вход по аккаунту



код для вставкиСкачать
Austral Ecology (2017) , –
Edge effects on small mammals: Differences between
arboreal and ground-dwelling species living near roads in
Brazilian fragmented landscapes
Instituto Fauna Selvagem, Street 3600, n 232, 88330-248 Balne
ario Cambori
u, Santa Catarina
ucleo em Ecologia e Desenvolvimento Socio-Ambiental de Macae,
(Email: [email protected]), N
Universidade Federal do Rio de Janeiro, Macae, Rio de Janeiro, 3Laboratorio de Estudos e Projetos em
Manejo Florestal, Universidade Federal de Lavras, Lavras, Minas Gerais, and 4Brazilian Center for
Road Ecology Research, Universidade Federal de Lavras, Lavras, Minas Gerais, Brazil
Abstract Habitat fragmentation often induces edge effects that can increase, decrease or have minimal effect
upon the population density of a species, depending upon environmental conditions and the requirements of the
species. Using a trapping study and generalized linear mixed models, we evaluated edge effects on small tropical
mammals living near roads, including two ground-dwelling (Akodon sp. and Cerradomys subflavus) and two arboreal (Marmosops incanus and Riphidomys sp.) species. We examined the relationship of these edge effects to environmental factors at both plot and patch scales. Generalist ground-dwelling species were attracted to edges, with
higher population densities recorded in habitats close to road or matrix edges where vegetation density was
lower. In contrast, populations of the generalist arboreal species avoided edge habitats, their populations were
found in greater density in habitats far from roads/matrix edges. Thus, our results show that patterns of edge
habitat utilization were related to the ecological requirements of each species. These findings are especially
important in the tropics, where demand for economic growth in many countries has accelerated the fragmentation process and has recently culminated in increased road construction and expansion. Fragmented habitats
promote an increase in edge environments, and consequently will reduce the abundance of arboreal small mammal species, such as those used as models in this study.
Key words: habitat selection, highway, matrix, road ecology.
Deforestation is a primary cause of global biodiversity
decline and has resulted in the loss of more than a
third of all forest cover worldwide (Hansen et al.
2013). Deforestation typically leads to habitat fragmentation, transforming large areas of native forest
into a number of smaller patches isolated by a matrix
of human-transformed land cover (Haddad et al.
2015). Habitat fragmentation induce an increase in
the amount of edge habitat, consisting of forest
patches exposed to abrupt abiotic (e.g. microclimate)
and biotic (e.g. vegetation structure) changes due to
the condition of the human-transformed matrix (Fahrig 2003; Haddad et al. 2015).
Induced edges initiate long-term changes to the
structure and function of fragments, and may affect
the survival, reproduction, behaviour and interspecific relationships (e.g. competition and predation) of
organisms (Fahrig 2003; Jaeger et al. 2005). On the
*Corresponding author.
Accepted for publication September 2017.
© 2017 Ecological Society of Australia
behavioural level, changes at edges trigger three possible responses: (i) edge attraction, which occurs
when the number of individuals increases at habitat
edges; (ii) edge avoidance, when individuals avoid
the edges, reducing the number of individuals in
these areas or (iii) neutral edge effects, when there is
no significant effect on the individual, population or
community level (Jaeger et al. 2005; Rosa & Bager
Most vertebrates demonstrate edge avoidance or
no edge effect, although edge attraction is a common
response for small mammals, especially in temperate
zones (McGregor et al. 2008; Fahrig & Rytwinski
2009; Rosa & Bager 2013; Michal & Rafal 2014).
However, in the tropics, some rainforest-dependent
species of small mammals avoid edges (Goosem
2000, 2007). A number of studies suggest that in the
tropics, ground-dwelling species are prone to edge
attraction behaviour (Bergallo et al. 2005; Gomez
et al. 2010), whereas arboreal species are more
dependent on the forest interior and show edge
avoidance behaviour (Pardini 2004; P€
uttker et al.
2008). Apart from the edge itself, the habitat
C . A . D A R O S A ET AL.
structure (e.g. size, isolation and vegetation structure) as well as the type and quality of the landscape
matrix are also important factors for the distribution
of small mammals (Passamani & Ribeiro 2009; Brady
et al. 2011; Delciellos et al. 2015).
Due to increased deforestation over the past halfcentury, many tropical regions are now dominated by
a heterogeneous landscape composed of restricted
natural habitats, crops, pastures, urban areas and
roads (Hansen et al. 2013; Haddad et al. 2015). Proliferation of roads over recent decades is also contributing to lethal and sub-lethal edge effects
(Laurance et al. 2009; Rosa & Bager 2012; NavarroCastilla et al. 2014), especially within certain habitat
types (Fuentes-Montemayor et al. 2009). In contrast
to a crop matrix, which provides habitat for some
small mammals (Passamani & Ribeiro 2009), road
surfaces commonly lead individuals to avoid the
road-forest edge due to paving and changes in the
microclimate and vegetation (Jaeger et al. 2005;
McGregor et al. 2008).
The presence of roads acts as an intensifying agent
in the fragmentation process (Laurance et al. 2009).
With ongoing forest fragmentation and projected
economic growth in many countries in the tropics
(Bager et al. 2015), understanding the road edge
effect and its interaction with the matrix edge effect
is fundamental to assessing mitigation measures for
biodiversity living near roads. In this sense we are
able to plan to reduce the human-biodiversity conflict
(e.g. prioritize expansion of the road network in areas
that are already cleared and which have unexplored
economic potential) (Laurance et al. 2014); and
implement mitigation strategies (e.g. improve the
environment of wildlife passages with habitat surrogates such as rocks and wood, or shelves and ramps
that separate species) that consider the behavioural
response of the target species to different disturbances (Ascens~ao et al. 2015, 2016; D’Amico et al.
Rapid population replacement and greater abundance of small mammal species (Gentile et al. 2000)
facilitate studies evaluating how species respond to
habitat and landscape features (Pardini 2004; Brady
et al. 2011). In addition, responses of species to edge
effects can be identified and related to species ecological requirements. In this context, we evaluated
the number of individuals of four taxa of tropical
small mammals living near roads. We used two
ground-dwelling, insectivorous–omnivorous species,
Akodon sp. and Cerradomys subflavus (Wagner 1842)
(Paglia et al. 2012); and two arboreal insectivorous–
frugivorous species, Marmosops incanus (Lund 1841)
and Riphidomys sp. (Paglia et al. 2012). We chose
these species because they are widespread and common in fragmented areas of Brazil (Bonvicino et al.
We aimed to evaluate how the abundance of these
ground-dwelling and arboreal small mammals are
affected by distance to the nearest road and to the nearest matrix edge, by vegetation density at the sampling
point, and by the matrix type surrounding the sampled
forest fragment. We hypothesized that the grounddwelling species, being more tolerant of open environments (Bergallo et al. 2005), would be positively
affected by road edges (would show edge attraction
behaviour), whereas the arboreal species, being more
specialized to forest environments (Pardini 2004;
uttker et al. 2008), would be negatively affected by
road edges (i.e. would show edge avoidance behaviour).
Study area and selection of sampling areas
We conducted our study in southeastern Brazil in a transition zone between two conservation hotspots, the Atlantic
Forest and the Brazilian Savannah. The study area ranges
from 920 to 1180 m in altitude, with an average annual
temperature of 19.4°C and 1500 mm of rainfall annually,
concentrated in a rainy season in spring and summer (Brasil 1992). The regional landscape is characterized by a
mosaic of farmland, grassland, savanna vegetation, gallery
forests and semi-deciduous forest patches.
We sampled forest fragments within two fragmented
landscapes intersected by the roads BR-383 and BR-354.
Both roads have a width of 12 m, with no canopy connectivity and road edges undergo periodic maintenance to cut
herbaceous vegetation in the first 5 to 10 m. We sampled
BR-383 between the cities of S~ao Jo~ao Del Rei (21°130 44″
S, 44°220 32″W) and S~ao Vicente de Minas (21°410 50″S,
44°260 20″W), and BR-354 between the cities of Luminarias
(21°300 22″S, 44°540 57″W) and Bom Sucesso (21°020 03″S,
44°460 33″W) (Fig. 1). In each road section, we selected five
pairs of forest fragments created by the road (Fig. 2), 5–
25 km from each other. Each pair of fragments was separated by one of the roads, which resulted in 10 fragments
in each road section and a total of 20 fragments of different
sizes (26–4423 ha).
Small mammal sampling
We carried out five field campaigns at quarterly over
15 months during 2010 and 2011. We sampled small
mammal species of the orders Rodentia and Didelphimorphia by trapping, and used two types and sizes of traps:
Sherman traps had dimensions of 27 9 12 9 12 cm and
can capture the smallest mammals (<40 g), and Tomahawk
traps had dimensions of 45 9 17.5 9 15 cm and can capture small mammals from 100 g to 2 kg in body mass (e.g.
Didelphis sp.). At the end of the trapping study, we selected
the most common species – Akodon sp., C. subflavus,
Riphidomys sp. and M. incanus – for further study.
In each fragment, we established a transect perpendicular
to the road. In each transect, we established 16 capture
points that were 20 m apart, starting at the edge of the
© 2017 Ecological Society of Australia
Fig. 1. Sampled areas (black dots) on highway BR-354, between Bom Sucesso and Luminarias and highway BR-383,
between S~ao Jo~ao Del Rei and S~ao Vicente de Minas.
fragment (10 m from the road). At each point, we
installed two traps, one on the ground (Shermann and
Tomahawk traps interspersed) and another in a tree (Shermann traps), between 1 and 2 m above the ground. We
baited the traps with a mixture of banana, sardines, ground
peanuts and cornmeal. We checked the traps each morning
during the following days, totalling 5 days from trap installation to removal in each fragment, resulting in an overall
sampling effort of 12 800 trap-nights. We ear-tagged
(National Band & Tag Co.) and released all individuals
captured at the point of capture.
Sampling of environmental variables
We evaluated environmental parameters at the plot (local
habitat) and patch scales (fragment surroundings) that could
influence small mammals (Rytwinski & Fahrig 2007). At the
plot scale, we considered each capture point as a sampling
unit, whose geographical position we collected using a
© 2017 Ecological Society of Australia
handheld GPS with a maximum error of 5 m. From a mosaic
of georeferenced images with a spatial resolution of 5 m from
the RapidEye satellite, we measured the distance from the
road edge and the distance from the matrix edge in ArcGIS
9.3 (ESRI, Redlands, CA). We measured the distance from
the road edge as the closest distance between the road and
the capture point. We considered the distance from the
matrix edge to be the distance between the capture point and
the nearest non-road edge of the fragment that contacted a
matrix of any type (we used this same matrix for the classification of the patch scale variable). We measured vegetation
density (basal area of trees, m2 ha1) in circular plots 3 m in
radius at each capture point. We measured all living trees
within plots with a DBH (diameter at breast height, 1.30 m)
greater than or equal to 5 cm. We calculated the vegetation
density of each plot as the sum of the basal area (BA = p * r2)
of trees divided by the plot area (0.0028 ha) (Table 1).
We represented environmental parameters at the patch
scale using the matrix types identified from the shortest distance between the sampling point and the nearest edge with
C . A . D A R O S A ET AL.
Fig. 2. Sampled fragments of highways BR-354, between Bom Sucesso and Luminarias and BR-383, between S~ao Jo~ao Del
Rei and S~ao Vicente de Minas.
the matrix. We generated a land use map from the same
mosaic of images used for the plot scale. We classified the type
of matrix present at the matrix edge of each forest fragment
into four types: permanent coffee plantation, crop rotation,
pasture and eucalyptus (Table 1). In cases where there were
different types of matrix at the edge closest to each sampling
point, we classified the matrix by the combination of matrix
types (e.g. Crop rotation/Pasture; Coffee/Pasture; Eucalyptus/
Pasture). As such, all sampling points within a single fragment
(patch) always had the same matrix classification.
Data analysis
We performed analyses separately for each species, considering each capture point as a sample unit and the number
© 2017 Ecological Society of Australia
Table 1. Plot and patch variables used in generalized linear mixed models for predicting the number of individuals of
ground-dwelling and arboreal small mammal living near road edges
Variable class
Range of values
Distance from sampling point to the nearest road
Distance from sampling point to the nearest fragment edge
Vegetation density at sampling point (m2 ha1)
Matrix type surrounding the sampled forest fragment
Crop rotation/Pasture
of small mammals captured at each point as a Poisson-distributed response variable. To avoid pseudo-replication, we
defined the number of individuals as including only the first
capture of an individual at each sample point (excluding
recaptures at that point) in each separate field campaign.
We used generalized linear mixed models (GLMM) for
Poisson-distributed data (Zuur et al. 2009) to evaluate the
environmental variables that influence the number of individuals of each species.
We defined candidate models using plot and patch variables as fixed factors (Table 1). We used site and season as
random effects to control for pseudo-replication in space
and time respectively (Millar & Anderson 2004). Controlling for these random effects was important due to our lack
of knowledge of the history of the sampled areas and
because of a change in seasonal capture success of small
mammals between dry and rainy seasons in our region
(Gentile et al. 2000). We evaluated the multicollinearity of
continuous predictor variables using Spearman’s correlation. Among pairs with a correlation coefficient above 0.5,
we included only the variable that showed highest correlation coefficient with the response variable in the GLMM
analysis. We defined 13 candidate models, each representing a hypothesis that could explain the number of individuals of each species.
Considering plot-scale variables (vegetation density, road
distance and matrix edge distance), we tested the following
models as predictors of species abundance: (i) road distance; (ii) matrix edge distance; (iii) vegetation density; (iv)
road distance + matrix edge distance; (v) vegetation density + road distance; (vi) vegetation density + matrix edge
distance; (vii) vegetation density + road distance + matrix
edge distance; (viii) interaction between matrix edge distance and road distance; (ix) interaction between matrix
edge distance and vegetation density; (x) interaction
between road distance and vegetation density and (xi) null
model. Considering models that include the type of matrix
surrounding forest fragments as a variable at the patch
scale, we tested the following hypothetical models: (xii)
matrix type and (xiii) plot variables that integrate the best
predictor model + matrix type. We included the interactions in the candidate models due to the narrow shape of
our patches that result in a higher density of edge habitat
per unit area, exposing the species to edge effects from
multiple edges in close proximity to one another. This
results in an additive increase in the overall magnitude of
edge-related changes (Fletcher 2005), that in our work was
represented by vegetation density.
© 2017 Ecological Society of Australia
We used the Akaike information criterion (AIC) to rank
the candidate models. We considered models with
DAIC ≤ 2 as having the best descriptive capacity (Burnham
& Anderson 2002) for the number of individuals of each
species. We computed the Akaike weights (Wi) of all models, and for each variable we summed the weights of models
where the variable was included (Burnham & Anderson
2002). When there was more than one model with
DAIC ≤ 2, we used model averaging to calculate parameter
estimates and standard errors for the best models. We used
quantile–quantile plots to evaluate model accuracy. Modelling was performed using the function ‘glmer’ in the ‘lme4’
suite of the R environment (R Development Core Team
2009; Bates 2010).
We recorded 330 individuals and 153 useful recapture events (made in different field campaigns). We
registered Akodon sp. 163 times (female = 63,
male = 86), C. subflavus 80 times (female = 47,
male = 29), M. incanus 131 times (female = 34,
male = 64) and Riphidomys sp. 109 times (female = 62, male = 45), including all captures and
recaptures. From all captures, we recorded only one
individual on opposite sides of a road, the grounddwelling species C. subflavus.
Both plot and patch variables influence the number
of individuals of the ground-dwelling species Akodon
sp. and C. subflavus, with road distance, matrix type
and vegetation density as predictors in the same
model for each species, the only one ranked with
DAIC ≤ 2 (Table 2), and all three predictors had
similar relative importance (Table 4). For Akodon
sp., distance from the road and vegetation density
were significantly and negatively related to the number of individuals (Table 3; Fig. 3). The matrix type
was not statistically significant (Table 3). For C. subflavus, the distance from the road was significantly
and negatively related to the number of individuals.
Vegetation density had a negative relationship with
the number of individuals of C. subflavus, though
without statistical significance (Table 3; Fig. 3). The
matrix type also was not significant (Table 3).
C . A . D A R O S A ET AL.
Table 2. Ranking of the best generalized linear mixed models for predicting the number of individuals of ground-dwelling
and arboreal small mammals as functions of plot and patch variables: K (number of parameters estimated + intercept), AIC
(Akaike information criterion), DAIC (AICi minAIC), Wi (Akaike weight)
AIC parameters
Type of candidate models
Candidate models
Ground-dwelling species Akodon sp.
Plot + Patch
DistRoad + vegetD + Matrix
Ground-dwelling species Cerradomys subflavus
Plot + Patch
DistRoad + vegetD + Matrix
Arboreal species Marmosops incanus
distRoad + distEdge
vegetD + distRoad + distEdge
Plot + Patch
DistRoad + distEdge + Matrix
Arboreal species Riphidomys sp.
distRoad + distEdge
distRoad + vegetD
The number of individuals of the arboreal species
M. incanus was explained mainly by plot variables,
with distance from the road and distance from the
matrix edge composing the model with lowest
DAIC value (Table 2). The distance from the edge
appeared in all models ranked (DAIC ≤ 2), and
had the highest relative importance among predictive variables (Table 4). The distance from the road
appeared in three of the four best models
(Table 2), and had the second highest importance
among predictor variables (Table 4). Model-averaging showed that distance from the matrix edge
and distance from the road were significantly positively related to the number of M. incanus individuals, while the coffee plantation/pasture matrix
mixture was significantly negatively related to the
number of individuals of this species (Table 5;
Fig. 3). For Riphidomys sp., only plot variables
explained the number of individuals; all models
had distance from the road or distance from the
matrix edge as a predictor, except for the null
model (Table 2). The distance from the road
appeared in five of the seven best models, including
the model with lowest DAIC (Table 2), and had
the highest relative importance value (Table 4).
The distance from the edge appeared in three of
the seven best models (Table 2), and had the second highest importance value (Table 4). Modelaveraging showed that both the distance from the
road and the interaction between the distance from
the road and the distance from the matrix edge
were significant and positively related to the number of Riphidomys sp. individuals (Table 5, Fig. 3).
Table 3. Parameter estimates of the variables included in
the best model of each ground-dwelling species explaining
the number of individuals living near road edges
(DAIC ≤ 2)
Z-value P-value
Akodon sp.
0.005 0.001 3.949 <0.001
0.008 0.003 2.232 0.026
Crop rotation/Pasture
0.872 0.649
1.343 0.179
0.889 0.549
1.620 0.105
0.780 0.656
1.190 0.234
0.114 0.635 0.180 0.857
0.714 0.549 1.299 0.194
Cerradomys subflavus
0.008 0.002 4.706 <0.001
0.008 0.005 1.454 0.146
Crop rotation/Pasture 13.102 0.694 0.019 0.985
1.107 1.175
0.942 0.346
0.208 1.599 0.130 0.896
0.076 1.380
0.055 0.956
1.649 1.141
1.455 0.148
b = estimated coefficients, SE = standard error,
value = Z test, P-value = significance in the Z test.
Use of forest edge habitats by small mammals was
influenced mainly by local (plot) characteristics. As
hypothesized, the species that we evaluated showed
different behaviours with respect to edges: the
ground-dwelling species tended to be attracted to
edges, whereas the arboreal species tended to avoid
edges. In addition, our results indicate that both
© 2017 Ecological Society of Australia
Fig. 3. Differences among environmental conditions that influence the number of individuals of ground-dwelling (AKO,
Akodon sp., CER, Cerradomys subflavus) and arboreal (MAR, Marmosops incanus; RIP, Riphidomys sp.) small mammal species.
The horizontal black line represents theoretically optimal conditions for each species according to final models (Tables 3 and
5). If the horizontal black line occurs at either end of the bar, the result is statistically significant (positively if at the top of the
bar, negatively if at the bottom of the bar) (P ≤ 0.05). If the horizontal black line occurs in the middle of the bar the variable
is absent from the best models.
Table 4. Relative importance of the predictor variables in
models for ground-dwelling and arboreal small mammals
distEdge distRoad vegetD Matrix
Akodon sp.
Cerradomys subflavus
Marmosops incanus
Riphidomys sp.
For each variable, the Akaike weights (Wi) of the models
in which the variable was present were summed (higher values represent higher importance).
types of edges (forest-road and forest-matrix) affected
the number of individuals of these species.
Our models show that the ground-dwelling species
prefer habitats with lower vegetation density, corroborating the pattern observed for ground-dwelling
© 2017 Ecological Society of Australia
species of the Atlantic Forest (Delciellos et al. 2015).
In tropical forests, reduced fragment size and an
increased proportion of edge habitat cause shifts in
the physical environment that lead to the loss of
large old trees in favour of pioneer trees (Haddad
et al. 2015). In addition, forbs and grasses quickly
colonize disturbed areas due to the lack of competition from mature trees in conjunction with better
light conditions (Peterken 2008). These new habitat
structures provide shelter for habitat-generalists (Fischer et al. 2011) such as Cerradomys sp. and Akodon
sp. (Dalmagro & Vieira 2005; Paglia et al. 2012).
Thus, they may be better adapted for colonizing disturbed areas such as road edges, consistent with our
models which showed a negative relationship
between distance from the road and the number of
individuals of these species.
C . A . D A R O S A ET AL.
Table 5. Weighted parameters of variables from modelaveraging of the best models explaining the number of individuals of arboreal small mammals living near road edges
(DAIC ≤ 2)
Marmosops incanus†
Crop rotation/Pasture
Riphidomys sp.†
<0.001 <0.001
<0.001 <0.001
0.002 0.004
<0.001 <0.001
0.001 0.001
Z-value P-value
b = estimated
SE = standard
Z-value = Z test, P-value = significance in the Z test. †Averaged model when we obtained more than one most supported model.
The arboreal species demonstrated preferences for
a contrasting set of habitats, farther from roads and
matrix edges. In fragmented areas, habitat quality has
been shown to be even more important than previously considered for the conservation of small mammals, and even for large mammals (Brodie et al.
2015; Delciellos et al. 2015). In the tropics, arboreal
species tend to be more abundant in fragments with
greater overstory and understory vegetation density
(Michal & Rafal 2014; Delciellos et al. 2015). Even
in temperate zones, arboreal species of small mammals take refuge in core forested areas where mature
trees exist (Michal & Rafal 2014). Arboreal species
can be especially affected by fragmentation and road
edge effects if they generally avoid descending to
ground level (Goosem et al. 2008; van der Ree et al.
2010). The connectivity of suitable forest patches has
also been recognized as an important landscape feature for arboreal mammals (Reunanen et al. 2002;
Pardini et al. 2005; Umetsu & Pardini 2007). Efforts
to improve habitat quality and maintain vegetation
structure, or even the construction of overpasses
across roads to connect forested areas, can assist
arboreal species conservation (Weston et al. 2011;
Soanes & van der Ree 2015).
Roadsides vegetation can act as a refuge for small
mammals, particularly in fragmented areas where
roadsides may provide the majority of remaining
habitat (Ruiz-Capillas et al. 2013). This can be true
at the community or a species level (e.g. McGregor
et al. 2008; Fahrig & Rytwinski 2009). However,
even in temperate zones the responses of specialized
arboreal species to the different effects of fragmentation are more varied than the responses of generalist
ground-dwelling species (Michal & Rafal 2014).
Small mammals in temperate zones show differing
responses to the road barrier effect: some experience
deleterious effects to the genetic structure of their
populations (Ascens~ao et al. 2016), whereas effects
are more neutral for species able to move through
the landscape and maintain interconnected population (Grilo et al. 2016). These different patterns
exemplify species-specific relationships to roads and
the landscape in general, and show the importance of
considering small mammal species individually
according to their habits and behaviours.
The Atlantic forest is now dominated by fragments
smaller than 100 ha (Ribeiro et al. 2009). This is an
extreme example of the fragmentation process in the
tropics and has been used as an example of fragmentation in forested landscapes worldwide (e.g. Haddad
et al. 2015). The species we studied are widespread and
have been exposed to the long-term changes historically
imposed on the Atlantic Forest and its ecotonal areas
(Bonvicino et al. 2008; Haddad et al. 2015). Consequently, we believe the patterns shown here can be
extrapolated to other fragmented landscapes. However,
the next step is to evaluate the response of specialist species of arboreal and ground-dwelling species to assess
whether they show similar patterns of those found in
this study, or if there are other species traits (e.g. diet)
that trigger distinctive edge effects. It is also important
to consider that even though generalist species may not
show obvious behavioural changes at road edges, there
may be stress reactions resulting from trade-offs that
species make in a degraded ecosystem (Navarro-Castilla
et al. 2014). Further research will involve coordinating
a network of experiments on species with different ecological requirements across ecosystems and landscape
We thank S.R. Freitas and C.B. Grilo for their suggestions on the first version of this paper. We are
grateful for the financial support provided by FAPEMIG (Process CRA – PPM-00139-14; 453 and CRA
– APQ-03868-10), CNPq (Process 303509/2012 0),
Fundac~ao Grupo Boticario (Process 0945-20122),
and the Tropical Forest Conservation Act – TFCA
(through Fundo Brasileiro para Biodiversidade –
Ascens~ao F., LaPoint S. & van der Ree R. (2015) Roads,
traffic and verges: big problems and big opportunities for
© 2017 Ecological Society of Australia
small mammals. In: Handbook of Road Ecology (eds R. van
der Ree, D. J. Smith & C. Grilo) pp. 325–33. John Wiley
& Sons, Chichester.
Ascens~ao F., Mata C., Malo J. E. et al. (2016) Disentangle the
causes of the road barrier effect in small mammals through
genetic patterns. PLoS One 11, e0151500.
Bager A., Borghi C. E. & Secco H. (2015) The influence of
economics, politics, and environment on road ecology in
South America. In: Handbook of Road Ecology (eds R. van
der Ree, D. J. Smith & C. Grilo) pp. 407–13. John Wiley
& Sons, Chichester.
Bates D. M. (2010) lme4: Mixed-Effects Modeling with R.
Springer, Madison.
Bergallo H. G., Luz J. L., Raıces D. S., Hatano F. H. &
Martins-Hatano F. (2005) Habitat use by Oryzomys
subflavus (Rodentia) in an open shrubland formation in
restinga de Jurubatiba National Park, RJ, Brazil. Braz. J.
Biol. 65, 583–8.
Bonvicino C. R., Oliveira J. A. & D’Andrea P. S. (2008) Guia
dos roedores do Brasil. Centro Pan-Maericano de Febre
Aftosa, Rio de Janeiro, Brazil.
Brady M. J., Mcalpine C. A., Possingham H. P., Miller C. J. &
Baxter G. S. (2011) Matrix is important for mammals in
landscapes with small amounts of native forest habitat.
Landscape Ecol. 26, 617–28.
Brasil (1992) Normais climatol
ogicas (1961–1990). Ministerio
da Agricultura. Departamento Nacional de Metereologia,
Brodie J. F., Giordano A. J. & Ambu L. (2015) Differential
responses of large mammals to logging and edge effects.
Mamm. Biol. 80, 7–13.
Burnham K. P. & Anderson D. R. (2002) Model Selection and
Inference: A Practical Information-Theoretic. Springer-Verlag,
New York.
Dalmagro A. D. & Vieira E. M. (2005) Patterns of habitat
utilization of small rodents in an area of Araucaria forest in
Southern Brazil. Austral Ecol. 30, 353–62.
D’Amico M., Clevenger A. P., Roman J. & Revilla E. (2015)
General versus specific surveys: estimating the suitability of
different road-crossing structures for small mammals. J.
Wildl. Manag. 79, 854–60.
Delciellos A. C., Vieira M. V., Grelle C. E., Cobra P. &
Cerqueira R. (2015) Habitat quality versus spatial variables
as determinants of small mammal assemblages in Atlantic
Forest fragments. J. Mammal. 97, 253–65.
Fahrig L. (2003) Effects of habitat fragmentation on
biodiversity. Ann. Rev. Ecol. Syst. 34, 487–515.
Fahrig L. & Rytwinski T. (2009) Effects of roads on animal
abundance: an empirical review and synthesis. Ecol. Soc.
14, 1–19.
Fischer C., Thies C. & Tscharntke T. (2011) Small mammals
in agricultural landscapes: opposing responses to farming
practices and landscape complexity. Biol. Conserv. 144,
Fletcher R. J. J. (2005) Multiple edge effects and their
implications in fragmented landscapes. J. Anim. Ecol. 74,
Fuentes-Montemayor E., Cuar
on A. D., Vazquez-Domınguez
E., Benıtez-Malvido J., Valenzuela-Galvan D. & Andresen
E. (2009) Living on the edge: roads and edge effects on
small mammal populations. J. Anim. Ecol. 78, 857–65.
Gentile R., D’andrea P. S., Cerqueira R. & Maroja L. S.
(2000) Population dynamics and reproduction of
marsupials and rodents in a Brazilian rural area: a five-year
study. Stud. Neotrop. Fauna Environ. 35, 1–9.
© 2017 Ecological Society of Australia
Gomez D., Sommaro L., Steinmann A., Chiappero M. &
Priotto J. (2010) Movement distances of two species of
sympatric rodents in linear habitats of Central Argentine
agro-ecosystems. Mamm. Biol. 76, 58–63.
Goosem M. (2000) Effects of tropical rainforest roads on small
mammals: edge changes in community composition. Wildl.
Res. 27, 151–63.
Goosem M. (2007) Fragmentation impacts caused by roads
through rainforests. Curr. Sci. 93, 1587–95.
Goosem M., Wilson R., Weston N. & Cohen M. (2008) Highway
overpass evaluation of effectiveness: Kuranda Range Road
Upgrade Project. Cairns: Commonwealth Environment
Research Facilities-Marine and Tropical Sciences Research
Facility. School of Earth and Environmental Sciences, James
Cook University, Queensland.
Grilo C., Del Cerro I., Centeno-Cuadros A. et al. (2016)
Heterogeneous road networks have no apparent effect on
the genetic structure of small mammal populations. Sci.
Total Environ. 565, 706–13.
Haddad N. M., Brudvig L. A., Clobert J. et al. (2015) Habitat
fragmentation and its lasting impact on Earth’s
ecosystems. Sci. Adv. 1, e1500052.
Hansen M. C., Potapov P. V., Moore R. et al. (2013) Highresolution global maps of 21st-century forest cover change.
Science 342, 850–3.
Jaeger J. A. G., Bowman J., Brennan J. et al. (2005) Predicting
when animal populations are at risk from roads: an
interactive model of road avoidance behavior. Ecol. Model.
185, 329–48.
Laurance W. F., Goosem M. & Laurance S. G. W. (2009)
Impacts of roads and linear clearings on tropical forests.
Trends Ecol. Evol. 24, 659–99.
Laurance W. F., Clements G. R., Sloan S. et al. (2014) A
global strategy for road building. Nature 513, 229–32.
McGregor R. L., Bender D. J. & Fahrig L. (2008) Do small
mammals avoid roads because of the traffic? J. Appl. Ecol.
45, 117–23.
Michal B. & Rafal Z. (2014) Responses of small mammals to
clear-cutting in temperate and boreal forests of Europe: a
meta-analysis and review. Eur. J. Forest Res. 133, 1–11.
Millar R. B. & Anderson M. J. (2004) Remedies for
pseudoreplication. Fish. Res. 70, 397–407.
Navarro-Castilla A., Mata C., Ruiz-Capillas P., Palme R.,
Malo J. E. & Barja I. (2014) Are motorways potential
stressors of roadside Wood mice (Apodemus sylvaticus)
Populations? PLoS One 9, e91942.
Paglia A. P., Fonseca G. A. B., Rylands A. B. et al. (2012)
Annotated Checklist of Brazilian Mammals. Occasional
Papers in Conservation Biology No. 6, Conservation
International, Arlington.
Pardini R. (2004) Effects of forest fragmentation on small
mammals in an Atlantic Forest landscape. Biodiv. Conserv.
13, 2567–86.
Pardini R., de Souza S. M., Braga-Neto R. & Metzger J. P.
(2005) The role of forest structure, fragment size and
corridors in maintaining small mammal abundance and
diversity in an Atlantic forest landscape. Biol. Conserv. 124,
Passamani M. & Ribeiro D. (2009) Small mammals in a
fragment and adjacent matrix in southeastern Brazil. Braz.
J. Biol. 69, 305–9.
Peterken G. F. (2008) Structural dynamics of forest stands and
natural processes. In: The Forests Handbook Volume 1: An
Overview of Forest Science (ed. J. Evans) pp. 83–104.
Blackwell Science Ltd, Oxford, Malden.
C . A . D A R O S A ET AL.
uttker T., Pardini R., Meyer-Lucht Y. & Sommer S. (2008)
Responses of five small mammal species to micro-scale
variations in vegetation structure in secondary Atlantic
Forest remnants, Brazil. BMC Ecol. 8, 1188–9.
R Development Core Team (2009) R: A Language and
Environment for Statistical Computing. R Foundation for
Statistical Computing. [Cited 16 November 2015.]
Available from URL:
Reunanen P., Nikula A. & Monkkonen M. (2002) Regional
landscape patterns and distribution of the Siberian flying
squirrel Pteromys volans in northern Finland. Wildl. Biol. 8,
Ribeiro M. C., Martensen J. P., Ponzoni A. C. & Hirota M.
M. (2009) The Brazilian Atlantic Forest: how much is left,
and how is the remaining forest distributed? Implications
for conservation Biol. Conserv. 142, 1141–53.
Rosa C. A. & Bager A. (2012) Seasonality and habitat types affect
roadkill of neotropical birds. J. Environ. Manage. 97, 1–5.
Rosa C. A. & Bager A. (2013) Review of the factors underlying
the mechanisms and effects of roads on vertebrates.
Oecologia Australis 17, 6–19.
Ruiz-Capillas P., Mata C. & Malo J. E. (2013) Road verges are
refuges for small mammal populations in extensively
managed Mediterranean landscapes. Biol. Conserv. 158,
Rytwinski T. & Fahrig L. (2007) Effect of road density on
abundance of white-footed mice. Landscape Ecol. 22,
Soanes K. & van der Ree R. (2015) Reducing Road impacts on
tree-dwelling animals. In: Handbook of Road Ecology (eds
R. van der Ree, D. J. Smith & C. Grilo) pp. 334–40.
Wiley-Blackwell, London.
Umetsu F. & Pardini R. (2007) Small mammals in a mosaic of
forest remnants and anthropogenic habitats—evaluating
matrix quality in an Atlantic forest landscape. Landscape
Ecol. 22, 517–30.
van der Ree R., Cesarini S. P., Moore S. J. L. & Taylor A.
(2010) Large gaps in canopy reduce road crossing by a
gliding mammal. Ecol. Soc. 15, 35.
Weston N., Goosem M., Marsh H., Cohen M. & Wilson R.
(2011) Using canopy bridges to link habitat for arboreal
mammals: successful trials in the Wet Tropics of
Queensland. Aust. Mammal. 33, 93–105.
Zuur A. F., Ieno E. N., Walker N. J., Saveliev A. & Smith G.
M. (2009) Mixed Effects Models and Extensions in Ecology
with R. Springer Science & Business Media, New York.
© 2017 Ecological Society of Australia
Без категории
Размер файла
496 Кб
12549, aec
Пожаловаться на содержимое документа