Anuran assemblage composition and distribution at a modified environment in Três Marias reservoir, south-eastern Brazil

This article was downloaded by: [Fondren Library, Rice University ]On: 01 October 2012, At: 14:11Publisher: Taylor & FrancisInforma Ltd Registered in England and Wales Registered Number: 1072954 Registeredoffice: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Journal of Natural HistoryPublication details, including instructions for authors andsubscription information:http://www.tandfonline.com/loi/tnah20

Anuran assemblage composition anddistribution at a modified environmentin Três Marias reservoir, south-easternBrazilPriscilla Ferreira Torres a & Paula Cabral Eterovick aa Programa de Pós Graduação em Zoologia de Vertebrados,Pontifícia Universidade Católica de Minas Gerais, 30535-610, BeloHorizonte, Minas Gerais, Brazil

Version of record first published: 20 Oct 2010.

To cite this article: Priscilla Ferreira Torres & Paula Cabral Eterovick (2010): Anuran assemblagecomposition and distribution at a modified environment in Três Marias reservoir, south-easternBrazil, Journal of Natural History, 44:43-44, 2649-2667

To link to this article: http://dx.doi.org/10.1080/00222933.2010.501529

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions

This article may be used for research, teaching, and private study purposes. Anysubstantial or systematic reproduction, redistribution, reselling, loan, sub-licensing,systematic supply, or distribution in any form to anyone is expressly forbidden.

The publisher does not give any warranty express or implied or make any representationthat the contents will be complete or accurate or up to date. The accuracy of anyinstructions, formulae, and drug doses should be independently verified with primarysources. The publisher shall not be liable for any loss, actions, claims, proceedings,demand, or costs or damages whatsoever or howsoever caused arising directly orindirectly in connection with or arising out of the use of this material.

http://www.tandfonline.com/loi/tnah20

http://dx.doi.org/10.1080/00222933.2010.501529

http://www.tandfonline.com/page/terms-and-conditions

http://www.tandfonline.com/page/terms-and-conditions

Journal of Natural HistoryVol. 44, Nos. 43–44, November 2010, 2649–2667

ISSN 0022-2933 print/ISSN 1464-5262 online© 2010 Taylor & FrancisDOI: 10.1080/00222933.2010.501529http://www.informaworld.com

TNAH0022-29331464-5262Journal of Natural History, Vol. 1, No. 1, Jul 2010: pp. 0–0Journal of Natural HistoryAnuran assemblage composition and distribution at a modified environment in Três Marias reservoir, south-eastern BrazilJournal of Natural HistoryP.F. Torres and P.C. EterovickPriscilla Ferreira Torres and Paula Cabral Eterovick*

Programa de Pós Graduação em Zoologia de Vertebrados, Pontifícia Universidade Católica de Minas Gerais, 30535-610, Belo Horizonte, Minas Gerais, Brazil

(Received 24 November 2009; final version received 18 May 2010)

We assessed the influence of environmental features on anuran species distributionin Pirapitinga, an island formed by the filling of a reservoir in south-easternBrazil. We investigated whether the species exhibited preferences towards certainhabitat types and the existence of niche partitioning. We recorded 474 individualsfrom seven species. Although considered as generalists, these species were notdistributed randomly, but according to specific habitat preferences that happenedto be fulfilled by the shores of the reservoir. Niche overlaps were higher thanexpected by chance during the wet season, when most species were active, but notdifferent from random during the dry season. Environmental variables seemed tobe important and promote aggregation of several anuran species at the samehabitats. Interspecific signals may be used for this purpose, and competition isprobably small or less important than the benefits provided by choosing the mostsuitable habitats in this impacted environment.

Keywords: Anura; reservoir; anthropogenic impact; habitat preference; niche

Introduction

Threats to amphibian conservation in Brazil include habitat destruction or alterationas a result of deforestation, agriculture expansion, mining, illegal fires and urbanization(Silvano and Segalla 2005). Habitat degradation and fragmentation are among themain causes of amphibian declines worldwide (Young et al. 2001; Becker et al. 2007).Although the construction of hydroelectric plants constitutes a more localized impact,the impact is no less important, can also be detrimental to anuran communities and iswidespread in Brazil (Brandão and Araújo 2008). Anthropogenic changes at the land-scape level may alter the availability of bodies of water, ultimately impoverishinganuran assemblages (Ernst and Rödel 2008). Reservoirs constitute an extreme example;they transform lotic environments into lentic bodies of water, thus representing a pro-found change in habitat for anurans. However, some species persist, and their abilitiesto use the modified habitats available may play a key role in their survival, althoughthere is little information available on the topic. Considering that a great number ofamphibian species are threatened with extinction worldwide (Stuart et al. 2004), there isan increasing need for studies that focus on the effects of anthropogenic disturbance onamphibian community structure and composition (Ernst and Rödel 2008).

Amphibians respond to a multitude of environmental variables, such as forestcover (Herrmann et al. 2005), canopy cover and leaf litter depth (McLeod and Gates

*Corresponding author. Email: [email protected]

Dow

nloa

ded

by [

Fond

ren

Lib

rary

, Ric

e U

nive

rsity

] a

t 14:

11 0

1 O

ctob

er 2

012

2650 P.F. Torres and P.C. Eterovick

1998), topographic and edaphic variables (Menin et al. 2007), vegetation type andmicroclimate (Parris 2004). Thus, changes in such variables are likely to inducechanges in anuran distribution and maybe even exclude some species from alteredareas (e.g. Herrmann et al. 2005). Patterns of amphibian species distribution andabundance are dependent on environmental conditions including humidity, rain-fall, temperature, photoperiod, available nutrients, and habitat structure (Austinet al. 1994; Saenz et al. 2006; Both et al. 2008). Patterns may also reflect, at somelevel, effects of biotic processes such as competition, predation, diseases, andmigration (Parris 2004; Ernst and Rödel 2008). However, environmental factorsseem to assume greater importance under several circumstances (e.g. Parris 2004),including harsh climates (e.g. Van Buskirk 2005) and anthropogenic impact (Ernstand Rödell 2005). Thus, relationships between species, as well as relationshipsbetween species and their environment, are reflected in distribution and abun-dance, in time and space (MacNally 1979; Brown et al. 1995; Canavero et al.2009).

Distribution patterns can be described through the relationship between theproportion of available spaces and their effective use (Mackenzie et al. 2002) definingthe niche spatial dimension and its role in species co-existence. According to classicalniche theory, species can differ in three major niche dimensions: food, habitat andactivity period (Hutchinson 1957; Pianka 1973; Schoener 1974), which can be furtherdivided into six: macrohabitat, microhabitat, food type, food size, daily activity andseasonal activity periods (Schoener 1974). For amphibians, habitat has been consid-ered as the main dimension partitioned, followed by food and activity period(Schoener 1974; Toft 1985). Habitat differential use can be influenced by morphology,physiology, natural history or the interactions between species (Caldwell 1996), andthe response of the organisms in different environments is an essential component ofthe niche (Colwell and Futuyma 1971).

Anthropogenic changes are expected to make modified habitats harsher onamphibians than their original natural habitats, thus increasing the importance ofenvironmental factors in limiting the ability of species to survive locally (Ernst andRödel 2005). We aimed to assess the composition of a potentially impoverishedanuran fauna in the drastically modified landscape of a reservoir island and under-stand how the prevailing conditions shaped surviving anuran assemblages. We studiedhabitat preferences of the anuran species that occur at the Pirapitinga EcologicalStation, an island formed by the filling of Três Marias reservoir, 50 years ago, in themunicipality of Morada Nova de Minas in Minas Gerais state, south-eastern Brazil.If environmental features are important in determining species distribution, weexpected to find species that are adapted to use the available habitats, being distributedin a non-random way. We expected species able to survive in this human alteredenvironment to have broad niches and to be opportunistic, thus tolerating great nicheoverlap, given the low habitat heterogeneity of human altered environments. Wecompared niche overlap among co-existing species both in the wet and in the dryseason. We hypothesized that environmental factors would be more important thanbiotic interactions, thus species niches should show greater overlap in the period ofgreater activity of most species (i.e. the wet season). Alternatively, greater nicheoverlap could also indicate positive biotic interactions, with species with similarmicrohabitat preferences attracting each other to suitable breeding sites (Pupin et al.2007; Seppänen et al. 2007).

Dow

nloa

ded

by [

Fond

ren

Lib

rary

, Ric

e U

nive

rsity

] a

t 14:

11 0

1 O

ctob

er 2

012

Journal of Natural History 2651

Methods

Study site

The Ecological Station (ESEC) of Pirapitinga (18° 20′–18° 23′ S and 45° 17′–45° 20′ W),located in the municipality of Morada Nova de Minas, is the only Federal ESEC inMinas Gerais state (Figure 1). It encompasses a land-bridge island formed during thecreation of the Três Marias reservoir by damming of the São Francisco River forelectricity generation in 1960. The Station was established in 1984 and legally recog-nized in 1987, to preserve the remaining cerrado of São Francisco River, present inthe area before creation of the reservoir (Carolsfeld and Johnsen 2005). The ESECPirapitinga has an area of about 1000 ha which varies with the level of the reservoir.When the reservoir is full, it attains a maximum depth of 568 m, when Pirapitingabecomes an island. At its minimum depth of 550 m, the island assumes peninsularcharacteristics, its southern portion connecting to the adjacent margins of the reser-voir for a short period in most years. In Pirapitinga, varied formations of the cerradobiome can be observed, including “cerradão”, “cerrado sensu stricto” and “camposujo”, from the most closed to the most open formation. Trees can reach up to 20 mheight (Azevedo et al. 1987). The average annual temperature of the area varies from20 to 22°C and the average annual precipitation is 1600 mm (Goodland and Ferri1979; Eiten 1993). The first studies conducted in Pirapitinga focused on vegetation(Azevedo et al. 1987), galling insects (Gonçalves-Alvim and Fernandes 2001) andinsect-plant interactions (Cornelissen and Fernandes 2001; Maia and Fernandes2005). The vertebrate fauna, although more likely affected by the landscape-levelchanges in the area, remains little known.

Sampling procedures

We sampled anurans on the island shores, since preliminary searches and observationsshowed that they were restricted to these habitats, probably due to the absence ofwater bodies within the island. We visually searched for frogs and potential breedinghabitats walking across the island, always inspecting lower sections of the relief thatmight store water or increased humidity. Although we did not find any anuransinland, they may occasionally use habitats beyond the shores. However, constantavailability of water is restricted to the shores, so anurans are unlikely to avoid thesehabitats year-round and our method should detect all the species on the island. Weconsidered as “shore” the area available between the water line and the inlandoriginal vegetation (cerrado formations beyond the area affected by variations of thewater level). We first described the whole island shore extension by walking around itand placing wood poles with reflective bands (that could be later located from a boat,at night) every 2 km, resulting in 13 points (Figure 1). At each of these points werecorded geographic coordinates and characterized the type of shore substrate (mudor rocks) and vegetation presence or absence. Shore vegetation, when present, wascomposed of herbs and some sparse shrubs (Figure 2). We marked 26 additionalpoints 300 m away at each side of the 13 original points, resulting in a total of 39points that we used for anuran sampling.

Every month, from October 2006 to October 2007, we randomly picked one sam-pling point within each of the four habitat types recorded: (1) mud with vegetation(MV); (2) mud without vegetation (M); (3) rocks with vegetation (RV); and (4) rocks

Dow

nloa

ded

by [

Fond

ren

Lib

rary

, Ric

e U

nive

rsity

] a

t 14:

11 0

1 O

ctob

er 2

012


without vegetation (R; Figure 2) to sample anurans. At each sampling point, wemarked 16 m2 quadrats within which we recorded all anurans present. Thus our sam-ples were standardized per habitat and area sampled. Since our sampling pointscould be repeated throughout the year for habitat types which were rarer, we might

Figure 1. Map of Pirapitinga in Três Marias Reservoir, Minas Gerais state, south-eastern Brazil,showing the 13 anuran sampling points separated by 2 km of shore line between each other.

Dow

nloa

ded

by [

Fond

ren

Lib

rary

, Ric

e U

nive

rsity

] a

t 14:

11 0

1 O

ctob

er 2

012


potentially sample individual frogs repeatedly in different months. However, anattempt to conduct a mark-recapture study on Leptodactylus ocellatus, the most abun-dant species on the island, rendered no recaptures during a whole year (I Lazarotti andPC Eterovick, unpublished data), so we believe our recapture rate in the presentstudy is likely to be zero.

Sampling lasted up to 40 min at each sampling point, and up to 3 h per night,starting after sunset. The quadrats were located at three different distances fromwater, with three replicates each, to encompass the whole shore and its moisturegradient (Figure 3). The first quadrat was always adjacent to the water, the third onewas always next to the original inland vegetation and the second one in between theother two. The distance perpendicular to the water line between quadrats variedaccording to shore width, which depended on the level of the reservoir. When theshore was narrower than 10 m, the distance between quadrats was 1 m; when it wasbetween 10 and 20 m, the distance was 2 m; when it was between 20 and 30 m, the dis-tance was 3 m; and when it was broader than 30 m, the distance between quadrats was4 m. Parallel to the water line, distance between quadrats was always 4 m (Figure 3).

One of the species recorded, Pseudopaludicola sp., has snout–vent length of about1 cm and occurs in high local densities. Thus, whenever we recorded any individual ofthis species in a quadrat, we sampled it within a 1 m2 sub-section of the quadrat andestimated its density in the whole 16 m2 quadrat. We collected a few voucher speci-mens, euthanized them with benzocaine 20% gel, preserved them according to Heyeret al. (1994), and deposited them in the Museu de História Natural of the PontifíciaUniversidade Católica de Minas Gerais (MCNAM).

Figure 2. Examples of the four habitat types recognized at the shores of Pirapitinga in TrêsMarias Reservoir, south-eastern Brazil: (A) mud with vegetation (MV); (B) mud without vege-tation (M); (C) rocks with vegetation (RV); and (D) rocks without vegetation (R).

Dow

nloa

ded

by [

Fond

ren

Lib

rary

, Ric

e U

nive

rsity

] a

t 14:

11 0

1 O

ctob

er 2

012


Data analyses

We first built a rarefaction curve based on 1000 randomizations to show whether oursample was enough to reflect species richness at Pirapitinga using EcoSim (Gotelliand Entsminger 2005).

We used Chi-square tests (Zar 1996) to compare anuran distribution between seasons(wet season: October to March; dry season: April to September) and among the four hab-itat types identified on the shore. We first compared, for each species recorded in bothseasons, the number of individuals in each habitat type in each season, under the nullassumption that patterns of habitat use would not be influenced by season, resulting in nosignificant differences between seasons. We then compared numbers of individuals ofeach species recorded in each habitat type in both seasons under the null assumption ofno habitat preference, which would result in equivalent proportions of each species ineach habitat type (considering that the area sampled was the same for each habitat type).

We used Pianka’s (1973) niche ovelap index (Ojk) in EcoSim (Gotelli andEntsminger 2005) to characterize niche overlap among co-occurring species in eachseason. In this analysis we considered 12 categories of habitats, given by the four hab-itat types recorded and the three distances from the water sampled at each of them.We performed 1000 simulations to assess whether the observed pattern of niche over-lap was greater than expected by chance. We used the randomization algorithm RA3that retains niche breadth but reshuffles zero states, that is, it allows the species topotentially use other habitats that were available but were not recorded as used in thefield. Overlap varies between 0 (when there is no habitat in common used by bothspecies being compared) and 1 (complete overlap in habitat use). We compared nicheoverlap values obtained for wet and dry seasons using the two-sample t-test in Systat(2007).

Figure 3. Distribution of quadrats for anuran sampling at the shores of Pirapitinga in TrêsMarias Reservoir, south-eastern Brazil.

Dow

nloa

ded

by [

Fond

ren

Lib

rary

, Ric

e U

nive

rsity

] a

t 14:

11 0

1 O

ctob

er 2

012


To test whether anuran species showed preference for any of the four habitattypes or the 12 habitat categories (given by four habitat types versus three distancesfrom the water), and whether their preferences varied between seasons, which couldhelp explain patterns of niche overlap, we calculated electivity for each habitat typeand category for each anuran species in each season using Jacobs’ (1974) electivityindex: D = (Rk – Pk)/[(Rk + Pk) – 2RkPk], where Rk represents the proportion of hab-itat type k used by recorded individuals and Pk represents the proportion of this hab-itat type in the island (which we considered as the proportion of our 39 samplingpoints that represented this habitat). The value of D varies between +1 (completepreference for habitat k) and –1 (habitat k is never used). Zero indicates that the hab-itat is used in the proportion that it is available.

Results

We recorded 474 individuals of seven anuran species (Ameerega flavopicta, Dendrop-sophus rubicundulus, Leptodactylus fuscus, Leptodactylus ocellatus, Physalaemuscentralis, Pseudopaludicola sp. and Rhinella schneideri) distributed in five families(Table 1). Although the rarefaction curve we built using data from quadrat samplingshowed stabilization in species richness (Figure 4), we still observed four additionalspecies at Pirapitinga (R. gr. granulosa, Scinax fuscovarius, S. longilineus, P. cuvieri).These species occurred in very low densities and were not recorded inside our quadrats.Most individuals (n = 346) and most species (n = 6) were recorded in MV habitats,followed by RV (64 individuals and five species), R (52 individuals and three species)and M (15 individuals and three species) habitats. The most abundant species wasLeptodactylus ocellatus (239 individuals), followed by Pseudopaludicola sp. (156 indi-viduals). Species composition varied between seasons, with Rhinella schneideri,Dendropsopus rubicundulus and Physalaemus centralis being recorded only during thewet season, Ameerega flavopicta only during the dry season, and Leptodactylus ocellatus,L. fuscus and Pseudopaludicola sp. in both seasons. Distribution among habitat types

Figure 4. Rarefaction curve (based on 1000 simulations) showing species accumulation atPirapitinga in Três Marias Reservoir, south-eastern Brazil, based on number of individualsrecorded.

Dow

nloa

ded

by [

Fond

ren

Lib

rary

, Ric

e U

nive

rsity

] a

t 14:

11 0

1 O

ctob

er 2

012


Tab

le 1

.N

umbe

r of

ind

ivid

uals

of

each

spe

cies

rec

orde

d in

eac

h ha

bita

t ty

pe (

R,

RV

, M

and

MV

) at

the

sho

res

of P

irap

itin

ga E

colo

gica

l St

atio

n,so

uth-

east

ern

Bra

zil,

and

resu

lts

of C

hi-s

quar

e te

sts

com

pari

ng s

peci

es d

istr

ibut

ion

amon

g ha

bita

t ty

pes

and

betw

een

seas

ons.

R =

roc

ks w

itho

utve

geta

tion

, RV

= r

ocks

wit

h ve

geta

tion

, M =

mud

wit

hout

veg

etat

ion,

MV

= m

ud w

ith

vege

tati

on. A

ster

isks

indi

cate

sig

nifi

cant

val

ues

at th

e le

vel o

fp

< 0

.001

.

Tax

aR

(dr

y/w

et)

RV

(dr

y/w

et)

M (

dry/

wet

)M

V (

dry/

wet

)T

otal

Den

sity

(ha

-1)c2

hab

itat

sc

2 sea

sons

Buf

onid

aeR

hine

lla s

chne

ider

i (W

erne

r, 1

894)

0 (0

/0)

0 (0

/0)

0 (0

/0)

2 (2

/0)

22.

9—

—D

endr

obat

idae

Am

eere

ga f

lavo

pict

a (L

utz,

192

5)7

(0/7

)17

(0/

17)

2 (0

/2)

0 (0

/0)

2637

.626

.61*

—H

ylid

aeD

endr

opso

phus

rub

icun

dulu

s (R

einh

ardt

an

d L

ütke

n, 1

862)

0 (0

/0)

0 (0

/0)

0 (0

/0)

22 (

22/0

)22

31.8

66.0

0*—

Lei

uper

idae

Phy

sala

emus

cen

tral

is (

Bok

erm

ann,

196

2)0

(0/0

)1

(1/0

)1

(1/0

)4

(4/0

)6

8.7

6.00

—P

seud

opal

udic

ola

sp.

0 (0

/0)

16 (

4/12

)0

(0/0

)14

0 (1

08/3

2)15

622

5.8

353.

13*

29.6

4*L

epto

dact

ylid

aeL

epto

dact

ylus

fus

cus

(Sch

neid

er, 1

799)

6 (5

/1)

2 (2

/0)

0 (0

/0)

15 (

15/0

)23

33.3

23.0

8*23

.17*

Lep

toda

ctyl

us o

cella

tus

(Lin

naeu

s, 1

758)

39 (

1/38

)25

(4/

21)

12 (

1/11

)16

3 (8

7/76

)23

934

5.9

244.

00*

11.7

5*T

otal

5261

1534

647

468

6—

—

Dow

nloa

ded

by [

Fond

ren

Lib

rary

, Ric

e U

nive

rsity

] a

t 14:

11 0

1 O

ctob

er 2

012


varied between seasons for all the three species that were recorded in both seasons in theshores of Pirapitinga. Species distribution was non random, varying significantly amonghabitat types for all species except for Physalaemus centralis, which may have been aneffect of the small number of records obtained for this species (Table 1). One species(Rhinella schneideri) was excluded from the analyses because it had only two records.

During the wet season, niche overlaps between species pairs were significantlylarger than expected by chance (range observed Ojk: 0.780–0.981; average observedOjk = 0.858; average simulated Ojk= 0.220, p < 0.0001; Table 2). During the dryseason, there were no significant differences between observed and simulated nicheoverlaps (range observed Ojk = 0.000–0.707; average observed Ojk = 0.312; averagesimulated Ojk = 0.257, p = 0.235; Table 2). Niche overlap between species pairs wassignificantly higher during the wet season (t = 6.593, df = 14, p < 0.001).

Most species (R. schneideri, D. rubicundulus, L. fuscus, L. ocellatus, P. centralisand Pseudopaludicola sp.) showed preference (positive electivity values) for MVhabitats, especially during the wet season. During the dry season, L. fuscus preferredR habitats, Pseudopaludicola sp. preferred MV habitats, and Ameerega flavopictapreferred RV habitats (Table 3). Preferences regarding distance from water variedbetween seasons. During the wet season, most species showed preference for the sectionsclosest to the water line throughout all habitat types: R (L. fuscus and L. ocellatus),RV (L. ocellatus and Pseudopaludicola sp.), M (L. ocellatus) and MV habitats(D. rubicundulus, L. fuscus, L. ocellatus, P. centralis and Pseudopaludicola sp.). Pref-erence for intermediate distances from water was observed for Leptodactylus fuscusand P. centralis in RV habitats. Physalaemus centralis showed preference for areasnear the inland vegetation at M habitats (Table 3).

During the dry season, Ameerega flavopicta preferred areas intermediate ordistant from water in R habitats, or areas distant from water in RV and M habitats.Leptodactylus fuscus preferred areas near the water in R habitats. Leptodactylusocellatus preferred areas near the water in RV, R, and MV habitats, and areas closeto the inland vegetation in M habitats. Pseudopaludicola sp. preferred short to inter-mediate distances from water in RV and MV habitats (Table 3).

Discussion

Habitat destruction imposes a strong selective pressure on specialist species, whilespecies with high behavioural plasticity are more likely to survive in impacted habitats(Grandinetti and Jacobi 2005). Indeed, species that survive in habitats that sufferedanthropogenic disturbance are more resistant, meaning that they can tolerate abroader range of environmental conditions (Ernst and Rödel 2008). Most speciesrecorded in Pirapitinga can be considered as generalists, having broad distributionand occurring in other modified environments (Vasconcelos and Rossa-Feres 2005;Silveira 2006; Santos et al. 2007; Giaretta et al. 2008; Vasconcelos et al. 2009).Dendropsopus rubicundulus is considered endemic to the cerrado biome (Colli et al.2002), but none of the species in Pirapitinga’s shores is threatened according to theIUCN (2008). Still, anuran distribution on the shores of Pirapitinga varied in spaceand time, corroborating the existence of specific preferences for habitats types andactivity periods even in these species considered as “generalists”. It is also interestingto notice that habitat preferences varied between seasons, most species being closer tothe water during their breeding period.

Dow

nloa

ded

by [

Fond

ren

Lib

rary

, Ric

e U

nive

rsity

] a

t 14:

11 0

1 O

ctob

er 2

012


Tab

le 2

.P

iank

a’s

(197

3) o

verl

ap in

dex

(Ojk

) re

gard

ing

use

of 1

2 ha

bita

t ca

tego

ries

by

co-o

ccur

ring

anu

ran

spec

ies

pair

s in

the

wet

(up

per

half

of

the

tabl

e) a

nd t

he d

ry s

easo

n (l

ower

hal

f of

the

tab

le)

at t

he s

hore

s of

Pir

apit

inga

Eco

logi

cal S

tati

on, s

outh

-eas

tern

Bra

zil.

Spec

ies

A. f

lavo

pict

aD

. rub

icun

dulu

sP

. cen

tral

isP

seud

opal

udic

ola

sp.

L. f

uscu

sL

. oce

llatu

s

A. f

lavo

pict

a—

——

——

D. r

ubic

undu

lus

—0.

788

0.98

10.

938

0.82

6P

. cen

tral

is—

—0.

834

0.82

50.

780

Pse

udop

alud

icol

a sp

.0.

115

——

0.92

50.

838

L. f

uscu

s0.

232

——

0.00

00.

848

L. o

cella

tus

0.37

3—

—0.

707

0.44

5

Dow

nloa

ded

by [

Fond

ren

Lib

rary

, Ric

e U

nive

rsity

] a

t 14:

11 0

1 O

ctob

er 2

012


Tab

le 3

.Ja

cobs

’ (19

74)

elec

tivi

ty in

dexe

s (D

) fo

r an

uran

spe

cies

rec

orde

d at

wet

and

dry

sea

sons

at

the

shor

es o

f P

irap

itin

ga E

colo

gica

l Sta

tion

, sou

th-

east

ern

Bra

zil,

rega

rdin

g fo

ur h

abit

at t

ypes

(R

, RV

, M, a

nd M

V)

and

12 h

abit

at c

ateg

orie

s (i

nclu

ding

dis

tanc

e to

wat

er).

Hab

itat

typ

es a

re r

epre

sent

edas

in T

able

1. D

ista

nces

to

wat

er a

re c

lose

(1)

, int

erm

edia

te (

2) o

r fa

r (3

; see

tex

t fo

r sa

mpl

ing

deta

ils).

ab

= s

peci

es a

bsen

t (n

ot r

ecor

ded

in t

he c

orre

-sp

ondi

ng s

easo

n or

in t

he c

orre

spon

ding

hab

itat

).

Hab

itat

typ

eD

ista

nce

to w

ater

Ele

ctiv

ity

(D):

wet

/dry

sea

son

A. f

lavo

pict

aD

. rub

icun

dulu

sP

. cen

tral

isP

seud

opal

udic

ola

sp.

L. f

uscu

sL

. oce

llatu

s

R—

ab /

0.08

ab /

abab

/ ab

ab /

ab−0

.50

/ 1.0

0−0

.96

/ 0.0

5R

V—

ab /

0.70

ab /

ab−0

.21

/ ab

−0.7

7 / 0

.07

−0.5

2 / a

b0.

40 /

−0.3

1M

—ab

/ −0

.61

ab /

ab−0

.19

/ ab

ab /

abab

/ ab

−0.9

2 / −

0.67

MV

—ab

/ ab

1.00

/ ab

0.47

/ ab

0.97

/ 0.

790.

72 /

ab0.

94 /

−0.9

2R

1ab

/ −0

.46

ab /

abab

/ ab

ab /

ab0.

45 /

1.00

1.00

/ 0.

162

ab /

0.31

ab /

abab

/ ab

ab /

ab−0

.28

/ −1.

00−1

.00

/ −0.

413

ab /

0.31

ab /

abab

/ ab

ab /

ab−0

.36

/ −1.

00−1

.00

/ 0.0

3R

V1

ab /

−0.1

2ab

/ ab

−1.0

0 / a

b1.

00 /

0.49

−1.0

0 / a

b0.

60 /

0.72

2ab

/ −1

.00

ab /

ab1.

00 /

ab−1

.00

/ 0.8

01.

00 /

ab−1

.00

/ −1.

003

ab /

0.47

ab /

ab–1

.00

/ ab

–1.0

0 / −

1.00

–1.0

0 / a

b–0

.07

/ −0.

52M

1ab

/ −1

.00

ab /

ab−1

.00

/ ab

ab /

abab

/ ab

1.00

/ −0

.22

2ab

/ −1

.00

ab /

ab−1

.00

/ ab

ab /

abab

/ ab

−1.0

0 / −

1.00

3ab

/ 1.

00ab

/ ab

1.00

/ ab

ab /

abab

/ ab

−1.0

0 / 0

.42

MV

1ab

/ ab

0.71

/ ab

0.28

/ ab

0.50

/ 0.

120.

66 /

ab0.

44 /

0.36

2ab

/ ab

−0.2

8 / a

b−0

.08

/ ab

−0.0

5 / 0

.44

−0.4

4 / a

b−1

.00

/ −0.

223

ab /

ab−0

.86

/ ab

−0.2

3 / a

b−0

.63

/ −1.

00−0

.56

/ ab

0.13

/ −0

.31

Dow

nloa

ded

by [

Fond

ren

Lib

rary

, Ric

e U

nive

rsity

] a

t 14:

11 0

1 O

ctob

er 2

012


As observed in other cerrado localities (e.g. Giaretta et al. 2008), most anuranspecies recorded at Pirapitinga were active during the wet season, when we estimatedthe greatest niche overlaps. This result is expected, since even in less seasonal tropicalforest sites, a greater number of species is usually recorded during the wet season (e.g.Ernst and Rödel 2008; Canavero et al. 2009). The higher niche overlap in the wetseason could be due to more data (frogs) obtained in this season. Although we foundmore frogs per area, the whole area available was also larger (broader shores resultedfrom decreased water level due to water release through the dam in the wet season),so if they concentrated in certain habitats, increasing niche overlap, it means theyreally prefer such habitats. In addition, the lower number of frogs in the dry seasonprobably means they are less active and thus less likely to use resources in common.

According to Schoener (1974), species with the same niche requirements cannotco-occur if available resources are limited, unless they are flexible and niche parti-tioning occurs (Rosenzweig 1981). Considering the niche overlap indices obtained forco-occurring species pairs at Pirapitinga, competition for space is unlikely. Theresources (e.g. shelters, calling sites) are probably sufficiently abundant to allow allspecies to remain in their preferred habitats. Considering that higher densities ofanurans in preferred habitats may attract more predators and/or facilitate diseasetransmission, thus being potentially detrimental to anurans, it is reasonable toconsider that biotic interactions in general are less likely to influence anuran distribu-tion than environmental features. However, in the case of increased predation risk,species can use both conspecific and heterospecific signals as cues to detect predatorpresence (Phelps et al. 2006), in which case large aggregations might bring benefits atthe individual level (Downes and Hoefer 2004).

Anurans are likely to choose breeding sites guided by earlier breeders’ calls whentheir habitat preferences coincide (see Pupin et al. 2007), and this phenomenon is alsolikely to contribute to the higher niche overlaps observed during most species’ breedingseason at Pirapitinga. For aggregation to be beneficial, the costs of competing forbreeding habitats should be less than the benefits of using information provided byother individuals (Seppänen et al. 2007); thus, in this scenario, high chances of inter-specific competition at Pirapitinga would be unlikely, since they would increase costsrelative to benefits of heterospecific attraction (Pupin et al. 2007).

The shores of the reservoir can be considered as an altered habitat less complexthan the original cerrado habitats used by anuran species (see Figure 2). Added to theimpoverishment of the anuran assemblage, they may provide a higher amount of areduced variety of resources per species than in the original richer assemblages.Anuran species require breeding habitats with specific characteristics which aredependent on their reproductive modes. For instance, some species depend on vege-tation, some needing large leaves to lay eggs on; others need humid microhabitats foregg laying at the forest floor; some attach egg clutches under highly oxygenated flowingwater whereas others lay loose eggs in standing water; some need mud to build nestsfor eggs and young tadpoles to develop (Haddad and Prado 2005). Breeding siteswere shown to be strongly associated with the evolution of anuran lineages so thatlocal species composition can be predicted from habitat composition (Zimmermanand Simberloff 1996). That means certain species, and even certain lineages, may beexcluded from impacted areas when their breeding habitats are altered or destroyed.

Indeed, Pirapitinga had fewer species than many other cerrado localities previ-ously studied (see Table 4). Comparison with other sites inventoried for amphibians

Dow

nloa

ded

by [

Fond

ren

Lib

rary

, Ric

e U

nive

rsity

] a

t 14:

11 0

1 O

ctob

er 2

012


at similar altitudes showed that even considering varied sampling efforts and areasizes anuran assemblages were always richer than at Pirapitinga. Impoverishment ofanuran communities is expected in human altered landscapes, and has also beenreported for disturbed forests (Ernst and Rödel 2005; 2008) and wetlands (Herrmanet al. 2005). However, even sites impacted by agriculture and pastures sheltered richeranuran assemblages than Pirapitinga (Table 4), reinforcing the magnitude of theimpact caused by reservoir formation on anuran assemblages. Besides the destructionof breeding habitats, Pirapitinga has also become relatively more isolated fromadjacent areas, causing chances of colonization by new species to decrease, anexpected effect of increased isolation (MacArthur and Wilson 1967). According toBrandão and Araújo (2008), reservoirs are expected to cause species loss boththrough habitat loss and fragmentation/isolation.

Anuran samplings conducted on other islands and shores of the Três Mariasreservoir resulted in a total of 12 species (IM Barata and PC Eterovick, unpublisheddata), showing that even being a single island, Pirapitinga still contributes to anuranconservation at the reservoir, sheltering almost all of its remaining species. The othershores are frequently impacted by pastures and human settlements, which may reducehabitat heterogeneity even more. Such shores are also the potential colonization sourcesof Pirapitinga, which reinforces the low chance of its colonization by new species.

Besides harsher habitat conditions and reduced heterogeneity, anurans are forcedto cope with new barriers to dispersal that can alter the whole dynamics of habitatoccupancy in anthropogenically altered habitats (Ernst and Rödel 2005). Suchchanges are especially strong in reservoirs, and may provide opportunities for speciesthat are more efficient in exploring the available resources to attain high densities, aswe observed for Leptodactylus ocellatus (398 individuals/ha) and Pseudopaludicola sp.(226 individuals/ha). Increases in abundance after disturbance are common in specieswith wide ecological tolerance and broad geographical range (Hamer et al. 1997;Spitzer et al. 1997; Ernst and Rödel 2005). On the other hand, species whose needsare not likely to be completely fulfilled at Pirapitinga showed lower densities, such asspecies that usually breed in temporary bodies of water (e.g. Rhinella schneideri;Vasconcelos et al. 2009; Scinax fuscovarius; Eterovick and Sazima 2004; Leptodactylusfuscus; Brasileiro et al. 2005), or streams (e.g., Ameerega flavopicta; Haddad andMartins 1994; Eterovick and Sazima 2004; Scinax longilineus; PCE, pers. obs.).

Local factors are known to affect amphibian distribution by attracting specieswith a preference for them (Van Buskirk 2005). Most anuran species that survived inPirapitinga after the formation of Três Marias reservoir showed preference for shoreswith mud and vegetation, where we recorded the highest frog densities. Someresources or environmental conditions are important in promoting the aggregation ofindividuals in areas suitable for survival and reproduction (Brown et al. 1995). In thecase of anurans in Pirapitinga, for instance, MV habitats may provide a certain heter-ogeneity of microhabitats such as cavities for shelter, vegetation for perching, and asoft substrate, which may be an advantage for species that build subterranean nests,such as Leptodactylus fuscus (Martins 1988). Pseudopaludicola sp. and Physalaemuscentralis usually call on flooded soil with many little puddles (as we observed duringfieldwork), a microhabitat that was very common in MV habitats. The fine silt thatformed the mud in such shores retains a lot of moisture and prevents fast water drainage,maintaining high humidity levels that constitute a suitable condition for anurans ingeneral (Shoemaker et al. 1992). Vasconcelos et al. (2009) also recorded more frog

Dow

nloa

ded

by [

Fond

ren

Lib

rary

, Ric

e U

nive

rsity

] a

t 14:

11 0

1 O

ctob

er 2

012


Tab

le 4

.P

ublis

hed

amph

ibia

n sp

ecie

s in

vent

orie

s co

nduc

ted

in a

reas

bel

ow 1

000

m a

bove

sea

lev

el i

n th

e C

erra

do b

iom

e, f

or c

ompa

riso

n w

ith

the

impo

veri

shed

fau

na o

f P

irap

itin

ga,

sout

h-ea

ster

n B

razi

l. M

issi

ng d

ata

wer

e no

t pr

ovid

ed b

y th

e au

thor

s. *

= a

ltit

udes

est

imat

ed b

ased

on

loca

lity

info

rmat

ion.

Site

Coo

rdin

ates

; alt

itud

eR

efer

ence

to

anth

ropo

geni

c im

pact

Are

a (h

a)Sa

mpl

ing

effo

rt

(mon

ths)

Num

ber

of a

nura

n sp

ecie

sSo

urce

Est

ação

Eco

lógi

ca d

e It

irap

ina

22°0

0′–2

2°15

′S, 4

7°45

′–48

°00′

W; 7

20–7

50 m

Wel

l pre

serv

ed23

0043

28B

rasi

leir

o et

al.

2005

Nov

a It

apir

ema

21°0

4′S,

49°

32′W

; 460

m*

Inte

nsel

y im

pact

ed

by a

gric

ultu

re—

1527

Vas

conc

elos

and

R

ossa

-Fer

es 2

005

João

Pin

heir

o m

unic

ipal

ity

17°4

4′S,

46°

10′W

; 720

m—

1.07

4737

Silv

eira

200

6

Faz

enda

Par

aíso

17°8

7′S,

49°

24′W

; 735

m*

Impa

cted

by

past

ures

214

1326

Bor

ges

and

Julia

no

2009

Ube

rlân

dia

(tw

o lo

calit

ies)

18°5

5′S,

48°

17′W

; 750

mP

rese

rved

are

as71

025

30G

iare

tta

et a

l. 20

08

Per

dize

s19

°12′

S, 4

7°10

′W; 7

00–1

000

mP

rese

rved

are

a28

4036

18G

iare

tta

et a

l. 20

08P

irap

itin

ga18

°20′

–18°

23′S

, 45°

17′–

45°

20′W

; 585

mIm

pact

ed b

y re

serv

oir

form

atio

n

1000

1311

Pre

sent

wor

k

Dow

nloa

ded

by [

Fond

ren

Lib

rary

, Ric

e U

nive

rsity

] a

t 14:

11 0

1 O

ctob

er 2

012


species in ponds with wet soils at their margins. On the other hand, the vegetation caninfluence anuran distribution in several ways and in different formations, from forest(Parris 2004; Vasconcelos et al. 2009) to cerrado (Oliveira and Eterovick 2009) and“restinga” (Bastazini et al. 2007), including human altered areas (Herrman et al.2005) such as the margins of Pirapitinga (present work). For instance, we onlyrecorded Dendropsophus rubicundulus in MV habitats and perching on vegetationthat was probably the habitat feature that attracted it to this habitat type. However,the sparse vegetation and altered breeding sites may impair high breeding success,resulting in a low density for this species (about 32 individuals/ha).

Ameerega flavopicta is a species that has shown local population declines at anothersite in south-eastern Brazil (Eterovick et al. 2005) and is known to call from rocks inpermanent or temporary streams, where its tadpoles develop (Haddad and Martins1994; Eterovick and Sazima 2004). Its presence in Pirapitinga was unexpected, sincestreams are no longer available. We only recorded this species at the shores of Pirapit-inga during the dry season, although it is expected to breed during the rainy season(Eterovick and Sazima 2004). In an ongoing study, we observed indications of breedingactivity (tadpoles and froglets) during the rainy season in a ditch temporarily filled bywater floods after heavy rains, about 200 m from the shore (NGS Lima and PC Eter-ovick, unpublished data). The use of banks along rain drainages was also reported forthis species by Costa et al. (2006). These frogs may be using the shores just as a moistrefuge during the dry season, thus showing low niche overlap with all the other species.Unlike other dendrobatids, this species occupies open habitats and even toleratesperiods of high temperatures and low humidity (Costa et al. 2006), features that mayhave allowed its permanence in Pirapitinga, although in low densities.

The anuran assemblages that survived in Pirapitinga are noticeably impoverishedcompared to those of other cerrado areas (e.g., Brasileiro et al. 2005; Vasconcelosand Rossa-Feres 2005; Silveira 2006; Giaretta et al. 2008; Borges and Juliano 2009).The habitats adjacent to water bodies, usually occupied by amphibians, wererestricted to the shores of the reservoir, which are relatively homogeneous but stillharbour some different habitat types used by anurans. Although there is great overlapin habitat use among many species, preferences still vary, confirming the importanceof the existing variation in habitat types. When environmental features are importantin determining anuran distribution, which is more likely in impacted areas (Ernst andRödel 2008), it is important to preserve the whole array of environmental conditionsin order to provide suitable habitat for the most species (Parris 2004). Habitat hetero-geneity, including substrate and plant types on lake shores, influences anuran speciescomposition (Vasconcelos et al. 2009). Reservoirs produce highly impacted land-scapes, but the remaining amphibian species still need protection because they can beamong the most abundant vertebrates on land and represent an important part ofmany food chains (Duellman and Trueb 1994). Conservation of the habitat surroundingbreeding wetlands is fundamental to successful amphibian conservation (Herrman et al.2005), and here we show that this also applies to human made reservoirs.

Acknowledgements

We are thankful to Izabela M. Barata, Isabela Lazarotti and Júlia O. L. da Cruz for help duringfieldwork, to M. Wachlevski, Luciana B. Nascimento, Henrique Paprocki, Hélio da Silva andtwo anonymous reviewers for helpful suggestions in previous versions of this manuscript, and

Dow

nloa

ded

by [

Fond

ren

Lib

rary

, Ric

e U

nive

rsity

] a

t 14:

11 0

1 O

ctob

er 2

012


to the Instituto Brasileiro do Meio Ambiente e dos Recursos Naturais Renováveis (Ibama) forlogistics and permits. A Research Productivity grant (305889/2007-9) was provided to P. C.Eterovick by CNPq. We are especially grateful to the whole staff of Ibama who invited us to dothis study and welcomed us and did everything to help in every field trip.

References

Austin MP, Nicholls AO, Doherty MD, Meyers JA. 1994. Determining species response func-tions to an environmental gradient by means of a b-function. J Veg Sci. 5(2):215–228.

Azevedo LG, Barbos AAA, Oliveira ALC, Gorgonio AS, Bedretchuk AC, Siqueira FB, RizzoHG, Silva IS, Moura LC, Araújo-Filho M, Santos RV. 1987. Ensaio metodológico deidentificação e avaliação de Unidades de Conservação: A Estação Ecológica de Pirapitinga,MG. Brasília: SEMA/EMBRAPA – CPAC/MDU.

Bastazini CV, Mundurucá JFV, Rocha PLB, Napoli MF. 2007. Which environmental variablesbetter explain changes in anuran community composition? A case study in the Restinga ofMata de São João, Bahia, Brazil. Herpetologica 63(4):459–471.

Becker CG, Fonseca CR, Haddad CFB, Batista RF, Prado PI. 2007. Habitat split and the glo-bal decline of amphibians. Science 318(5857):1775–1777.

Borges FJA, Juliano RF. 2009. Temporal and spatial distribution in an anuran community inmunicipality of Morrinhos, Goiás, Brazil (Amphibia: Anura). Neotrop Biol Conserv.2(1):21–27.

Both C, Kaefer IL, Santos TG, Cechin STZ. 2008. An austral anuran assemblage in theNeotropics: seasonal occurrence correlated with photoperiod. J Nat Hist. 42(3–4):205–222.

Brandão RA, Araújo AFB. 2008. Changes in anuran species richness and abundance resultingfrom hydroelectric dam flooding in Central Brazil. Biotropica 40(2):263–266.

Brasileiro CA, Sawaya RJ, Kiefer MC, Martins M. 2005. Amphibians of an open Cerradofragment in southeastern Brazil. Biota Neotrop. 5(2):93–109.

Brown JH, Mehlman DW, Stevens GC. 1995. Spatial variation in abundance. Ecology76(7):2028–2043.

Caldwell JP. 1996. The evolution of myrmecophagy and its correlates in poison frogs (familyDendrobatidae). J Zool. 240(11):75–101.

Canavero A, Arim M, Brazeiro A. 2009. Geographic variations of seasonality and coexistencein communities: The role of diversity and climate. Austral Ecol. 34:741–750.

Carolsefeld C, Johnsen, B. Educação ambiental no Vale do Rio São Francisco: uma revisãodos recursos, requisitos e recomendações para o desenvolvimento de programas deeducação ambiental de apoio à pesca e comunidades sustentáveis no Vale do Rio São Fran-cisco [Internet]. 2005. Victoria, Canada: Canadian International Development Agency;[cited 2008 Sept 16]. Available from: http://www.worldfish.org/CIDA%20Brazil%20PDFs/SemiAnnual%20V%20Portuguese/5th%20s.a.%20port_F.5.pdf

Colli GR, Bastos RP, Araujo AFB. 2002. The Cerrados of Brazil: Ecology and natural historyof a neotropical savanna. New York: Columbia University Press. Chapter 12, The characterand dynamics of the Cerrado Herpetofauna; p. 223–241.

Colwell RK, Futuyma DJ. 1971. On the measurement of niche breadth and overlap. Ecology52(4):567–576.

Cornelissen TG, Fernandes GW. 2001. Induced defences in the neotropical tree Bauhinia bre-vipes (Vog.) to herbivory: effects of demage-induced changes on leaf quality and insectattack. Trees 15(4):236–241.

Costa RC, Facure KG, Giaretta AA. 2006. Courtship, vocalizations, and tadpole descriptionof Epipedobates flavopictus (Anura, Dendrobatidae) in southern Goiás, Brazil. BiotaNeotrop. 6(1) doi: 10.1590/S1676-06032006000100006.

Downes S, Hoefer AM. 2004. Antipredatory behaviour in lizards: interactions between groupsize and predation risk. Anim Behav. 67:485–492.

Dow

nloa

ded

by [

Fond

ren

Lib

rary

, Ric

e U

nive

rsity

] a

t 14:

11 0

1 O

ctob

er 2

012


Duellman W, Trueb L. 1994. Biology of amphibians. New York: McGraw-Hill.Eiten G. 1993. Cerrado: Caracterização, ocupação e perspectivas. Brasília: Editora da Univer-

sidade de Brasília. Vegetação de cerrado; p. 17–73.Ernst R, Rödel MO. 2005. Anthropogenically induced changes of predictability in tropical

anuran assemblages. Ecology 86(11):3111–3118.Ernst R, Rödel MO. 2008. Patterns of community composition in two tropical tree frog assem-

blages: separating spatial structure and environmental effects in disturbed and undis-turbed forests. J Trop Ecol. 24(2):111–120.

Eterovick PC, Carnaval ACOQ, Borges-Nojosa DM, Silvano DL, Segalla MV, Sazima I.2005. Amphibian declines in Brazil: an overview. Biotropica 37(2):166–179.

Eterovick PC, Sazima I. 2004. Anfíbios da Serra do Cipó – Amphibians from the Serra doCipó, Minas Gerais, Brasil. Belo Horizonte, Brazil: Ed. PUC Minas.

Giaretta AA, Menin M, Facure KG, Kokubum MNC, Oliveira-Filho JC. 2008. Species rich-ness, relative abundance, and habitat of reproduction of terrestrial frogs in the TriânguloMineiro region, Cerrado biome, southeastern Brazil. Iheringia, Sér Zool. 98(2):181–188.

Gonçalves-Alvim SJ, Fernandes GW. 2001. Comunidades de insetos galhadores (Insecta) emdiferentes fisionomias do cerrado em Minas Gerais, Brasil. Rev Bras Zool. 18 (Suppl.1):289–305.

Goodland R, Ferri MG. 1979. Ecologia do Cerrado. São Paulo: Itatiaia, Editora da Universi-dade de São Paulo.

Gotelli NJ, Entsminger GL. 2005. EcoSim: Null models software for ecology. Version 7.7.2.Acquired Intelligence Inc. and Kesey-Bear. Available from: http://homepages.together.net/∼gentsmin/ecosim.htm.

Grandinetti L, Jacobi CM. 2005. Distribuição estacional e espacial de uma taxocenose deanuros (Amphibia) em uma área antropizada em Rio Acima – MG. Lundiana 6(1):21–28.

Haddad CFB, Martins M. 1994. Four species of Brazilian poison frogs related to Epipedobatespictus (Dendrobatidae): taxonomy and natural history observations. Herpetologica50(3):282–295.

Haddad CFB, Prado CPA. 2005. Reproductive modes in frogs and their unexpected diversityin the Atlantic forest of Brazil. BioScience 55(3):207–217.

Hamer KC, Hill JK, Lace LA, Langhan AM. 1997. Ecological and biogeographical effects offorest disturbance on tropical butterflies of Sumba, Indonesia. J Biogeogr. 24(1):67–75.

Herrmann HL, Babbitt KJ, Baber MJ, Congalton RG. 2005. Effects of landscape characteris-tics on amphibian distribution in a forest-dominated landscape. Biol Conserv.123(1):139–149.

Heyer WR, Donnelly MA, McDiarmid RW, Hayek LAC, Foster MS. 1994. Measuring andmonitoring biological diversity. Standard methods for amphibians. Washington and London:Smithsonian Institution Press.

Hutchinson GE. 1957. Concluding remarks. Cold Spring Harbor Symposium on QuantitativeBiology 22:415–427.

IUCN (International Union for Conservation of Nature and Natural Resources). Red listof threatened species [Internet]. 2008. [cited 2008 Nov 8]. Available from: http://www.iucnredlist.org

Jacobs J. 1974. Quantitative measurement of food selection: a modification of the forageration and Ivlev’s electivity index. Oecologia 14(4):413–417.

MacArthur RH, Wilson EO. 1967. The theory of island biogeography. Princeton 481 Univer-sity Press, Princeton.

Mackenzie DI, Nichols JD, Lachman GB, Droege S, Royle A, Langtimm CA. 2002. Estimat-ing site occupancy rates when detection probabilities are less than one. Ecology83(8):2248–2255.

MacNally RC. 1979. Social organization and interspecific interactions in two sympatric speciesof Ranidella (Anura). Oecologia 42(3):293–306.

Dow

nloa

ded

by [

Fond

ren

Lib

rary

, Ric

e U

nive

rsity

] a

t 14:

11 0

1 O

ctob

er 2

012


Maia VC, Fernandes GW. 2005. Two new species of Asphondylinii (Diptera: Cecidomyiidae)associated with Bauhinia brevipes (Fabaceae) in Brazil. Zootaxa 1091:27–40.

Martins M. 1988. Biologia reprodutiva de Leptodactylus fuscus em Boa Vista, Roraima(Amphibia: Anura). Rev Bras Biol. 48(4):969–977.

McLeod RF, Gates JE. 1998. Response of herpetofaunal communities to forest cutting andburning at Chesapeake Farms, Maryland. Am Midl Nat. 139(1):164–177.

Menin M, Lima AP, Magnusson WE, Waldez F. 2007. Topographic and edaphic effects on thedistribution of terrestrially reproducing anurans in Central Amazonia: mesoscale spatialpatterns. J Trop Ecol. 23(5):539–547.

Oliveira FFR, Eterovick PC. 2009. The role of river longitudinal gradients, local and regionalattributes in shaping frog assemblages. Acta Oecol 35(5):727–738.

Parris KM. 2004. Environmental and spatial variables influence the composition of frogassemblages in sub-tropical eastern Australia. Ecography 27(3):392–400.

Phelps SM, Rand AS, Ryan MJ. 2006. The mixed-species chorus as public information:túngara frogs eavesdrop on a heterospecific. Behav Ecol. 18:108–114.

Pianka ER. 1973. The structure of lizard communities. Ann Rev Ecol Syst. 4:53–74.Pupin F, Sacchi R, Gentilli A, Galeotti P, Fasola M. 2007. Discrimination of toad calls by

smooth newts: support for the heterospecific attraction hypothesis. Anim Behav74:1683–1690.

Rosenzweig ML. 1981. A theory of habitat selection. Ecology 62(2):327–335.Saenz D, Fitzgerald LA, Baum KA, Conner RN. 2006. Abiotic correlates of anuran calling

phenology: the importance of rain, temperature, and season. Herpetol Monogr. 20:64–82.Santos TG, Rossa-Feres DC, Casatti L. 2007. Diversidade e distribuição espaço-temporal de

anuros em região com pronunciada estação seca no sudeste do Brasil. Iheringia, Sér Zool.97(1):37–49.

Schoener TW. 1974. Resource partitioning in ecological communities. Science 185(4145):27–39.Seppänen J-T, Forsman JT, Mönkkönen M, Thomson RL. 2007. Social information use is

a process across time, space, and ecology, reaching heterospecifics. Ecology88(7):1622–1633.

Shoemaker VH, Hillman SS, Hillyard SD, Jackson DC, McClanahan LL, Withers PC,Wygoda ML. 1992. Environmental physiology of the amphibians. Chicago: The Univer-sity of Chicago Press. Exchange of water, ions, and respiratory gases in terrestrialamphibians; p 125–150.

Silvano DL, Segalla, MV. 2005. Conservação de anfíbios no Brasil. Megadiversidade 1:79–86.Silveira AL. 2006. Anfíbios do município de João Pinheiro, uma área de Cerrado no Noroeste

de Minas Gerais, Brasil. Arq Mus Nac, Rio de Janeiro 64(2):131–139.Spitzer K, Jaros J, Havelka J, Leps J. 1997. Effects of small-scale disturbance on butterfly

communities of an Indochinese montane rainforest. Biol Cons. 80(1):9–15.Stuart SN, Chanson JS, Cox NA, Young BE, Rodrigues ASL, Fischman DL, Waller RW.

2004. Status and trends of amphibian declines and extinctions worldwide. Science306(5702):1783–1786.

SYSTAT for Windows, 2007. Statistics, Version 12.0 Edition. Evanston, IL: Systat, Inc.Toft CA. 1985. Resource partitioning in amphibians and reptiles. Copeia 1985(1):1–21.Van Buskirk J. 2005. Local and landscape influence on amphibian occurrence and abundance.

Ecology 86(7):1936–1947.Vasconcelos TS, Rossa-Feres DC. 2005. Diversidade, distribuição espacial e temporal de anfí-

bios anuros (Amphibia: Anura) na região noroeste do estado de São Paulo, Brasil. BiotaNeotrop. 5(2):137–150.

Vasconcelos TS, Santos TG, Rossa-Feres DC, Haddad CFB. 2009. Influence of the environ-mental heterogeneity of breeding ponds on anuran assemblages from southeastern Brazil.Can J Zool. 87(8):699–707.

Dow

nloa

ded

by [

Fond

ren

Lib

rary

, Ric

e U

nive

rsity

] a

t 14:

11 0

1 O

ctob

er 2

012


Young BE, Lips KR, Reaser JK, Ibáñez R, Salas AW, Cedeño JR, Coloma LA, Ron S, LaMarca E, Meyer J, Muñoz A, Bolaños F, Chaves G, Romo D. 2001. Population declinesand priorities for amphibian conservation in Latin America. Conserv Biol. 15(5):1213–1223.

Zar JH. 1996. Biostatistical analysis. 3rd ed. New Jersey: Prentice Hall, Inc.Zimmerman BL, Simberloff D. 1996. An historical interpretation of habitat use by frogs in a

Central Amazonian Forest. J Biogeogr. 23:27–46.

Dow

nloa

ded

by [

Fond

ren

Lib

rary

, Ric

e U

nive

rsity

] a

t 14:

11 0

1 O

ctob

er 2

012

Documents

Anuran assemblage composition and distribution at a modified environment in Três Marias reservoir, south-eastern Brazil