Quantifying the effects of deforestation and fragmentation on a range-wide conservation plan for jaguars

Abstract

The impact of extensive changes in land use and climate on species has led to an increasing focus on large-scale conservation planning. However, these plans are often static conservation prescriptions set against a backdrop of rapidly changing environments, which suggests that large-scale information on threats can improve the func-tionality of planning efforts. Jaguars (Panthera onca) are the focus of a range-wide conservation strategy extending from Mexico to Argentina that consists of jaguar conservation units (JCUs) and modeled corridors. Recent deforestation is a major threat to jaguar populations, but forest loss has not been systematically assessed across the entire jaguar network. In this study, we quantified the amount and rate of deforestation in JCUs and corridors between 2000 and 2012. JCUs lost 37,780 km 2 forest (0.93%) at an increasing rate of 149.2 km 2 yr −2 , corridors lost 45,979 km 2 (4.43%) at a decreasing rate of 40.1 km 2 yr −2 , and levels of forest fragmentation increased in corridors. Protected sections of JCUs and corridors lost less forest than unprotected sections, suggesting efforts to increase protected status of jaguar conservation areas are warranted. Higher deforestation in corridors indicates difficulties in maintaining connectivity of jaguar populations, and suggests the need for increased engagement with communities within corridor landscapes. Assessment of spatial variability of anthropogenic threats within the jaguar network may improve jaguar conservation by informing network prioritization and function.

Full text

Quantifying the effects of deforestation and fragmentation on a
range-wide conservation plan for jaguars
Peter J. Olsoy
a,
⁎, Kathy A. Zeller
b
, Jeffrey A. Hicke
c
, Howard B. Quigley
d
,
Alan R. Rabinowitz
d
, Daniel H. Thornton
a,d
a
School of the Environment, Washington State University, 100 Dairy Road/1228 Webster, Pullman, WA 99164-2812, USA
b
Department of Environmental Conservation, University of Massachusetts Amherst, 160 Holdsworth Way, Amherst, MA 01003-9285, USA
c
Department of Geography, University of Idaho, Moscow, ID 83844-3021, USA
d
Panthera USA, 8 West 40th Street, 18th Floor, NY 10018, USA
abstractarticle info
Article history:
Received 21 April 2016
Received in revised form 25 August 2016
Accepted 29 August 2016
Available online xxxx
The impact of extensive changes in land use and climate on specieshas led to an increasing focus on large-scale
conservation planning. However, these plans are often static conservation prescriptions set against a backdrop of
rapidly changing environments, which suggests that large-scale information on threats can improve the func-
tionality of planning efforts. Jaguars (Panthera onca) are the focus ofa range-wide conservation strategy extend-
ing from Mexico to Argentina that consists of jaguar conservation units (JCUs) and modeled corridors. Recent
deforestation is a major threat to jaguar populations, but forest loss has not been systematically assessed
across the entire jaguar network. In this study, we quantified the amount and rate of deforestation in JCUs and
corridors between 2000 and 2012. JCUs lost 37,780 km
2
forest (0.93%) at an increasing rate of 149.2 km
2
yr
−2
,
corridors lost 45,979 km
2
(4.43%) at a decreasing rate of 40.1 km
2
yr
−2
, and levels of forest fragmentation in-
creasedin corridors. Protectedsections of JCUs and corridors lost lessforest than unprotected sections, suggesting
efforts to increase protected status of jaguar conservation areas are warranted. Higher deforestation in corridors
indicates difficulties in maintaining connectivity of jaguar populations, and suggests the need for increased en-
gagement with communities within corridor landscapes. Assessment of spatial variability of anthropogenic
threats within the jaguar network may improve jaguar conservation by informing network prioritization and
function.
© 2016 Elsevier Ltd. All rights reserved.
Keywords:
Conservation planning
Connectivity
Deforestation
Fragmentation
Panthera onca
Protected areas
1. Introduction
The pervasive impacts of anthropogenic change on species and eco-
systems has led to increasing emphasis on developing conservation
strategies across broad spatial scales. These conservation plans cross
landscapes, regions, and political boundaries, and are often designed
to conserve species in the face of global change processes such as
large-scale habitat loss, fragmentation, and climate change (Lawler et
al., 2010). Such large-scale plans recognize the need to go beyond man-
agement of one or a few key areas to consider the wider landscapes and
regions that provide connections and conserve ecological processes
such as disturbance and connectivity (Guerrero et al., 2015). The devel-
opment of efforts such as the Yellowstone to Yukon Conservation Initia-
tive (Chester, 2015), growth of organizations that encourage cross-
landscape coordination (e.g., Landscape Conservation Cooperatives in
the US), and increasing emphasis on landscape or region-wide system-
atic conservation efforts (e.g., Klein et al., 2009; Pressey and Bottrill,
2009) attest to the new emphasison ambitious large-scale conservation
action.
Once a conservation strategy has been designed, periodic updates
are required to assesssuccess and failure and adapt theplan to prioritize
conservation action (Pressey, 2004). One key factor in assessing large-
scale conservation efforts is an understanding of how ongoing human
activities impact the components of a plan (e.g., protected areas),
which may be highly variable in intensity and spatial distribution. How-
ever, incorporating information on landscape change, or other forms of
anthropogenic impacts, into conservation strategy remains difficult
at large scales and is often ignored (Heller and Zavaleta, 2009; Pressey
et al., 2007). Knowledge of these impacts informs scheduling and
Biological Conservation 203 (2016) 8–16
Abbreviations: JCU, jaguar conservation unit; PD, patch density; CLUMPY, clumpiness
index.
⁎Corresponding author.
E-mail addresses: peterolsoy@gmail.com (P.J. Olsoy), kzeller@gmail.com (K.A. Zeller),
jhicke@uidaho.edu (J.A. Hicke), hquigley@panthera.org (H.B. Quigley),
arabinowitz@panthera.org (A.R. Rabinowitz), daniel.thornton@wsu.edu (D.H. Thornton).
http://dx.doi.org/10.1016/j.biocon.2016.08.037
0006-3207/© 2016 Elsevier Ltd. All rights reserved.
Contents lists available at ScienceDirect
Biological Conservation
journal homepage: www.elsevier.com/locate/bioc

prioritizing of conservation actions (Pressey, 2004; Pressey et al., 2007;
Visconti et al., 2010). It is therefore important to analyze the effects of
anthropogenic pressure on conservation plans to assess their perfor-
mance and inform priority setting decisions.
Range-wide conservation strategies for single species are one form
of large-scale conservation planning (Redford et al., 2011; Sanderson
et al., 2008, 2002). Typically, these plans are constructed by identifying
priority habitats or sites for conservation across the range of the species,
based on some combination of expert opinion, modeling, and ecological
literature (e.g., Rabinowitz and Zeller, 2010; Sanderson et al., 2002;
Thorbjarnarson et al., 2006; Viña et al., 2010), and often span political
and ecological boundaries. When the species is wide-ranging, conserva-
tion efforts designed around the needs of that species can serve as an
umbrella for many other species (Roberge and Angelstam, 2004;
Thornton et al., 2016).
Jaguars (Panthera onca) are an ideal candidate for range-wide con-
servation planning, given declining populations across a large range
that encompasses much of Latin America (Sanderson et al., 2002).
Major threats to jaguars include habitat loss and fragmentation due to
conversion of forest to agriculture (De Angelo et al., 2011; Sanderson
et al., 2002), declining prey base (Espinosa, 2012; MacDonald and
Loveridge, 2010), and persecution by humans (Azevedo and Murray,
2007; Crawshaw and Quigley, 1991; De Angelo et al., 2013). Identifica-
tion of major threats to jaguars led to the development of a large-scale
conservation plan consistingof core habitat, termed jaguar conservation
units (JCUs) (Sanderson et al., 2002), and corridors connecting those
JCUs (Rabinowitz and Zeller, 2010). This conservation strategy has
been a major driver of conservation action and research across the
jaguar's range (e.g., Petracca et al., 2014; Rabinowitz, 2014; Thornton
et al., 2016; Zeller et al., 2011). Approximately 34% of JCU's total area
is protected (IUCN categories I-VI), while only 11% of the corridor's
total area is protected. The goal of large-scale species conservation
plans are genetically robust, healthy, representative and resilient popu-
lations (Redford et al., 2011). Accomplishing that goal requires assess-
ment of the plan's performance and corresponding actions to address
weak points.
The importance of habitat to jaguars and the threat of habitat loss
suggest that documenting the extent of habitat change within the spe-
cies' range can improve conservation efforts by prioritizing targeted
conservation actions to maintain populations. Satellite remote sensing
is a useful tool for documenting land cover change through deforesta-
tion activities (e.g., DeFries et al., 2005; Willis, 2015). The recent devel-
opment of a global database of forest cover change using Landsat
imagery (30-m spatial resolution) (Hansen et al., 2013) provides the
opportunity to evaluate changes in forest cover across large areas and
at fine spatial resolution. Although jaguars may use habitat other than
forest for some activities (Tôrres et al., 2012; Vynne et al., 2011), jaguars
generally select more heavily forested landscapes andjaguar population
persistence is tied to forest cover (Azevedo and Murray, 2007;
Cavalcanti and Gese, 2009; De Angelo et al., 2013, 2011).
Our goal was to examine recent deforestation and fragmentation
within jaguar core areas (i.e., JCUs) and corridors using this dataset on
forest loss (Hansen et al., 2013). Specifically, we sought to: (1) quantify
rates of deforestation and fragmentation in JCUs and modeled corridors
between 2000 and 2012; and (2) identify JCUs, corridors, and broader
geographic regions with the most deforestation. Given that protected
areas are often the backbone of conservation planning and form a key
component of the jaguar conservation network, yet are threatened by
a variety of processes that may limit their effectiveness (Curran et al.,
2004; DeFries et al., 2005; Leverington et al., 2010; Mascia and Pailler,
2011), we also sought to (3) determine intercountry variability in defor-
estation within protected and unprotected sections of JCUs and corri-
dors. By applying a consistent dataset of recent deforestation across
the entire range of the jaguar (from Argentina to Mexico), our analysis
is a first step in incorporating anthropogenic threats in range-wide con-
servation planning for this threatened species.
2. Materials and methods
2.1. Materials
We used an established range-wide conservation network for jag-
uars in our analysis (Sanderson et al., 2002). This network was devel-
oped through consultation with jaguar experts that identified core
habitat containing stable jaguar populations, termed jaguar conserva-
tion units (JCUs) (Sanderson et al., 2002). These JCUs were defined as
areas with a stable prey community that contained a population of at
least 50 breeding jaguars, or as areas with fewer than 50 breeding jag-
uars but with sufficient habitat and prey base such that jaguar popula-
tions could increase under favorable conditions (Sanderson et al.,
2002). These JCUs were updated in 2006 (Zeller, 2007), and have been
further modified to result in the most recent map of JCUs (Fig. 1). Al-
though some local-level refinements to the JCUs have been and will
continue to be made (e.g., De Angelo et al., 2013), the JCU network pro-
vides a consistent, expert-validated, range-wide model with which to
conduct a forest loss comparison.
Evaluation of potential jaguar movement, or dispersal, between
JCUs, was missing from the initial range-wide plan. Accordingly,
Rabinowitz and Zeller (2010) estimated dispersal corridors between
JCUs via least-cost path modeling (Adriaensen et al., 2003). Given a
lack of empirical information on how jaguars disperse across land-
scapes, expert opinion was used to derive the resistance values for the
creation of jaguar least-cost corridors. The end result of this analysis
was a map representing landscape resistance to jaguar movements
and identification of potential least-cost corridors linking JCUs (Fig. 1).
We estimated deforestation in JCUs and jaguarcorridors using global
maps of forest change between 2000 and 2012 from Hansen et al.
(2013). This dataset quantifies recent forest change at high resolution
(30 m), which informs habitat loss across large scales. Percent forest
cover in 2000, forest loss, loss-year, and forest gain are available online
(Hansen et al., 2013). Percent forest cover was defined as canopy cover
over 5 m, with loss representing a stand-clearing event (Hansen et al.,
2013). The year that the forested pixel was lost is given by an integer
value between 1 (lost in 2001) and 12 (lost in 2012). Gain was then de-
fined as the inverse of loss. By defining forest cover as a structural mea-
sure, Hansen et al. (2014) acknowledged that their map of forest cover
does not differentiate functional forests from agricultural plantations,
nor true forest regeneration from agricultural conversion (Tropek et
al., 2014). However, in calculating net change in forest cover, we mini-
mize problems related to misclassifications in the initial forest cover
layer by focusing on the loss and gain in forest cover. Misclassified re-
growth or “gain”could still influence our analysis; however, there was
very little forest gain within JCUs and corridors when compared to for-
est loss.
We generated pixel-level binary layers of forest cover in 2000
(fc
2000
) and 2012 (fc
2012
) to analyze forest change and fragmentation.
We used a 50% threshold to convert the Hansen et al. (2013) data to a
binary map of forest/non-forest. Non-forested pixels that had forest
gain were changed to forest for fc
2012
; similarly, forested pixels that
had forest loss were changed to non-forested for the fc
2012
layer. By in-
corporating both forest gain and loss into the fc
2012
dataset, we were
able to look at net forest change between 2000 and 2012. We performed
a sensitivity analysis by testing a cutoff of 20% forest cover, which was
used by Heino et al. (2015) in a global analysis of the same dataset.
Our selection of 50% did not qualitatively change the results, with
most corridors and JCUs only showing an additional forest loss of
0.10% or less (Appendix A). Moreover, a threshold of 50% cover provides
a better indication of functional forest for jaguars, which preferentially
move through and select for more heavily forested habitat across their
range (Azevedo and Murray, 2007; Cavalcanti and Gese, 2009;
Crawshaw and Quigley, 1991; Davis et al., 2011; De Angelo et al.,
2011). We included all corridors and JCUs of the jaguar range in our
analysis, even those that fall within non-forest biomes (savannas and
9P.J. Olsoy et al. / Biological Conservation 203 (2016) 8–16

drylands). We determined the majority biome for each JCU and corridor
with Climate Change Initiative land cover (CCI-LC) data from the Euro-
pean Space Agency (ESA, 2014). Although forest loss better reflects or
represents the impact of anthropogenic disturbance in forested biomes,
we included all biomes in the analysis to have a comparable dataset
across the range. We note, however, that many corridors and JCUs with-
in non-forested biomesstarted with relatively low levels of forest cover
in 2000.
To examine forest loss and fragmentation within protected and un-
protected areas of the jaguar conservation plan, we used a database
from the International Union for Conservation of Nature (IUCN) con-
taining polygons for every protected area in Central and South America
(IUCN and UNEP-WCMC, 2015). Protected area designation s range from
a strict nature reserve (Ia) and wilderness areas (Ib) to protected areas
with sustainable use of natural resources (VI). For our analysis, we com-
bined all designations (I-VI) into a single category of protected, while all
other land was categorized as unprotected.
2.2. Forest change analysis
We calculated forest change between 2000 and 2012 for each JCU
and corridor. We calculated change as areal extent (i.e., forest change
between 2000 and 2012 [km
2
]) and as a percentage of the total land
area. To determine change as percentage of the total area, we subtracted
the percent of the JCU or corridor that was forest in 2012 from the per-
cent of the JCU or corridor area that was forest in 2000. To assess range-
wide trends in deforestation, we determined the slope of the regression
line between loss year and forest loss for JCUs andcorridors. We used an
unequal variance t-test to determine whether corridors lost forest at a
faster rate than JCUs each year. To assess the effectiveness of designated
protected areas for jaguar habitat, JCUs and corridors were pooled at the
country level and separated into protected and unprotected sections.
Values calculated for each JCU or corridor therefore include total area,
fc
2000
, area change between 2000 and 2012 (km
2
), and percent change
between 2000 and 2012. Finally, we calculated range-wide trends in
forest change for JCUs and corridors over the 12-year period.
2.3. Fragmentation analysis
We used FRAGSTATS 4 (McGarigal and Marks, 1995) to calculate
fragmentation metrics for corridors in 2000and 2012. We did not calcu-
late these metrics for JCUs or the 12 largest corridors (of 83 corridors)
due to computer memory limitations in FRAGSTATS 4. Examining frag-
mentation metrics likely wouldn't improve our understanding of the
status of JCUs beyond forest loss metrics as JCUs maintained forest
cover in large central blocks. In contrast, fragmentation metrics are
more meaningful for corridors, which tend to be narrow, start with
less forest cover, lose forest at a higher rate, and are supposed to func-
tion as landscape connections. Although omitting the largest corridors
from the analysis was not ideal, it is unlikely that this would heavily in-
fluence our results. We tested the relationship between corridor
size and fragmentation metrics and found no correlation (Pearson's
rb0.1).
The two metrics we focused on were patch density (PD) and
clumpiness index (CLUMPY). We chose these metrics as indices of
changes in fragmentation and connectedness across the landscape, irre-
spective of corridor size, which varied greatly. PD is the number of
patches divided by total landscape area (units are patches per 100 ha),
with low PD indicating a more connected landscape and high PD indi-
cating a more fragmented landscape. CLUMPY ranges from −1to1,
with −1 representing a maximally disaggregated landscape (greater
dispersion), 0 having randomly distributed patches, and 1 representing
Fig. 1. Percentforest loss (2000−2012) in jaguar conservation units (JCUs) (A, C) and corridors (B, D) in Central America (A, B) and South America (C, D) with warmer colors indicating
more deforestation and cooler colors indicating less deforestation. To distinguish between JCUs and corridors, each are grayed out when focusing on the other.
10 P.J. Olsoy et al. / Biological Conservation 203 (2016) 8–16

maximally clumped (greater contagion). CLUMPY isolates the configu-
ration component from the area component, thereby giving an effective
index of fragmentation that is not confounded by changes in area. We
determined if fragmentation metrics significantly changed between
2000 and 2012 with a paired t-test.
3. Results
3.1. Forest change analysis
Between 2000 and 2012, deforestation was higher in corridors than
JCUs (Fig. 1; Appendix B). JCUs lost 37,780 km
2
of forest and corridors
lost 45,979km
2
, representing 0.93% and4.43% of their total area,respec-
tively. The average JCU lost 522.8 km
2
of forests (σ= 1579.1 km
2
,
range: −13,001–+2.8 km
2
) or 1.81% (σ= 1.99%, range: −
11.37%–+0.03%), while the average corridor lost 574.9 km
2
(σ=
1351.1 km
2
, range: −6428–+ 161 km
2
) or 3.88% (σ= 4.53%, range:
−23.18%–+2.00%). Six out of 75 JCUs (8%) lost N5% forest, and only
one (1%) surpassed 10% loss. Twenty-three out of 83 corridors (28%)
lost N5% of forest cover, and eight of those (10%) surpassed 10% loss.
One JCU (1%) and five corridors (6%) in the Atlantic coastal forests of
Brazil gained forest cover from 2000 to 2012, while all other JCUs and
corridors lost forestover that time frame. See Appendix B for a list of for-
est change in each JCU and corridor.
The highest deforestation in corridors by area was in South America
where the Atlantic Coastal forests of Brazil connect to the Amazon. In
particular, Brazil contained the five corridors that lost the most forest
(N3000 km
2
)(Fig. 1; Appendix B). The heavy deforestation in South
American corridors extended to the southern part of the jaguar's
range. In Central America, JCUs and corridors from the Yucatan
Peninsula in Mexico south through Guatemala, Honduras, and
Nicaragua lost the most forest. The two corridors in Central America
with the highest deforestation by area were in the Yucatan corridor in
Mexico with 1617.9 km
2
(6.77%) and the Bosawas-Cerro Silva corridor
in Nicaragua with 1615.7 km
2
(10.58%).
Forest loss within JCUs and corridors demonstrated substantial
intercountry variability (Fig. 2). For example, the Yucatan Peninsula
showed evenly dispersed deforestation throughout most of the JCUs
and corridors, and significant deforestation at the southern edges of
JCUs in Guatemala (Fig. 2B). Corridors on both sides of the Honduras–
Nicaragua border were heavily deforested, and deforestation in their
shared JCU occurs in both countries (Fig. 2C). In Colombia, recent defor-
estation is expanding from non-forested sections along forest bound-
aries (Fig. 2D). Across the shared borders between Bolivia, Argentina
and Paraguay, large contiguous patches of forest are being clear-cut
(note regularly shaped blocks of forest loss), particularly in Paraguay
where the corridor to Argentina and surrounding areas have been al-
most completely deforested (Fig. 2E). Lastly, the group of JCUs and cor-
ridors near the Atlantic Coastal forests of Brazil were deforested earlier
Fig. 2. (A) Central and SouthAmerica with jaguar conservationunits (JCUs; outlined in gray) and corridors (outlined in black), and red boxes indicating the location of close-upforest loss
figures.Close-up locationsillustrating forestarea (cyan) and loss (2000–2012; red) within JCUs and corridors are: (B) Yucatan Peninsulain Mexico, Guatemala,and Belize; (C) Honduras–
Nicaragua border; (D) Colombia; (E) borders between Bolivia, Argentina, and Paraguay; and (F) central Brazil and part of the Atlantic Coastal forests.
11P.J. Olsoy et al. / Biological Conservation 203 (2016) 8–16

than 2000 (white areas within JCUs and corridors), and that deforesta-
tion subsequently spread westward (Fig. 2F).
In addition to total area lost, we also computed rates of loss. JCUs
generally lost around 0.15%–0.25% forest yr
−1
, while corridors lost
0.3%–0.5% forest yr
−1
(Fig. 3). Rates of deforestation were higher in cor-
ridors than JCUs in all years (P b0.05 in all years; Fig. 3). In terms of for-
est area lost, JCUs lost an average 3499 km
2
yr
−1
, while corridors lost
4084 km
2
yr
−1
(Fig. 3). JCUs showed a slight trend of increasing defor-
estation (R
2
= 0.55, P = 0.006) at a rate of 149.2 km
2
yr
−2
, while corri-
dors on average showed a decreasing trend of 40.1km
2
yr
−2
(R
2
=0.03,
P = 0.578), especially noticeable after 2005 (Fig. 3).
3.2. Fragmentation analysis
In addition to deforestation, most corridors exhibited increased frag-
mentation based on both metrics examined. Eighty-six percent of corri-
dors showed an increase in PD from 2000 to 2012. Similarly, most
corridors (77%) had lower CLUMPY in 2012 than 2000, indicating a
shift towards more dispersed or disaggregated patches. When only con-
sidering those corridors that fall within the forest biome (excluding sa-
vanna and dryland biomes), the results are even more striking, with 93%
of corridors increasing in PD and 79% decreasing in CLUMPY. On
average, corridors became more highly fragmented with increases in
PD for both Central America (x= 0.68 patches per 100 ha; P b0.001)
and South America ( x= 0.19 patches per 100 ha; P = 0.005)
(Table 1). The change in CLUMPY was more modest, with Central
America displaying a larger change (x= 0.03; P b0.001) than South
America (x= 0.01; P = 0.33) (Table 1). Taken together, these results
show forest habitat in corridors is becoming more fragmented, particu-
larly in Central America.
3.3. Deforestation in protected areas
Collectively, protected sections of JCUs and corridors experienced
lower rates of deforestation than unprotected sections (Tables 2 and
3). This pattern is clearest in South America where deforestation oc-
curred at twice the rate in unprotected sections of corridors (−4.85%)
compared to protected sections (−2.26%) (Table 3). An example of
the effectiveness of protected areas in reducing deforestation is shown
in Fig. 4 for two corridors on the border between Brazil and Bolivia. In
Corridor 21, there is heavy deforestation on the Brazilian side of the b or-
der, except within theprotected section of the corridor, and that pattern
extends into the protected sections of Corridor 54 in Brazil. However,
protected areas did not always reduce deforestation: protected sections
of Honduran JCUs lost forest at 2.74 times the rate of unprotected sec-
tions (−11.94% vs. −4.36%), and protected sections of Guatemalan cor-
ridors lost forest at 1.37 times therate of unprotected sections (−9.16%
vs. −6.67%) (Table 2). Significantly, protected sections of corridors in
Central America experienced 1.27 times the rates of loss than protected
sections of JCUs (−3.56% vs. −2.80%; Table 2) and 5.14 times the rates
of loss in South America (−2.26% vs. −0.44%; Table 3), indicating that
the overall higher rates of loss in corridors cannot be solely attributed
to lower rates of protection.
We performed individual analyses for each protected area IUCN des-
ignation, but no range-wide pattern emerged (Appendix C). While level
of protection may be important at a finer scale, not every country uti-
lizes all levels of protection. For example, less than half of CentralAmer-
ican and South American countries have any land within JCUs or
corridors as a strict nature reserve (IUCN Ia) (Appendix C). Therefore,
the relationship between deforestation and protection status is likely
confounded by intercountry variability in forest loss.
4. Discussion
4.1. Forest change analysis
Our results demonstrate that jaguar corridors are experiencinghigh
rates of deforestation and fragmentation of forest. Numerous jaguar
core areas(JCUs) also experienced substantial forest loss, and JCUs dem-
onstrated accelerating forest loss between 2000 and 2012. These forest
loss rates and increased fragmentation of forests were generally higher
in Central America and the southern edge of the jaguar range, where
JCUs tend to be smaller, which suggests that long-term viability of
some core areas for jaguars may be threatened.
Compared to JCUs, forest loss was higher in corridors for both
protected and unprotected sites, suggesting that human pressure on re-
maining forest in corridors is high regardless of protection status. This
finding is alarming, considering that maintaining connectivity of jaguar
populations across the range is one of the key goals for their conserva-
tion (Rabinowitz and Zeller, 2010; Zeller et al., 2013), and that work
on determining the functionality of jaguar corridors forms the backbone
of much recent research (Cuyckens et al., 2014; Petracca et al., 2014;
Rodríguez-Soto et al., 2013; Silveira et al., 2014; Zeller et al., 2011).
Given their substantial movement capabilities, jaguars may be able to
move across some types of non-forested habitat during dispersal, and
therefore minor loss of forest may not always lead to reduced connec-
tivity. However, jaguars are often absent from smaller forest patches
(Thornton et al., 2011; Urquiza-Haas et al., 2009), persist better in
areas of more forest cover, and are vulnerable to increased human per-
secutionin less forested and more fragmented landscapes (De Angelo et
al., 2013), strongly suggesting that forest loss in corridors will be prob-
lematic for jaguar connectivity and the persistence of jaguar residents
within corridor landscapes. Therefore, these results suggest that in-
creased engagement with communities in key corridors is needed to
maintain connectivity for jaguars in the face of rapid land-use change
across the jaguar's range. For example, working with communities to
minimize human-wildlife conflict, reduce forest loss, or protect private
0.0%
0.1%
0.2%
0.3%
0.4%
0.5%
0.6%
0.7%
2000 2002 2004 2006 2008 2010 2012
Deforestation (% yr-1)
A
B
0
2000
4000
6000
8000
10000
2000 2002 2004 2006 2008 2010 2012
Deforestation (km2yr-1)
Year
Corridor
JCU
Total
*
**
***
***
***
***
***
***
**
**
*
*
Fig. 3. Average rates of deforestation for jaguar conservation units (JCUs) and corridors
between 2000 and 20 12. (A) Deforestation was higher in cor ridors on a percentage
basis in every year (*P b0.05; **P b0.01; ***P b0.001); error bars represent 95%
confidence intervals. (B) On a total area basis, deforestation was similar in corridors and
JCUs.
12 P.J. Olsoy et al. / Biological Conservation 203 (2016) 8–16

reserves within corridors may be productive approaches (Hoogesteijn
and Hoogesteijn, 2010; Salom-Pérez et al., 2010).
We also documented higher rates of loss in unprotected versus
protected sections of corridors and JCUs. This finding confirms other
large-scale research regarding the utility of protected areas to buffer
against forest loss and other anthropogenic threats (Butsic et al., 2015;
Figueroa and Sánchez-Cordero, 2008; Pfaff et al., 2015), particularly in
Latin America (Heino et al., 2015). Given differential rates of loss in
protected versus unprotectedJCUs and corridors,we suggest that efforts
should be increased to secure protection for more jaguar habitat, partic-
ularly in corridors where only 5% of corridor area is protected in Central
America and 12% in SouthAmerica. For example, Mexicocontains 69% of
the Central American corridors in terms of total area, with b3% of that
area being protected. Further, Mexico is at the current northern edge
of jaguar range, providing the only possible link to reestablishing a pop-
ulation in the United States. Based on our results, bringing more forest-
ed sections of jaguar corridors under officialprotection in Mexico would
be highly beneficial.
Despite most countries showing less deforestation in protected areas
than unprotected, Honduran JCUs and Guatemalan corridors show
higher forest loss in protected sections. Moreover, most protected sec-
tions of JCUs and corridors in all countries were affected to some extent
by deforestation, suggesting that protected area status only slows the
rate of deforestation, but does not fully halt habitat loss and fragmenta-
tion. Indeed, some protected areas approached or exceeded 5% forest
loss, which was approximately equivalent to a global average of forest
loss (across all forest including non-protected forest tracts) calculated
from the Hansen et al. (2013) dataset (Heino et al., 2015). A full analysis
of why certain protected areas were more functional than others is be-
yond the scope of this paper, but may be related to differential country
or site level drivers such as population density, enforcement, or level of
development (Butsic et al., 2015; Geldmann et al., 2015; Leverington et
al., 2010).
At a regional level, forest loss is quite uneven across the jaguar
range; however, some patterns do emerge. Within the core of the
range, theAmazon forest and surrounding areas, there is relatively little
forest loss. In contrast, heavy deforestation extends from the edges of
the jaguar's range to the north, south, and east. Of particular concern
is the rapid decline in forest of the jaguar corridors of Central America,
where there is already a natural geographic bottleneck and loss of a sin-
gle corridor could disconnect the entire northern population from South
America. Loss of connectivity between jaguar populations in Central and
South America could be problematic for maintaining range-wide genet-
ic diversity as well as potentially long-term metapopulation stability
(Brown and Kodric-Brown, 1977; Frankham et al., 2002; Soulé and
Mills, 1998).
4.2. Limitations
Our use of a global-scale dataset of forest loss and regrowth derived
from satellite imagery has some limitations with implications for our re-
sults. Some regrowth and areas classified as forest by Hansen et al.
(2013) are young regenerating stands or agricultural areas such as
palm oil plantations (Tropek et al., 2014), with relatively unknown im-
pacts on jaguar movement or habitat use. Inclusion of such areas as “for-
est”in our analysis may bias our findings towards lower rates of habitat
loss. Refinement of the Hansen et al. (2013) dataset at a smaller, local
scale with data from higher-resolution sensors could help alleviate
these issues, but would be challenging to implement across the jaguar
range. On the other hand, non-forested areas not accounted for in this
Table 1
Summary of fragmentation statistics for jaguar corridors in Central America and South America in 2000 and 2012, and the difference between those years(2012−2000). Patch density
(PD; patchesper 100 ha) measuresconnectedness of thelandscape, and clumpiness index (CLUMPY; −1to 1) measures the extent to which the landscape is aggregated orclumped. As-
terisks rep resent sign ificance levels from a paired t-test (*P b0.05; **P b0.01; ***P b0.001).
Central America (n = 21) South America (n = 50)
2000 2012 Difference 2000 2012 Difference
PD
x1.86 2.55 0.68*** 1.87 2.06 0.19**
SEx0.33 0.37 0.16 0.28 0.27 0.06
Min/max 0.09/6.41 0.40/7.24 0.03/3.37 0.00/6.00 0.00/5.89 −1.10/1.72
CLUMPY
x0.73 0.70 −0.03*** 0.67 0.69 0.01
SEx0.02 0.02 0.01 0.05 0.03 0.01
Min/max 0.49/0.87 0.57/0.87 −0.09/0.08 −0.86/0.94 −0.27/0.93 −0.06/0.60
Table 2
Country-level analysis of forest change (2000–2012) in protected and unprotected areas of conservation concern for jaguars (Panthera onca) in Central America.Protected areas in this
table include areas with International Unionfor the Conservationof Nature (IUCN)designations I-VI.Areas where the designation wasnot reported are included in theunprotected group.
fc
2000
= forest cover in 2000.
Country Status Corridors JCUs
Total area (km
2
)fc
2000
Area change (km
2
) Percent change Total area (km
2
)fc
2000
Area change (km
2
) Percent change
Belize Protected 1007 76.12% −19.3 −1.91% 6141 97.37% −125.5 −2.04%
Unprotected 3602 76.51% −234.8 −6.52% 1788 94.48% −66.6 −3.72%
Costa Rica Protected 375 89.75% −5.8 −1.56% 9292 93.15% −48.1 −0.52%
Unprotected 3144 76.90% −108.7 −3.46% 8263 79.28% −184.7 −2.24%
Guatemala Protected 1874 82.11% −171.8 −9.16% 11,969 97.04% −783.4 −6.55%
Unprotected 4647 66.88% −309.9 −6.67% 6417 93.53% −715.9 −11.16%
Honduras Protected 572 82.57% −18.5 −3.25% 4105 95.95% −489.9 −11.94%
Unprotected 11,133 77.72% −673.7 −6.05% 14,176 80.94% −618.3 −4.36%
Mexico Protected 3162 31.78% −39.8 −1.26% 27,340 64.89% −206.5 −0.76%
Unprotected 105,113 58.69% −2593.1 −2.47% 113,794 61.78% −2183.8 −1.92%
Nicaragua Protected 312 80.43% −11.6 −3.71% 6674 98.02% −363.9 −5.45%
Unprotected 14,220 82.53% −1539.4 −10.83% 8940 93.30% −646.3 −7.23%
Panama Protected 252 95.55% −2.4 −0.92% 8655 98.25% −59.8 −0.69%
Unprotected 6670 91.43% −202.5 −3.04% 18,965 93.11% −659.3 −3.48%
Central America totals Protected 7554 61.03% −269.2 −3.56% 74,176 84.90% −2077.1 −2.80%
Unprotected 148,529 64.94% −5662.1 −3.81% 172,343 70.80% −5074.9 −2.94%
13P.J. Olsoy et al. / Biological Conservation 203 (2016) 8–16

study may be suitable habitat for jaguars. Savanna and shrubland habi-
tat that is otherwise unaffected by humans can be used by jaguars
(Tôrres et al., 2012; Vynne et al., 2011). However, Sanderson et al.
(2002) categorized most savanna, grasslands, and shrublands as serving
low to medium probability of sustaining jaguar populations long-term,
and thus these areas may only be suitable for dispersal. We also do
not consider potential impacts to habitat that can occur without defor-
estation, such as overhunting of prey species, which may impact jaguar
movement and survival. However, our analysis provides a first step in
developing an understanding of the most important driver of jaguar de-
cline–deforestation–and how that relates to a range-wide conservation
plan.
Another potential concern with our analysis is that jaguars may uti-
lize additional corridors for dispersal not accounted for in this analysis.
It is important for conservation prioritization plans to account for this,
particularly in rapidly changing environments such as the Neotropics.
Mapping alternative corridors should be explored in future research,
for example via circuit theory (McRae et al., 2008), which identifies
multiple potentialdispersal pathways. More than one forestedroute be-
tween JCUs may be viable for jaguars, but least-cost corridors (such as
the ones used in the design of the jaguar network) will only identify
the single “best”corridor, and therefore may underrepresent connectiv-
ity. If alternative corridors are present that are not experiencing rapid
deforestation, overall network connectivity may be relatively unaffect-
ed, but in Central America, where the options for connectivity are al-
ready limited, alternative pathways are unlikely.
4.3. Implications for conservation planning
Our results generally align with the Zeller et al. (2013) prioritization
of conservation areas, who found that Mexico, Central America, and the
very southern and eastern areasof jaguar range were the highest prior-
ity for maintaining a range-wide conservation network based on graph
theory indices. The addition of land-use change to identify priority areas
bolstersthe case for the importance of these areas,indicating which cor-
ridors and JCUs are most affected by deforestation, and therefore most
urgently in need of conservation action. Systematic conservation plan-
ning literature suggests that accounting for anthropogenic threats
such as habitat loss in the planning process can result in more effective
conservation prioritization (Moilanen and Cabeza, 2007; Pressey et al.,
2007; Visconti et al., 2010). For example, given that not all sites can be
protected or managed at the same time, information on threats can be
used to aid in deciding which sites receive attention first (Pressey et
al., 2007). Our results clarify which areas are under greatest threat
from habitat loss, and therefore where conservation intervention
should be considered or alternative corridors advanced. More broadly,
our study demonstrates the utility of combining spatially explicit infor-
mation about protected areas and habitat change when considering
strategies for conserving species. Given the rapid pace of landscape
and climate change in many parts of the world, this approach may be
Table 3
Country-level analysis of forestchange (2000–2012) in protectedand unprotectedareas of conservation concernfor jaguars (Panthera onca) in SouthAmerica. Protected areas in thistable
includeareas with International Unionfor the Conservation of Nature(IUCN) designations I-VI.Areas where the designation wasnot reported are included in theunprotected group. fc
2000
= forest cover in 2000.
Country Status Corridors JCUs
Total area (km
2
)fc
2000
Area change (km
2
) Percent change Total area (km
2
)fc
2000
Area change (km
2
) Percent change
Argentina Protected 8 62.48% −0.2 −3.35% 5923 87.88% −30.6 −0.52%
Unprotected 9240 44.74% −430.8 −4.66% 28,657 69.47% −945.0 −3.30%
Brazil Protected 58,726 67.44% −1978.8 −3.37% 902,583 89.17% −3611.8 −0.40%
Unprotected 488,637 65.00% −30,637.2 −6.27% 1,549,950 86.79% −7856.4 −0.51%
Bolivia Protected 11,651 97.15% −149.9 −1.29% 116,625 79.22% −890.2 −0.76%
Unprotected 57,618 90.37% −2431.8 −4.22% 103,578 76.83% −1727.6 −1.67%
Colombia Protected 878 97.49% −9.3 −1.06% 51,405 94.68% −561.4 −1.09%
Unprotected 7081 80.11% −219.7 −3.10% 652,023 79.60% −10,987.0 −1.69%
Ecuador Protected 908 93.61% −14.4 −1.59% 22,997 98.87% −74.1 −0.32%
Unprotected 14,373 88.82% −360.6 −2.51% 41,935 98.74% −193.5 −0.46%
French Guiana Protected ––– –194 96.20% −1.1 −0.56%
Unprotected ––– –3 99.17% 0.0 −1.74%
Guyana Protected 2230 99.71% −1.3 −0.06% 4534 99.82% −3.6 −0.08%
Unprotected 55,482 99.29% −55.5 −0.10% 8023 88.17% −32.3 −0.40%
Paraguay Protected 803 94.14% −3.8 −0.47% 7557 66.44% −77.1 −1.02%
Unprotected 27,563 44.33% −3112.7 −11.29% 32,108 57.99% −2277.2 −7.09%
Peru Protected 8541 99.22% −7.0 −0.08% 75,784 98.62% −173.2 −0.23%
Unprotected 73,112 97.91% −220.8 −0.30% 53,769 96.25% −661.8 −1.23%
Suriname Protected 18 100.00% 0.0 −0.02% 11,680 99.76% −6.7 −0.06%
Unprotected 24,792 99.31% −10.9 −0.04% 4935 98.99% −3.9 −0.08%
Venezuela Protected 20,121 95.39% −186.9 −0.93% 108,630 89.34% −321.9 −0.30%
Unprotected 19,813 55.77% −216.0 −1.09% 18,504 56.71% −191.8 −1.04%
South America totals Protected 103,884 80.19% −2351.6 −2.26% 1,307,912 89.23% −5751.7 −0.44%
Unprotected 777,711 72.88% −37,696.0 −4.85% 2,493,485 84.14% −24,876.5 −1.00%
Fig. 4. An example of the effectiveness of protected areas in reducing deforestation in two
corridors on the border between Brazil and Bolivia. In Corridor 21 (left) there is heavy
deforestation on the Brazilian side of the border, except within the protected section of
the corridor (orange boundary). The pattern extends into Corridor 54 (right), where
protected sections of the corridor contain relatively little deforestation compared to
unprotected sections.
14 P.J. Olsoy et al. / Biological Conservation 203 (2016) 8–16

important for assessing the functionality of a wide variety of large-scale
conservation plans and keeping them current in a changing world.
Acknowledgments
Support for this research was provided by a Washington State Uni-
versity college-funded graduate research assistantship to P. Olsoy.
Supplementary data
Supplementary data to this article can be found online at http://dx.
doi.org/10.1016/j.biocon.2016.08.037.
References
Adriaensen, F., Chardon, J.P., De Blust, G., Swinnen, E., Villalba, S., Gulinck, H., Matthysen,
E., 2003. The application of ‘least-cost’modelling as a functional landscape model.
Landsc. Urban Plan. 64 (4), 233–247. http://dx.doi.org/10.1016/S016 9-
2046(02)00242-6.
Azevedo, F.C., Murray, D.L., 2007. Evaluation of potential factors predisposing livestock to
predation by jaguars. J. Wildl. Manag. 71 (7), 2379–2386. http://dx.doi.org/10.2193/
2006-520.
Brown, J.H., Kodric-Brown, A., 1977. Turnover rates in insular biogeography: effect of immi-
gration on extinction. Ecology 58 (2), 445–449. http://dx.doi.org/10.2307/1935620.
Butsic, V., Baumann, M., Shortland, A., Walker, S., Kummerle, T., 2015. Conservation and
conflict in the Democratic Republic of Congo: the impacts of warfare, mining, and
protected areas on deforestation. Biol. Conserv. 191, 266–273. http://dx.doi.org/10.
1016/j.biocon.2015.06.037.
Cavalcanti, S.M.C., Gese, E.M., 2009. Spatial ecology and social interactions of jaguars
(Panthera onca) in the southern Pan tanal, Brazil. J. Mammal. 90 (4), 935–945.
http://dx.doi.org/10.1644/08-MAMM-A-188.1.
Chester, C., 2015. Yellowstone to Yukon: transborder conservation across a vast interna-
tional landscape. Environ. Sci. Pol. 49, 75–84. http://dx.doi.org/10.1016/j.envsci.
2014.08.009.
Crawshaw,P.G., Quigley, H.B., 1991. Jaguar spacing,activity and habitat use in a seasonally
flooded environment in Brazil. J. Zool. 223 (3), 357–370. http://dx.doi.org/10.1111/j.
1469-7998.1991.tb04770.x.
Curran, L.M., Trigg, S.N., McDonald,A.K., Astiani, D., Hardiono, Y.M., Siregar, P., Caniago, I.,
Kasischke, E., 2004. Lowland forest loss in protected areas of Indonesian Borneo. Sci-
ence 303 (5660), 1000–1003. http://dx.doi.org/10.1126/science.1091714.
Cuyckens, G.A.E., Falke, F., Petracca, L., 2014. Jaguar Panthera onca in its southernmost
range: Use of a corridor between Bolivia and Argentina. Endanger. Species Res. 26
(2), 167–177. http://dx.doi.org/10.3354/esr00640.
Davis, M.L., Kelly, M.J., Stauffer, D.F., 2011. Carnivore co-existence and habitat use in the
mountain pine ridge forest reserve, Belize. Anim. Conserv. 14 (1), 56–65. http://dx.
doi.org/10.1111/j.1469-1795.2010.00389.x.
De Angelo, C., Paviolo, A., Di Bitetti, M.S., 2011. Differential impact of landscape transfor-
mation on pumas (Puma concolor) and jaguars (Panthera onca) in the upper Paraná
Atlantic Forest. Divers. Distrib. 17 (3), 422–436. http://dx.doi.org/10.1111/j.1472-
4642.2011.00746.x.
De Angelo, C., Paviolo, A., Wiegand, T., Kanagaraj, R., Di Bitetti, M.S., 2013. Understanding
species persistence for defining conservation action: a management landscape for
jaguars in the Atlantic Forest. Biol. Conserv. 159, 422–433. http://dx.doi.org/10.
1016/j.biocon.2012.12.021.
DeFries, R., Hansen, A., Newton, A.C., Hansen, M.C., 2005. Increasing isolation of protected
areas in tropical forests over the past twenty years. Ecol. Appl. 15 (1), 19–26. http://
dx.doi.org/10.1890/03-5258.
ESA, 2014. Climate Change Initiative Land Cover (CCI-LC) Dataset 2012 (Version 1.4).
Available online bwww.esa-landcover-cci.org/N. Accessed 26 May 2015.
Espinosa, S., 2012. Road Development, Bushmeat Extraction and Jaguar Conservation in
Yasuní Biosphere Reserve - Ecuador PhD dissertation University of Florida, Gaines-
ville, FL.
Figueroa,F., Sánchez-Cordero, V., 2008.Effectiveness of natural protected areas toprevent
land use and land cover change in Mexico.Biodivers. Conserv. 17, 3223–3240. http://
dx.doi.org/10.1007/s10531-008-9423-3.
Frankham, R., Briscoe,D.A., Ballou, J.D., 2002. Introduction to Conservation Genetics. Cam-
bridge University Press, Cambridge, UK.
Geldmann, J., Coad, L., Barnes, M., Craigie, I.D., Hockings, M., Knights, K., Leverington, F.,
Cuadros, I.C., Za mora, C., Woodley, S., Burgess, N.D., 2015. Changes in protected
area management effectiveness over time: a global an alysis. Biol. Co nserv. 191,
692–699. http://dx.doi.org/10.1016/j.biocon.2015.08.029.
Guerrero, A.M., Mcallister, R.R.J., Wilson, K.A., 2015. Achieving cross-scale collaboration
for large-scale conservation initiatives. Conserv. Lett. 8 (2), 107–117. http://dx.doi.
org/10.1111/conl.12112.
Hansen,M.C.,Potapov,P.V.,Moore,R.,Hancher,M.,Turubanova,S.A.,Tyukavina,A.,Thau,D.,
Stehman,S.V.,Goetz,S.J.,Loveland,T.R.,Kommareddy,A.,Egorov,A.,Chini,L.,Justice,
C.O., Townshend, J.R.G., 2013. High-resolution global maps of 21st-century forest cover
change. Science 342 (6160), 850–853. http://dx.doi.org/10.1126/science.1244693.
Hansen, M.C., Potapov, P.V., Margono, B., Stehman, S.V., Turubanova, S.A., Tyukavina, A.,
2014. Response to comment on “high-resolution global maps of 21st-century forest
cover change.”. Science 344 (6187), 981. http://dx.doi.org/10.1126/science.1248817.
Heino, M., Kummu, M., Makkonen, M., Mulligan, M., Verburg, P.H., Jalava, M., Räsänen,
T.A., 2015. Forest loss in protected areas and intact forest landscapes: a global analy-
sis. PLoS One 10 (10), e0138918. http://dx.doi.org/10.137/journal.pone.0138918.
Heller, N.E., Zavaleta, E.S., 2009. Biodiversity management in the faceof climate change: A
review of 22 years of recommendations. Biol. Conserv. 142 (1), 14–32. http://dx.doi.
org/10.1016/j.biocon.2008.10.006.
Hoogesteijn, A., Hoogesteijn, R., 2010. Cattle ranching and biodiversity conservation as
allies in South America's flooded savannas. Great Plains Res. 20, 37–50.
IUCN and UNEP-WCMC, 2015. The World Database on Protected Areas (WDPA) [On-
Line], Downloaded June 2015. UNEP- WCMC, Cambridge, UK Available at: www.
protectedplanet.net.
Klein, C., Wilson, K., Watts, M., Stein, J., Berry, S., Carwardine, J., Smith, M.S., Mackey, B.,
Possingham, H., 2009. Incorporating ecological and evolutionary processes into con-
tinental-scale conservation planning. Ecol. Appl. 19 (1), 206–217. http://dx.doi.org/
10.1890/07-1684.1.
Lawler, J.J., Tear, T.H., Pyke, C., Shaw, M.R., Gonzalez, P., Kareiva, P., Hansen, L., Hannah, L.,
Klausmeyer, K., Aldous, A., Bienz, C., Pearsall, S., 2010. Resource management in a
changing and uncertain climate. Front. Ecol. Environ. 8 (1), 5–43. http://dx.doi.org/
10.1890/070146.
Leverington, F., Lemos Costa, K., Pavese, H., Lisle, A., Hockings, M., 2010. A global analysis
of protected area management effectiveness. Environ. Manag. 46 (5 ), 685–698.
http://dx.doi.org/10.1007/s00267-010-9564-5.
MacDonald, D.W., Loveridge, A.J., 2010. Dramatis personae: an introduction to thewild fe-
lids. In: MacDonald, D.W., Loveridge, A.J. (Eds.), Biology and Conservation of Wild Fe-
lids. Oxford University Press.
Mascia, M.B., Pailler, S., 2011. Protected area downgrading, downsizing, and
degazettement (PADDD) and its conservation implications. Conserv. Lett. 4 (1),
9–20. http://dx.doi.org/10.1111/j.1755-263X.2010.00147.x.
McGarigal, K., Marks, B.J., 1995. FRAGSTATS: Spatial Pattern Analysis Program for Quanti-
fying Landscape Structure. Gen. Tech. Rep. PNW-GTR-351. Pacific Northwe st Re-
search Station.
McRae, B.H., Dickson, B.G., Keitt, T.H., Shah, V.B., 2008. Using circuit theory to model con-
nectivity in ecology, evolution, and conservation.Ecology 89 (10), 2712–2724. http://
dx.doi.org/10.1890/07-1861.1.
Moilanen, A., Cabeza, M., 2007. Accounting for habitat loss rates in sequential reserve se-
lection: Simple methods for large problems. Biol. Conserv. 136 (3), 470–482. http://
dx.doi.org/10.1016/j.biocon.2006.12.019.
Petracca, L.S., Hernandez-Potosme, S., Obando-Sampson, L., Salom-Perez, R., Quigley, H.,
Robinson, H.S., 2014. Agricultural encroachment and lack of enforcement threaten
connectivity of range-wide jaguar (Panthera onca) corridor. J. Nat. Conserv. 22 (5),
436–444. http://dx.doi.org/10.1016/j.jnc.2014.04.002.
Pfaff, A., Robalino, J., Herrera, D., Sandoval, C., 2015. Protected areas' impacts on Brazilian
Amazon deforestation: examining conservation-development interactions to inform
planning. PLoS One 10 (7), e0129460. http://dx.doi.org/10.1371/journal.pone.
0129460.
Pressey, R.L., 2004. Conservation planning and biodiversity: assembling the best data for
the job. Conserv. Biol. 18 (6), 1677–1681. http://dx.doi.org/10.1111/j.1523-1739.
2004.00434.x.
Pressey, R.L., Bottrill, M.C., 2009. Approaches to landscape- and seascape-scale conserva-
tion planning: convergence, contrasts and challenges. Oryx 43 (4), 464–475. http://
dx.doi.org/10.1017/S0030605309990500.
Pressey, R.L., Cabeza, M., Watts, M.E., Cowling, R.M., Willson, K.A., 2007. Conservation
planning in a changing world. Trends Ecol. Evol. 22 (11), 583–592. http://dx.doi.
org/10.1016/j.tree.2007.10.001.
Rabinowitz, A., 2014. An Indomitable Beast: The Remarkable Journey of the Jaguar. Island
Press, Washington, D.C.
Rabinowitz, A., Zeller, K.A., 2010. A range-wide model of landscape connectivity and con-
servation for the jaguar, Panthera onca. Biol. Conserv. 143 (4), 939–945. http://dx.doi.
org/10.1016/j.biocon.2010.01.002.
Redford, K.H., Amato, G., Baillie, J., Beldomenico, P., Bennett, E .L., Clum, N., Coo k, R.,
Fonseca,G., Hedges, S., et al.,2011. What does it meanto successfully conserve a (ver-
tebrate)species? Bioscience61 (1), 39–48. http://dx.doi.org/10.1525/bio.2011.61.1.9.
Roberge, J.-M., Angelstam, P., 2004. Usefulness of the umbrella species concept as a con-
servation tool. Conserv. Biol. 18 (1), 76–85. http://dx.doi.org/10.1111/j.1523-1739.
2004.00450.x.
Rodríguez-Soto, C., Monroy-Vilchis, O., Zarco-González, M.M., 2013. Corridors for jaguar
(Panthera onca) in Mexico: conservation strategies. J. Nat. Conserv. 21 (6), 438–443.
http://dx.doi.org/10.1016/j.jnc.2013.07.002.
Salom-Pérez, R., Polisar, J., Quigley, H., Zeller, K., 2010. Iniciativa del corredor del jaguar:
un corredor biol ógico y un compro miso a largo plazo para la conservacion.
Mesoamericana 14 (3), 25–34.
Sanderson, E.W. , Redford, K.H., Chetkiewicz, C.L.B., Medellin, R.A., Rabinowitz, A.R.,
Robinson, J.G., Taber, A.B., 2002. Planning to save a species: the jaguar as a model.
Conserv. Biol. 16 (1), 58–72 doi:0.1046/j.1523-1739.2002.00352.x.
Sanderson, E.W., Redford, K.H., Weber, B., Aune, K., Baldes, D., Berger, J., Carter, D., Curtin,
C., Deer, J., et al., 2008. The ecological future of the North American bison: conceiving
long-term, large-scaleconservation of wildlife. Conserv. Biol. 22 (2), 256–266. http://
dx.doi.org/10.1111/j.1523-1739.2008.00899.x.
Silveira,L., Sollmann, R., Jácomo, A.T.A., Diniz Filho, J.A.F., Tôrres, N.M., 2014. The potential
for large-scale wildlifecorridors between protected areas in Brazil using the jaguar as
a model species. Landsc. Ecol. 29 (7), 1213–1223. http://dx.doi.org/10.1007/s10980-
014-0057-4.
Soulé, M.E., Mills, L.S., 1998. No need to isolate genetics. Science 282 (5394), 1658–1659.
http://dx.doi.org/10.1126/science.282.5394.1658.
Thorbjarnarson, J., Mazzotti, F., Sanderson, E., Buitrago, F., Lazcano, M., Minkowski, K.,
Muňiz, M., Ponce, P., Sigler, L., Soberon, R., Trelancia, A.M., Velasco, A., 2006. Regional
15P.J. Olsoy et al. / Biological Conservation 203 (2016) 8–16

habitat conservation priorities for the American crocodile. Biol. Conserv. 128 (1),
25–36. http://dx.doi.org/10.1016/j.biocon.2005.09.013.
Thornton, D.H., Branch, L.C., Sunquist, M.E., 2011. Passive sampling effects and landscape
locationalter associationsbetween species traits and response to fragmentation.Ecol.
Appl. 21 (3), 817–829. http://dx.doi.org/10.1890/10-0549.1.
Thornton, D., Zeller, K., Rondinini, C., Boitani, L., Crooks, K., Burdett, C., Rabinowitz, A.,
Quigley, H., 2016. Assessing the umbrella value of a range-wide conservation net-
work for jaguars (Panthera onca). Ecol. Appl. 26 (4), 1112–1124. http://dx.doi.org/
10.1890/15-0602.
Tôrres, N.M., De Marco, J.P., Santos, T., Silveira, L., de Almeida Jácomo, A.T., Diniz-Filho,
J.A.F., 2012. Can species distribution modelling provide estimates of population den-
sities? A case study with jaguars in the Neotropics. Divers. Distrib. 18 (6), 615–627.
http://dx.doi.org/10.1111/j.1472-4642.2012.00892.x.
Tropek, R.,Sedláček, O., Beck, J., Keil, P., Musilova, Z.,Šímová, I., Storch, D.,2014. Comment
on “High-resolution global maps of 21st-century forest cover change.”. Science 344
(6187), 981. http://dx.doi.org/10.1126/science.1248753.
Urquiza-Haas, T., Peres, C.A., Dolman, P.M., 2009. Regional scale effects of human density
and forest disturbance on large-bodied vertebrates throughout the Yucatán peninsu-
la, Mexico. Biol. Conserv. 142 (1), 134–148. http://dx.doi.org/10.1016/j.biocon.2008.
10.007.
Viña, A., Tuanmu, M.-N., Xu, W., Li, Y., Ouyang, Z., DeFries, R., Liu, J., 2010. Range-wide
analysis of wildlife habitat: Implications for conservation. Biol. Conserv. 143,
1960–1969. http://dx.doi.org/10.1016/j.biocon.2010.04.046.
Visconti,P., Pressey, R.L., Bode, M., Segan, D.B., 2010. Habitat vulnerability in conservation
planning—when it matters and how much. Conserv. Lett. 3 (6), 404–414. http://dx.
doi.org/10.1111/j.1755-263X.2010.00130.x.
Vynne, C., Keim, J.L., Machado, R.B.,Marinho-Filho, J., Silveira, L.,Groom, M.J., Wasser, S.K.,
2011. Resource selection and its implication forwide-ranging mammals of the Brazil-
ian Cerrado. PLoS One 6 (12), e28939. http://dx.doi.org/10.1371/journal.p one.
0028939.
Willis, K.S., 2015. Remote sensing change detection for ecological monitoring in United
States protecte d areas. Biol. Co nserv. 182, 233–242. http://dx.doi.org/10.1016/j.
biocon.2014.12.006.
Zeller, K ., 2007. Jaguarsin the New Millennium Data Set Update: The State of the Jaguar in
2006. WCS Report. Wildlife Conservation Society, New York, pp. 1–82.
Zeller, K.A., Nijhawan, S., Salom-Perez, R., Hernandez-Potosme, S., Hines, J.E., 2011. Inte-
grating occupancy modeling and interview data for corridor identification: A case
study for jaguars in Nicaragua. Biol. Conserv. 144 (2), 892–901. http://dx.doi.org/10.
1016/j.biocon.2010.12.003.
Zeller, K.A., Rabinowitz, A., Salom-Perez, R., Quigley, H., 2013. The Jaguar Corrido r Initia-
tive: A range-wide conservation strategy. In: Ruiz-Garcia, M., Shostell, J.M. (Eds.), Mo-
lecular Population Genetics, Evolutionary Biology and Biological Conservation of
Neotropical Carnivores. Nova Publishers, pp. 629–658.
16 P.J. Olsoy et al. / Biological Conservation 203 (2016) 8–16