Content uploaded by Sônia Talamoni
Author content
All content in this area was uploaded by Sônia Talamoni on Jan 30, 2016
Content may be subject to copyright.
Introduction
Al though op por tu nis tic and gen er a list in
their feed ing hab its, squir rels feed pri mar ily on
fruits and seeds. Fruit pro duc tion nor mally
var ies ac cord ing to spe cies, lo ca tion and year,
and this may in flu ence the num ber of an i mals in
a pop u la tion, as well as their dis tri bu tion in
dif fer ent for ested hab i tats (Gurnell 1996). Be -
sides, dif fer ences in food avail abil ity and qual ity
be tween hab i tats may ex plain vari a tions in the
ac tiv ity pat tern and for ag ing be hav iour of squir -
rels (Wauters et al. 1992). Ar eas with poor food
sup plies are avoided by squir rels, which in stead
con cen trate their ac tiv i ties in hab i tats with
more abun dant food re sources (Lurz et al. 2000).
In this study, we in ves ti gated the for ag ing
be hav iour of a pop u la tion of Sciurus aestuans
Linnaeus, 1776, in a sec ond-growth for est in the
Serra do Caraça Re serve, in which there is a
high con cen tra tion of the palm Syagrus roman-
zoffiana (Arecaceae). Some stud ies in di cated
that the fruits of this palm tree were an im -
[69]
Acta Theriologica 51 (1): 69–74, 2006.
PL ISSN 0001–7051
For ag ing be hav iour of the Bra zil ian squir rel Sciurus aestuans
(Rodentia, Sciuridae)
Cibele A. ALVARENGA and Sônia A. TALAMONI
Alvarenga C. A. and Talamoni S. A. 2006. For ag ing behaviour of the squir rel Sciurus aestuans
(Rodentia, Sciuridae). Acta Theriologica 51: 69–74.
We de scribe the for ag ing behaviour of the Bra zil ian squir rel Sciurus aes-
tuans Linnaeus, 1776 in a nat u ral free-liv ing pop u la tion, in an area with a
high con cen tra tion of palm fruits Syagrus romanzoffiana (Chamisso) Glassman.
Our ob jec tive was to in ves ti gate if the abun dance of S. romanzoffiana fruits,
an im por tant food item for squir rels, tem po rally in flu enced the for ag ing
behaviour of this pop u la tion. The ob ser va tions were car ried out over an
eight-month (April–No vem ber 2001) pe riod. Fruit ing phenology of S. roman-
zoffiana was also mon i tored. The most con sumed food item for this pe riod was
the fruit of S. romanzoffiana (70%), fol lowed by foods of hu man or i gin (14%).
The per cent age of time spent feed ing in creased in Sep tem ber, and the time
spent trav el ling de clined in Sep tem ber, Oc to ber and No vem ber, when fe males
were ob served show ing ev i dence of re pro duc tive ac tiv ity. At the same time
S. romanzoffiana fruits be come avail able. The con sump tion of human items
il lus trates the op por tu nis tic feed ing habit of these squir rels, but vari a tions in
for ag ing behaviour de tected prob a bly were more strongly re lated to avail -
abil ity of S. romanzoffiana fruits.
Post-graduate Programme in Vertebrate Zoology. Pontifícia Universidade Católica de Minas
Gerais. Av. Dom José Gaspar, 500, Belo Horizonte, MG, 30535-610, Brazil,
e-mail: cibele_alvarenga@yahoo.com.br, talamoni@pucminas.br
Key words: Sciurus aestuans, Syagrus romanzoffiana, for ag ing be hav iour
por tant food re source for this spe cies (Galetti et
al. 1992, Paschoal and Galetti 1995, Bordignon
and Monteiro Filho 1999, 2000). Thus, the high
avail abil ity of this fruit in the area of study
al lowed us to in ves ti gate if the abun dance of
pre ferred food tem po rally af fected for ag ing be -
hav iour of this free-living population.
Methods
The study was con ducted in the Serra do Caraça Pri vate
Nat u ral Her i tage Re serve (20°05’S, 43°29’W), lo cated on the
slopes of the Serra do Espinhaço Moun tains sys tem, in a
re gion con sid ered to be a tran si tion zone be tween the
Cerrado and the At lan tic For est biomes, in the state of
Minas Gerais, Brazil. The Re serve mea sures 10 187 ha,
with el e va tion rang ing from 850 to 2070 m a.s.l. and a
cli mate marked by a rainy sum mer (Oc to ber– March) and a
dry win ters (April–Sep tem ber). Var i ous floristic for ma tions
char ac ter ize the Re serve and in clude semideciduous for ests,
high-al ti tude fields, and rupestrian fields (grass lands sur -
rounded by rock out crops, as well as shrubs and small trees).
The study was car ried out in sec ond-growth for est lo -
cated near the Re serve’s head quar ters, and in a gar den
around the head quar ters where there are other ex otic tree
spe cies, such as Cupressus (20–30 trees, Cupressaceae) and
Quercus sp. (5 trees, Fagaceae). In the study area tour ists
are con stantly pres ent and food of hu man or i gin (bread,
pop corn, bis cuits, cake, pea nuts, ba nanas) is fre quently
sup plied to the an i mals. From April to No vem ber 2001, for -
ag ing be hav iour of squir rels was in ves ti gated at nine ob ser -
va tion points, en com pass ing an area of 4.0 ha. In each
month a set route for mak ing ob ser va tions was trav eled
dur ing 6 days to en sure that all nine points were sam pled
at all times of the day. Observations were made us ing 8–17
´ 25 mm bin oc u lars be tween 6:00 to 18:00 hours. Thirty
min utes were spent at each ob ser va tion point, dur ing which
fo cal-an i mal sam pling with con tin u ous re cords, and mod i -
fied scan sam pling were car ried out (Altmann 1974). The
thirty-min ute pe riod was di vided into five ini tial min utes of
scan sam pling, twenty min utes of fo cal-an i mal sam pling
and the last five min utes again of scan sam pling. Time al lo -
cated to each be hav iour cat e gory was quan ti fied us ing data
ob tained from fo cal-an i mal sam pling. Data from scan sam -
pling was ex pressed as the to tal num ber of oc cur rences of
the dif fer ent be hav iour cat e go ries (Mar tin and Bateson
1986). In each month, ob ser va tion time to taled 60 hours.
Af ter pre lim i nary ob ser va tions, five be hav iour cat e go ries
were re corded as shown in the ethogram be low: Trav el ing –
walk ing or run ning in trees or on the ground; Feed ing – con -
sump tion of food, in clud ing food han dling prior to eat ing,
and drink ing of wa ter; Search ing for food – in trees or on
the ground; Rest ing – sit ting or ly ing on a tree branch, or
stand ing still on the ground; Other be hav iours – Food
caching, food retrieval, grooming, vocalizing, nest building,
agonistic interactions.
The be hav ioural ob ser va tions were car ried out on 15
squir rels (8 fe males and 7 males) pre vi ously trapped and
marked be tween Feb ru ary and April 2001, and three non
marked squir rels, who were iden ti fi able from in di vid ual
marks (ie, one fe male who had a big white spot on her right
leg; one male with out a right fore foot and blind in one eye;
and an other male with a torn ear). These squir rels were
in cluded in our data col lec tion be cause they were fre quently
ob served in the months af ter the cap ture of the marked an i -
mals. The cap ture was done us ing 12 live-traps (gal va nized
wire cages mea sur ing 15 ´ 15 ´ 32 cm) in the study area.
S. romanzoffiana fruits or pea nuts were used as bait. In di -
vid u als were marked us ing plas tic col lars with two col ored
beads: one color for sex iden ti fi ca tion and the other for
in di vid ual iden ti fi ca tion. The mark ing and re lease of the
squir rels oc curred im me di ately af ter cap tures.
The phenology of S. romanzoffiana was an a lyzed ev ery
two weeks. Al to gether, 105 trees dis trib uted in the ob ser va -
tion area were in di vid u ally marked. The num ber of ripe
fruit bunches, im ma ture fruit bunches and flow ers in each
tree were re corded. To cal cu late the avail abil ity of fruit, an
av er age of 796.5 ± 331.6 fruits (n = 20, Galetti et al. 1992)
per bunch was mul ti plied by the num ber of ob served ripe
fruit bunches.
Spearman’s rank cor re la tion (Zar 1999) was used to test
the hy poth e sis that time spent by squir rels in the dif fer ent
be hav iour cat e go ries and num ber of oc cur rences of each cat e -
gory may be as so ci ated with avail abil ity of S. romanzoffia-
na fruits. Also in ves ti gated was whether fruit con sump tion
cor re lated with the es ti mated avail abil ity of S. romanzoffia-
na fruits in the study area.
Results
Feeding and trav el ing had the great est per -
cent ages of oc cur rence us ing both fo cal-an i mal
sam pling (Fig. 1) and scan sam pling (Fig. 2). The
most in fre quent be hav iours were search ing for
food and rest ing (Figs 1 and 2). Feeding in creased
in the month of Sep tem ber, while trav el ing
de clined in Sep tem ber, Oc to ber and No vem ber,
when com pared with pre vi ous months (Fig. 1).
At the be gin ning of the study, most of the
an i mals ob served were sub-adults. Fe males
show ing ev i dence of re pro duc tive ac tiv ity (eg,
large nip ples and lac ta tion) were ob served in
Sep tem ber and Oc to ber. Dur ing this time, three
new fe males with signs of re pro duc tive ac tiv ity
ap peared in study area; whereas, a to tal of 2
males dis ap peared, one in June and one in
July, re spec tively. Fur ther more, 2 new ju ve nile
squir rels were ob served in the study area in
No vem ber, at the end of the study.
70 C. A. Alvarenga and S. A. Talamoni
The food most of ten con sumed by the squir -
rels was the fruit of S. romanzoffiana, to tal ing
70.7% of the to tal food con sump tion (Ta ble 1).
Food of hu man or i gin was also eaten by squir rels,
and rep re sented 14.1% of to tal food con sump tion.
S. romanzoffiana fruits were con sumed most
of ten in the month of April, when 57 oc cur rences
of squir rels feed ing on these fruits were re corded,
af ter which the num ber of oc cur rences was vari -
able. In No vem ber, 66 oc cur rences of squir rels
feed ing on dif fer ent food types were ob served
(Ta ble 1). Phenological anal y sis of S. roman-
zoffiana showed that avail able ripe fruits in the
trees were in creas ing from 796.5 (1 bunch), 2389
(3 bunches), 3186 (4 bunches) and 3982 (5
bunches) for the months of Au gust, Sep tem ber,
October and November, respectively.
Squir rels were se lec tive in their con sump tion
of S. romanzoffiana fruits; they would eat some
and dis card oth ers. Squir rels re quired 220.2 sec
± 84.6 (mean ± SD; n = 48) to re move the exocarp,
open the endocarp and eat the en do sperm. One
young squir rel, ob served in No vem ber, spent
al most one hour on this pro cess. The act of
cach ing fruit of S. romanzoffiana, bread and
pea nuts be tween rocks, un der leaves, fallen
trunks, or in side bro meli ad (Aechmea blanche-
tiana), or bury ing S. romanzoffiana fruit in
sand, was ob served 27 times dur ing the study.
An adult squir rel was seen re cov er ing S. roman-
Foraging of Brazilian squirrel 71
0
20
40
60
80
Feeding
Travelling
Searching for food
Resting
Oct Nov
Aug Sep
Apr May Jun Jul
Behaviour categories (%)
Fig. 1. Per cent ages of time al lo cated for be hav iour cat e go ries ob served for Bra zil ian squir rels, Sciurus aestuans, in the Serra
do Caraça Re serve dur ing April to No vem ber 2001. Data ob tained by fo cal-an i mal sam pling.
Feeding
Travelling
Searching for food
Resting
0
20
40
60
80
Behaviour categories (%)
Oct Nov
Aug Sep
Apr May Jun Jul
Fig. 2. Per cent ages of oc cur rences of be hav iour cat e go ries ob served for Bra zil ian squir rels, Sciurus aestuans, in the Serra do
Caraça Re serve dur ing April to No vem ber 2001. Data ob tained by scan sam pling.
zoffiana fruit from the ground and cach ing
them, one by one, at a dis tance of ap prox i mately
3 m from the tree. Twelve fruits were cached in
11 min utes.
Ex cept for an in verse cor re la tion be tween the
per cent age of time spent search ing for food and
the con sump tion of hu man items (r
s
= –0.85, p =
0.007), Spearman’s rank cor re la tions showed no
sig nif i cant re la tion ship be tween the main be -
hav iour cat e go ries ob tained by both sam pling
types and the con sump tion of S. romanzoffiana
fruit (p > 0.05). A sig nif i cant neg a tive cor re -
la tion was ob served be tween the con sump tion of
S. romanzoffiana fruits and the con sump tion of
anthropic items (r
s
= –0.73, p = 0.037). The
num ber of oc cur rences of the cat e gory feed ing
cor re lated pos i tively with the avail abil ity of S.
romanzoffiana fruits (r
s
= 0.74, p = 0.05).
Discussion
Lac tat ing fe males were cap tured in Sep tem -
ber and Oc to ber, so the young must have been
born around this time. Young ap pear to be
in de pend ent at the same time that fruit of
S. ro manzoffiana be come avail able. Dur ing our
study, some males moved out of the study area
dur ing the re pro duc tive pe riod and never re -
turned. Con versely, non-res i dent lac tat ing fe -
males were re corded in the study area only in
the fi nal months of the study, when we ob served
the great est abun dance of food. These facts may
be due to be hav ioural dif fer ences be tween the
sexes, with males tend ing to travel lon ger dis -
tances to find re cep tive fe males for mat ing and
fe males look ing for ar eas with more abun dant
food sup plies to meet their en ergy needs dur ing
72 C. A. Alvarenga and S. A. Talamoni
Ta ble 1. Oc cur rences and per cent ages (in brack ets) of Bra zil ian squir rels, Sciurus aestuans, ob served feed ing on the dif fer ent
food items, in the Serra do Caraça Re serve dur ing April to No vem ber 2001. Data ob tained by scan sampling.
Food item
Apr May Jun Jul Aug Sep Oct Nov Total
Syagrus romanzoffiana 57 (93.4) 36 (66.7) 25 (61.0) 28 (52.8) 26 (74.3) 23 (82.1) 43 (84.3) 37 (56.1) 275 (70.7)
fruit (Palmae)
Lafoensia pacari fruit 1 (1.6) 3 (5.6) 6 (14.6) 11 (20.8) – – 1 (2.0) – 22 (5.7)
(Lythraceae)
Cupressus fruit – – 2 (4.9) 3 (5.7) – 1 (3.6) – 1 (1.5) 7 (1.8)
(Cupressaceae)
Bark – 1 (1.9) – – 1 (2.9) – 3 (5.9) 1 (1.5) 6 (1.5)
Rubus brasiliensis fruit 1 (1.6) 2 (3.7) – 1 (1.9) – – 1 (2.0) – 5 (1.3)
(Moraceae)
Musa fruit – – – – 2 (5.7) 2 (7.1) – – 4 (1.0)
(Musaceae)
Dasyphyllum fruit – – – 3 (5.7) – – – – 3 (0.8)
(Asteraceae)
Salacia brachypoda – – – – – – – 3 (4.5) 3 (0.8)
flower (Lauraceae)
Psidium guajava fruit 1 (1.6) 1 (1.9) – – – – – – 2 (0.5)
(Myrtaceae)
Moss 1 (1.6) – – – – – – 1 (1.5) 2 (0.5)
Prunus persica fruit – – – – – – – 2 (3.0) 2 (0.5)
(Rosaceae)
Myrciaria delicatua – – – – – – – 2 (3.0) 2 (0.5)
fruit (Myrtaceae)
Lichen – – – – – – – 1 (1.5) 1 (0.3)
Anthropogenic items – 11 (20.4) 8 (19.5) 7 (13.2) 6 (17.1) 2 (7.1) 3 (5.9) 18 (27.3) 55 (14.1)
Total 61 (100) 54 (100) 41 (100) 53 (100) 35 (100) 28 (100) 51 (100) 66 (100) 389 (100)
lac ta tion (Wauters and Dhondt 1992, Andrén
and Delin 1994, Lurz et al. 2000, Ro dri guez and
Andrén 1999). In the area stud ied, sev eral nests
were found near food re sources (Alvarenga 2002),
thus nurs ing fe males would not have to search
for food in other ar eas and could thus spend less
time away from their young.
Ac cord ing to Gurnell (1996), re pro duc tive
suc cess of Sciurus carolinensis Gmelim, 1788 fe -
males cor re late pos i tively with the avail abil ity
of food in the area. Wauters and Dhondt (1989),
in a study of the red squir rel Sciurus vulgaris
Linnaeus, 1758, re ported that lac tat ing fe males
have to ad just their for ag ing hab its, ac tiv ity pat -
tern and nest uti li za tion, be cause they can not
leave the nest to search for food, to meet their
en ergy re quire ments dur ing lac ta tion. Thus, the
abun dance of fruit of S. romanzoffiana, the main
food re source in the area, as so ci ated with the
prob a ble birth of young, may ex plain the in -
creased time spent feed ing and the de cline in the
time spent mov ing around in the months of Sep -
tem ber, Oc to ber and No vem ber. Our data sug -
gest that lac tat ing fe males can op ti mize their
time spent feed ing by se lect ing hab i tats with
greater food avail abil ity, thereby reducing the
need to spend time traveling in search of food.
Phenological anal y sis of S. romanzoffiana
in di cated that the fruits of this palm tree were
avail able from Au gust to No vem ber, yet were
con sumed dur ing all months be cause of the food
cach ing habit of squir rels. The cach ing and
dis card ing of fruit, of ten ob served in our study,
sug gests that squir rels ef fi ciently dis perse S.
romanzoffiana as re ported for S. aestuans (Maia
et al. 1987, Galetti et al. 1992, Paschoal and
Galetti 1995, Bordignon and Monteiro Filho
1999). Sev eral seed lings were ob served at a
dis tance from the mother-tree, sug gest ing that
dis persal had oc curred. How ever, re search is
needed to de ter mine the ac tual role played by
squir rels as agents of dis persal of this tree
spe cies in the area. Quercus fruit is usu ally con -
sumed by squir rels in the north ern hemi sphere
(Koprowski 1994, Gurnell 1996), how ever the
trees that oc cur in the study area had not
developed fruits for a long time.
The ab sence of sta tis ti cal sig nif i cance be -
tween feed ing oc cur rences and the avail abil ity
of fruit of S. romanzoffiana is prob a bly due to
the fact that the abun dance of fallen and cached
fruits, and the quan tity of anthropogenic foods,
the sec ond most im por tant food re source in the
Re serve, were not eval u ated. The neg a tive as -
so ci a tion be tween the con sump tion of fruit of
S. romanzoffiana and the con sump tion of an-
thropogenic food, to gether with the neg a tive
cor re la tion be tween the time spent search ing for
food and the con sump tion of anthropogenic
items, pro vides fur ther ev i dence of the op -
por tu nis tic feed ing hab its of these an i mals.
How ever, the vari a tions in for ag ing be hav iour of
squir rels prob a bly re flect avail abil ity of fruits of
S. romanzoffiana.
Ac knowl edge ments: We thank J. L. Koprowski and two
anon y mous re view ers for help ful com ments. To R. J. Young
and A. Linghorn for manu script re vi sion. The Caraça Re -
serve Staff pro vided us with all fa cil i ties to com plete our
study. FIP PUC Minas pro vided fi nan cial sup port.
References
Altmann J. 1974. Ob ser va tion study of be hav iour: sam pling
meth ods. Be hav iour 49: 227–265.
Alvarenga C. A. 2002. [Study of a Sciurus aestuans ingrami
Thomas, 1901 (Rodentia, Sciuridae) pop u la tion at Serra
do Caraça Pri vate Nat u ral Re serve, Minas Gerais,
Brasil]. MSc the sis, Pontifícia Universidade Católica de
Minas Gerais, Belo Horizonte: 1–44. [In Por tu guese]
Andrén H. and Delin A. 1994. Hab i tat se lec tion in the Eur -
asian red squir rel Sciurus vulgaris in re la tion to for est
frag men ta tion. Oikos 70: 43–48.
Bordignon M. and Monteiro-Filho E. L. A. 1999. Sea sonal
food re sources of the squir rel Sciurus ingrami in a sec -
ond ary Arau ca ria for est in south ern Brazil. Stud ies
Neo tropi cal Fauna & En vi ron ment 34: 137–140
Bordignon M. and Monteiro-Filho E. L. A. 2000. Be hav iour
and daily ac tiv ity of the squir rel Sciurus ingrami in a
sec ond ary Arau ca ria For est in south ern Brazil. Ca na -
dian Jour nal of Zo ol ogy 78: 1732–1739.
Galetti M., Paschoal M. and Pedroni F. 1992. Pre da tion on
palm nuts (Syagrus romanzoffiana) by squir rels (Sciu-
rus ingrami) in south-east Brazil. Jour nal of Trop i cal
Ecol ogy 8: 121–123.
Gurnell J. 1996. Ef fects of food avail abil ity and win ter
weather on the dy nam ics of a grey squir rel pop u la tion
in south ern Eng land. Jour nal of Ap plied Ecol ogy 33:
325–338.
Koprowski J. L. 1994. Sciurus carolinensis. Mam ma lian
Spe cies 480: 1–9.
Lurz P. W. W., Garson P. J. and Wauters L. A. 2000. Ef fects
of tem po ral and spa tial vari a tions in food sup ply on the
Foraging of Brazilian squirrel 73
space and hab i tat use of red squir rels. Jour nal of Zo ol -
ogy, Lon don 251: 167–178.
Maia A., Serran F., Fernándes H. Q., Oliveira R. R., Oliveira
R. F. and Pen na T. M. 1987. [In fer ence fau nal by re -
mains veg e tal III: In ter ac tions of squirrel (Sciurus aes-
tuans ingrami, Thomas 1901) with the palm “baba-
de-boi” (Syagrus romanzoffiana) (Chamisso)]. Atas da
Sociedade Botânica do Brasil 3: 89–96. [In Por tu guese]
Mar tin P. and Bateson P. 1986. Mea sur ing be hav iour. An
in tro duc tory guide. Cam bridge Uni ver sity Press: 1–200.
Paschoal M. and Galetti M. 1995. Sea sonal food use by the
Neo tropi cal squirrel Sciurus ingrami in south east ern
Brazil. Biotropica 27: 268–273.
Ro dri guez A. and Andrén H. 1999. A com par i son of Eur -
asian red squir rel dis tri bu tion in dif fer ent frag mented
land scapes. Jour nal of Ap plied Ecol ogy 36: 649–669.
Wauters L. A. and Dhondt A. A. 1989. Body weight, lon gev -
ity and re pro duc tive suc cess in red squir rel Sciurus
vulgaris. Jour nal of An i mal Ecol ogy 58: 637–651.
Wauters L. A. and Dhondt A.A. 1992. Spac ing be hav iour of
red squir rel Sciurus vulgaris: vari a tion be tween hab i -
tats and the sexes. An i mal Be hav iour 43: 297–311.
Wauters L. A., Swinnen A. A. and Dhondt A. A.1992. Ac tiv -
ity budget and for ag ing be hav iour of red squir rel
Sciurus vulgaris in co nif er ous and de cid u ous hab i tats.
Jour nal of Zo ol ogy, Lon don 227: 71–86.
Zar J. H. 1999. Biostatistical anal y sis. Fourth Edi tion.
Prentice-Hall, Up per Sad dle River, NJ, New Jer sey:
1–718.
Received 29 November 2004, accepted 23 August 2005.
Associate Editor was Joseph F. Merritt.
74 C. A. Alvarenga and S. A. Talamoni