Diving Behavior and Foraging Areas of the Neotropic Cormorant at a Marine Colony in Patagonia, Argentina

2004·4.1 MB·English

by  Flavio Quintana

Wilson Bulletin, 116(1), 2004, pp. 83—88
DIVING BEHAVIOR AND FORAGING AREAS OF THE NEOTROPIC
CORMORANT AT A MARINE COLONY
IN PATAGONIA, ARGENTINA
FLAVIO QUINTANA,!5 pabLO YORIO,' NORA LISNIZER,^
ALEJANDRO GATTO,^ AND CASPAR SORIA^
—
ABSTRACT. The Neotropic Cormorant {Phalacrocorax brasilianus) is a widespread and abundant species
distributed throughout the Neotropics. We reporton diving behaviorand foraging areasofNeotropicCormorants
in inshore marine waters of Patagonia, Argentina. Mean dive duration was 18.9 ± 5.3 sec and differed signifi-
cantly among birds. Maximum dive duration was 43 sec. Birds spent 58-79% of their time at sea underwater.
Mean recovery time at the surface between successive dives was 6.7 ± 1.5 sec, and was also significantly
different among individuals. Mean diving efficiency (mean dive duration time/mean recovery time) was 2.6 ±
0.5 (range; 0.5-4.6). We were able to locate birds while feeding on 112 of 215 occasions. Almost 90% of
mapped locations (68 feeding trips) were within 2.5 km ofthe colony. However, we did not receive a signal on
103 occasions, suggesting that foraging may also occur in waters outside the inlet where the colony was located.
High variability in dive duration and recovery time is congruent with flexible foraging techniques and diet, as
Neotropic Cormorants feedon both bottom and pelagic schooling fish. Received8July2003, accepted26March
2004.
The Neotropic Cormorant {Phalacrocorax Argentina (De la Pena 1980, Navas 1993). On
brasilianus), is widely distributed in the Neo- the marine coasts of Patagonia, it breeds at
tropics. The species ranges from the southern twelve sites with a total breeding population
'
United States to Cape Horn at the southern estimated at 1,200 pairs (Yorio et al. 1999).
I extreme of South America (Orta 1992, Telfair Despite its abundance and broad distribution,
and Morrison 1995). Neotropic Cormorants the Neotropic Cormorant has been little stud-
I
are one of the few cormorant species that oc- ied throughout its range (Telfair and Morrison
1 cupy both freshwater and marine environ- 1995). In Argentina, general aspects of its
ments, breeding in inland wetlands, fast-flow- breeding biology and behavior have been de-
ing rivers, high-altitude lakes, and on marine scribed for freshwater populations (Bo 1956,
I
' shores and islands (Orta 1992). Like other Daneri 1960, Aramburu and Bo 1961, De la
i cormorants and shags (see Johnsgard 1993 for Pena 1980, Mosqueira et al. 1987) but rela-
a review), they are foot-propelled pursuit-di- tively little has been published on marine pop-
;
vers, although they can feed by plunge-diving ulations (Yorio et al. 1994, Quintana et al.
under special sea conditions (Duffy et al. 2002b). To date, there is little information on
!
1986, Humphrey and Rasmussen 1988). the feeding ecology of the Neotropic Cormo-
In Argentina, the Neotropic Cormorant is a rant, and no available data on diving param-
I widespread and abundant species, with colo- eters of birds feeding in freshwater or marine
nies reaching up to several thousand individ- environments. Here, we report on the di\ing
Lials at some wetlands in northern and central behavior and use of I'oraging areas of this spe-
I
cies in inshore marine waters of Patagonia,
Argentina.
' ' Centro Nacional Patagonico (CONICP7f) and
Wildlife Conservation Society, Blvd. Brown 3500, MITHODS
(9120) Puerto Madryn, Chubut. Argentina.
^ F^'acultad deCiencias Naturalesy Museo. Univ. Na- During the 2001 breeding .season, we stutl-
, cional de I.a Plata (1900), La Plata, Buenos Aires, Ar- ietl the diving behavit)r and determined the
gentina. feeding areas of the Neotrojiic C'ormorant at
I
I ' Lacultad de C'iencias Lxactas y Naturales. Univ. tie a colony of 120 breeding pairs at Isla Vernaci
3B7u0e*0nU.onis(v9.A1i2r0Ne)ascPi(uoCen1rat4lo28MdLaeHdArl)ya,n.PBa(utdeaungioobnsuita,A.iArerBsgI,evntAIt.rigneaBn.tmiwnan. .WS)u,doe(sftleui,bu(t',aleAtragMeantliansapindai(g.451).1 1T'h,iS,s66i°sla3i0u'l
'C'orrespontling author; e-mail; comi^lex is located at the mouth of an inlet
quintana^Aenpat.etlu.ar (35 km'; Herrera 1997) where 9 of 16 seabirtl
83
1
84 THE WILSON BULLETIN • Vol. 116, No. I, March 2004
LIG. 1. Loraging locations and density contours (resulting from kernel estimation oflocations) forNeotropic
Cormorants breeding at Isla Vernaci Sudoeste (Caleta Malaspina, Argentina), 2001. Contoursencompass 50, 75,
and 95% oflocations.
species that breed along the Argentine Pata- Receivers were single channel (R161A model.
gonian coast have colonies (Yorio et al. 1998). Advanced Telemetry Systems, MN) connected
A VHP radio transmitter (Standard model. to hand held 3-element Yagi aerials (Ad-
Advanced Telemetry Systems, Bethel, MN) vanced Telemetry Systems). Radio signals
was attached to eight adult birds during the disappeared when birds were underwater and
late incubation period (29 November). Radios resumed when cormorants resurfaced. It was
were fixed to the two central tail feathers us- therefore possible to identify a sequence of
ing waterprooftape (Wanless et al. 1998). The dives as a series ofregular breaks in the signal
procedure was completed in less than 5 min (see Wanless et al. 1991). Dive duration cor-
and the released birds flew directly to sea be- responded to the time interval over which the
fore returning to their nests. Mean weight of signal was interrupted and recovery times (the
radios was 18.4 g (SD = 0.2, n = 6), approx- interval at the surface between successive
imately 1% of adult body mass (mean = 1.65 dives) to the time during which a signal was
kg, SD = 0.2, /? = 8 birds). Transmitters did audible. Dive efficiency was defined as the ra-
not appear to affect cormorant breeding and tio between mean dive duration and mean re-
feeding activities. covery time (Dewar 1924). Birds were radio-
Radio signals were monitored from 1 tracked throughout daylight hours between 29
tracking stations separated by 0.7-9.3 km, lo- November and 3 December 2001.
cated on shore at 10-50 m above sea-level. Birds were treated individually because dif-
Quintana et al. • DIVING BEHAVIOR OF NEOTROPIC CORMORANT 85
TABLE 1. Dive parameters ofNeotropic CormorantsatCaletaMalaspina, Argentina, duringthe 2001 breed-
ing season.
Meandive Meanrecovery
Numberof Trips duration(sec) Maximumdive time^(sec) %time
Bird dives recorded ± SD duration(sec) ± SD underwater± SD
14 213 4 15.2 ± 6.9 33 5.0 ± 2.3 69.2 ± 9.2
54 113 2 16.9 ± 4.1 28 6.2 ± 2.2 72.6 ± 3.6
104 6 1 22.8 ± 7.0 30 9.8 ± 1.2 69.9
133 251 5 15.9 ± 8.5 35 6.5 ± 4.0 58.4 ± 17.7
143 95 2 13.4 ± 8.2 30 5.5 ± 3.5 70.2 ± 3.7
153 176 5 28.6 ± 8.9 43 7.1 ± 2.4 79.4 ± 4.4
163 65 1 19.3 ± 6.1 32 6.8 ± 2.5 71.5
Total 919 20
Mean ± SD 18.9 ± 5.3 6.7 ± 1.5 70.2 ± 6.2
“Excludesrecoverytimes>20sec.
ferences between birds were significant (one- characterize the spatial distribution of feeding
way ANOVA, P < 0.05). Means of diving locations following Wood et al. (2000). We
parameters were determined for each individ- defined three categories of activity ranges as
ual and then pooled to calculate overall mean the areas encompassing 95, 75, and 50% of
values. The use of non-independent observa- locations at sea.
tions is valid if the replicates are pooled to
estimate a mean value (Hurlbert 1984). RESULTS
—
Feeding areas were dehned as the locations Diving behavior We recorded 919 dives
.
where birds dived repeatedly; these were es- from seven incubating birds during 20 feeding
timated by triangulation from the direction of trips (Table 1). We were unable to obtain data
the radio signals. During feeding trips, and from one of the eight instrumented birds due
while birds were at the surface between con- to radio failure. Mean dive duration was 18.9
secutive dives, at least two stations simulta- sec (SD = 5.3, n = 1 birds; Table 1) and mod-
neously recorded the direction ofthe most sta- al dive duration was 15-20 sec. Only 3% of
ble and strongest signal. Observers at the two the dives {n = 919) were longer than 35 sec
tracking stations were in radio contact. To (Fig. 2). Maximum dive duration was 43 sec
control for errors, an observer with binoculars and birds spent between 58 and 79% of their
aponsditiaonhsa.ndW-eheulsdedaekreiarlnelchdeecnkseidtyaesstaimmaptlees otof ttiiomnedaitffseeraeduanmdeornwgatienrdiv(iTdauballes 1()K.ruDsikvael-dWuarla--
lis H = 252.5, n = 887, P < 0.001 ). Recovery
time, excluding prolonged surface intervals
(PSI) longer than 20 sec, was 6.7 sec (SD =
1.5, /? = 7 birds; Table 1). PSls ranged from
21 to 199 sec and were infrecpient (3% oftotal
surface intervals). Recovery time differed
among individuals (Kruskal-Wallis // = 72.3,
n = 871, P < O.OOl). Mean diving efficiency
was 2.6 ± 0.5 (// = 6 birds). The correlation
between dive duration and sub.sequent recov-
ery time was low (H = 0.00()3, n = 892.
> 0.05). Recovery time increased linearly
0 5 10 15 20 25 30 35 40 45 with duration of the previous dive (Recovery
Diveduration(sec) time = 7.0 + 0.02 X Dive duration) but the
= Jy'iICni.d2i.ves)Frteocr|uNeenoctyrodpiistcriCb’uotrimoonroafntdsivaet dCuarlaettiaonMsal(/-; sPlo>pe0.w0a5s).low and non-signilicant (/^.m = 0.5,
aspina, Argentina. 2001. (.luring the late incubation pe- Both ilive duration and recovery time dif-
riod. fered among trips for three birds, while only
86 THE WILSON BULLETIN • Vol. H6, No. I, March 2004
one (either dive duration or recovery time) dives, followed by short intervals at the sur-
differed for two other birds (Kruskal-Wallis face. Morrison et al. (1978) reported a similar
and Mann-Whitney L/-tests, // > 14 and Z > diving pattern for adult Neotropic Cormorants
3.5, n = 94-236 dives per bird, all P < 0.05). foraging in a freshwater pond in Texas, USA.
We excluded two birds (nos. 104 and 163) Diving durations and recovery times of the
from the analysis, as only one trip was re- Neotropic Cormorant were markedly shorter
corded for each of these birds. than those ofRock Shags (Phalacrocoraxma-
In a diving sequence, the birds performed gellaniciis) and Imperial Cormorants {Phala-
a general pattern of immersions longer than crocorax atriceps) breeding in the same study
the intervals on the surface, except forone trip area. Neotropic Cormorants performed shorter
of bird no. 133. Time underwater during for- dives (18.9 sec compared to 47.2 and 94.8 sec
aging trips varied from 31 to 84% o—f the time for Rock Shags and Imperial Cormorants, re-
at sea, but in 60% of the trips {n 20) the spectively) and spent shorter intervals at the
birds foraged underwater more than 70% of surface (6.7 sec compared to 15.5 and 163.6
the time. Within a feeding trip, dive duration sec for the Rock Shag and Imperial Cormo-
and recovery time were variable, with CV rant, respectively) (Quintana 1999, Quintana
ranging from 13 to 54% and 20 to 252% for et al. 2002a, Sapoznikow and Quintana 2003;
dive duration and recovery time, respectively. FQ unpubl. data).
—
Foraging areas. Both telemetry and ob- While Rock Shags and Imperial Cormo-
servational data indicated that Neotropic Cor- rants show a diving pattern that is character-
morants from Isla Vernaci Sudoeste fed in in- istic of bottom feeders (Quintana 1999, Quin-
shore and shallow waters (<1.3 km from the tana et al. 2002a, Sapoznikow and Quintana
coast and <10 m depth) both inside and out- 2003; FQ unpubl. data), the Neotropic Cor-
side the inlet. We were able to locate birds morant shows a more flexible diving strategy.
while feeding inside the inlet on 112 occa- Wilson and Wilson (1988) suggested that be-
sions (52%, 11 = 215). However, we did not sides physiological constraints related to body
receive a signal on 103 occasions (48%, n = mass, differences in diving behavior could re-
215) while the birds were absent from the col- flect variation in features of foraging areas
ony. Both tagged and untagged individuals such as bathymetry, bottom topography, water
were observed outside the inlet diving in in- turbidity, and type and/or availability of prey.
shore waters in small bays located at both Because our data from Rock Shags and Im-
sides of its mouth. perial Cormorants also come from adult birds
Inside the inlet, 71 feeding areas were iden- feeding within Caleta Malaspina during the
tified by means of radio telemetry for 68 for- same part of the season, differences in envi-
aging trips (three foraging trips included two ronmental characteristics of foraging areas do
locations; Fig. 1). We also observed four of not seem to be a plausible explanation for dif-
the eight instrumented birds feeding inside the ferences in diving behavior. Rather, differenc-
inlet on a total of six occasions (Fig. 1). es in diving depths, foraging techniques, and
Eighty seven percent of mapped locations {n diet may best explain dissimilar diving pat-
= 71) were within 2.5 km of the colony, near terns found among these three sympatric spe-
the mouth ofthe inlet (Fig. 1). During a single cies.
trip, cormorants almost invariably used the In our study area. Neotropic Cormorants
same foraging area; on only 2 of the 68 trips have been observed foraging individually as
did the birds change foraging sites. well as in flocks (FQ pers. obs.); Imperial Cor-
morants seldom engage in group foraging and
DISCUSSION
Rock Shags always forage alone (Quintana et
—
Diving behavior. Our study presents the al. 2002a, Punta et al. 2003). In addition, even
first description of the diving behavior of the when pursuit diving appears to be the usual
Neotropic Cormorant in marine waters. Our feeding behavior. Neotropic Cormorants have
results show that Neotropic Cormorants dive been observed foraging in flocks by plunge-
near the coast in shallow waters. The diving diving in very shallow and coastal waters
pattern consisted of extremely short dives not (Duffy et al. 1986, Humphrey and Rasmussen
exceeding 20 sec in 97% of the recorded 1988; FQ pers. obs). Similar foraging tech-
Quintana et al. • DIVING BEHAVIOR OF NEOTROPIC CORMORANT 87
niques for Neotropic Cormorants also have birds (H. Hernandez pers. comm.). This and
been observed in Puerto Deseado, Santa Cruz, other activities, such as mariculture and oc-
Argentina (R A. Gandini and E. Frere pers. casional guano harvesting from Imperial Cor-
comm.). In accordance with the more flexible morant colonies, also likely affect the foraging
foraging strategy shown by Neotropic Cor- behavior and spatial distribution of Neotropic
morants in our study area, preliminary data Cormorants. Our study on foraging patterns
suggest that their diet includes a higher pro- and locations should help in the definition of
portion of surface-schooling fish than that of spatial and temporal harvest zoning schemes
the other two cormorant species (Malacalza et and help prevent modification of key feeding
al. 1994, 1997; Gosztonyi and Kuba 1998; habitats.
Punta et al. 2003; FQ and PY unpubl. data). ACKNOWFEDGMENTS
The high within-trip variability in dive dura-
tion and recovery time found in Neotropic Research was funded by grants from the Wildlife
Cormorants in this study is congruent with Conservation Society, Consejo Nacional de Investi-
their flexible foraging techniques and diet. In gaciones CientificasyTecnicas, and Agenciade Prom-
Caleta Malaspina, Neotropic Cormorants ap- ocion Cientifica y Tecnologica. We thank R. Vera and
M. A. Diaz for field assistance and P. Dell’Arciprete
pear to spend time searching for sheltered, but for spatial analysis ofdata. We thank R. C. Telfair, II,
predictable, bottom-dwelling prey as well as and two anonymous reviewers for comments that im-
pelagic schooling fish when they occur in the proved the manuscript. We also thankCentro Nacional
area. This probably results in the variability in Patagonico (CONICET) for institutional support and
their diving parameters. A theoretical model Soriano S.A. for logistical support.
developed by Gremillet and Wilson (1999) for FITERATURE CITED
the Great Cormorant (Phalacrocorax carbo)
predicts that cormorant diving parameters are Aramburu, R. H. and N. A. Bo. 1961. Descripcidn de
most strongly influenced by prey availability. colonias de nidificacion (Deltadel Paranay Golfo
— San Jose, Chubut) y estudio de los estados juve-
Foraging areas Neotropic Cormorants
. niles de Phalacrocorax hrasilianus (Gmelin). Re-
often foraged within Caleta Malaspina in an vistadel Museo de Ciencias Naturales de La Plata
area close to the colony at the mouth of the 7:107-121.
inlet. We did not detect radio signals almost Bo, N. A. 1956. Observacionesecoldgicasyetoldgicas
half of the time birds were absent from the sobre el Bigua. Hornero 10:147-157.
Daneri, C. a. 1960. La nidihcacidn del Bigua Phal-
colony, suggesting that the cormorants also
acrocorax olivaceus olivaceus (H.) en Puerto De-
regularly forage outside the inlet. In some of seado. Physis 21:273-277.
these cases, however, birds may have been De La Pena, M. R. 1980. Notas nidoldgicas sobre big-
feeding in areas behind islands or in small uaes y cormoranes (Aves: Anhingidae y Phalacro-
bays hidden behind elevated terrain, which in- coracidae). Historia Natural 1:109-1 12.
terfered with signal transmission. This was Dewar, J. M. 1924. The bird as a diver: acontribution
confirmed in several cases through direct ob- to the natural history of diving birds. H. E &. G.
Witherby, London, United Kingdom.
sneerovuastiroandioof-ttraacgkgiendg.inOdbisveirdvuaatlsionasndalsoinmgulttah-e DiJtJVYE,EADs.gUCF.-,zRR..P1.98W6i.iP.slouNn,geM-.divP.inWgiEb.ysoONl,ivaancdeouC.s
outer coasts of the inlet revealed that birds Cormorants in Chile. Wilson Bulletin98:607-(i08.
forage in inshore waters of small bays. Gosztonyi, A. E. and L. Ki'ba. 1998. I'ishcs in the
Foraging activity was concentrated in spe- diet of the Imperial Cormorant Phalacrocorax
cific coastal sections of Caleta Malaspina, atricepsat Punta Loben'a, Chubut, Argentina. Ma-
mostly at the mouth of the inlet; we rarely GRt(Mirii.in.ei;Orr,nIi)t.hAolNoDgyR.26P.:5W9i-i6s1o.n. 1999. A life in the
observed foraging at the inland end of the in- fast lane: energetics and foraging strategies ofthe
let. Birds there generally foraged in mixed Great ('ormorant. Behavioral licology 10:516
feeding Hocks, likely taking advantage of pe- 524.
lagic prey which had moved into that part of Ill KRIRA, G. O. 1997. Dieta reproducti\a ile la ga\it)ta
the inlet. Neotropic Cormorants also regularly tie Olrog Ixiru.satlanticus en la pro\inciailel ('hu-
occur in association with the commercial har- but. Undergratluate thesis. Universitlad Nacional
de la Patagonia San Juan Bosco. Puerto Matlryn.
vesting activities of macroalgae, as the har- Argentina.
vesting sledge stirs the ocean bottom and like- lllMI’IIKIA. P. S.. P. C. RaSMISSIN. AND N. I.OPIZ.
ly makes bottom lishes more accessible to the 1988. I'ish surface acti\ity aiul pursuit-plunging
.
88 THE WILSON BULLETIN • Vol. 116, No. I, March 2004
by Olivaceous Cormorants. Wilson Bulletin 100; Cormoran Cuello Negro, Phalacrocorax magel-
27-328. lanicus, en dos colonias de la costa Patagonica.
Hurlbert, S. H. 1984. Pseudoreplication and the de- Ecologia Austral 12:19-28.
sign of ecological field experiments. Ecological Quintana, E, P. Yorio, and P. Garcia Borboroglu.
Monographs 54:187-21 1 2002b. Aspects of the breeding biology of the
JoHNSGARD, P. A. 1993. Cormorants, darters, and pel- Neotropic Cormorant Phalacrocorax olivaceus at
icans of the world. Smithsonian Institution Press, Golfo San Jorge, Patagonia, Argentina. Marine
Washington, D.C. Ornithology 30:25-29.
Malacalza, V. E., N. M. Bertellotti, and T. I. Po- Sapoznikow, a. and F. Quintana. 2003. Foraging be-
RETTi. 1997. Variacion estacional de la dieta de havior and feeding locations of Imperial Cormo-
Phalacrocorax magellanicus (Aves: Phalacrocor- rants and Rock Shags breeding sympatrically in
acidae) en Punta Loma (Chubut, Argentina). Neo- Patagonia, Argentina. Waterbirds 26:184-191.
tropica 43:35-37. Telfair, R. C., II, and M. L. Morrison. 1995. Neo-
Malacalza, V. E., T. I. Poretti, and N. M. Bertel- tropic Cormorant (Phalacrocorax brasilianus).
LOTTI. 1994. Ladietade Phalacrocoraxalbiventer The Birds ofNorth America, no. 137.
en Punta Leon (Chubut, Argentina) durante la Wanless, S., D. Gremillet, and M. P. Harris. 1998.
temporada reproductiva. Ornitologia Neotropical Foraging activity and performance ofShagsPhal-
5:91-97. acrocorax aristotelis in relation to environmental
Morrison, M. L., R. D. Slack, and E. Shanley, Jr. characteristics. Journal of Avian Biology 29:49-
1978. Age and foraging ability relationships of 54.
Olivaceous Cormorants. Wilson Bulletin 90:414- Wanless, S., M. P. Harris, and J. A. Morris. 1991.
422. Foraging range and feeding locations of Shags
Mosqueira, M. E., M. V. Albeza, andG. M. DeGon- Phalacrocorax aristotelis during chick rearing.
zo. 1987. Biologiareproductivade Phalacrocorax Ibis 133:30-36.
olivaceus (Humboldt, 1905) en el valle de Lerma. Wilson, R. P. and M. P. T. Wilson. 1988. Foraging
Salta. Revista de la Asociacion de Ciencias Na- behavior in four sympatric cormorants. Journal of
turales del Litoral 18:163-173. Animal Ecology 57:943-955.
Navas, J. R. 1993. Aves: Podicipediformesy Pelecan- Wood, A. G., B. Naef-Daenzer, P. A. Prince, and J.
iformes. Eauna de agua dulce de la Republica Ar- P. Croxall. 2000. Quantifying habitat use in sat-
gentina, vol. 43 (U). Profadu (CONICET), Museo ellite-tracked pelagic seabirds: application ofker-
de la Plata, La Plata, Argentina. nel estimation to albatross locations. Journal of
Orta,J. 1992. Family Phalacrocoracidae(cormorants). Avian Biology 31:278-286.
Pages 326-353 in Handbook of the birds of the Yorio, P, E. Frere, P. Gandini, and W. Conway.
world, vol. 1: ostrich to ducks (J. del Hoyo, A. 1999. Status and conservation of seabirds breed-
Elliott, and J. Sargatal, Eds.). Lynx Edicions, Bar- ing in Argentina. Bird Conservation International
celona, Spain. 9:299-314.
Punta, G., P. Yorio, andG. Herrera. 2003. Temporal Yorio, P, E. Frere, P Gandini, andG. Harris (Eds.).
patterns in the diet and food partitioning in Im- 1998. Atlas de la distribucion reproductiva de
perial Cormorants (Phalacrocorax atriceps) and aves marinas en el litoral Patagonico Argentino.
Rock Shags (P. magellanicus) breeding at Bahia Plan de manejo integrado de la zona costera Pa-
Bustamante, Argentina. Wilson Bulletin 115:307- tagonica. Fundacion Patagonia Natural and Wild-
315. life Conservation Society. Instituto Salesiano de
Quintana, E 1999. Diving behavior ofRock Shags at Artes Graficas, Buenos Aires, Argentina.
a Patagonian Colony ofArgentina. Waterbirds22: Yorio, R, F. Quintana, C. Campagna, andG. Harris.
466-471. 1994. Diversidad, abundanciaydinamicaespacio-
Quintana, E, F. Morelli, and Y. Benedetti. 2002a. temporal de la colonia mixta de aves marinas en
Buceo eficiente en aguas poco profundas: com- PuntaLeon, Patagonia. OrnitologiaNeotropical5;
portamientodebuceoypatrondealimentaciondel 69-77.