Jernvall and
Wright, 1998). Studies of extinction probabilities often
invoke some aspect of species-area relationships. Speciesarea relationships predict a positive relationship between
number of species and size of an area (Rosenzweig, 1995).
This relationship is expressed as the equation S ¼ CAz,
where S is species richness, C is a fitted constant that
varies among taxa and types of ecosystems, and z is a constant that tends to range from 0.10–0.50. It is surprising
in light of recent theoretical work on thresholds of forest
loss and fragmentation that dramatic landscape changes
have not resulted in the extinction of lemur species in
Madagascar (Fahrig, 2002), although there have been
many species extirpations (Godfrey et al., 1999). The ability of lemurs to tolerate edge effects may have enabled
them to survive such dramatic landscape changes.
Lemurs may not be the only tropical taxa unaffected by
edge effects. Malcolm (1997) found that the abundance of
many species of arboreal mammals was not affected by
forest fragmentation, matrix conditions, or edge effects in
Brazil.
The generally positive or negligible influences that forest edges have on lemurs are being offset by deteriorating
characteristics of their forest habitats. Only one regression model returned a positive relationship between a botanical measure and proximity to the forest edge (density
of H. g. griseus food trees). All other significant botanical
regression models indicate that the density and dendrometrics of food trees are being negatively influenced by
edge effects in Vohibola III. Thus, both the density and
size of lemur food trees are decreasing near the forest
edge. Continued deforestation may result in forest fragments being composed entirely of edge habitats. Although
Ganzhorn (1995) noted that low-intensity logging increased light levels and the quality of food resources in
dry forests, he also documented that more intense logging
resulted in a decline in lemur populations. For example,
increased temperatures near forest edges may be deleterious to nocturnal lemurs, such as Microcebus murinus,
that enter energy-saving torpor during the dry season in
dry forests (Ganzhorn and Schmid, 1998). These data indicate that there may be a threshold of habitat disturbance
and possibly edge effects for lemurs. Determining what
this edge threshold is may hold important answers for
questions about extirpation and extinction patterns in
lemurs.
Finally, edge-related variations in predation on lemurs
may influence their distribution and density in Vohibola
III. Numerous studies of predator-prey relationships and
edge effects indicated variations in predation pressures in
forest fragments (reviewed in Lahti, 2001). In many studies, predation rates were heightened for nesting birds
near forest edges. All lemurs species experience predation
pressures from raptors (e.g., Accipter henstii and Polyboroides radiatus), reptiles (e.g., Acrantophis madagascariensis), and/or carnivores (e.g., Cryptoprocta ferox) in
southeastern Madagascar (Goodman et al., 1993, 1998;
Wright et al., 1997; Rakotondravony et al., 1998; Burney,
2002; Goodman, 2003). Furthermore, Karpanty (2003)
documented that lemurs can distinguish between different types of predators (aerial or terrestrial) and then alter
their activity patterns to avoid predation. Thus, if there
are edge effects associated with the distribution and abundance of predators, then lemurs may be responding to this
effect in conjunction with the distribution and quality of
food resources. One possible method of testing for varia-
239
tions in predation rates as a function of edge effects would
be to conduct a series of mark-and-release experiments
with M. rufus at different distances from the forest edge.
Although such an experiment would not directly measure
predation, it should provide data on edge-related variations in mortality. Following Karpanty (2003), experiments could also be conducted using playbacks of predator
vocalizations for conspecific lemurs at differing distances
from the forest edge. These data may reveal important differences in how lemurs respond to predation risks relative
to proximity to forest edges.
CONCLUSIONS
The first question addressed in this paper related to
lemur and food tree responses to edge effects. Although
lemurs responded either positively or not at all to edge
effects, there was an overwhelming negative influence of
edge effects on the density and dendrometrics of lemur
food trees. The second question related to ecological correlates between the distribution of lemurs and their food
trees. These correlates were found only for the density of
M. rufus and its food trees. The predictions tested herein
related to food abundance as a function of edge effects.
However, edge-related variations in food quality and predation pressures may occur in Vohibola III. Thus, relationships between edge effects and lemur biogeography reflect
complex and varying causalities in Vohibola III, and perhaps in all humid forests in southeastern Madagascar.
Increased sample sizes, chemical data on food quality, and
experimental studies of predation pressures in forest fragments will provide a greater understanding of how lemurs
respond to edge effects in southeastern Madagascar.
Finally, data are needed on how forests and lemurs are
influenced by edge effects in other biogeographic regions
in Madagascar. These data will be critical to an increased
understanding of how edge effects influence the biogeography and conservation biology of lemurs in the rapidly
vanishing forest landscapes of Madagascar.
ACKNOWLEDGMENTS
We thank the Association Nationale pour la Gestion des
Aires Protégées, le Ministère de l’Eau et de Forêt, l’Office
National pour l’Environnement à Madagascar, and the
University of Antananarivo for permission to conduct our
research in Madagascar. We thank Patricia Wright, Benjamin Andriamahaja, and the staff at Institute for the
Conservation of Tropical Environments (ICTE) and Malagasy Institute for the Conservation of Tropical Environments (MICET) for their support, advice, and hospitality.
We are extremely grateful to Andriamihanta ‘‘Lekely’’
Harison, Zafimamonjy, Andriaharizaka Johnson, Andriamampiakatra ‘‘Ndrema’’ Rajaoarivony, Andriaharizaka
‘‘Tsiamidy’’ Andry, Randrianjandrimalalason Celestine,
Razatharimalaladraimy William, Razafimananjara Samuel, and Randriaharimanana Joelson for sharing their
knowledge of the forest, for assisting us with data collection, and for their friendship and support. We greatly
appreciate the hospitality and kindness of the Mayor and
people of the villages of Ambohimitombo and Sahanato.
We thank the Geographic Information System (GIS) unit of
the Royal Botanic Gardens at Kew for access to the GIS database on forest cover in Madagascar, Angel Vats for assistance
with data collection, and Natasha Bijelich and Serena Park
for data entry. A previous draft of the manuscript benefited
240
S.M. LEHMAN ET AL.
greatly from the comments of Colin Chapman, Robert Sussman, and two reviewers.
LITERATURE CITED
Atsalis S. 1998. Feeding ecology and aspects of life history in
Microcebus rufus (family Cheirogaleidae, order Primates).
Ph.D. dissertation, City University of New York, New York.
Atsalis S. 1999. Diet of the brown mouse lemur (Microcebus
rufus) in Ranomafana National Park. Int J Primatol 20:193–
229.
Burney DA. 2002. Sifaka predation by a large boa. Folia Primatol (Basel) 73:144–145.
Calder W. 1984. Size, function, and life history. Cambridge, MA:
Harvard University Press.
Chen J, Franklin JF, Spies TA. 1992. Vegetation responses to
edge environments in old-growth Douglas-fir forests. Ecol
Appl 2:387–396.
Corbin GD, Schmid J. 1995. Insect secretions determine habitat
use patterns by a female lesser mouse lemur (Microcebus
murinus). Am J Primatol 37:317–324.
Curran LM, Caniago I, Paoli GD, Astiani D, Kusneti M,
Leighton M, Nirarita CE, Haeruman H. 1999. Impact of El
Niño and logging on canopy tree recruitment in Borneo. Science 286:2184–2188.
Davies KF, Margules CR, Lawrence JF. 2000. Which traits of
species predict population declines in experimental forest
fragments? Ecology 81:1450–1461.
Du Puy D, Moat J. 1998. Vegetation mapping and classification
in Madagascar (using GIS): implications and recommendations for the conservation of biodiversity. In: Huxley CR, Lock
JM, Cutler DF, editors. Chorology, taxonomy and ecology of
the floras of Africa and Madagascar. Kew: Royal Botanic Gardens. p 97–117.
Fahrig L. 2002. Effect of habitat fragmentation on the extinction threshold: a synthesis. Ecol Appl 12:346–353.
Faulkner A. 2005. The relationship between leaf chemistry and
feeding patterns in a small-bodied nocturnal primate (Avahi
laniger). M.Sc. thesis, University of Toronto, Toronto.
Fortin M-J, Drapeau P, Jacquez GM. 1996. Quantification of
the spatial co-occurrences of ecological boundaries. Oikos 77:
51–60.
Ganzhorn J. 1995. Low-level forest disturbances effects on primary production, leaf chemistry, and lemur populations. Ecology 76:2084–2096.
Ganzhorn JU, Schmid J. 1998. Different population dynamics of
Microcebus murinus in primary and secondary deciduous dry
forests of Madagascar. Int J Primatol 19:785–796.
Garbut N. 1999. Mammals of Madagascar. Sussex: Pica Press.
Gaston KJ, Blackburn TM. 1996. Rarity and body size: importance of generality. Conserv Biol 10:1295–1298.
Gittleman JL, Purvis A. 1998. Body size and species-richness in
carnivores and primates. Proc R Soc Lond [Biol] 265:113–119.
Godfrey LR, Jungers WL, Simons EL, Chatrath PS, Rakotosamimanana B. 1999. Past and present distributions of lemurs
in Madagascar. In: Rakotosamimanana B, Rasamimanana H,
Ganzhorn JU, Goodman SM, editors. New directions in lemur
studies. New York: Kluwer Academic/Plenum. p 19–53.
Goodman SM. 2003. Predation on lemurs. In: Goodman SM,
Benstead JP, editors. The natural history of Madagascar. Chicago: University of Chicago Press. p 1221–1228.
Goodman SM, O’Connor S, Langrand O. 1993. A review of predation on lemurs: implications for the evolution of social
behavior in small, nocturnal primates. In: Kappeler PM,
Ganzhorn JU, editors. Lemur social systems and their ecological basis. New York: Plenum Press. p 51–66.
Goodman SM, de Roland LAR, Thorstrom R. 1998. Predation on
the eastern woolly lemurs (Avahi laniger) and other vertebrates by Henst’s goshawk (Accipiter henstii). Lemur News
3:14–15.
Green GM, Sussman RW. 1990. Deforestation history of the
eastern rain forests of Madagascar from satellite images.
Science 248:212–215.
Hutchinson GE. 1957. Concluding remarks. Cold Spring Harbor
Symp Quant Biol 22:415–427.
Jernvall J, Wright PC. 1998. Diversity components of impending primate extinctions. Proc Natl Acad Sci USA 95:11279–
11283.
Johnson S, Overdorff DJ. 1999. Census of brown lemurs (Eulemur fulvus spp.) in southeastern Madagascar: methods-testing and conservation implications. Am J Primatol 47:51–60.
Karpanty SM. 2003. Behavioral and ecological interactions of
raptors and lemurs in Madagascar: a multiple-predator
approach. Ph.D. dissertation, Stony Brook University, Stony
Brook.
Lahti DC. 2001. The ‘‘edge effect on nest predation’’ hypothesis
after twenty years. Biol Conserv 99:365–374.
Laurance WF. 1999. Reflections on the tropical deforestation crisis. Biol Conserv 91:109–117.
Laurance WF, Yensen E. 1991. Predicting the impacts of edge
effects in fragmented habitats. Biol Conserv 57:205–219.
Laurance WF, Laurance SG, Ferreira LV, Rankin-de Merona
JM, Gascon C, Lovejoy TE. 1997. Biomass collapse in Amazonian forest fragments. Science 278:1117–1118.
Lehman SM. 2000. Final report on the rapid biological and socioeconomic assessments conducted in the Fandriana-Marolambo forest corridor in eastern Madagascar. Antananarivo,
Madagascar: National Office of the Environment. p 156.
Lehman SM, Ratsimbazafy JH. 2000. Lemurs of the FandrianaMarolambo forest corridor, Madagascar. Antananarivo, Madagascar: National Office of the Environment.
Lehtinen RM, Ramanamanjato J-B, Raveloarison JG. 2003.
Edge effects and extinction proneness in a herpetofauna from
Madagascar. Biodivers Conserv 12:1357–1370.
Lidicker WZ. 1999. Responses of mammals to habitat edges: an
overview. Land Ecol 14:333–343.
Lovejoy TE, Bierregaard RO, Rylands AB, Malcolm JR, Quintela CE, Harper LH, Brown KS, Powell AH, Powell VN, Shubart HOR, Hays MB. 1986. Edge and other effects of isolation
on Amazon forest fragments. In: Soule ME, editor. Conservation biology: the science of scarcity and diversity. Sunderland,
MA: Sinauer Associates. p 257–285.
Lowry PP, Schatz GE, Phillipson PB. 1997. The classification of
natural and anthropogenic vegetation in Madagascar. In:
Goodman SM, Patterson BD, editors. Natural change and
human impact in Madagascar. Washington, DC: Smithsonian
Institution Press. p 93–123.
Malcolm JR. 1994. Edge effects in central Amazonian forest
fragments. Ecology 75:2438–2445.
Malcolm JR. 1997. Biomass and diversity of small mammals in
Amazonian forest fragments. In: Laurance WF, Bierregaard
O, editors. Tropical forest remnants: ecology, management,
and conservation of fragmented communities. Chicago: University of Chicago Press. p 207–221.
Malcolm JR. 2001. Extending models of edge effects to diverse
landscape configurations, with a test case from the neotropics.
In: Bierregaard RO, Gascon C, Lovejoy TE, Mesquita RM, editors. Lessons from Amazonia: the ecology and conservation
of a fragmented forest. New Haven: Yale University Press.
p 346–357.
Mbora DNM, Meikle DB. 2004. Forest fragmentation and the
distribution, abundance and conservation of the Tana River
red colobus (Procolobus rufomitratus). Biol Conserv 118:67–77.
Mittermeier RA, Tattersall I, Konstant WR, Meyers DM, Mast
RB. 1994. Lemurs of Madagascar. Washington, DC: Conservation International.
Murcia C. 1995. Edge effects in fragmented forests: implications
for conservation. TREE 10:58–62.
Nicoll ME, Langrand O. 1989. Madagascar: revue de la conservation et des aires protégées. Gland: WWF.
Norconk MA, Grafton BW. 2003. Changes in forest composition
and potential feeding tree availability on a small land-bridge
island in Lago Guri, Venezuela. In: Marsh LK, editor. Primates in fragments: ecology and conservation. New York:
Kluwer Academic/Plenum Publishers. p 211–227.
Overdorff DJ. 1993. Similarities, differences, and seasonal patterns in the diets of Eulemur rubriventer and Eulemur fulvus
LEMUR RESPONSES TO EDGE EFFECTS
rufus in the Ranomafana National Park, Madagascar. Int J
Primatol 14:721–753.
Passamani M, Rylands AB. 2000. Home range of a Geoffroy’s
marmoset group, Callithrix geoffroyi (primates, Callitrichidae)
in south-eastern Brazil. Rev Bras Biol 60:275–281.
Powzyk JA. 2003. Dietary and feeding differences between sympatric Propithecus diadema diadema and Indri indri. Int J
Primatol 24:1143–1162.
Rakotondravony D, Goodman SM, Soarimalala V. 1998. Predation on
Hapalemur griseus griseus by Boa manditra (Boidae) in the littoral
forest of eastern Madagascar. Folia Primatol (Basel) 69:405–408.
Ratsimbazafy J. 2002. On the brink of extinction and the process of recovery: responses of black-and-white ruffed lemurs
(Varecia variegata variegata) to disturbance in Manombo Forest, Madagascar. Unpublished Ph.D. dissertation, Stony Brook
University, Stony Brook, NY.
Rosenzweig ML. 1995. Species diversity in space and time.
Cambridge: Cambridge University Press.
Schmidt-Nielsen K. 1997. Animal physiology: adaptation and
environment. Cambridge: Cambridge University Press.
Smith RJ, Jungers WL. 1997. Body mass in comparative primatology. J Hum Evol 32:523–559.
Spironello WR. 2001. The brown capuchin monkey (Cebus apella):
ecology and home range requirements in central Amazonia. In:
241
Bierregaard RO, Gascon C, Lovejoy TE, Mesquita R, editors.
Lessons from Amazonia: the ecology and conservation of a fragmented forest. New Haven: Yale University Press. p 271–283.
Tan C. 1999. Group composition, home range size, and diet of
three sympatric bamboo lemur species (genus Hapalemur) in
Ranomafana National Park, Madagascar. Int J Primatol
20:547–566.
Tan C. 2000. Behavior and ecology of three sympatric bamboo
lemur species (genus Hapalemur) in Ranomafana National
Park, Madagascar. Ph.D. dissertation, State University of
New York at Stony Brook, Stony Brook, NY.
Tweheyo M, Lye KA, Weladjic RB. 2004. Chimpanzee diet and
habitat selection in the Budongo Forest Reserve, Uganda.
Forest Ecol Manage 188:267–278.
Whitesides GH, Oates JF, Green SM, Kluberdanz RP. 1988. Estimating primate densities from transects in a West African rain
forest: a comparison of techniques. J Anim Ecol 57:345–367.
Wright PC. 1999. Lemur traits and Madagascar ecology: coping
with an island environment. Yrbk Phys Anthropol 42:31–42.
Wright PC, Heckscher SK, Dunham AE. 1997. Predation on
Milne-Edward’s sifaka (Propithecus diadema edwardsi) by the
fossa (Cryptoprocta ferox) in the rain forest of southeastern
Madagascar. Folia Primatol (Basel) 68:34–43.