You are on page 1of 12

Animal Conservation.

Print ISSN 1367-9430

Defaunation and biomass collapse of mammals in the


largest Atlantic forest remnant
M. Galetti1,2, C. R. Brocardo1, R. A. Begotti1, L. Hortenci1, F. Rocha-Mendes1, C. S. S. Bernardo1,
R. S. Bueno1, R. Nobre1, R. S. Bovendorp1, R. M. Marques1, F. Meirelles1, S. K. Gobbo1, G. Beca1,
G. Schmaedecke1 & T. Siqueira1
1 Departamento de Ecologia, Instituto de Biocie^ncias, Universidade Estadual Paulista (UNESP), Rio Claro, Brazil
2 Section for Ecoinformatics and Biodiversity, Department of Bioscience, Aarhus University, Ny Munkegade 114, DK-8000 Aarhus C,
Denmark

Keywords Abstract
Brazilian Atlantic forest; hunting; line-transect
method; distance sampling; extinction risk; Large continuous rainforests are the main hope for sustaining the population of
protected areas; defaunation; mammalian large-bodied vertebrates that cannot cope with fragmentation or unsustainable hunt-
communities. ing. The Brazilian Atlantic forest is considered a biodiversity hotspot and although
highly fragmented, it still contains large forest patches that may be important for
Correspondence the conservation of mammals that require large areas. Here, we estimated species
Mauro Galetti, Departamento de Ecologia, richness, density and biomass of medium- and large-sized mammals along the lar-
^ncias, Universidade
Instituto de Biocie gest remnant of the Atlantic rainforest, Brazil (the Serra do Mar bioregion), an esti-
Estadual Paulista (UNESP), CP 199, 13506- mated area of 8000 km2. We recorded 44 species based on 4090 km of diurnal
900 Rio Claro, SP, Brazil. line transects and camera traps, animal tracks and interviews in 11 continental
Email: mgaletti@rc.unesp.br regions and two large land-bridge islands. We found high levels of similarity in
mammalian composition between pairs of sites in the continental forest sites (0.84
Editor: Res Altwegg 1), but much lower similarity between pairs from the continental forest sites and
Associate Editor: Goncalo Ferraz the two large land-bridge islands (0.290.74) indicating potential local extinctions
or poor dispersal of continental mammals to these islands. In addition, we found
Received 08 February 2016; accepted 13 that the density and biomass varied 16- and 70-fold among sites, respectively.
September 2016 Mammalian biomass declined by up to 98% in intensively hunted sites and was
53-fold lower than in other Neotropical non-fragmented forests. Although this large
doi:10.1111/acv.12311 forest remnant is able to maintain a high diversity of medium- and large-bodied
mammal species, their low density and biomass may affect the long-term persis-
tence of these populations and the ecosystem services they provide.

loss because small-bodied species may be favored once


Introduction large-bodied species disappeared (Peres & Palacios, 2007), a
Natural habitats worldwide are suffering from animal species phenomenon named density compensation (MacArthur, Dia-
loss and population decline, particularly on large-bodied ver- mond & Karr, 1972).
tebrates (Galetti et al., 2013; Dirzo et al., 2014). This con- Therefore, the use of species richness to evaluate conser-
servation crisis is so pronounced that some authors have vation importance of a given area may be a weak measure
been considered that we are amidst the sixth mass extinction of ecosystem health (Gonzalez-Maya et al., 2015) because
(Barnosky et al., 2011; Dirzo et al., 2014; Ceballos et al., many species may have populations so low that their effect
2015). Human occupation on natural environments and con- on the ecosystem is marginal (Novaro, Funes & Walker,
sequent land-use change directly reduce the number of spe- 2000; McConkey & OFarrill, 2015; Valiente-Banuet et al.,
cies (Ceballos & Ehrlich, 2002; Dirzo et al., 2014). In 2015). Thus, it is paramount to understand the effects of
addition, even relatively pristine habitats have experienced human-made disturbances, such as forest fragmentation
species losses and populations are declining due to hunting (Rylands & Keuroghlian, 1988; Chiarello, 1999), logging
activities for commercial, subsistence or animal trade (Peres (Johns, 1986; Meijaard et al., 2007) and exotic plantations
& Palacios, 2007; Young et al., 2016) with an estimate of 5 (Barlow et al., 2007; Azhar et al., 2014) on mammal persis-
ton harvest each year in the Amazon and Afrotropical rain- tence and abundance. Large continuous forests are usually
forests (Fa, Peres & Meeuwig, 2002). Indeed, the loss of poorly investigated in terms of the conservation status of
animal biomass from tropical ecosystems may be more dra- mammals, perhaps because of the false impression that these
matic than changes in overall population density and species areas suffer low human impact and that they are safe from

Animal Conservation  (2016)  2016 The Zoological Society of London 1


Defaunation in the Atlantic forest M. Galetti et al.

major anthropogenic disturbances (Rodrigues et al., 2004; been focused on few isolated forest fragments (Chiarello,
Beaudrot et al., 2016). 2000; Cullen, Bodmer & Padua, 2000; Bernardo & Galetti,
The Atlantic Forest in South America originally extended 2004; Bovendorp & Galetti, 2007; Ara ujo, Souza & Ruiz-
from 3 to 31S, and from 35 to 60W, covering more than Miranda, 2008) or in a few species (Price, Piedade & Wor-
150 million ha, extending along the Brazilian coast (92%), mell, 2002; Passamani, 2008; Norris et al., 2011). Here, we
but also reaching into Paraguay and Argentina. The high provide the rst comprehensive study to assess the species
levels of endemism of this forest have intrigued scientists for richness, population density, biomass and community compo-
many years (Leite et al., 2016) which considered a biodiver- sition of medium and large mammals in the largest Atlantic
sity hotspot (Myers et al., 2000). The Brazilian part of the forest remnant, the Serra do Mar subregion. We were partic-
Atlantic forest nowadays covers ~12% of the original area, ularly interested in understanding spatial variation in species
distributed in 245 173 forest fragments (Ribeiro et al., richness, density and biomass along the largest continuous
2009). The largest forest patch is located in the Serra do Atlantic forest and compare with studies that used similar
Mar subregion, mainly along the coastal mountains of the methods in the Atlantic forest fragments and continuous
states of S~ao Paulo, Parana and Santa Catarina, and Amazon forests. In addition, we expected to nd higher
extends into the southern part of the state of Rio de composition dissimilarity between pairs of sites located in
Janeiro. This single patch contains ~1 109 546 ha of con- the mainland sites versus island sites.
tinuous forests, which represents 7% of what remains
(Ribeiro et al., 2009). This forest patch suffered and con-
Study areas
tinues suffering from illegal palm harvesting (Galetti &
Fernandez, 1998), hunting (Galetti et al., 2009), human- We chose 13 sites across 8000 km2 in the Serra do Mar sub-
made infrastructure (pipeline, roads) and human encroach- region (see Ribeiro et al., 2009) for our mammal census
ment that negatively affect animal movement and increase (Fig. 1). This area was divided into three major regions that
mortality (Miotto et al., 2012; Bogoni et al., in press). Sev- differ in human density, historical occupation and hunting
eral studies point out that this region as the largest strong- pressure. The Paranapiacaba massif, Serra do Mar massif
hold for several endemic species of mammals in critical and the two largest land-bridge islands that ank the south
conservation status (Duarte et al., 2012; Keuroghlian et al., and southeast coast of Serra do Mar that were formed by
2012; Ingberman, Fusco-Costa & Monteiro-Filho, 2016). historical marine incursions due to glaciation events, tectonic
However, we do not have much information on the popula- activities and coastal erosion dynamics (Suguio & Martin,
tion status of most these species. 1987). The Paranapiacaba and Serra do Mar differ in the
Studies on the population density and biomass of med- density of human occupation, which reects in differences in
ium- and large-sized mammals in the Atlantic forest have hunting pressure.

Figure 1 Location of the sampled study areas along the Atlantic forest in southeastern Brazil. In light green the protected parks and in dark
green the areas sampled in this study (IT, Itamambuca-Cunha; PI, Picinguaba; VG, Vargem Grande; IB, Ilhabela; CA, Caraguatatuba; JU, Juru-
par ia; BH, Carlos Botelho-Highland; BL, Carlos Botelho-Lowland; IN, Intervales; PE, Petar; JA, Jacupiranga; IC, Ilha do Cardoso).
a; JR, Jure
All these areas are protected parks.

2 Animal Conservation  (2016)  2016 The Zoological Society of London


M. Galetti et al. Defaunation in the Atlantic forest

The Paranapiacaba massif consists of ve conservation minimum distance of 200 m between them (Figure 1). Sam-
protected areas (Parque Estadual Carlos Botelho Highland pling was carried out from September 2011 to January 2012
(BH) and lowland parts (BL) , Parque Estadual Intervales at IT, September 2011 to June 2012 at VG, August 2010 to
(IN), Parque Estadual Jacupiranga (JA), Parque Estadual September 2011 at IC and September 2009 to June 2010 at
Jurupara (JU) and Parque Estadual Turstico do Alto do BL. Cameras operated over 24 h a day and were active for
Ribeira (JA)), located at 48240 W24180 S, with an area of 232 days at IT, 440 days at VG, 528 days at IC and 6
460 000 ha. The primary forests occur in step terrains, 16 days at BH per sampling. The other areas were not sam-
whereas secondary forests occur in at areas that have been pled due to the lack of protection and safety. We installed
used by mining and selective logging in the past. The annual the cameras in trees at a height of c. 40 cm above the
average temperature is around 24C, but in the highlands ground with a small angle of inclination. No bait was used.
frost is common during the winter. Annual rainfall can reach
up to 4000 mm.
Population census
The Serra do Mar massif consists of two conservation
units: Estac~ao Ecol
ogica da Jureia (JR), and Parque Estadual Line transects are the most common method to estimate the
da Serra do Mar. The Parque Estadual da Serra do Mar is population of forest-dwelling mammals (Burnham, Anderson
divided into 10 administrative zones and we sampled four of & Laake, 1980; Peres, 1999). We carried out diurnal line-
them [Caraguatatuba (CA), Itamambuca-Cunha (IT), Picin- transect surveys for mammals >0.20 kg that include from
guaba (PI) and Vargem Grande (VG)]. These areas are small-bodied arboreal species (squirrels, marmosets), to mid-
located at coordinates between 45250 W and 23350 S to sized (agoutis, pacas) and large-bodied species (monkeys,
44500 W and 23220 S.These areas are covered by Atlantic peccaries, deer and tapirs).
ombrophilous dense forest, with the annual average tempera- We established a total of 119 transects (mean = 7.6 tran-
ture around 1824C, and annual average rainfall of sects per site) that ranged from 0.3 to 14.4 km in length
2200 mm. (mean = 3.53 km) depending on local topography (Support-
In addition, we sampled the two largest land-bridge ing Information Table S1). All observers were previously
islands: Parque Estadual da Ilha do Cardoso (IC) and Par- trained and followed the standardized methodology for line-
que Estadual Ilhabela (IB). IC is located in Cananeia transect censuses (Peres, 1999). The cumulative distance
(48050 W and 25030 S). It has an area of 13 500 ha, of sampled in each site varied from 77 to 572 km summing
which 11 000 ha are formed by Atlantic Forest in the strict 4090 km (mean = 292.1 km per site). The surveys were
sense, and the remainder is composed of mangroves, bea- done between 2002 and 2012, and each site was sampled
ches and sand dunes. The average annual temperature is monthly for at least 1 year, so we do not expect changes
21C with annual average rainfall between 1673 and due to potential seasonal variation.
3014 mm. Ilhabela is the largest and most populated island
in the southeastern Brazilian coast. It is located at Ilha de Data analyses
S~ao Sebasti~ao (45180 W and 2380 S), and has 33 600 ha
with mean temperature of 22C, and around 1500 mm of
Beta diversity
annual rainfall.
We estimated pairwise similarity in assemblage composition
Methods by constructing resemblance matrices from the sites-by-spe-
cies dataset based on the Srensen (presenceabsence) coef-
Field efforts cient. This coefcient incorporates both taxon replacement
and taxon richness differences, and gives equal weight to
both rare and abundant species (Koleff, Gaston & Lennon,
Species richness and composition
2003). We then plotted values of similarity among all pairs
We estimated the species richness of each area based on of sites as a function of geographical distances. This allowed
tracks, occasional sightings and interviews with local resi- us to investigate how beta-diversity patterns varied according
dents and rangers (Supporting Information Table S2). We to the region a site was located in (Paranapiacaba massif and
included information from the literature, but we were careful Serra do Mar massif vs. the two islands) and to the geo-
to not include data from management plans because the graphical distance between pairs of sites.
quality of the data is usually poor (Supporting Information
Table S2). In addition, we used a series of camera traps
Density and biomass estimates
(Reconyx model RC55 RapideFire, www.reconyx.com) to
sample medium- and large-sized mammals (>1 kg, except The densities of species were estimated using the Distance
Brazilian squirrels and marmosets with ~200 g) in 4 of the 6.0 Software (Thomas et al., 2010). We grouped the visual-
13 sites. We used 18 cameras at Itamambuca-Cunha (IT) 16 ization events of same species in all sites to obtain most pre-
at Vargem Grande (VG) 30 at Ilha do Cardoso (IC) and 49 cise effective strip width (ESW), but density was estimated
at Carlos Botelho - lowland area (BL) that were distributed based on site-specic sampling effort (Peres, 2000). We
across preexisting transects and trails throughout the areas pooled all data by sites, not transects, due to the low number
including secondary and primary forest habitats, with a of sightings.

Animal Conservation  (2016)  2016 The Zoological Society of London 3


4
Table 1 Density and biomass of eight mammal species in 13 Atlantic forest sites

Density (ind. km2) (biomass kg km2)


Brachyteles Guerlinguetus
Alouatta guariba arachnoides Callithrix aurita Sapajus nigritus Dasyprocta leporina brasiliensis Nasua nasua Tayassu pecari
Ilha do Cardoso NR/NO NO NR
Defaunation in the Atlantic forest

8.91  2.47 4.29  1.62 1.38  0.58 2.64  1.31 13.55  6.25
(50.07  13.88) (18.01  6.80) (0.23  0.09) (13.46  6.68) (474.2  218.7)
Ilhabela NR NR NO 2.49  2.04 NR 6.99  1.99 NR NR
(8.71  7.14) (1.18  0.33)
Carlos Botelho 2.70  0.73 12.15  3.79 NO 8.50  2.15 0.30  0.21 4.26  0.71 2.20  1.39 NR
Highland (15.17  4.10) (130.61  40.74) (29.75  7.52) (1.26  0.88) (0.72  0.12) (11.2  7.08)
Carlos Botelho 1.03  0.53 NR NO 13.94  3.99 NR 4.34  1.27 6.37  4.86 NR
Lowland (5.78  2.97) (48.79  13.96) (0.73  0.21) (32.48  24.78)
Intervales 0.42  0.33 3.04  1.62 NO 2.59  1.09 NR NR NR NR
(2.36  1.85) (32.68  17.41) (9.06  3.81)
Jacupiranga 4.18  2.73 NR NO 1.09  1.11 0.82  0.83 NR 3.38  3.27 NR
(23.49  15.34) (3.81  3.88) (3.44  3.48) (17.23  16.77)
a
Jurupar 0.57  0.48 NR NO 0.90  0.55 0.33  0.28 0.95  0.32 NR NR
(3.20  2.69) (3.15  1.92) (1.38  1.17) (0.16  0.05)
PETAR 0.28  0.25 0.52  0.44 NO 5.51  3.74 NR 0.62  0.39 NR NR
(1.57  1.40) (5.59  4.73) (19.28  13.09) (0.10  0.06)
Caraguatatuba 1.82  0.95 NR NR 7.37  2.39 0.34  0.24 0.16  0.14 NR NR
(10.22  5.33) (25.79  8.36) (1.42  1.00) (0.02  0.02)
Cunha-Itamambuca 0.87  0.43 0.44  0.44 11.45  3.83 2.64  1.30 NR 4.38  1.64 1.19  1.18 3.35  3.09
(4.88  2.41) (4.73  4.73) (4.58  1.53) (9.24  4.55) (0.74  0.27) (6.06  6.01) (117.2  111.6)
eia
Jur 0.98  0.59 1.53  1.65 NO 32.04  7.94 NR 9.42  1.81 NR NR
(5.50  3.31) (16.44  17.73) (112.14  27.79) (1.60  0.30)
Picinguaba NR NR NR NR 1.19  0.79 1.56  0.76 3.45  2.29 NR
(4.99  3.31) (0.26  0.12) (17.59  11.67)
Vargem Grande NR NR 11.06  1.62 2.69  0.54 0.67  0.03 28.74  2.54 2.77  0.13 NR
(4.42  0.64) (9.41  1.89) (2.81  0.12) (4.88  0.43) (14.12  0.66)
ESW 13.9  1.6 ESW 14.9  2.7 ESW 8.1  0.2 ESW 17.7  2.0 ESW 5.9  0.3 ESW 6.3  0.4 ESW 8.6  0.4 ESW 9.2  2.4
m TD 35 m m TD 30 m m TD 16 m m TD 35 m m TD 20 m m TD 15 m m TD 15 m m TD 22 m

NR, not recorded in line transect survey; NO, species does not occur in the area; ESW, effective strip width; TD, truncation distance used in model; PETAR, Parque Estadual Turstico do
Alto do Ribeira.
M. Galetti et al.

Animal Conservation  (2016)  2016 The Zoological Society of London


M. Galetti et al. Defaunation in the Atlantic forest

We used Akaike information criterion for choosing the composition between Paranapiacaba and Serra do Mar is the
best model for ESW denition (truncation distance and ESW occurrence of Buffy-Tufted-Ear Marmoset and Three-Toed
are presented in Table 1). Using density values (D) multi- Sloth in Serra do Mar massif.
plied by average body weight (BW), we calculated the bio- For the four areas that we used camera traps, we obtained
mass of species (D 9 BW = kg km2). BW was based on a total of 416 camera-days at IT (9995 h), 416 camera-days
Paglia et al. (2012). at VG (9988 h), 555 camera-days at IC (13 320 h) and 497
camera-days at BL (11 749 h). From this effort we identied
9 species at IT, 10 species at VG, 11 species at IC and 13
Ecological and human variables
species at BL (Supporting Information Table S3). The mean
We tested if ecological (forest type: lowland or montane rarefaction curve that estimated species richness per survey-
forest) or human inuence (hunting pressure) explains the day did not reach an asymptote in any site (Supporting Infor-
densities and biomass of mammals through Linear Mixed- mation Fig. S1), with species continuing to accumulate at a
Effects Model analysis using the package nlme (Pinheiro rate of 0.04 species at IT, 0.06 at VG and IC and 0.09 at BL
et al., 2007) in R program (R Core Team 2015). To do the after 45 survey days. Based on the rst-order jackknife the
analysis we summed the average density and biomass for extrapolated species richness was between about 8 and 14
game species only and all mammal sampled per site species at IT (estimate  95% CI = 10.96  2.72), 10 and
(n = 13). We used forest type and hunting pressure as xed 16 species at VG (estimate  95% CI = 12.93  3.32), 13
effects and region [Paranapiacaba (n = 6 sites), Islands and 21 species at CB (estimate  95% CI = 16.93  3.86)
(n = 2) and Serra do Mar (n = 5)] as random effects. Forest and 10 and 16 species at IC (estimate  95%
type was based on Oliveira-Filho & Fontes (2000). The CI = 12.96  2.72).
hunting pressure was estimated with same parameters
adopted in Galetti et al. (2009), where:
Beta diversity
1 Lightmoderate hunting. Forest sites located in areas with
low-to-moderate densities of human residents and weak The community composition reected the location of sites
evidence of hunting. Usually, these sites used to be pro- within the Paranapiacaba massif and Serra do Mar massif
tected by professional rangers who regularly patrolled the versus the two islands (Fig. 2), but was not related to the
sites. geographical distance between pairs. Pairs of sites located in
2 Heavy hunting. Forest sites exhibiting considerable evi- the mainland were very similar in terms of species composi-
dence of human disturbances, such as hunting, domestic tion (Srensen similarity values: min = 0.85, max = 1.0,
dogs and palm harvesting. Usually these areas are easily mean = 0.91, SD = 0.03). Just below those were the pairs
accessible by roads and have small residential human pop- that compare IC and the mainland sites (min = 0.74,
ulations inside the park. max = 0.84, mean = 0.78, SD = 0.02). Pairs comparing conti-
nental sites versus Ilhabela island were all very dissimilar
(min = 0.29, max = 0.36, mean = 0.32, SD = 0.02).
Results

Species richness and composition


We recorded 44 non-volant mammal species in the largest
Atlantic forest remnant based on tracks, indirect sightings,
interviews or literature review that varied from 9 to 40 spe-
cies per site (Supporting Information Table S2). However,
only 19 species were recorded through observations in line-
transect censuses. From these species, we were able to have
a robust density and biomass estimate of only eight species
(Table 1).
Of these 44 mammal species recorded, 2 are classied as
Endangered, 6 as Vulnerable, 7 as Near Threatened, 27 spe-
cies as Least Concern and 1 as Data Decient, according to
the IUCN Red List Status (Supporting Information
Table S2). In addition, two exotic species were recorded (do-
mestic dogs and European hare). At least three species are
typical from open savanna habitats (White-Eared Opossum,
Maned wolf and Giant ant-eater) and may experience an
Figure 2 Mammal (>0.20 kg) species composition similarity,
increase in their distribution. Some species are typical from
expressed as Srensen index, between continental forest pairs
semi-deciduous forests (Black-Fronted Titi Monkey, Black (black, green in online), Ilha do Cardoso and other sites (Grey, blue
Lion Tamarin, Azaraes Agouti) that occurred in the high- in online) and Ilhabela and other sites (light grey, red in online) in
land ecotone region. The major difference between species relation to their distances (km).

Animal Conservation  (2016)  2016 The Zoological Society of London 5


Defaunation in the Atlantic forest M. Galetti et al.

Neotropical primate (mean body mass ~12 kg) (Table 1).


Density and biomass estimates
The smallest primate species, Callithrix aurita, was present
The densities of mammals based on line transects varied 16- only in two sites, and although it made a large contribution
fold among sites, from 2.75 to 45.93 ind. km2, whereas to overall density in Cunha-Itamambuca and VG, its contri-
biomass varied 70-fold from 7.9 to 556 kg km2 (Table 1). bution to overall biomass is relatively small (Table 1). In
The species with the highest density was capuchin monkey both islands sampled, the density and biomass of primates
(Sapajus nigritus) at Jureia with 32.04 ind. km2, and the was lower than in continental sites. However, in IC we
lowest density was 0.16 ind. km2 for the Brazilian squirrel recorded the highest density and biomass of seed predators
(Guerlinguetus brasiliensis) at Caraguatatuba (Table 1). (D. leporina, G. brasiliensis and T. pecari; Table 1).
Interestingly both sites were in the Serra do Mar massif,
although the area was just 145 km apart from each other.
Ecological and human variables
The greatest density of primates was found at Jureia Eco-
logical Station, but the highest biomass was documented for The average density and especially the average biomass were
CBHighland, due to high density of muriquis, the largest higher in sites with lightmoderate hunting pressure.

Figure 3 Average (SD) of density and biomass of all survey species and all game (large-bodied) species in relation to hunting impact in 13
study sites the Brazilian Atlantic forest.

6 Animal Conservation  (2016)  2016 The Zoological Society of London


M. Galetti et al. Defaunation in the Atlantic forest

However, density was not signicantly inuenced by hunting The second largest island, IC, had at least two species
(F = 0.01; P = 0.92) or forest type, whereas biomass was locally extinct, the tapir and the jaguar (probably around
affected by hunting (F = 39.09; P < 0.01) and an interaction 1960s based on local interviews). The island is too small to
between hunting and forest type (F = 27.48, P < 0.01), but maintain a population of jaguars and they used to reach the
not by forest type alone (F = 2.73, P = 0.14; Fig. 3). When island crossing a shallow channel. Tapirs probably were resi-
we analyzed game species apart, the same pattern was dent and we recommend reintroducing this megaherbivore in
observed. Density was not related to any predictor variable, the island if hunting and illegal palm harvesting is con-
whereas biomass was related to hunting pressure and an trolled.
interaction between hunting pressure and forest type on bio- The diversity of mammals that we found in the continen-
mass (hunting F = 44.68, P < 0.01, hunting 9 forest type tal forests of Serra do Mar and Paranapiacaba is higher than
F = 29.5, P < 0.01; Fig. 3, Table 2). Lowland forest with in other regions of the Atlantic forest. The absence of local
light hunting pressure presented higher average biomasses extinction along the continental forests contrasts with other
than montane forest with light hunting pressure. well-studied regions in the northern Atlantic forest where up
to 80% of the mid and large mammal species have been
Discussion extirpated (da Silva & Pontes, 2008; Canale et al., 2012;
Pontes et al., 2016). This resulted in extremely low beta-
diversity values among continental sites and a lack of rela-
Species composition
tionship between similarity and geographical distance, which
Our results indicate a high diversity of mammals in the lar- contrasts with previous evidence from tropical areas (e.g.
gest Atlantic forest remnant (from 31 to 38 native species), Condit et al., 2002), especially from tree data. The wide-
with no extinction at the regional level, except in the two spread distribution of most mammal species across the conti-
land-bridge islands. However, the exact species richness of nental forests of Serra do Mar and Paranapiacaba likely
our study areas may be underestimated due to cryptic identi- explains this lack of distance decay relationship.
ties of some species. For instance, recently a new species of
brocket deer (Mazama bororo) and a small felid (Leopardus
Species density and biomass
guttulus) were described from the southeastern Atlantic for-
ests (Gonzalez et al., 2009; Trigo et al., 2013). The results of our census showed that the densities of mam-
The local extinction in the islands is intriguing. Ilhabela is mals varied strongly within continental Atlantic Forest sites,
2.3 bigger than IC (3475 km vs. 151 km2, respectively) and from around 3 mammals km2 (Jurupara) to 46 mam-
presented the lowest diversity of medium- and large-sized mals km2 (VG). Surprisingly the two sites with higher
mammals (eight species). We did not detect even common spe- mammalian density have the heaviest hunting pressure (VG
cies that usually thrive in highly fragmented forests, such as and Jureia). However, the mammalian assemblage of these
raccoon (Procyon cancrivorous) or agoutis (Dasyprocta lepor- areas is composed mostly of small-bodied species (squirrels,
ina). Although we have no historical information on the mam- marmosets and capuchin monkeys) that are less affected by
malian composition of Ilhabela, it is likely that the extensive hunting and it is likely that these smaller species are favored
deforestation for sugar cane plantations in the previous cen- in hunted sites (Peres & Dolman, 2000).
turies, intensive hunting and low immigration from the main- The overall mammalian density recorded in both Paranapi-
land explain this low diversity (Olmos, 1996). This is likely the acaba and Serra do Mar is 1.53.1 times lower than small
cause for the lower values of pairwise compositional similarity, Atlantic forest fragments (Table 3). The high density of pri-
that is, high beta diversity, when one of the pair of sites was mates and ungulates in Atlantic forest fragments seems to
from Ilhabela. Therefore, zoo archeological studies in this reect a crowding effect with expected high home range
island should be necessary for future wildlife reintroductions. overlap among individuals or groups.

Table 2 Summarized results of linear mixed-effects model analyses about the influence of forest type (lowland or highland) and hunting
pressure on total density and biomass of all mammals and game mammals

All mammals Game mammals


numDF denDF F-value P-value F-value P-value
Density
Intercept 1 7 14.09 0.007 14.57 0.006
Forest type 1 7 0.01 0.92 0.91 0.37
Hunting pressure 1 7 0.01 0.91 2.42 0.16
Forest type 9 hunting pressure 1 7 0.84 0.38 0.23 0.64
Biomass
Intercept 1 7 41.93 <0.001 38.10 <0.001
Forest type 1 7 2.73 0.14 2.58 0.15
Hunting pressure 1 7 39.09 <0.001 44.68 <0.001
Forest type 9 hunting pressure 1 7 27.48 0.001 29.55 0.001

Animal Conservation  (2016)  2016 The Zoological Society of London 7


Defaunation in the Atlantic forest M. Galetti et al.

Table 3 Comparison between three continuous Atlantic Forest (AF) regions, Atlantic forest fragments and Amazon forest (density:
ind. km2, biomass: kg km2; athis work, bChiarello, 2000; Chiarello & de Melo, 2001; Cullen et al., 2001; Arau
jo et al., 2008; cHaugaasen &
Peres, 2005)

All medium- and large-


Primates Ungulates sized mammals
Region Density Biomass Density Biomass Density Biomass
a
AF Paranapiacaba massif 9.57 55.72 0.00 0.00 13.50 67.18
AF Serra do Mar massifa 14.58 41.48 0.67 23.45 26.02 75.84
AF Islandsa 5.70 29.40 6.75 237.12 20.13 282.98
AF fragments (<2000 ha)b 22.85 65.30 1.12 29.27 41.28 103.31
AF fragments (200035 000 ha)b 53.45 196.87 4.74 178.20 78.84 401.12
Amazonian terra firmec 122.6 328.9 5.04 208.38 142.6 596.8
Amazonian varzeac 202.1 410.23 0.16 5.76 211.8 454.8

Our analyses showed the mammal density was not signi- biomass, Ilhabela, had 53 times smaller than non-hunted
cantly affected by hunting, whereas the overall crude mam- Amazon sites. The absence of white-lipped peccaries popula-
malian biomass was, and also by combination of forest type tions from most of our sites (84%) was responsible for the
and hunting pressure. Lowland forests with light hunting low biomass in most of our sites, except Itamambuca-Cunha
pressure, in average, presented higher biomass than montane and IC. White lipped peccaries correspond to 8085% of the
forest with same hunting pressure. These results probably whole mammalian biomass in these sites respectively, but
reect positive effects of forest productivity on mammal bio- drops to <34% in slightly hunted Atlantic forest fragments
mass, when hunting was limited (Peres, 2000; Galetti et al., (Cullen et al., 2001). Based on interview with local rangers,
2009). Similar results were obtained in continuous sites of white-lipped peccaries disappear from Paranapiacaba region
Amazonian and fragmented Atlantic forest, where hunting in the late 80s (last seen in 1986), and were not recorded in
had no effect on mammalian density, but on biomass (Chiar- our census. This species was recently rediscovered only in
ello, 2000; Peres, 2000). Thus, Neotropical forests exposed 2011 (Beisiegel, Nakano-Oliveira & Jorge, 2014) and hope-
to heavy hunting pressure show more pronounced change in fully the population will recover if hunting and palm har-
mammalian biomass than density because density compensa- vesting is controlled.
tion processes may be occurring (MacArthur et al., 1972). Although we could not calculate the densities for all
The low density of ungulates and large rodents in most of ungulate species, such as collared peccaries, or deer, the
our study areas may be related to hunting preferences of low encounter rate may represent lower densities too, as the
colonists in the Atlantic forest, who normally do not hunt encounter rates are correlated with density (Hill et al.,
primates, but strongly prefer ground-dwelling frugivorous 1997). The low density of ungulates may result in a
herbivorous mammals (Chiarello, 2000; Cullen, Bodmer & reduced perturbation of the forest oor by trampling and
Valladares-Padua, 2001; Brocardo et al., 2012) and the pre- seed predation and dispersal (Brocardo et al., 2013), which
dation by feral dogs that became highly abundant in most may affect plant composition by increasing the dominance
areas (Galetti & Sazima, 2006; Srbek-Araujo & Chiarello, of some species (Dirzo & Miranda, 1991; Wright, Her-
2008). We were able to obtain robust density estimate for nandez & Condit, 2007; Galetti & Dirzo, 2013). For
only one ungulate, the white-lipped peccaries in only two instance, the foraging behavior of large herds of white-
sites. Our density estimates for one site (Cardoso Island, lipped peccaries feeding on seeds, trampling seedlings and
13.55 ind. km2) is twice the average density reported for rooting of seedling and saplings may have a profound effect
non-hunted forests elsewhere (Keuroghlian et al., 2012). on plant composition (Beck, 2006) and rodent abundance
Based on tracks and latrines found, tapirs are apparently and diversity (Galetti et al., 2015). In addition, the defauna-
common in all continental sites (CB, Intervales, Caraguata- tion of large seed dispersers, such as the primates and
tuba; Norris et al., 2012), but the method that we used (diur- agouties, may also represent a threat for the dispersal of
nal line transect) is not appropriate to estimate their density. large-seeded species. These species play a fundamental role
The low density of primates, especially muriquis, can be as seed dispersers and their local extirpation may translate
related to historic and current hunting pressure by traditional into the lower recruitment and dispersal distances, leading
populations (Indians, Quilombolas and Caicaras), which do to long-term declines in plant diversity (Galetti et al., 2006;
not have taboo to eat primates (Galetti et al., 2009). Arroyo-Rodrguez et al., 2007; Donatti, Guimaraes &
Changes in the Atlantic Forest habitats (due selective log- Galetti, 2009; Stevenson, 2011; Bueno et al., 2013) with
ging, palm harvest, roads) coupled with extensive human potential consequences for carbon storage (Bello et al.,
density, and consequently high levels of hunting may be the 2015).
reason why our sites have low mammalian density and bio- Because most of the Atlantic forest is restricted to small
mass (Galetti et al., 2009; Table 3). For example, the mean patches, a lot of effort and resources have been invested to
biomass of Paranapiacaba massif is seven times smaller than restore and connect isolated forest patches (Melo et al.,
some Amazonia terra rme forests. The site with lowest 2013). However, the decline in animal populations is evident

8 Animal Conservation  (2016)  2016 The Zoological Society of London


M. Galetti et al. Defaunation in the Atlantic forest

even in the large forest patches such as the Central Atlantic Conservation Units of the State of Rio de Janeiro, Brazil.
Corridor in south Bahia with 50 000 ha of protected area Iher Ser. Zool. 98, 391396.
(Canale et al., 2012), the Misiones Corridor with Arroyo-Rodrguez, V., Aguirre, A., Bentez-Malvido, J. &
412 000 ha (De Angelo, Paviolo & Di Bitetti, 2011) and the Mandujano, S. (2007). Impact of rain forest fragmentation
Serra do Mar subregion. It is estimated that 96% of the on the population size of a structurally important palm
entire Atlantic forest biome is likely to be under some species: Astrocaryum Mexicanum at Los Tuxtlas, Mexico.
effects of trophic cascades due to mammalian defaunation Biol. Conserv. 138, 198206.
(Jorge et al., 2013). Azhar, B., Lindenmayer, D.B., Wood, J., Fischer, J. &
The Serra do Mar and Paranapiacaba region has been Zakaria, M. (2014). Ecological impacts of oil palm
pointed out as the main stronghold for several endemic spe- agriculture on forest mammals in plantation estates and
cies (Brachyteles arachnoides, Callithrix aurita, Mazama
smallholdings. Biodivers. Conserv. 23, 11751191.
bororo, Leopardus guttulus) or species that have vanished
Barlow, J., Gardner, T.A., Araujo, I.S., Avila-Pires, T.C.,
from most of the Atlantic forest (Tayassu peccari, Speothos
Bonaldo, A.B., Costa, J.E., Esposito, M.C., Ferreira, L.V.,
venaticus, Panthera onca) (Duarte et al., 2012; Keuroghlian
Hawes, J., Hernandez, M.M., Hoogmoed, M.S., Leite, R.N.,
et al., 2012; Galetti et al., 2013; Ingberman et al., 2016).
However, our data show that even in this large forest block Lo-Man-Hung, N.F., Malcolm, J.R., Martins, M.B., Mestre,
their population size may be too low to ensure their long- L.A.M., Miranda-Santos, R., Nunes-Gutjahr, A.L., Overal,
term persistence. Therefore, the main conservation strategies W.L., Parry, L., Peters, S.L., Ribeiro-Junior, M.A., da Silva,
for these regions are to promote effective conservation by M.N.F., Motta, C.D. & Peres, C.A. (2007). Quantifying the
reducing illegal hunting, palm harvesting and logging. These biodiversity value of tropical primary, secondary, and
strategies also involve hiring and empowering forest guards plantation forests. Proc. Natl Acad. Sci. 104, 1855518560.
to ght against well-structured militias of palm harvesters Barnosky, A.D., Matzke, N., Tomiya, S., Wogan, G.O.,
and hunters who are becoming common in Serra do Mar. Swartz, B., Quental, T.B., Marshall, C., McGuire, J.L.,
These war-zones of conservation are usually omitted by Lindsey, E.L. & Maguire, K.C. (2011). Has the earths sixth
decision makers who ignore that hunting is a widespread mass extinction already arrived? Nature 471, 5157.
practice in protected areas and by many conservation biolo- Beaudrot, L., Ahumada, J.A., OBrien, T., Alvarez-Loayza, P.,
gists who deny the clear global defaunation trend of mam- Boekee, K., Campos-Arceiz, A., Eichberg, D., Espinosa, S.,
mal populations in tropical forests (Beaudrot et al., 2016). In Fegraus, E., Fletcher, C., Gajapersad, K., Hallam, C.,
fact, the ongoing defaunation scenario in the largest Atlantic Hurtado, J., Jansen, P.A., Kumar, A., Larney, E., Lima,
forest remnant may be exacerbating due to the recent eco- M.G.M., Mahony, C., Martin, E.H., McWilliam, A.,
nomical crisis in Brazil. For instance, about 30% of all forest Mugerwa, B., Ndoundou-Hockemba, M.,
guards (rangers) were dismissed from the protected parks in Razamahaimodison, J.C., Romero-Saltos, H., Rovero, F.,
S~ao Paulo state in the last 4 years. Therefore, the future of Salvador, J., Santos, F., Sheil, D., Spironello, W.R., Willig,
the mammalian diversity in the largest Atlantic forest rem-
M.R., et al. (2016). Standardized assessment of biodiversity
nant may soon mirror the impoverished rainforests world-
trends in tropical forest protected areas: the end is not in
wide.
sight. PLoS Biol. 14, e1002357.
Beck, H. (2006). A review of peccary-palm interactions and
their ecological ramications across the neotropics. J.
Acknowledgements Mammal. 87, 519530.
This work was supported by the Fundac~ao de Amparo a Beisiegel, B., Nakano-Oliveira, E. & Jorge, M.L.S.P. (2014).
Pesquisa do Estado de S~ao Paulo (FAPESP) through the pro- Are white-lipped peccaries back in the Paranapiacaba Forest,
cess BIOTA FAPESP 2007/03392-6 and 2014/01986-0. S~ao Paulo, Brazil? Suiform Sound 29, 2934.
We thank the directors of the protected areas and Fundac~ao Bello, C., Galetti, M., Pizo, M.A., Magnago, L.F.S., Rocha,
and Instituto Florestal do Estado de S~ao Paulo for allowing M.F., Lima, R.A.F., Peres, C.A., Ovaskainen, O. & Jordano,
us to carry out this study. We thank S. Nazareth for assis- P. (2015). Defaunation affects carbon storage in tropical
tance in the eld. C.R.B., M.C.R. and M.G. receive a CNPq forests. Sci Adv 1, e1501105.
fellowship (Conselho Nacional para o Desenvolvimento Bernardo, C.S.S. & Galetti, M. (2004). Densidade e tamanho
Cientco e Tecnologico). R.A.F.L. received a CAPES fel- populacional de primatas em um fragmento orestal no
lowship (Coordenac~ao de Aperfeicoamento de Pessoal de sudeste do brasil. Revista Brasileira de Zoologia 21, 827
Nvel Superior). To Paulo Rubim and Carla Stefer for help 832.
in the eld work.
Bogoni, J.A., Cherem, J.J., Giehl, E.L.H., Oliveira-Santos,
L.G., de Castilho, P.V., Picinatto Filho, V., Fantacini, F.M.,
References Tortato, M.A., Luiz, M.R. & Rizzaro, R. (in press).
Landscape features lead to shifts in communities of
Araujo, R.M.D., Souza, M.B.D. & Ruiz-Miranda, C.R. (2008). medium-to large-bodied mammals in subtropical Atlantic
Density and population size of game mammals in two Forest. J. Mammal.

Animal Conservation  (2016)  2016 The Zoological Society of London 9


Defaunation in the Atlantic forest M. Galetti et al.

Bovendorp, R.S. & Galetti, M. (2007). Density and population possible consequences of contemporary defaunation. In
size of mammals introduced on a land-bridge island in Plant-animal interactions: evolutionary ecology in tropical
southeastern Brazil. Biol. Invasions 9, 353357. and temperate regions: 273287. Lewinsohn, T.M.,
Brocardo, C.R., Rodarte, R., Bueno, R.D.S., Culot, L. & Fernandes, G.W. & Benson, W.W. (Ed). New York: Wiley.
Galetti, M. (2012). Mamferos n~ao voadores do Parque Dirzo, R., Young, H.S., Galetti, M., Ceballos, G., Isaac, N.J.
Estadual Carlos Botelho, continuum orestal do & Collen, B. (2014). Defaunation in the Anthropocene.
Paranapiacaba. Biota Neotrop. 12, 198208. Science 345, 401406.
Brocardo, C.R., Zipparro, V.B., de Lima, R.A., Guevara, R. & Donatti, C.I., Guimaraes, P.R. Jr & Galetti, M. (2009). Seed
Galetti, M. (2013). No changes in seedling recruitment dispersal and predation in the endemic Atlantic rainforest
when terrestrial mammals are excluded in a partially palm Astrocaryum aculeatissimum across a gradient of seed
defaunated Atlantic rainforest. Biol. Conserv. 163, disperser abundance. Ecol. Res. 24, 11871195.
107114. Duarte, J.M.B., Vogliotti, A., dos Santos Zanetti, E., de
Bueno, R.S., Guevara, R., Ribeiro, M.C., Culot, L., Bufalo, Oliveira, M.L., Tiepolo, L.M., Rodrigues, L.F. & de
F.S. & Galetti, M. (2013). Functional redundancy and Almeida, L.B. (2012). Avaliac~ao do risco de extinc~ao do
complementarities of seed dispersal by the last neotropical veado-mateiro-pequeno Mazamo bororo Duarte, 1996, no
megafrugivores. PLoS ONE 8, e56252. Brasil. Biodiversidade Brasileira 1, 4249.
Burnham, K.P., Anderson, D.R. & Laake, J.L. (1980). Fa, J.E., Peres, C.A. & Meeuwig, J. (2002). Bushmeat
Estimation of density from line transect sampling of exploitation in tropical forests: an intercontinental
biological populations. Wildl. Monogr. 24, 3202. comparison. Conserv. Biol. 16, 232237.
Canale, G.R., Peres, C.A., Guidorizzi, C.E., Gatto, C.A.F. & Galetti, M. & Dirzo, R. (2013). Ecological and evolutionary
Kierulff, M.C.M. (2012). Pervasive defaunation of forest consequences of living in a defaunated world. Biol.
remnants in a tropical biodiversity hotspot. PLoS ONE 7, Conserv. 163, 16.
e41671. Galetti, M. & Fernandez, J.C. (1998). Palm heart harvesting in
Ceballos, G. & Ehrlich, P.R. (2002). Mammal population the Brazilian Atlantic forest: changes in industry structure
losses and the extinction crisis. Science 296, 904907. and the illegal trade. J. Appl. Ecol. 35, 294301.
Ceballos, G., Ehrlich, P.R., Barnosky, A.D., Garca, A., Galetti, M. & Sazima, I. (2006). Impact of feral dogs in an
Pringle, R.M. & Palmer, T.M. (2015). Accelerated modern urban Atlantic forest fragment in southeastern Brazil. Nat e
humaninduced species losses: entering the sixth mass Conserv 4, 146151.
extinction. Sci. Adv. 1, e1400253. Galetti, M., Donatti, C.I., Pires, A.S., Guimar~aes, P.R. &
Chiarello, A.G. (1999). Effects of fragmentation of the Jordano, P. (2006). Seed survival and dispersal of an
Atlantic forest on mammal communities in south-eastern endemic Atlantic forest palm: the combined effects of
Brazil. Biol. Conserv. 89, 7182. defaunation and forest fragmentation. Bot. J. Linn. Soc. 151,
Chiarello, A.G. (2000). Density and population size of 141149.
mammals in remnants of Brazilian Atlantic forest. Conserv. Galetti, M., Giacomini, H.C., Bueno, R.S., Bernardo, C.S.S.,
Biol. 14, 16491657. Marques, R.M., Bovendorp, R.S., Stefer, C.E., Rubim, P.,
Chiarello, A.G. & de Melo, F.R. (2001). Primate population Gobbo, S.K., Donatti, C.I., Begotti, R.A., Meirelles, F.,
densities and sizes in Atlantic forest remnants of northern Nobre, R.D.A., Chiarello, A.G. & Peres, C.A. (2009).
Espirito Santo, Brazil. Int. J. Primatol. 22, 396. Priority areas for the conservation of Atlantic forest large
Condit, R., Pitman, N., Leigh, E.G., Chave, J., Terborgh, J., mammals. Biol. Conserv. 142, 12291241.
Foster, R.B., Nu~nez, P., Aguilar, S., Valencia, R., Villa, G., Galetti, M., Eizirik, E., Beisiegel, B., Ferraz, K.M.P.,
Muller-Landau, H.C., Losos, E. & Hubbell, S.P. (2002). Cavalcanti, S., Srbek-Araujo, A.C., Crawshaw, P., Paviolo,
Beta-diversity in tropical forest trees. Science 295, A., Galetti, P.M. Jr, Jorge, M.L., Marinho-Filho, J.,
666669. Vercillo, U. & Morato, R. (2013). Atlantic rainforests
Cullen, L., Bodmer, R.E. & Padua, C.V. (2000). Effects of jaguars in decline. Science 342, 930.
hunting in habitat fragments of the Atlantic forests, Brazil. Galetti, M., Guevara, R., Neves, C.L., Rodarte, R.R.,
Biol. Conserv. 95, 4956. Bovendorp, R.S., Moreira, M., Hopkins, J.B. & Yeakel, J.D.
Cullen, L., Bodmer, E.R. & Valladares-Padua, C. (2001). (2015). Defaunation affects the populations and diets of
Ecological consequences of hunting in Atlantic forest rodents in Neotropical rainforests. Biol. Conserv.
patches, S~ao Paulo, Brazil. Oryx 35, 137144. 190, 27.
De Angelo, C., Paviolo, A. & Di Bitetti, M. (2011). Gonzalez, S., Maldonado, J.E., Ortega, J., Talarico, A.C.,
Differential impact of landscape transformation on pumas Bidegaray-Batista, L., Garcia, J.E., Duarte, J. & BarbantI,
(Puma concolor) and jaguars (Panthera onca) in the Upper M. (2009). Identication of the endangered small red
Parana Atlantic Forest. Divers. Distrib. 17, 422436. brocket deer (Mazama bororo) using noninvasive genetic
Dirzo, R. & Miranda, A. (1991). Altered patterns of herbivory techniques (Mammalia; Cervidae). Mol. Ecol. Resour. 9,
and diversity in the forest understory: a case study of the 754758.

10 Animal Conservation  (2016)  2016 The Zoological Society of London


M. Galetti et al. Defaunation in the Atlantic forest

Gonzalez-Maya, J.F., Vquez-R, L.R., Belant, J.L. & Ceballos, Myers, N., Mittermeier, R.A., Mittermeier, C.G., Da Fonseca,
G. (2015). Effectiveness of protected areas for representing G.A. & Kent, J. (2000). Biodiversity hotspots for
species and populations of terrestrial mammals in Costa conservation priorities. Nature 403, 853858.
Rica. PLoS ONE 10, e0124480. Norris, D., Rocha-Mendes, F., Marques, R., de Almeida
Haugaasen, T. & Peres, C.A. (2005). Primate assemblage Nobre, R. & Galetti, M. (2011). Density and spatial
structure in Amazonian ooded and unooded forests. Am. distribution of buffy-tufted-ear marmosets (Callithrix aurita)
J. Primatol. 67, 243258. in a continuous Atlantic Forest. Int. J. Primatol. 32, 811
Hill, K., Padwe, J., Bejyvagi, C., Bepurangi, A., Jakugi, F., 829.
Tykuarangi, R. & Tykuarangi, T. (1997). Impact of hunting Norris, D., Ramrez, J.M., Zacchi, C. & Galetti, M. (2012). A
on large vertebrates in the Mbaracayu Reserve, Paraguay. survey of mid and large bodied mammals in Nucleo
Conserv. Biol. 11, 13391353. Caraguatatuba, Serra do Mar State Park, Brazil. Biota
Ingberman, B., Fusco-Costa, R. & Monteiro-Filho, E.L. 2016. Neotrop. 12, 127133.
A current perspective on the historical geographic Novaro, A.J., Funes, M.C. & Walker, R.S. (2000). Ecological
distribution of the endangered muriquis (Brachyteles spp.): extinction of native prey of a carnivore assemblage in
implications for conservation. PLoS ONE, 11: e0150906. Argentine Patagonia. Biol. Conserv. 92, 2533.
Johns, A.D. (1986). Effects of selective logging on the Oliveira-Filho, A.T. & Fontes, M.A.L. (2000). Patterns of
behavioral ecology of West Malaysian primates. Ecology oristic differentiation among Atlantic Forests in
67, 684694. southeastern Brazil and the inuence of Climate1.
Jorge, M.L.S., Galetti, M., Ribeiro, M.C. & Ferraz, K.M.P. Biotropica 32, 793810.
(2013). Mammal defaunation as surrogate of trophic Olmos, F. (1996). Missing species in S~ao Sebasti~ao island,
cascades in a biodiversity hotspot. Biol. Conserv. 163, 49 southeastern Brazil. Papeis Avulsos Zoologia, S~ao Paulo 18,
57. 329349.
Keuroghlian, A., Desbiez, A.L.J., Beisiegel, B.M., Medici, Paglia, A.P., da Fonseca, G.A., Rylands, A.B., Herrmann, G.,
E.P., Gatti, A., Pontes, A.R.M., Campos, C.B., Tofoli, C.F., Aguiar, L.M., Chiarello, A.G., Leite, Y.L.R., Leonora, P.C.,
Moraes, E.A. Jr, de Azevedo, F.C., de Pinho, G.M., Siciliano, S., Kierulff, M.C.M., Mendes, S.L., Tavares,
Cordeiro, J.L.P., Santos-Junior, T.S., de Morais, A.A., V.C., Mittermier, R.A. & Patton, J.L. (2012). Lista anotada
Mangini, P.R., Flesher, K., Rodrigues, L.F. & de Almeida, dos mamferos do Brasil 2 Edic~ao Annotated checklist of
L.B. (2012). Avaliac~ao do risco de extinc~ao do queixada Brazilian mammals. Occasional Papers in Conservation
Tayassu pecari Link, 1795, no Brasil. Biodiversidade Biology 6, 76 pp.
Brasileira 1, 84102. Passamani, M. (2008). Density and group size of primates in
Koleff, P., Gaston, K.J. & Lennon, J.J. (2003). Measuring beta an Atlantic hill forest of southwest Espirito Santo, Brazil.
diversity for presenceabsence data. J. Anim. Ecol. 72, 367 Rev Brasil de Zooc. 10, 2934.
382. Peres, C.A. (1999). General guidelines for standardizing line-
Leite, Y.L., Costa, L.P., Loss, A.C., Rocha, R.G., Batalha- transect surveys of tropical forest primates. Neotropical
Filho, H., Bastos, A.C., Quaresma, V.S., Fagundes, V., primates 7, 1116.
Paresque, R., Passamani, M. & Pardini, R. (2016). Peres, C.A. (2000). Effects of subsistence hunting on
Neotropical forest expansion during the last glacial period vertebrate community structure in Amazonian forests.
challenges refuge hypothesis. Proc. Natl Acad. Sci. 113, Conserv. Biol. 14, 240253.
10081013. Peres, C.A. & Dolman, P.M. (2000). Density compensation in
MacArthur, R.H., Diamond, J.M. & Karr, J.R. (1972). Density neotropical primate communities: evidence from 56 hunted
compensation in island faunas. Ecology 53, 330342. and nonhunted Amazonian forests of varying productivity.
McConkey, K.R. & OFarrill, G. (2015). Cryptic function loss Oecologia 122, 175189.
in animal populations. Trends Ecol. Evol. 30, Peres, C.A. & Palacios, E. (2007). Basin-wide effects of game
182189. harvest on vertebrate population densities in Amazonian
Meijaard, E., Sheil, D., Nasi, R. & Stanley, S.A. (2007). forests: implications for animal-mediated seed dispersal.
Wildlife conservation in Bornean timber concessions. Ecol. Biotropica 39, 304315.
Soc. 11, 47. Pinheiro, J., Bates, D., DebRoy, S. & Sarkar, D. (2007).
Melo, F.P., Pinto, S.R., Brancalion, P.H., Castro, P.S., Linear and nonlinear mixed effects models. R package
Rodrigues, R.R., Aronson, J. & Tabarelli, M. (2013). version 3. Vienna: R Foundation for Statistical
Priority setting for scaling-up tropical forest restoration Computing.
projects: early lessons from the Atlantic Forest Restoration Pontes, A.R.M., Beltr~ao, A.C.M., Normande, I.C., Malta,
Pact. Environ. Sci. Policy 33, 395404. A.D.J.R., da Silva Junior, A.P. & Santos, A.M.M. (2016).
Miotto, R.A., Cervini, M., Begotti, R.A. & Galetti, P.M. Jr (2012). Mass extinction and the disappearance of unknown mammal
Monitoring a puma (Puma concolor) population in a fragmented species: scenario and perspectives of a biodiversity hotspots
landscape in southeast Brazil. Biotropica 44, 98104. hotspot. PLoS ONE 11, e0150887.

Animal Conservation  (2016)  2016 The Zoological Society of London 11


Defaunation in the Atlantic forest M. Galetti et al.

Price, E.C., Piedade, H.E.L.M. & Wormell, D. (2002). data reveal complex hybridization and a cryptic species of
Population densities of primates in a Brazilian Atlantic Neotropical wild cat. Curr. Biol. 23, 25282533.
forest. Folia Primatol. 73, 5456. Valiente-Banuet, A., Aizen, M.A., Alcantara, J.M., Arroyo, J.,
R Core Team. (2015). R: a language and environment for Cocucci, A., Galetti, M., Garcia, M.B., Garcia, D., Gomez,
statistical computing. Vienna: R Foundation for Statistical J.M., Jordano, P., Medel, R., Navarro, L., Obeso, J.R., Oviedo,
Computing. Available at: http://www.R-project.org/ R., Ramrez, N., Rey, P.J., Traveset, A., Vedu, M. & Zamora,
Ribeiro, M.C., Metzger, J.P., Martensen, A.C., Ponzoni, F.J. & R. (2015). Beyond species loss: the extinction of ecological
Hirota, M.M. (2009). The Brazilian Atlantic Forest: how much interactions in a changing world. Funct. Ecol. 29,
is left, and how is the remaining forest distributed? 299307.
Implications for conservation. Biol. Conserv. 142, 11411153. Wright, S.J., Hernandez, A. & Condit, R. (2007). The
Rodrigues, A.S.L., Andelman, S.J., Bakarr, M.I., Boitani, L., bushmeat harvest alters seedling banks by favoring lianas,
Brooks, T.M., Cowling, R.M., Fishpool, L.D.C., da large seeds, and seeds dispersed by bats, birds, and wind.
Fonseca, G.A.B., Gaston, K.J., Hoffmann, M., Long, J.S., Biotropica 39, 363371.
Marquet, P.A., Pilgrim, J.D., Pressey, R.L., Schipper, J., Young, H.S., McCauley, D.J., Galetti, M. & Dirzo, R. (2016).
Sechrest, W., Stuart, S.N., Underhill, L.G., Waller, R.W., Patterns, causes, and consequences of anthropocene
Watts, M.E.J. & Yan, X. (2004). Effectiveness of the global defaunation. Annu. Rev. Ecol. Evol. Syst. 47,
protected area network in representing species diversity. 333358.
Nature 428, 640643.
Rylands, A.B. & Keuroghlian, A. (1988). Primate populations
in continuous forest and forest fragments in central Supporting information
Amazonia. Acta Amazon 18, 291307.
da Silva, A.P. Jr & Pontes, A.R.M. (2008). The effect of a Additional Supporting Information may be found in the
mega-fragmentation process on large mammal assemblages online version of this article at the publishers web-site:
in the highly-threatened Pernambuco Endemism Centre,
north-eastern Brazil. Biodivers. Conserv. 17, 14551464. Figure S1. Mean rarefaction curve showing the upper
Srbek-Araujo, A.C. & Chiarello, A.G. (2008). Domestic dogs bounds of the 95% CI of the expected number of mid-and
in Atlantic forest preserves of south-eastern Brazil: a large-sized mammals recorded with camera traps in Itamam-
buca (IT) and Vargem Grande (VG), Carlos Botelho (BH)-
camera-trapping study on patterns of entrance and site
Highland, Ilha do Cardoso (IC).
occupancy rates. Brazil J of Biol 68, 771779.
Table S1. Physical attributes and sample size of medium
Stevenson, P.R. (2011). The abundance of large Ateline
and large sized mammals in 13 sites long the of Serra do
monkeys is positively associated with the diversity of plants
Mar subregion.
regenerating in Neotropical forests. Biotropica 43, 512519. Table S2. Species occurrence of medium and large sized
Suguio, K. & Martin, L. (1987). Classicac~ao de costas e mammals in each sampling region in the continuous Atlantic
evoluc~ao geologica das plancies litor^aneas quaternarias do Forest, based on tracks, sightings, interviews and literature
sudeste e sul do Brasil. Simposio sobre Ecossistemas da (NR, not recorded in line transect survey; NO, species does
Costa Sul e Sudeste Brasileira 1, 128. not occur historically; O, species observed; Ex, probably
Thomas, L., Buckland, S.T., Rexstad, E.A., Laake, J.L., extinct).
Strindberg, S., Hedley, S.L., Bishop, J.R.B., Marques, T.A. Table S3. Terrestrial mammals recorded by camera trapping
& Burnham, K.P. (2010). Distance software: design and (IT, Itamambuca; VG, Vargem Grande; BH, Carlos Botelho-
analysis of distance sampling surveys for estimating Highland; IC, Cardoso Island).
population size. J. Appl. Ecol. 47, 514.
Trigo, T.C., Schneider, A., de Oliveira, T.G., Lehugeur, L.M.,
Silveira, L., Freitas, T.R. & Eizirik, E. (2013). Molecular

12 Animal Conservation  (2016)  2016 The Zoological Society of London