9

HARVARD UNIVERSITY

Library of the

Museum of

Comparative Zoology

OCCASIONAL PAPERS

of the

MUSEUM OF NATURAL HISTORY
The University of Kansas
Lawrence, Kansas

NUMBER 126, PAGES 1-41 MARCH 3. 1989

FEEDING MORPHOLOGY, FORAGING BEHAVIOR,

AND FOODS OF STEAMER-DUCKS

(ANATIDAE: TACHYERES)

By

WA.R 1 Q

Bradley C. Livezey^

Vo^^

Steamer-ducks are large, diving members of the shelducks (Tadorriinae;
Livezey, 1986) that are limited in distribution to southern South America.
Four species are recognized (Murphy, 1936; Humphrey and Thompson,
1981): Magellanic FUghtless Steamer-Duck (J. pteneres) of marine coasts
of southern Chile (south of 40*^ S), Tierra del Fuego, and Isla de los Estados;
White-headed Rightless Steamer-Duck (r. /eMC(9C£/?/ia/u5) of coastal Chubut,
Argentina; Falkland Flightless Steamer-Duck {T. brachypterus) of marine
shores of the Falkland Islands (Fig. 1); and Rying Steamer-Duck {T. pat-
achonicus), which occurs on marine coasts and freshwater lakes through-
out southern Argentina and Chile (south of 40° S), Isla de los Estados, and
the Falkland Islands. The three flightless species are mutually allopatric,
but each is sympatric, at least seasonally, with T. patachonicus.

Although steamer-ducks are relatively well known among ornithologists
for the permanent flightlessness that characterizes three of the species
(Livezey and Humphrey, 1986), as well as some individuals of the'flying'
species (Humphrey and Livezey, 1982a), little is known of their natural his-
tory. This is especially true of the feeding ecology of steamer-ducks. With
a few notable exceptions (e.g.. Murphy, 1936; Weller, 1972, 1976), only
a few brief anecdotal comments are scattered through a literature spanning
150 years.

During our recent research on the systematics, ecology, and flightless-
ness of the genus Tachyeres (Humphrey and Livezey, 1982a, 1982b, 1985;

'Museum of Natural History and Department of Systematics and Ecology, University of
Kansas, Lawrence, Kansas 66045 - 2454.

OCCASIONAL PAPERS MUSEUM OF NATURAL fflSTORY

m»4^-

Fig 1. Photographs of T. brachypterus feeding in marine habitat in the Falkland Islands.
Top: Pair foraging in kelp bed near shore, the head of the male is largely submerged. Bottom:
Male ingesting food in shallow water, a piece of plant material is visible in its bill.

Livezey and Humphrey, 1983, 1984a, 1984b, 1986), Philip S. Humphrey and
I collected data on the feeding morphology, foraging behavior, and stomach
contents of birds observed or collected. This information provided an oppor-
tunity to contribute to the knowledge of the natural history of the genus, to

STEAMER-DUCK FEEDING ECOLOGY 3

make ecological and behavioral comparisons of the flightless and flying
species of steamer-duck where sympatric, and to test for habitat-related
differences in feeding ecology within the relatively widespread T. patachoni-
cus. In this paper I present these analyses, summarize previously published
information, and compare the feeding ecology of steamer-ducks to that of
other waterfowl,

STUDY AREAS AND METHODS

Localities and Collecting Techniques

Sites and dates of collection of specimens and other data for each species
were as follows: T. pteneres at Ushuaia and vicinity, Tierra del Fuego
(Argentina) during November-December 1979 and December 1980-January
1981, and at Puerto Montt and vicinity, Region X (Chile) during December
1982-January 1983; T. leucocephalus at Puerto Melo, Chubut (Argentina)
during September-October 1979, February and December 198 1 , and January
1982; 7. /j<3/ac/i<?/j/cM^at Ushuaia and vicinity during November 1979 and De-
cember 1980-January 1981, at Puerto Deseado, Santa Cruz (Argentina)
during October 1979 and January-February 1981, at Puerto Melo during
September-October 1979 and December 1981, on Andean lakes in Santa
Cruz and Chubut (Argentina) during December 198 1-January 1982, and near
Puerto Montt during December 1982-January 1983; and T. brachypterus at
Port S tanley and Lively Island (east Falkland Islands) during January-February
1984. Excluding a few birds found dead on beaches, birds were collected by
shooting or, rarely, by hand capture. Collecting sites were chosen primarily
on the basis of the abundance of steamer-ducks and logistic considerations.
The cumulative sample sizes of specimens from this field work were T.
patachonicus 88, T. brachypterus 21,7. leucocephalus 34, and T. pteneres 20,
although associated data for the specimens taken in 1979 were limited.
Additional data on body mass, culmen length, and nail width were taken from
specimens available at institutions other than the University of Kansas (see
Acknowledgments).

Feeding Morphology

Within a few hours after collection, birds were weighed and the following
bill measurements were taken (to the nearest mm): length of exposed culmen,
width of nail, depth (height) of upper bill at base, length of upper bill along
torn ium , width of gape at base of upper bill , and the number of lamellae on one
side of the upper bill. Gizzards were detached, cleaned and weighed, and the
length and width (at orbit) of supraorbital glands were recorded. A few heads
of birds were saved in alcohol for description of bill morphology.

4 OCCASIONAL PAPERS MUSEUM OF NATURAL fflSTORY

Feeding Habitats and Behavior

When possible, I recorded the following data for each feeding bird
observed: species, sex, age, date, time, locality, weather, flock size, propor-
tion of flock feeding, macrohabitat (protected saltwater bay, exposed bay,
open saltwater, river mouth, inland brackish, or freshwater lake), feeding
behavior of adults and ducklings, shore type, and distance to shore, water, and
kelp beds. I distinguished four types of feeding behavior: dabbling on sur-
face with body parallel to water surface, sometimes with head partially or
completely submerged; tipping on surface with head and neck submerged and
with body and tail held roughly perpendicular to the water surface; picking of
food items from exposed mud, typically while walking; and diving below the
water surface. Durations ofdives (dive times) were measured in seconds. The
term duckling refers to members of apparent broods. A bout was defined as
a period of feeding that was temporally distinguished from other feeding
periods of a given bird by a period of loafing on land, or at least one day since
the last observation.

Stomach Contents

As a basis for comparison, I compiled the previously pubUshed data on
foods of Tachyeres by species. For all specimens collected after 1979,
contents of the esophagus and gizzard were removed separately, weighed, and
preserved in 10% formaUn. Later each of these samples was subdivided into
constituent prey taxa which then were weighed separately. Grit was separated
from other contents of gizzards (manually and by rinses and acid treatments);
these samples were dried and sifted into 1 1 size classes using Tyler sieves
(nos. 2-9, 14, 20, and 28), and then subsamples were weighed separately to
within 0.1 g.

Statistical Analyses

Morphological measurements of birds grouped by species and sex were
compared using two-way analyses of variance (ANOVA). Associated pairwise
comparisons were based on f-tests using pooled variance estimates unless
group variances were found to be significantly different {P < 0.05) by
Levene's test (LT; Brown and Forsythe, 1974), which necessitated the use of
approximate t *-tests. Lamellar density (the ratio of lamellar number to bill
length) was log-transformed before analysis. Approximate areas (length x
width) of the left and right supraorbital glands were estimated and averaged
within species-sex groups. Supraorbital areas and masses of gizzards were
log-transformed for statistical comparisons using ANOVA. An analysis of
covariance (ANCOVA), in which the effects associated with total body mass
were removed, also was used for species-sex comparisons of these characters.

STEAMER-DUCK FEEDING ECOLOGY 5

Finally, a canonical analysis (CA) was performed on the bill measurements,
grouping birds by species and sex.

Associations between birds and characteristics of sites were tested using
X^ tests with Yate's correction, and quantified using odds ratios (Agresti,
1984). Tests were performed on numbers of flocks rather than numbers of in-
dividuals because of the probable dependence among members of flocks.

Dive times were summarized by species, sex, and age classes using overall
means and ranges and by:

within-bout standard deviations,

S^=a,lj{x,j-^f/(n-m)y'';

among-bout standard deviations,

and total standard deviations,

5,= {Z,Z.^-3c..)V(n-l)}^'^

In these statistics, m is the number of bouts, n the number of dives in the
fth bout (L.n. = n), x.. is theyth observation in the zth bout, 3c .is the mean dive
time for the iih bout, and 3c .. is the grand mean of the dive times.

Average dive times for bouts were compared across species and sexes
using Kruskal-Wallis tests (KWT), and variances of times within bouts were
compared using Squared-Ranks Tests for Variances (SRTV). Sexual differ-
ences in average dive times were tested within species using Mann-Whitney
tests (MWT) on individual dives if bout differences were not significant,
MWTs on bout means if bout effects were significant, or Wilcoxon Signed-
Rank tests (WSRT) on bout means if bout effects were significant and a
substantial proportion of the bouts were 'paired.' Paired dives were those
performed by consorting birds at the same time and place; they often appeared
to have been initiated synchronously. Significance of the pairing effects was
tested using Spearman Rank correlation (SRC) of individual dives performed
simultaneously by paired birds and of paired bout means for a number of
different pairs of birds. SRCs were used to test for relationships between
consecutive dive times for individuals and between durations of dives and im-
mediately following rest periods but, like the tests for pairing effects, were
done only on T. brachypterus for which I had recorded adequate numbers of
uninterrupted, paired sequences of dives and rest periods. Interspecific dif-
ferences in mean dive times were tested using the KWT based on bout means
in species lacking significant sexual differences in dive times and for the un-
weighted average of the paired male and female bout averages in species

6 OCCASIONAL PAPERS MUSEUM OF NATURAL fflSTORY

showing significant sexual differences.

Foods of steamer-ducks were analyzed by calculating the percentages of
specimens of each species that contained each prey taxon. For those
specimens containing a given prey type, an average percentage of the mass of
the total stomach contents (wet mass of preserved contents without grit) was
calculated for each species.

Mean total masses of dry grit samples were compared by species, sex, and
habitat (T. patachonicus only) using MWTs for two groups, and with KWTs
and protected pairwise tests for comparisons of three or more groups.
Variances of total masses of grit were compared using SRTVs. Median grit
sizes (size intervals containing the sample median) were compared for sexes
within species using MWT, and among species and habitat groups with KWTs
and associated pairwise comparisons. Grit-size diversities (H' = - X,/>, In/?.;
p.= proportion in i\h size interval) of individual birds were compared by sex
and species-habitat groups using the same tests.

Tests and critical values used are described in standard references on
parametric (Neter and Wasserman, 1974) and nonparametric statistics
(Conover, 1982). Computer programs used for ANOVA, ANCOVA, and CA
are part of the Biomedical Computer Programs (Dixon, 1985).

RESULTS

Morphology

Total body masses of steamer-ducks (Table 1) differed significantly (two-
way ANOVA) among species (F = 330.47, df = 3, 181; P < 0.0001) and sex
classes (F= 189.43, df= 1, 181;F<0.0001). Significant interaction between
species and sex effects (ANOVA; F = 5.13, df = 3, 181; /' < 0.0005) indicated
differences among species in sexual dimorphism; ratios of male means to
female means were 1.26 for the relatively small T. patachonicus, 1.22 and

Table 1. Total body masses (g) of steamer-ducks by species and sex.

Standard
Species^ Sex*" n Mean Deviation Range

T. patachonicus
T. hrachypterus
T. leucocephalus
T. pteneres

t t

54

2958

296

2100-3600

22

49

2347

262

1800-2900

i$

10

4200

484

3300-4800

22

10

3450

264

2900-3900

$S

17

3944

288

3500-4400

22

16

3013

295

2450-3550

$i

16

5394

392

4950-6500

''■: '?

15

4228

478

3400-5000

^Species differed (P< 0.0001) in two-way ANOVA.
''Sexes ditTered (f< 0.0001) in two-way ANOVA.

STEAMER-DUCK FEEDING ECOLOGY 7

1 .26 for the intermediate-sized T. brachypterus and T. leucocephalus, respec-
tively, and 1.28 for the largest species, T. pteneres.

The powerful grasping bills of steamer-ducks (Fig. 2) differed in the five
linear measurements (Table 2) among species (ANOVA, P < 0.(XX)1) and, ex-
cept in culmen length and lamellar count, among sexes (ANOVA, P <
0.0001). Only bill width showed significant interaction effects (F = 2.79, df
= 3, 108; P < 0.05). Ranking of species by culmen length agreed with ranking
by average body mass. However, in the three flightless species, males, larger
overall than the females, had slightly smaller average culmen lengths than did
females. Nail width and bill width measures led to a different ranking; T.
pteneres and T. patachonicus had the largest and smallest means, respectively,
but means of male T. brachypterus and T. leucocephalus were more similar
to each other than to the means of their respective conspecific females; similar
overlap between species differences and sexual differences characterized the
rankings of bill length and bill depth (Table 2). Mean lamellar counts and den-
sities were ranked perfectly inversely with mean body size, i.e., small average
body size was associated with high lamellar densities and vice- versa (Table
2). In addition to greater lamellar density, T. patachonicus had longer, better-
developed lamellae than the flightless species (Fig. 2), a difference that was
noted previously by Bennett (1924, p. 280; 1926, p. 326).

Given that T. patachonicus is sympatric with each of the (mutually
allopatric) flightless species in parts of its distributional range, comparisons
between it and the flightless forms are of particular interest. Of all the bill
measurements and comparisons considered, only T. leucocephalus over-
lapped with T. patachonicus significantly (r-test, P > 0.05) in bill dimensions,
specifically in nail width, length and depth of bill, and count and density of
lamellae (Table 2). Among the flightless species there were many significant
differences in bill morphology as well; although complicated by varying
degrees of sexual dimorphism, bill measurements (except lamellar count) of
the flightless forms followed body size in interspecific rankings.

The C A achieved relatively high separation of species-sex groups (Wilks'
X = 0.0286, df = 6, 7, 107; approximate F = 13.01, df = 42, 482; P < 0.001;
74% correct jackknifed classification). The first canonical axis appeared to
be a contrast between overall bill size vs. number of lamellae and, to a lesser
extent, culmen length (Table 3). Scores on this axis differed among species
(P < 0.0001) and sexes (P < 0.0001), and groups essentially were ranked on
the first canonical axis by lamellar density (Fig. 3). The second canonical axis
evidently represented relative bill breadth; scores on this axis differed among
species (P < 0.001) and indicated that T. brachypterus had relatively narrower
bills than the other species (Table 3, Fig. 3).

Gizzards of steamer-ducks are large and muscular. Cunningham (1871a)
described and illustrated the digestive anatomy of Tachyeres including the
tongue, crop, and gizzard. Gizzard masses (Table 4) differed both among
species (ANOVA; F = 42.92, df = 3, 82; P < 0.0001) and sexes (ANOVA; F

OCCASIONAL PAPERS MUSEUM OF NATURAL HISTORY

"'^'^'^^^^^-^

mimrmMmrmmnTrf?,^,,,,^

E

2 cm

l3U«IMM««o Jy

Fig 2. Bill morphology of steamer-ducks. A: Lateral view of upper bill of female T.
palachonicus; B, C: Medial view of upper bill lamellae in female T. pteneres; D: Ventral view
of upper bUl of female T. pteneres; E: Medial view of left mandible of female T. pteneres; F:
Lateral view of right mandible of female T. pteneres.

STEAMER-DUCK FEEDING ECOLOGY

X

T3
C
03

O

<u
D.

x>
o

3
T3

1)

C
3
O
O

o

E
E

c
E

T3

o

c
o

>

T3

-a

c

C3

+1

C

-1

09

4J

5

E

^q

m

Ov

oo

r-

m

q

(^' ^^

d ^

— ^^

d

_;

_•

o

+1 vo

+1^

+i2

+1 VO
00 ^

OS

iri

r~

r»-i

Tt

d

E

IJU

_5J

E

-s:

Tl rr,

Tl r<-,
00 ^
v-j

OO

—

—

O)

o

ri ^

rj

ri

ri

d _

00

+1^

+1 VD

i/-i

U-;

\d

d

lo

r^

m

r^

rj-

d

— ;

On

00

—

t^

rn

s

ri ^

+12:

d —

+12

.4±1
(10)

0±1
(10)

0±0
(10)

6 + 2.
(10)

^2

o ^

r-'

ri

Ov

ON

^

_^

^

•/^

~™

\£)

rj

m

^

d

+1

^
m

00 ^^

+112

00 ^

ON

O;

OO

^

.5±0
(23)

^2

6±1,
(10)

+ 2

rn

1^

o

CO

rn

ON

d ^
+lCi

00 ^^

+ION

q "

NO

+ION
ON '~'

rg

+ION

q '~'

ON

OO

d ^
+1:2

ri

+i2

NO
NO

-r — 'I- o ^

m

00

TT

-*

ri ^

+iS

,8±0,
(64)

8 + 2.
(30)

+1
00

ri

NO

rj

+i2

ON —
ON

o

+i2

rj

NO

+1'^

Tl r<-)

\o

ON

d ^
+lS

NO

rj ^ — c

+1;

ON ■

Tl rn
\D ^

NO
rNi

+1

00

X)

03

Wl

c

(U

j;

-J

^

c
E

^

3

ca

U

z

00
c
u

PQ

13

D.
u

Q

n

NO

rj

+10^

+IOV
O ^

?

+i2

rj

rg ^^

Tl rn

NO

rJ ^^

Tl rn

C
3
O

U

CO

3

d ^
+10^
00 ^^

d

q
d

+ION
o ^

O

d ^

+i2

3

d

S

d —

Tl rr)

o ^

Tl fT)

o --

C
<U

Q

u

E
m

o
«

u

-o
E

00

_o

c
o

E

u
a.

c
o

Q.

C
(U

>

o

c

3
o
o

E

>*-

o

o

lU

s:
H

10 OZCASIONAL PAPERS MUSEUM OF NATURAL HISTORY

Table 3. Standardized canonical coefficients and related statistics for 115 steamer-ducks
grouped b> species and sex.

Canonical Coefficients-'

Variable

Axis I

Axis U

Culmen Lenaih
Nail Width "
Bill Lensth
BiU Width
Bill Depth
Lamellar Count

Eigenvalue
Vanance Explained (

0.238
-0.358
•0.470

0.166
■0.490

0.458

-0.(M5
0.893

-0.844
0.318

-0.258
0.322

9.193
i5.6

0.767

7.1

MZoefficients standardized by multiplication by pooled within-groups standard deviations of
original \ariables.

= 19.16, df= 1, 82; P < 0.0001), both in analyses involving all four species
or excluding ihe poorly represented T. pteneres; group variances also differed
(LT; r= 3.18, df = 6, 82; P <0.01). An ANCOVA, in which the effects of total
body mass were removed, showed that the relative masses of gizzards varied
among species {F = 29.08, df = 3, 80; P < 0.0001) but not between sexes (F
= 0.98, df = 1, 80; P > 0.95); interspecific differences also were significant if
the small sample for T. pteneres was excluded. Saltwater and freshwater T.
patachonicus did not differ in gizzard masses (ANOVA; F = 1 .34, df = 1 , 40;
P > 0.25). Pooling sexes, the mean relative gizzard masses (percentages of
total body mass) for the species are, in decreasing order: T. leucocephalus
(3.7%), T. patachonicus (2.8%), T. pteneres (2.5%), and T. brachypterus
(2.0%).

Sizes of supraorbital or 'salt' glands differed among species (ANOVA; F
= 42.15, df = 3, 97; P < 0.0001) and sexes (ANOVA; f = 9.11, df = 1, 47; P

y 3 7;

OQ

<

y z z
o

5 '■

Q -2 50 U

o

O .,7'

LU
CO

-5.00

»TEN i

PTEU

RATA 9

BRAC O RATA d

I I I

-6.00 -4 00 -3 00 - 50 ; 00

FIRST CANONCAL VARIABLE

Fig 3. Canonical p ot of bill morphology of steamer-ducks (n = 115) by species and sex;
only the extreme specur;»ns in each group are plotted.

STE.AMER-DUCK FEEDING ECOLOGY 1 1

T=iB LE - Masses of fresh empt> gizzards of non-juvenile steamer-ducks b>- species and sex.

Sr-e:.;- Sex N'lis of Giiz^r; z^

n

Mean

^ _' :_' i

18

"s.I
63.5

13.4

10
10

81.6
67.2

17.7
6.8

12
9

138.8
101.6

23.2

T. puiuLnoniLiLi
T. brachypterus
T. leucocephalus
T. pteneres

^Both species ar.; e ; :^e:r--c> sgnifcin: P<0.0001) in two-wa> .\NO\A.

< 0.005). As in rankings by body mass. T. piene^es had ihe largest glands, fol-
lowed by the approximately equal T. leucocephalus and T. brachypterus, and
T.patachonicush2Ld\hi smallest: average gland sizes of males exceeded those
of females in all species. Saltwater T. patachonicus had signific-andy larger
supraorbital glands than freshwater birds (two-way .■VS'OVA; F = 141.63, df
= 1, 60; P < 0.001): this difference also was significant in a size-correcting
ASCOW {F = 3.10, df = 1, 61: F < 0.0001). Consequendy, freshwater T.
patachonicus were excluded from interspecific comparisons of gland size.
Differences in gland size (excluding freshwater T. patachjonicus) were
significant among species (F = 37.09, df = 3, 72; P < 0.0001) and se.xes yF =
24.19. df = 1, 72: F< 0.005). Correcting for differences in body mass uith an
.\NCO\A, interspecific differences in glandular area (saltwater birds only)
were confirmed (F = 10.06, df= 3, 73; F < 0.0001) but intersexual dj^'erences
were not (F = 2.03, df= 1,73; F > 0.15); the former were due largely to the
relatively small glandular areas in T. pteneres and large glands in T. leu-
cocephalus.

Feeding Habitats

Steamer-ducks are particularly numerous in estuaries and near river
mouths (A'allentin, 1904; Weller, 1975; pers. obs.), and around islands.
Tacky ere s pteneres e\idently avoids highly tidal waterfronts, possibly to
avoid being stranded during low tide ("ReNiiolds in Lowe, 1934: Murphy,
1936: Weller. 1975, 1976\ T. leucocephalus, another flightless species,
inhabits similar but highly tidal habitats in Chubut. The fl\ing species. T.
patachonicus. also occurs on alkaline and freshwater lakes, especially in the
southern .\ndes and the Falkland Islands.

Steamer-ducks on both fresh and salt water were obsers-ed most frequendy
along shores dominated by rocky outcrops or stony beaches, and less
frequently on sandy or muddy shorelines (Table 5). Some interspecific
differences were susirested. however, from obsenations of feedine birds or

12 OCCASIONAL PAPERS MUSEUM OF NATURAL fflSTORY

Table 5. Percentages of birds observed adjacent to two major shore types and birds in or near
(less than 100 m away from) surface-visible kelp beds, by species of Tachyeres and age class.

Species

Adjacent Shore^

Near

Kelp

/!

Rocky

Other

n

%

T. patachonicus

saltwater

'adults'

271

16%

84%

Tl\

40%

ducklings

15

73

27

18

100

freshwater

'adults'

48

31

69

—

—

ducklings

20

10

90

—

—

T. brachypterus^
'adults'

_

34

100

ducklings

—

—

—

8

100

T leucocephalus

'adults'

572

22

78

305

69

ducklings

33

100

25

44

T pteneres
'adults'

157

85

15

142

62

ducklings

36

100

30

100

''Rocky indicates shore characterized by exposed, poorly weathered or solid rock. Other
includes shores covered by mud, sand, gravel, or cobbles.
''Shore types and method of observing and collecting birds not comparable.

birds encountered during collecting expeditions. Freshwater T. patachonicus
were found more often along muddy or gravel shores than were their marine
conspecifics (Table 5). This probably reflected a reduced availability of rocky
shores on freshwater lakes. Similarly, the relatively frequent use by T.
leucocephalus of waters adjacent to muddy and sandy shores may reflect the
preponderance of these shore types in the vicinity of Puerto Melo. However,
100% of the ducklings of T. leucocephalus were encountered near rocky
shores (Table 5) indicating that successful breeding birds prefer these rugged,
probably more productive shorelines for recruitment. An interspecific
difference in use of shoreUne type not attributable to differences in availabil-
ity was evident in syntopic saltwater populations of T. pteneres and T.
patachonicus in Chile and Tierra del Fuego. Tachyeres pteneres was encoun-
tered 14 times more often along rocky shores than the sympatric T. patachoni-
cus, which was observed more frequently along open beaches (x^ = 38.44
with Yate's correction, 121 flocks, P < 0.001; odds ratio = 14.4, with
continuity correction).

Marine steamer-ducks of all species, and especially broods, frequented
beds of kelp (Macrocystis pyrifera) (Table 5). The relatively low visitation
to kelp by saltwater T. patachonicus and T. leucocephalus may be due, in part,
to the sparseness of kelp in the estuary at Puerto Deseado and at Puerto Melo.
All species, regardless of habitat, typically were observed close to shore,
particularly members of broods (Fig. 4). T. brachypterus and T. leucocepha-
lus were characterized by somewhat higher frequencies of birds at moderate

STEAMER-DUCK FEEDING ECOLOGY

13

>-
O

UJ

Z)

o

LU

q:

LU

>

LU

02

02-

02-

02-

A
B

adults
broodlings

T. patachonicus

Saltwater
A=I29 B=I8

1 r

T. patachonicus

Freshwater
A = 37 B = I8

T. brochypterus

A=IOO BHO

W

mrn^

^

T. leucocephalus
A=I59 8=34

B^^^^^

T. pteneres

A=I09 B= 39

'mWM

10 30 50

DISTANCE FROM

SHORE (m)

Fig 4. Histograms of proximity to shore of steamer-ducks by species and age class; areas
of histograms reflect relative frequencies of individuals per 5 m interval. Vertical hatching
represents broodlings, stippling represents adults.

14 OCCASIONAL PAPERS MUSEUM OF NATURAL fflSTORY

distances from shore, probably because of the highly tidal, shallow water-
fronts at the study sites.

Methods of Feeding

Diving is the most important method of feeding for steamer-ducks, and
even was observed in downy young of all species. The details of this behavior
were described by Livezey and Humphrey (1984a) and an analysis of dive
times is presented below.

Dabbling was observed frequently in all species, especially in birds
feeding within 5 m of shore in water up to 0.5 m deep, or in exposed kelp beds
(Fig. 1). Dabbling birds searched the substrate with rapid movements of the
head, neck, and sometimes feet, and typically kept their heads submerged for
1-5 sec before resting. Dabbling bouts often lasted for 20-30 min or more,
and were observed most frequently near shore on low or rising tides.

Tipping was observed infrequently in T. pteneres, T. leucocephalus, and T.
brachypterus. Tipping was the method of foraging probably used by steamer-
ducks in removing mussels (Mytilus spp.) from aquacultural piles near
Chayahue, Chile (aquacultural personnel; pers. comm., 29 December 1982).
Picking of food items from exposed mud was observed only rarely, and only
in T. pteneres at low tide.

Durations of Dives

Steamer-ducks forage mostly by diving, and, as noted by Woods (1965, p.
121), generally remain submerged for about 30 sec (Table 6). However, dive
times are quite variable (most within-bout coefficients of variation approxi-
mated 20%; Table 6), and field observations prompted me to investigate
possible specific, sexual, and bout differences in diving. The few dive times
I recorded for freshwater Tachyeres patachonicus did not permit separate
analyses and were pooled with those for birds on saltwater.

Mean dive times differed (KWT; P < 0.005) among bouts within the eight
species-sex groups except male T. leucocephalus and T. pteneres. Within-
bout variances also differed (SRTV; P < 0.025) in female T. patachonicus,
male and female T. brachypterus, and male T. pteneres.

Differences between sexes in dive times were not significant (MWT or
WSRT; P > 0.20) in T. patachonicus, T. leucocephalus, or T. pteneres.
However, in T. brachypterus, for which I had dive times from numerous
'paired' bouts, males remained submerged significantly longer than females
(WSRT on individual paired dives; 7= 2.38, n=l9,P< 0.05). I found no sig-
nificant interspecific differences in dive times, but data for this test were
relatively few (KWT; 7= 4.65, n = 45; 0.10 <P< 0.25).

The relatively numerous sequences of dive times for pairs of T. bra-
chypterus permitted additional analyses. My impression in the field was that

STEAMER-DUCK FEEDING ECOLOGY

15

X

C/0

3
O
X)

c
u
u

00

c

3

-o
-a
E

>

-5
s:
c
o
•a

oil
c
'oi)

T3

a.

60

C
3

us

X) X)

^ £

3 O

•o P
' 5

i =

ca CO
oj „

t? 3

'*- o

x>

_ I

1 o

>.i

• w

U X

—J fl>

5 -

c
o

1)

00

c
ca
Qi

c
o

3

■3

pa

>
Q

00

c

i

T3

<

in

•0

c

«

00

1

r

00

I^

00
ON

fN

00

r<-i

^

m

;I^

3

!£

—

^

00

(N

t^

iri

00

r<-i

<N

r^

OS

irl

•-^

■0

</-!

lO

't

'^

'^

u->

■*

'^

T^

m

GO

t^

c

Q.

3
O
u

a

u

1

u

-«

u

to

D.

a

C/5

a

Ci.

III I I I I I I I I

5

>o O

I I I

00 r) 00

00

fN

00

ON

00

2;

00

(N

fN

f

3

m

1/-)

^

rt

ir-j

iri

— .

r-

Ov

m

'^

ON

u-i

00

ON

00

—

t--

vO

r~-

t^

^

r-

r-

lO

00

rn

>o

^

tN

tN

fN

00

5

i^

^

■*

<N

r^

m

00

>ri

—

t^

iri

[^

m

■T)

in

1 r-i

m

ro

m

'^

ON

ON

Tf

rt

Ov

ON in On

rs|

m

00

(N

t^

Ov

00

r-4

■^

m

(N

(N

m

fN

(N

(N

fN

fN

m fN

-H (N

fN

ON

rn \0 -H u-i in

(N — (N — fN

fN —1

ON

m

^

Ov

Ov

m

w->

r-~

■ri

fN

tT

m

<Ti

00

fN

^

tN

r<-i

'^

m

fN

IT)

ro

0+ -^
0+ c«

-

WD

00

3

c

Hn

CH

T3

»o

0+

^W

15

3
T3

ii 00

3 S

o

_ 3

0+ "O

CH U3

2

tj
s

Si

X

c

u
•a

c
3

o
IS

C3

c

1 6 OCCASIONAL PAPERS MUSEUM OF NATURAL HISTORY

dives of paired birds were initiated together, often led by the males. I did not
observe synchronous diving by larger flocks of birds, as reported by Cobb
(1933) for T. brachypterus. Only two of 12 paired sequences of dive times,
however, were significantly positively correlated with each other (SRC, one-
tailed, P < 0.05), perhaps due to independent termination of dives. A more
general pairing effect, the correlation of bout means for paired birds, was
highly significant (SRC; r = 0.71, n = 19; two-tailed, P < 0.001).

If birds were diving for periods at or near the limits of exhaustion, I
suspected that dive times would be related positively to subsequent rests and/
or related negatively to the following dive time. However, only one of 1 1 se-
quences of males and two of 12 sequences of females showed a positive cor-
relation between dive times and duration of following rests (SRC, one-tailed;
P < 0.05). Evidence was equally weak in sequential pairs of dives; none of
five sequences for males and only two of seven sequences for females showed
negative correlations (SRC, one-tailed; P < 0.05). Moreover, those few
sequences showing significant correlations between consecutive dives or
between dives and following rests were not characterized by exceptionally
long dives.

Stomach Contents

Published information on the diet of steamer-ducks indicates that molluscs
(principally mussels and limpets) and crustaceans (mostly crabs and shrimp)
are their chief food items (Appendix 1), although only seven genera in these
food groups had been identified prior to this study. Fish or their eggs rarely
were reported as food items, despite the widespread belief of local residents
that steamer-ducks are mainly piscivorous (pers. obs.). Plants were listed by
most authors as apparently incidental components of the diet. Stomach
contents from four young and six adult T. brachypterus collected by Weller
(1972, p. 34) indicated that young birds ate proportionately more snails, iso-
pods, and amphipods than did adults.

Brooks (1917, p. 156) found a crab with carapace "... two inches [5 cm]
long by one inch [2.5 cm] wide ..." in what was probably T. brachypterus.
Cobb (1933, p. 82) counted over 450 small mussel shells in a female steamer-
duck (probably T. brachypterus). Phillips (1925, p. 295) and Murphy (1936,
p. 962-963) reported that R. M. Beck found mussels (Mytilus) up to 55 mm
long in T. pteneres, and Beck (fide Murphy 1936) found in a single male T.
pteneres a snail (Euthria) 32 mm long and 14 mm in diameter, a chiton (Plax-
iphora setiger) about 27 mm long, and 12 imperforate limpets (Nacella
fuegiensis) up to 38 mm long and 27 mm wide.

In all species and habitat groups I sampled, molluscs and crustaceans were
the most important dietary components (Table 7). Molluscs were the primary
dietary constituent, occurring in 71% of the specimens containing food and
comprising an average of 7 1 % of their stomach contents. Mussels (especially

STEAMER-DUCK FEEDING ECOLOGY

17

4J ? g

2l5 o
ca *; a.
D.2 p

- s.>

o

Si

OO

o

E

u

L«

o

t^

en

en

U

"3

_i;j

tn

en

'2

(U

o

o

O.

3

en

on

e

00

en

^4—

x:

en

o

eg

X

«

<u

C

c

(U

E

Xi

lU

•o

a.

'o

u

c

u

en

D.

'C

c

'■n

Q.

u

^-^

E

o

>

^

'oij

!Z)

u

1)

en

u-

k.

c

ca

u

4J "O

IS

o

O

'«-'

O

P-,

CO
C/3

"«

c

ca

"2

o

00

c

O

en
C

'3

O

c
o
o

^-^

c

o

en
C

'S

U

?

E

u

"u

<u

00

y

u

•a

o.

<^-

u-

en

<u

<4-.

>

(0

O

en

3

(U

Ui

•O J=

U

*— •

Xi

T3

E

E

c

3
C

v5

en

a.

?

o

3

s

IZI

00 X)

.3^

>^

'o

<u

0)

b.

>

a.

00

3

■•-'

w

<S

00

S "O

o

en

3

N

N

C

c
o

00

>.

D

o -o

C

en

c

o.

o

O

C
E

03

o

SZ
^

S
e->

en

E

c

en

o

V5

o

o

3
en

Q

O

en

'^-

r^

X>

!.M

E

u

O

o

-J

E

(75

c
o

t^

3

C«

X

z

E

iS

T3

<U

eC

c

03
X

u-

u.

1)

X

f2

I §

1^

§

_•

3

g

§

c
^

3

-s

eri
1

1

2

s

s

"5,

a

2

s

.2

e?

s

.a
5

00

5

g^'^i

.0

§

f::o 1

1

S-2

5

O

.Ci.

^ -

.o

^

c

^

c

^

a

c

-5

p.

^

ej

V

~^

.e-3

a

d.

1

l>2

^

*>•

• ii

^

.1^

e^

c^

5
a

^

|§

^

r~

1

5 ^

1

r<-i

r^

1

1

1

'^

■^ >ri O O — <N

■O oo I lo 00 rn
00 in I t^ 00 -^

— [^ O n — o

Tt On O O Tl- O

O^

o
in

<^ O O O rn

0^

^

t^

■*

O OS fN O — O

■^ r^ I in I t^

r- m o ■* O rn

I 00 I 00

— O (N — _- o — o o

^ i^ I M i^ ^

OO Tf moOOr^ —

II I OO 00 — I I in

II I 00 00 II in

OO o r^vo— 'OO tN

^ I JS I 1 I ^ ^

OrJ O ONOOOr^ (N

J? s^ « ■"

c o .5 ;E
o j£:£ 3 y O

3 03USaQE-

— c

O C/5

^ "

U

T3 73

03

41 •- _

T3

K ,, 1)

ntifi

nari
idae

"S

en
>

ea U « <U
"O 43 "O =2

T3

1)

u ^2 =

T5 ni X (U

3

en

^^■5 i^

C

3 E t/^ 0.

en

"15
>

=a en 2, >^

c

ij

n

3

18

OCCASIONAL PAPERS MUSEUM OF NATURAL fflSTORY

T3

3
C

C
O

U

-J

-o

c

-o

>o

03
X

Q. 1^

1/] a

1^
s

3

i;Ca::c5

0\ — ' t^
<VJ — I

On r- — O

oo

O On O Tt

Tt -rt 00 ON

r~- 1^ ■* U-)

^ NO r»^ r~

NO O

— NO

NO

O m

D.
• (A
o

=" .2

~ s s

Q '»j Q

I I 2

o o —

60

i s - g

t:2 g § ^

-s: -S -i: i, ~.
Lq U a, CO to

d.

s
o

■s g
^f2

d.

<4)

§,

On
On

I 2

O ON

— ' O O CnI — '

rj ■*

ooo o o— — oo —

O O I

NO NO

>r) Vl O O

r- ON ' 00

ON — O On

OO

o

NO

O m

ooo r^ ooo oo O

OOO

OOO O O

O (N

NO

-H rJ O r<^

— — o

oo

c

■O en

So

u

1)

.^^

^

C T3

1)

^

u

2c

c3 -n

E
2

u

&.

on

-n

S2

mp

allianassid
Peneidae
Euphausii

a

o

Corallanii
Serolidae
Idoteidae

C/3

c
o

II

ll

'x: —
E^

■cu

«5

U

<

J=

lyj

STEAMER -DUCK FEEDING ECOLOGY

19

1)

s ^

&

^ Si

§ OS

o o o o o

o o o o o

o
so

o o —

IT)

O O fS

fN O O

O O (^

■^ en

o o o o o

O t^ ''^

1/-) — r~

0\ ro Tt — — ■

fN — O O O rt-

o o o

OS

o o o o o

oo —

O O ro

e

wis

D. C

o c

o "
U

O

-o

'5.
<u

w

•:?^

<yi ^ ^ -n

t; h C -S

lP o n c

^S 4J u S

.2

ca TS

c aj

'c

13 o «=

o

•g OJj-s

(U

</:

i: >»c

v:

c

C/5

-J

terres

Pol

unide

e

20 OCCASIONAL PAPERS MUSEUM OF NATURAL fflSTORY

Mytilidae) and snails were the most frequent molluscs found in the stomach
contents, and crabs (especially Grapsidae) were the predominant crustaceans.
Species of marine prey identified were generally among the most common
species in the South American littoral. Although species differences in diet
were not obvious, the data indicated that T. leucocephalus consumed more
crabs than did the other species. Freshwater specimens of T. patachonicus
contained comparatively high proportions of insects (mostly larvae of aquatic
species). Seeds of the introduced weed Rumex acetosella were unexpectedly
important in the diets oiT. patachonicus atLago Roca, Santa Cruz, Argentina;
these seeds were abundant on the surface of the lake. Related species of
Rumex are known foods of ducks in North America (Martin and Uhler, 1939).
A number of large prey items were found in specimens of all species
including: a 20 x 13 mm clam (Malletia sp.), 33 x 14 mm snail (Chilina
neuquenensis), and 24 x 21 mm crab (Cyrtograpsus sp.) in T. patachonicus;
a 34 X 15 mm snail in T. brachypterus; a 27 x 15 mm mussel {Mytilus edulis),
a 48 x 24 mm limpet {Fissurella oriens), and a crab {Munida sp.) with
carapace measuring 45 x 29 mm in T. pteneres. A male T. patachonicus
collected at Puerto Deseado contained 175 g of clams {Malletia sp.).

Analyses of Grit

As noted by several previous investigators (Sclater and Salvin, 1878, p.
437; Oustalet, 189 1 , p. B225; Bennett, 1924, p. 280; Reynolds in Lowe, 1934,
p. 477; Humphrey et al., 1970, p. 136), gizzards of steamer-ducks typically
contain grit, as do those of other waterfowl (e.g., Weller, 1972; Thomas et al.,
1977). Grit is acquired early in life; small quantities ofgrit (3-6 mm in longest
dimension) were present in gizzards of two of three very young downy T.
leucocephalus, one of which was collected on the nest and still contained yolk
reserves. Grit samples from non-juvenile steamer-ducks we collected (Table
8) were of approximately equal mass in males and females of all species-
habitat groups (MWT, P > 0.20). Masses of grit samples (Table 8) from
freshwater T. patachonicus were greater than those from their saltwater
conspecifics (MWT, sexes pooled; T= 1096, n = 40, 26; P < 0.005). These
two habitat groups of T. patachonicus, and the other three species, showed
substantial differences in the means (KWT, sexes pooled; T= 23.27, 5 groups,
n=nS;P< 0.00 1) and variances (SRTV; T = 1042, 5 groups, n = 1 1 8; 0.025
<P< 0.05) of the masses ofgrit in their gizzards. Average masses ofgrit were
highest in large T. pteneres, moderate in freshwater T. patachonicus and in T.
leucocephalus, and least in saltwater T. patachonicus and T. brachypterus
(Table 8); within-species variances in grit masses followed an identical
ranking.

Median grit size, like the total masses of the grit samples, did not differ
(MWT; P > 0.05) between the sexes in saltwater and freshwater T. patachoni-
cus, or in the three flightless species. Comparisons among the five species-

STEAMER-DUCK FEEDING ECOLOGY 21

Table 8. Summary statistics for masses (g) of clean dry grit samples from gizzards of non-
juvenile steamer-ducks by species and sex; samples from T. patachonicus collected on fresh
and salt water are summarized separately.

Species

Sex

n

Mean

Standard
Deviation

Range

T. patachonicus

/ /

21

5.50

3.94

0.43-13.99

(saltwater)

19

5.81

4.17

0.66-16.45

All

40

5.70

4.04

0.43-16.45

T. patachonicus

S$

16

10.57

6.57

3.10-23.29

(freshwater)

22

10

9.89

6.60

2.73-20.40

All

26

10.32

6.46

2.73-23.29

T. brachypterus

r f

10

2.93

2.86

0.11- 8.13

22

10

3.92

2.59

0.52- 8.31

All

20

3.43

2.71

0.11- 8.31

T. leucocephalus

SS

11

7.31

6.18

1.12-21.96

22

9

7.59

6.23

1.03-23.08

All

20

7.43

6.04

1.03-23.08

T. pteneres

$6

6

12.66

8.97

3.11-27.02

'i'i

6

10.22

8.99

1.66-27.33

All

12

11.44

8.66

1.66-27.33

habitat groups (sexes pooled), however, revealed significant differences
(KWT; T= 57.55, 5 groups, « = 118; Z' < 0.001). Pooled sample probabUity
densities for grit size in the species (Fig. 5) show that a ranking of groups by
decreasing grit size corresponded to that based on body masses and raw giz-
zard masses: T. pteneres, T. brachypterus, T. leucocephalus, saltwater T.
patachonicus, and freshwater T. patachonicus. One male T. pteneres con-
tained a piece of grit which measured 2.2 x 1.3 x 0.3 cm, comparable to the
largest piece of grit found in a White- winged Scoter (Melanittafusca) that
measured 2.5 x 1.8 x 1.2 cm (Cottam, 1939).

Within all five species-habitat groups, diversities of grit size (H') were
similar between the sexes (MWT; P >0A0 in all). There were differences,
however, among the five groups (sexes j>ooled) in grit-size diversity (KWT;
T= 11.14, 5 groups, n= 118; P < 0.05), largely due to the lower grit-size
diversity in T. pteneres.

DISCUSSION

Feeding Morphology

Steamer-ducks resemble other diving ducks, especially the Mergini, in the
low pneumaticity of their skull bones (Harrison, 1958), general structure of
the bill (Goodman and Fisher, 1962), and the large size of the supraorbital
processes (Raikow, 1970). Multivariate comparisons ofbill morphology and
qualitative differences in bill lamellae in Tachyeres (Figs. 2, 3) indicate that
both interspecific and intersexual differences largely lie along an axis reflect-
ing a contrast between bill size and number ofbill lamellae. This variation
probably is associated with the sizes of prey organisms retained (Jenkin,

22 OCCASIONAL PAPERS MUSEUM OF NATURAL fflSTORY

0.60

40-

0.30-

0.20-

0.10-

OOO

Freshwater Specimens
268.3 g from 20 birds

T. patachonicus

Saltwater Specimens
?<9. g from 40 birds

I I I

T. leucocephalus

148. 7 g from 20 birds .

T brachypterus

68- 6 g from 20 birds

1 1

T. pteneres

1373 g from 12 birds

-r
2

4

6

n —
10

rill

Ji

'1 1 1 'I
ill.il

I

-r
4

2 4 6

SIZE OF GRIT (mm)

8

10

Fig 5. Sample probability densities for grit-size distributions of steamer-ducks; sexes are
pooled in all groups.

1957; Zweers et al., 1977; Crome, 1985). In Tachyeres, T. patachonicus is
equipped to strain smaller prey than the flightless species, and females of each
species can retain smaller particles than their male conspecifics. Similarly,
interspecific and intersexual differences in bill-lamellar densities were re-
lated to prey sizes in seven species of Anas in North America (Nudds and

STEAMER-DUCK FEEDING ECOLOGY 23

Bowlby, 1984; Nudds and Kaminski, 1984), and Tremblay and Couture
(1986) found that density of bill lamellae was inversely correlated with body
size in eight species of Anas. POysa (1983a) concluded that interspecific dif-
ferences in neck length and bill morphology effected significant separation of
niches of six sympatric species of Anas. Koehl et al. (1984) documented
intersexual differences in sizes of mussels (Mytilus edulis) consumed by
Barrow's Goldeneyes (Bucephala islandica). The dietary importance of prey
organisms small enough to be strained by the bill lamellae in Tachyeres is not
well supported by my data on food items, however, with the exception of the
seeds oiRumex in freshwater T. patachonicus. It may be that the techniques
of preservation and examination employed here were not sophisticated
enough to determine accurately the importance of small, more fragile food
items.

Gizzards of steamer-ducks were more massive than those of five of nine
Australian anatids (Norman and Brown, 1985). Five North American species
oiAythya had much smaller mean gizzard masses than those oi Tachyeres; as
in Tachyeres, however, relative gizzard masses did not correspond in rank to
total body mass among species and probably were associated with dietary
differences (Kehoe and Ankney, 1985). Gizzards of Spur-winged Geese
{Plectropterus gambensis) had overall mean masses comparable to those of
T. leucocephalus (males, 121 g; females, 95 g) but which varied significantly
with molt and egg-laying (Halse, 1985). Barnes and Thomas (1987) and
Kehoe and Thomas (1987) described important relationships between diet
and morphology of the gut in North American anatids.

The capacity of the larger species and sex groups to handle larger food
items is supported further by the size distributions of grit in their gizzards (Fig.
5). The large quantities of relatively fine grit in freshwater T. patachonicus
correspond with the smaller sizes of prey items and larger proportion of
vegetable matter in their diets. The extent to which grit size is related func-
tionally to the sizes of food items digested or is limited by availabiUty at each
locality is unknown. Meinertzhagen (1954) detailed the importance of grit in
the digestion of vegetable matter in birds, especially seeds, but did not
consider species feeding on molluscs and crustaceans. Keith (1961) reported
that the masses and size distributions of grit in six species of Anas and two
species of Ay thya in Alberta were associated with dietary differences. Ellar-
son (1956) found grit of consistent size in Oldsquaws (Clangula hyemalis).
Based on data presented by Thomas et al. (1977), steamer-ducks carry greater
quantities of grit than 1 European anatids except for swans (Cygnus olor and
C. columbianus); of these 10 species, only Tufted Ducks {Aythya fuligula)
contained grit of comparable size (mode = 5 mm). Halse (1983) found
differences between sexes and age groups of Plectropterus gambensis in
masses of grit in their gizzards, which was associated with gizzard mass, but
that size distribution of grit was constant among groups. Masses of grit in
Tachyeres also exceeded that found in nine Australian anatids of diverse

24 OCCASIONAL PAPERS MUSEUM OF NATURAL fflSTORY

relationships (Norman and Brown, 1985). Trost (1981) found that consump-
tion of grit in Mallards (Anas platyrhynchos) was affected by diet and sex of
birds, season, and the size and mineral composition of grit available; grit
retention followed a negative exponential distribution. Weller (1972) in-
ferred a direct relationship between maximum food size and maximum grit
size in five waterfowl species in the Falkland Islands; T. brachypterus was
larger in both variables than the Crested Duck {Lophonetta specularioides).
Kelp Sheldgoose (Chloephaga hybrida). Upland Sheldgoose (C. pictd), and
Speckled Teal (Anas flavirostris).

The supraorbital or "salt" glands of waterfowl are known to function in the
concentration and excretion of salt, and size of the glands varies directly with
the salinity of the water consumed (Cooch, 1964; Schmidt-Nielsen and Kim,
1964; Holmes and Phillips, 1985). The occurrence of large glands in
flightless Tachyeres and in marine T. patachonicus and comparatively small
glands in freshwater T. patachonicus agrees with this generalization, even
though most marine T. patachonicus are able to fly to freshwater daily, and
both T. pteneres and T. brachypterus are known to walk to freshwater to drink
(Percy in Phillips, 1925, p. 292; Pettingill, 1965, p. 73; Weller, 1971, p. 108,
1976, p. 49; pers. obs.). NystrOm and Pehrsson (1988) suggested that drinking
fresh water may be important for prevention of salt-related dehydration in
marine waterfowl feeding largely on mussels. The small relative size of
glands in T. pteneres is explained less easily; it may be that the glands of T.
pteneres are disproportionately deeper than those of the other species.

Feeding Habitat and Behavior

High densities of feeding steamer-ducks at river mouths undoubtedly are
related to the rich supply of nutrients and associated benthic fauna character-
istic of such confluences (Pehrsson, 1984). The importance of kelp beds as
feeding sites was noted early in the study of the genus (Darwin, 1839, p. 258;
Cunningham, I87Ib, pp. 96, 154). WeUer (1972, p. 33) concluded that this
association between steamer-ducks and kelp is the basis for the popular
misconception that the birds feed directly on the plant. The distributional
association of kelp and steamer-ducks on the marine littoral of southern South
America (Kiihnemann, 1970) underscores this relationship.

The differences in shore types used by T. pteneres and T. patachonicus in
sympatry, in combination with the strongly developed interspecific territori-
ality shown by T. pteneres towards T. patachonicus (Livezey and Humphrey,
1985), strongly suggest the existence of competitively maintained habitat
segregation. Advantages of large body size in interference competition are
most likely in habitats with rich food resources (Persson, 1985), a condition
that probably is met in the marine littoral of southern South America. The
larger size of T. pteneres not only is advantageous in usurping the prime
marine feeding territories, but also may permit the utilization of larger prey

STEAMER-DUCK FEEDING ECOLOGY 25

items (Wilson, 1975). I suspect that occupancy of the rocky peninsulas and
islands defended by T. pteneres has at least three selective advantages: (1)
inclusion of nesting areas with reduced densities of predators; (2) provision
of a vantage point for territorial vigilance and early warning of approaching
dangers; and (3) a high ratio of shoreline to defended area, which permits
efficient defense of food resources. Where sympatric with flightless Tachyeres,
the more mobile T. patachonicus perhaps can tolerate displacement from the
more favorable locations by exploiting a larger area, less accessible feeding
sites, and a wider range of habitats. An hypothesis of habitat segregation is
supported by the widely held belief that T. patachonicus on the Falklands
breeds only on freshwater lakes, whereas larger T. brachypterus breeds on the
more fertile marine coasts (Bennett, 1926; pers. obs.). Also, it appears that T.
patachonicus does not breed on the coasts of Chubut with the morphologi-
cally similar T. leucocephalus (Table 2; Humphrey and Livezey, 1985;
Livezey et al., 1985), although T. patachonicus commonly occurs on the sea-
coasts of both the Falklands and the continent during the winter.

Evidence for interspecific competition between T. pteneres and T. pat-
achonicus might include: (1) data on availability, consumption, and disper-
sion of resources in feeding territories and evidence that resources are
limiting; (2) information on the habitats used by T. patachonicus in similar
environments where no flightless species occurs; and (3) demonstration of
density-dependence between the species. James and Boecklen (1984) and
Wiens (1984) rightly stressed the need to consider alternatives to competition
in studies of bird communities. Possible alternatives for steamer-ducks in
habitat selection might involve protection of broods, prevention of rape of
females, or differences in food requirements associated with body size. In
recent studies, some investigators attributed the ecological segregation of
sympatric waterfowl to habitat preferences or tradition (Nilsson, 1972;
WeUer, 1972; Stott and Olson, 1973; Siegfried, 1976), whereas others
acknowledged the possible importance of competition in producing these pat-
terns (Pehrsson, 1976;Oksanenetal., 1979; Nuddsetal., 1981; Savard, 1982,
1984;Toftetal., 1982; Nudds and Bowlby, 1984; Nudds and Kaminski, 1984;
Vermeer and Bourne, 1984; DuBowy, 1988); still other workers reserved
judgment on the cause(s) of the ecological differences they observed (White
and James, 1978; Eadieetal., 1979; Poysa, 1983a, 1983b, 1985; Sanger and
Jones, 1984). POysa (1986) suggested both harmful and beneficial interac-
tions in mixed-species foraging groups oiAnas.

The influence of tides on the feeding of steamer-ducks was evident but not
quantified in this study. Dabbling was especially common in the shallow
waters of a low or rising tide; Cawkell and Hamilton (1961) also noted this
for T. brachypterus. These observations contrast with the statement of
Johnsgard (1978: p. 137) that steamer-ducks "... forage at or near high tide
periods..." I suspect that low tides permitted birds to dive and dabble at greater
distances from (high-tide) shore, extending the area of benthos accessible to

26 OCCASIONAL PAPERS MUSEUM OF NATURAL HISTORY

foraging. Effects of tides on feeding have been noted for Common Shelducks
(Tadorna tadorna; Evans and Pienkowski, 1982), Black Ducks (Anas ru-
bripes; Grandy and Hagar, 1971), Eurasian Wigeons (Anas penelope; Owen
and Williams, 1976), Greater Scaup (Aythya marila; Burger, 1983), Common
Eiders (Somateria mollissima; Campbell, 1978), Steller's Eiders (Polysticta
stelleri; Petersen, 1980), Black Scoters (Melanitta nigra; Mudge and Allen,
1980), and Barrow's Goldeneyes (Koehl et al., 1984).

P. S. Humphrey and I observed flocks of T. pteneres in the Bahia de
Lapataia, Tierra del Fuego, and a flock of T. patachonicus at F*uerto Deseado,
which contained a high proportion of birds in wing molt. The molting T.
pteneres were feeding intensively at a river mouth; the molting T. patachoni-
cus were feeding in shallow water, and several birds we collected contained
up to 175 g of clams and mussels. Petersen (1980, 1981) found that feeding
methods, amount of time spent feeding, diet, and the spatial distribution of
flocks of Steller's Eiders were influenced significantly by the chronology of
wing molt. Reduction in average relative size of breast muscles (Mm.
pectoralis and supracoracoideus) in T. patachonicus (sexes equal and pooled),
from 17.6% of total body mass for birds with remiges to 11.9% for birds in
wing molt (ANOVA on log-transformed ratios; F = 210.44, df = 1, 58; P <
0.0001), suggests that wing molt may be energetically demanding. Female
T. pteneres showed an average change from 14.6% for non-molting birds to
11.2% for birds lacking remiges (F= 14.14,df= 1, 10; P< 0.005). These molt-
related reductions parallel the findings of Hanson (1962) for Canada Geese
(Branta canadensis) and Bailey (1985) for Redheads (Aythya americana), but
contrast with those of Ankney (1979, 1984) for Snow Geese (Anser caerules-
cens) and Brant (Branta bernicla); reductions in breast muscles may be
related to reduced exercise, however, and conclusive evidence for the
demands of molt would require tests for possible compensatory increases in
pelvic musculature (Young and Boag, 1982). Sjoberg (1988) concluded that
weight losses during wing molt in Green-winged Teal (Anas crecca), and
perhaps waterfowl generally, may be an adaptation to permit flight before the
new remiges are fully grown; this hypothesis may be relevant to weight losses
in molting T. patachonicus, but it does not explain those of molting flightless
steamer-ducks.

Differences among bouts in mean dive times probably reflect a combina-
tion of differences in densities of foods, diving abilities of individual birds,
levels of satiation and relative importance of alternate activities, and in the
turbidity, turbulence, and depth of the water. Dive times of waterfowl are
known to be related to water depth (De war, 1924;Nilsson, 1972). Differences
in within-bout variances are interpreted less easily, but may result partly from
local variations in benthic substrate or in patchiness in the dispersion of prey.

Dive times of adult T. brachypterus reported by Weller (1972, p. 34; ^ =
17.3 sec, n = 47) averaged much less than those recorded by me (Jc = 30 sec;
Table 6), probably because Weller timed birds diving in water no deeper than

STEAMER-DUCK FEEDING ECOLOGY 27

1 10 cm, although he suspected that some birds dived in water over 9 m deep.
Dive times for downy young (1-35 days old) averaged only 12.8 sec (« = 51),
with a maximum time of 17.4 sec (Weller, 1972, p. 34). Dive times are related,
in part, to depths of dives; interspecific differences in feeding depths contrib-
uted significantly to differentiation of niches of the downy young of sympa-
tric anatids in North America (Monda and Ratti, 1988). Mean dive times (sec)
of steamer-ducks 1 observed (Table 7) were comparable to or greater than
those for a number of other benthic-feeding diving ducks (Dow, 1964;
Nilsson, 1972; Siegfried, 1976): Redhead (Y = 11.0, « = 10), Canvasback (A.
valisineria; 7 = 16.6, n = 12), Greater Scaup (3c = 17.2, n = 27), Lesser Scaup
(A. qffinis; X = 11.9, n = 11), Tufted Duck (3c = 19.0, n = 35), European
Pochard (A.ferina; 3c = 15.1, n = 1), Common Eider (3c = 31.2, n = 74),
Oldsquaw (^ = 30.2, n = 65), Surf Scoter {Melanitta perspicillata; 3c = 32.7,
n = 20), Common Goldeneye {Bucephala clangula; 3c = 19.5, n = 1388), and
Ruddy Duck (Oxyurajamaicensis; 3c = 1 8.6, n = 1 1). Much longer dives have
been inferred for the Oldsquaw and King Eider (Somateria spectabilis), both
of which have been caught in nets at depths greater than 50 m (Nilsson, 1972;
Bellrose, 1976).

Schenkeveld and Ydenberg (1985) demonstrated that Surf Scoters in
flocks dive synchronously, and that this synchrony was most pronounced in
the presence of kleptoparasitic Glaucous-winged Gulls (Larus glaucescens).
These authors concluded that synchrony of dives acts to reduce foraging
efficiency of the ducks. Although my data for Tachyeres failed to document
synchrony of dives, my impression in the field was otherwise. Also, I
observed two instances of kleptoparasitism by Kelp Gulls {Larus dormnica-
nus) on diving T. pteneres near Ushuaia, Tierra del Fuego. Assessment of the
possible relevance of the hypothesis of Schenkeveld and Ydenberg (1985) to
diving by Tachyeres must await fiuther study.

The ecological implications of longer dive times in males than in females
of T. brachypterus are unclear. In a few previous studies, males were found
to dive for longer periods than females in the Common Goldeneye (Hartman,
1935; Mester and Prunte, 1966) and the Tufted Duck (Bauer and Blotzheim,
1969). It may be that the larger body size of males acts as ballast and allows
more efficient descent underwater which would conserve oxygen and permit
longer dives than in the smaller females. Reynolds (1987) found that, in the
Oldsquaw, males tended to drive for longer periods than females, a difference
he attributed to the physiological advantage of large body size of males.
Weller (1980, p. 57) hypothesized that the large size and rugged build of
steamer-ducks, along with their diving skills, permit foraging in deep,
turbulent waters. Calder (1984) showed that for homeotherms the physiologi-
cal upper hmit for duration of dives, and the time of submergence in cold
water before chilling of the body, increase allometrically with body mass. The
lack of correlation between dive times and subsequent pauses on the surface,
however, does not support the hypothesis that typical feeding dives of

28 OCCASIONAL PAPERS MUSEUM OF NATURAL fflSTORY

Steamer-ducks are near the limit of exhaustion. This result also may reflect
the fact that pauses after dives not only are rest periods but also include time
for handling and swallowing prey captured during the dive.

Diet

Murphy (1936, p. 972) concluded that T. patachonicus eat fewer thick-
shelled molluscs and more crustaceans than T. pteneres, a view repeated by
Goodalletal.(1951,p. 163-164), Johnson (1965, p. 197), Woods (1975, p.
123), and Johnsgard (1978, p. 135). Based on differences in intestinal color,
Bennett (1924) hypothesized that T. brachypterus and T. patachonicus
differed in diet. Weller (1976, p. 49) stated that flightless steamer-ducks tend
to eat larger prey in short, intense foraging bouts, whereas T. patachonicus
generally prefer smaller items that are taken in less punctuated feeding
periods.

Interspecific differences in diet were difficult to document in this study due
totherelatively small numbers of specimens having full stomachs. There was
evidence that T. leucocephalus fed comparatively intensively on crabs,
inasmuch as 16 of 21 specimens with food in their stomachs contained crabs
comprising an average of 74% of the contents by mass (Table 7). The extent
to which this reflects differences in availability of prey types is not known.
There also was some indication that large T. pteneres consumed somewhat
larger prey items than did its congeners, although larger samples are needed
to demonstrate interspecific differences in prey size. Goudie and Ankney
( 1 986) found that wintering Harlequin Ducks (//. histrionicus) and Oldsquaws
spend more time feeding and ate organisms of higher energy density than the
larger, syntopic Black Scoters and Common Eiders; the authors concluded
that these differences in energy intake reflect the thermodynamics of body
size. Draulans (1982, 1984) documented marked preferences for mussels of
certain sizes in Tufted Ducks, preferences which contribute to dietary
differences between congeners (Draulans, 1987) and which may represent a
compromise between profitabihty of prey items and handling capabilities.
Observations by Tome and Wrubleski (1988) suggest that pochards locate
prey visually. Nystrom and Pehrsson (1988) hypothesized that diving ducks
which feed on mussels can reduce diet-related salt stress by foraging in
habitats with lower salinity, ingesting items with lower salt content, selecting
smaller mussels, drinking fresh water, using energetically efficient foraging
techniques, and (on an evolutionary time scale) having comparatively large
body size. This hypothesis finds substantial support in the feeding ecology
of Tachyeres, a genus characterized by large body size, and in which members
heavily exploit mussels for food, show a preference for foraging near river
mouths, and have been observed to actively seek out fresh water for drinking
on a regular basis.

STEAMER-DUCK FEEDING ECOLOGY 29

Ecological Equivalence and Specialization

Lack (1974) and Weller (1976) commented that steamer-ducks are eco-
logical equivalents of the northern-hemisphere eiders, especially the genus
Somateria. It is true that, like Tachyeres, the diet oiSomateria on saltwater
is composed predominantly of molluscs (especially Mytilus), and also in-
cluded substantial quantities of crustaceans and some amphipods (Millais,
1913; Phillips, 1926;Cottam, 1939;McGilvrey, 1967;Player, 1971;BeUrose,
1 976). However, the same is true of a number of other diving ducks, including
Steller's Eider, Oldsquaw, all three species of scoter {Melanitta spp.), and
both goldeneyes; to a lesser extent the saltwater diets of the Common
Shelduck, Harlequin Duck, Bufflehead, and several pochards {Aythya spp.)
also resemble those of marine steamer-ducks (Millais, 1913; Bent, 1925;
PhiUips, 1926; Munro, 1934; Cottam, 1939; Delacour, 1954; Madsen, 1954;
Olney, 1963, 1965; McGilvrey, 1967;NUsson, 1969, 1972; Bryant and Leng,
1975; BeUrose, 1976; Vermeer, 1982).

It appears that steamer-ducks consume prey of smaller size than Common
Eiders, which frequently have been found to contain molluscs of 5 cm length
and once up to 25 cm long, or White-winged Scoters, which have contained
molluscs over 5 cm in length (Cottam, 1939). Unlike other diving ducks, both
Tachyeres (excluding freshwater T. patachonicus) and Somateria are Umited
to high-latitude seacoasts for nesting. The general similarity of diet and
breeding ecology of Somateria and Tachyeres, in combination with their
relatively large size, tends to corroborate the 'equivalence' hypothesis.
However, it is worth noting that marine Tachyeres, unlike Somateria, are
largely flightless and hence non-migratory, do not nest colonially, show less
sexual dichromatism , and typically do not use their wings for underwater pro-
pulsion (Humphrey and Livezey, 1985; Livezey and Humphrey, 1984a,
1985).

Weller (1976, p. 45) described steamer-ducks as "... probably the most
specialized diving ducks of the world..." While it is true that steamer-ducks
are capable divers and that diving is the most frequently employed foraging
method in the group, other feeding methods, especially dabbling and tipping,
often are important (Cawkell and Hamilton, 1961; Weller, 1972; Woods,
1975). The frequency of shallow- water foraging prompted Eyton (1869, p.
101) to conclude: "They are also bad divers, obtaining their Uving almost
entirely by breaking shell-fish along the water-mark..." Steamer-ducks also
lack the specialization of underwater wing-propulsion, typical of eiders and
the Oldsquaw, only employing wing-strokes at submergence, during turns,
and during high-speed escape and territorial attack dives (Livezey and
Humphrey, 1984a; pers.obs.; Snell, 1985). Hence steamer-ducks are no more
specialized for diving than eiders, scoters, and the Oldsquaw, and certainly
much less specialized morphologically (Livezey, 1986) than the mergansers
{Mergus spp. andLophodytes cucullatus) or the Musk Duck (Biziura lobata).

30 OCCASIONAL PAPERS MUSEUM OF NATURAL fflSTORY

ACKNOWLEDGMENTS

Throughout this research, Philip S. Humphrey, (Director, Museum of
Natural History, University of Kansas) was a constant source of inspiration,
enthusiasm, and assistance, for which I am truly indebted. This study was
supported by National Science Foundation grants DEB-80-12403, DEB-81-
17942, and BSR-83- 19900, the Kansas University Endowment Association,
Southwestern College, British Broadcasting Corporation, W. Saul, M. C.
Thompson, L. A, Osborne, R. T. Peterson, T. Mastin, R. Hamilton, and the
Humphrey family. Collecting permits were issued by E. O. Gonzalez Ruiz
(Direccion Nacional de Fauna Sylvestre, Argentina), L. O. Saigg de Chialva
(Proteccion Ambiental, Chubut, Argentina), D. H. Soria (Departmento de
Conservacion de la Fauna, Santa Cruz, Argentina), E. Cruz (Servicio Agricola
y Ganadero, Chile), and Sir R. Hunt (Civil Commissioner, Falkland Islands).
Field work in Argentine national parks was made possible by A. Tarak (Di-
reccion Nacional de Parques Nacionales). Logistic support was provided by
J. M. Gallardo, R. A. Bockel, and J. Navas (Museo Argentino de Ciencias
Naturales, Argentina), G. Pincheira and T, G. Iriarte (Universidad de Chile),
and M. M. L'Huillier, D. Bersalobec, S. Canessa, C. Montenar, and D, Nunez
D, (Instituto Professional de Osomo, Chile). Essential arrangements were
made by G. A. Giaroli, F. Lobbe and F. Villar. Housing and laboratory faciU-
ties were arranged by O. Kiihnemann, D. Nunez D., P. Medina, and J. Sesti;
B. Mayer and F. V. T. J. Fauring provided a home and field support at Puerto
Melo. Logistic help and friendship were provided in the Falkland Islands by
A. F. G. Douse (Stanley) and A. and Y. Davis (Lively Island). I also am
grateful for the assistance and hospitality of B. de Ferradas, L. Orquera, E.
Plana, G. Piacentino, J. E. Bonczak, R. E. Caferata, A. Fernandez, R. Landi-
var, G. C. Sarceda, F. Erize, R. Straneck, W. Conway, Y. Lucero, P. Canevari,
M. A. E. Rumboll, R. Hall, H. Niemeyer, M. Sallaberry A., A. Veloso, J. Bal-
lasteros C, and the Gibson Family. I especially wish to thank the following
persons for identification of stomach contents: D. Siegel-Causey (University
of Kansas; crustaceans), G. W. Byers (University of Kansas; insects), E. O.
Wiley (University of Kansas; fish), P. Raven (Missouri Botanical Garden;
plants), G. Davis (Philadelphia Academy of Natural Science; freshwater mol-
luscs), and J. McLean (Los Angeles County Museum; marine molluscs). I am
grateful to D. Grisafe (Kansas Geological Survey) for help in sieve analysis
of grit. Specimens were made available by the Field Museum of Natural
History (Chicago), British Museum of Natural History (Tring), American
Museum of Natural History (New York), San Diego Museum of Natural
History, and U. S. National Museum (Division of Birds). J. P. Angle and M.
A. E. Rumboll provided stomach contents of steamer-ducks. I also thank R.
Mengel, M. Jenkinson, and their colleagues for preparation of specimens; K.
McManness for typing; G. Arratia for translation of the summary; P. C.
Rasmussen for drawing Fig. 2; and M. B, Ibrahim for drafting Figs. 3-5.

STEAMER-DUCK FEEDING ECOLOGY 31

SUMMARY

Data collected during 1980-1984 and previously published information
are used to describe the feeding ecology of the four species of steamer-duck
(Anatidae: Tachyeres) of South America. The three largest species are
flightless and strictly marine whereas the fourth {T. patachonicus) is largely
flighted and occurs on both seacoasts and freshwater lakes. Mean body mass
was greatest in T. pteneres, intermediate in T. brachypterus and T. leucocepha-
lus, and least in largely flighted T. patachonicus; males were larger than
females in all species.

Most bill measurements followed body mass in rankings of sj)ecies and
sexes, although a diversity of univariate patterns were found. Numbers and
density of bill lamellae were ranked inversely with body mass between sexes
and among species. Mean masses of gizzards and sizes of supraorbital glands
followed body mass in rankings, but gizzard mass relative to body mass was
equal between the sexes within species and was largest in T. leucocephalus
and smallest in T. brachypterus, both species being of intermediate size.
Supraorbital glands were larger in saltwater T. patachonicus than in freshwa-
ter conspecifics.

Marine habitats of steamer-ducks are diverse; T. patachonicus also occurs
on freshwater, typically montane lakes. Steamer-ducks were observed most
frequently along rocky shores, and were especially abundant at river mouths
in protected bays and near kelp beds. Where sympatric, T. pteneres occupied
the evidently preferable rocky shorelines 14 times more often than T. pat-
achonicus, which instead frequented open stoney beaches; this difference in
habitats may be related to interspecific competition mediated by direct
interference.

Diving was the most common feeding technique, but birds also employed
the techniques of dabbling, tipping, and picking. Durations of dives averaged
30 sec overall, but were quite variable at several scales. Means differed
among bouts in most species-sex groups, and within-bout variances in dive
times also differed in some groups; both effects probably reflect differences
in water depth, dispersion of prey, and underwater visibility. Males of T. bra-
chypterus remained submerged longer than their female consorts.

Stomach contents consisted mostly of molluscs and crustaceans, with the
former occurring in 71% of specimens containing food and comprising an
average of 7 1 % of their stomach contents by mass. Mytilid mussels were the
most common mollusc in the diet, and grapsid crabs the most important
crustacean. Other prey were generally among the most commonly available
benthic organisms in the region.

Masses of grit in gizzards were greatest in T. pteneres, intermediate in
freshwater T. patachonicus and T. leucocephalus, and least in saltwater T.
patachonicus and T. brachypterus; within species, sexes did not differ in
masses of grit. Particle size of grit was the same for males and females of each

32 OCCASIONAL PAPERS MUSEUM OF NATURAL fflSTORY

species, but grit size (sexes pooled) followed the interspecific ranking by
body mass. In T. patachonicus, saltwater birds contained grit of larger size
than freshwater birds.

High feeding intensity and reduced masses of breast muscles of molting
steamer-ducks indicate that the wing-molt may be energetically demanding,
and this may affect flocking behavior, feeding schedules, and capability of
escape. Tachyeres differs from the superficially similar eiders (Somateria
spp.) in its flightlessness and related sedentary habits, modest sexual dichro-
matism, and method of underwater locomotion.

RESUMEN

Datos recolectados durante 1980-1984 e informacion previamente publi-
cada son utiUzados para describir la ecologi'a ahmentaria de las cuatro
especies de patos vapores (Anatidae: Tac/iyere^) de America del Sur. Lastres
especies mas grandes (T. brachypterus, T. leucocephalus, y T. pteneres) son
no voladoras y son estrictamente marinas; la mayoria de los individuos de la
cuarta especie (T. patachonicus) tienen el poder de volar y se hallan en las
costas Atlantica y Pacifica del sur de Chile y Argentina y en los lagos de la
region. El peso promedio es mas grande en T. pteneres, intermedio en T.
brachypterus y T. leucocephalus, y menor en T. patachonicus; los machos son
mas grandes que las hembras en todas las especies de Tachyeres.

La mayoria de las medidas del pico siguen los rangos del peso de las
especies y de los sexos, aunque se encuentra una diversidad de patrones
univariados. La abundancia y la densidad de las lamelas del pico tienen un
rango in verso al peso del cuerpo entre los sexos y en las especies. Los pesos
promedios del estomago muscular y los tamanos de las glandulas supraorbi-
tales siguen los rangos del peso del cuerpo, pero el peso del estomago
muscular en relacidn al del cuerpo es igual entre los sexos en cada una de las
especies; el mas grande se presenta en T. leucocephalus y el mas chico en T.
brachypterus; ambas especies son de tamafSo intermedio. Las glandulas
supraorbitales son mas grandes en las poblaciones de T. patachonicus que se
hallan en las aguas costeras marinas que en las poblaciones de las aguas
limneticas.

Los habitats marinos de los patos vapores son diversos; T. patachonicus se
encuentra tambien en las aguas continentales y tipicamente en los lagos de las
montanas, Los patos vapores se hallan mas frecuentemente en las orillas
rocosas; son especialmente abundante en la boca de los rios, en bahias
protegidas y en las cercanias de agrupaciones de algas. En lugares donde
ambas especies concurren, T. pteneres ocupa las orillas rocosas 14 veces mas
frecuentemente que T. patachonicus que en reemplazo ocupa playas abiertas
pedregosas. Esta diferencia en el habitat posiblemente se trata de competen-
cia interespecifica mediada por intervencion directa.

Bucear es el modo mas comiin de alimentarse pero los patos vapores

STEAMER-DUCK FEEDING ECOLOGY 33

tambien emplean las t6cnicas siguientes: "dabble" (mantener el cuerpo,
cuello y cabeza horizontal a la superficie del agua mientras van tomando las
presas de la superficie), "tip" (hundir la cabeza para buscar alimento bajo la
superficie del agua, inclinando el cuerpo y alzando la regi6n caudal) y "pick"
(picotear).

La duracion de las buceadas tiene un promedio de 30 segundos pero son
bastante variables. Los promedios varian entre cada sucesion de buceadas en
la mayoria de los grupos definidos para la especie y segiin el sexo. Las vari-
aciones dentro de cada sucesidn de buceadas difieren ademas en algunos
grupos. Probablemente ambos efectos son resultado de diferencias en la
profundidad del agua, la dispersion de las presas y la visibilidad subacuatica.
Los machos de T. brachypterus se sumergen mas tiempo que las hembras.

Los contenidos estomacales de los patos vapores consisten en su mayor
parte de moluscos y de crustaceos. Se encontro moluscos en 71% de los
estomagos conteniendo comida y ellos representan un peso promedio de 7 1 %
del peso de los contenidos estomacales. Los moluscos de la familia Mytilidae
fueron el alimento mas comiin y los cangrejos de la familia Grapsidae fueron
los crustaceos mas representados. Los organismos benticos, los mas comunes
en la region, fueron otro tipo de presa.

El peso de la arenilla encontrada en los estomagos musculares fue mayor
en T. pteneres, intermedio en T. patachonicus de las aguas limn6ticas y en T.
leucocephalus, y menor en T. patachonicus de las aguas marinas y en T.
brachypterus. No se encontraron diferencias en el peso promedio de la
arenilla entre cada una de las especies. El tamano promedio de las particulas
de la arenilla fue similar en los machos y las hembras de cada una de las
especies. El tamafio promedio de las particulas de la arenilla por cada especie
(en ambos sexos) siguieron los rangos interespecificos del peso del cuerpo.
En los individuos del T. patachonicus de las aguas marinas, se encontraron
particulas de arenilla de un tamafio mayor que en los individuos de las aguas
limn6ticas.

Los patos vapores en la muda del plumaje tienen una alimentacidn intensa
y un peso reducido de los miisculos pectorales; esto indica que la muda de las
alas podria ser energ^ticamente exigente y 6so podria influir el comportam-
iento de agniparse, los horarios de alimentacion, y la capacidad de escaparse.
Tachyeres se distingue del genero Somateria, superficialmente semejante, en
su falta de capacidad de volar, en los habitos sedentarios, y en el modo de
locomocion subacuatico.

LITERATURE CITED

Agresti, A. 1984. Analysis of Ordinal Categorical Data. John WUey and Sons, New York.
Ankney, C. D. 1979. Does the wing molt cause nutritional stress in Lesser Snow Geese? Auk

96:68-72.
Ankney, C. D. 1984. Nutrient reserve dynamics of breeding and molting Brant. Auk

101:361-370.

34 OCCASIONAL PAPERS MUSEUM OF NATURAL fflSTORY

Bailey, R. O. 1985. Protein reserve dynamics in postbreeding adult male Redheads. Condor

87:23-32.
Bames, G. G. and V. G. Thomas. 1987. Digestive organ morphology, diet, and guild structure

of North American Anatidae. Canad. J. Zool. 65:1812-1817.
Bauer, K. M. and U. N. G. von Blotzheim. 1964. Handbuch der Vogels Mitteleuropus. Band

3. Frankfurt am Main.
Bellrose, F. C. 1976. Ducks, Geese and Swans of North America. Stackpole Books, Harrisburg.
Bennett, A. G. 1924. Notas sobre el pato vapor Tachyeres palachonicus (King) llamado

"Canvas-back" en las Falklands. Homero 3:280-282.
Bennett, A. G. 1926. A list of the birds of the Falkland Islands and dependencies. Ibis

68:306-333.
Bent, A. C. 1925. Life Histories of North American Wild Fowl. Part 2. U. S. Nat. Mus. Bull.

No. 130.
Blaauw, F. E. 1916. Field-notes on some of the waterfowl of the Argentine Republic, Chile,

and Tierra del Fuego. Ibis 58:478-492.
Brooks, W.S. 1917. Notes on some Falkland Islands birds. Harvard Mus. Comp. Zool. Bull.

61:135-160.
Brown, M. B. and A. B. Forsythe. 1974. Robust tests for the equality of variances. J. Amer.

Stat. Assoc. 69:364-367.
Bryant, D. M. and J. Leng. 1975. Feeding distribution and behaviour of Shelduck in relation

to food supply. WUdfowl 26:20-30.
Burger, J. 1983. Jamaica Bay Studies FV. Factors affecting distribution of Greater Scaup Aythya

marila in a coastal estuary in New York, USA. Omis Scand. 14:309-316.
Calder, W. A., EI. 1984. Size, Function, and Life History. Harvard Univ. Press, Cambridge.
Campbell, L. H. 1978. Diurnal and tidal behavior patterns of Eiders wintering at Leith. Wild-
fowl 29:147-152.
Cawkell, E. M. and J. E. HamQton. 1961. The birds of the Falkland Islands. Ibis 103a:l-27.
Cobb, A. E 1933. Birds of the Falkland Islands. H. F. and G. Witherby, London.
Conover, W. J. 1980. Practical Nonparamelric Statistics, 2nd ed. John Wiley and Sons, New

Yoric.
Cooch, F. G. 1964. A preliminary study of the survival value of a funclitmal salt gland in prai-
rie Anatidae. Auk 81:380-393.
Cottam, C. 1939. Food habits of North American diving ducks. U. S. Dept. Agr. Tech. BuU.

643.
Crome, F. H. J. 1985. An experimental investigation of filter-feeding on zooplankton by some

specialized waterfowl. Austr. J. Zool. 33:849-862.
Cunningham, R. O. 1871a. On some points in the anatomy of the steamer duck {Micropterus

cinereus). Trans. Zool. Soc. London 7:493-501.
Cunningham, R. O. 1871b. Notes on the Natural History of the Strait of Magellan and West

Coast of Patagonia made During the Voyage of H. M. S. "Nassau" in the Years 1866, 67,

68 and 69. Edmonstcm and Douglas, Edinburgh.
Daciuk, J. 1976. Notas faunisticas y bioecologicas de peninsula Valdes y Patagonia. XVin.

Comportaminento del pato vapor volador observado durante el ciclo reproductivo en costas

e islas de Chubut (Rep. Argentina) — (Anserif., Anatidae). Neotropica 22:27-29.
Darwin, C. 1 839. Journal of Researches into the Geology and Natural History of the Various

Countries Visited by H. M. S. Beagle, Under the Command of Captain Fitzroy, R. N. from

1832-1836. Henry Colbum, London.
Delacour, J. 1954. The Waterfowl of the World, Vol. 1. Country Life, London.
Delacour, J. 1959. The Waterfowl of the World, Vol. 3. Country Life, London.
Dewar, J. M. 1924. The Bird as a Diver. Witherby, London.

Dixon. W. J. (chief ed.). 1985. BMDP Statistical Programs. Univ. Calif. Press, Berkeley.
Dow, D. D. 1964. Diving times of wintering water birds. Auk 81:556-558.
Draulans, D. 1 982. Foraging and size selection of mussels by the Tufted Duck, Aythya fuligula.

LAnim.Ecol. 51:943-956.
Draulans, D. 1984. Sub-optimal mussel selection by Tufted Ducks Aythya fuligula: a test of

STEAMER-DUCK FEEDING ECOLOGY 35

a hypothesis. Anim. Behav. 32:1192-1196.
Draulans, D. 1987. Do Tufted Duck and Pochard select between differently sized mussels in

a similar way? Wildfowl 38:49-54.
DuBowy, P. J. 1988. Waterfowl communities and seasonal environments: temporal variability

in interspecific competition. Ecology 69:1439-1453.
Eadie, J. McA., T. D. Nudds, and C. D. Ankney. 1979. Quantifying interspecific variation in

foraging behavior of syntopic Anas (Anatidae). Canad. J. Zool. 57:412-415.
Ellarson, R. S. 1956. A Study of the Oldsquaw Duck on Lake Michigan. Ph.D. Dissertaticxi,

Univ. of Wisconsin, Madison.
Evans, P. R. and M. W. Pienkowski. 1982. Behaviour of Shelducks Tadorna ladorna in a winter

flock: does regulation occur? J. Anim. Ecol. 51:241-262.
Eyton, T. C. 1 869. A Synopsis on the Anatidae, or Duck Tribe. R. Hobson, Wellington, Salop.
GoodaU, J. D., A. W. Johnson, and R. A. Phillipi B. 1951. Las Aves de Chile. Su Conocimiento

y sus Costumbres, Vol. 2. Piatt Esubl. Graf. S. A., Buenos Aires.
Goodman, D. G. and H. I. Fisher. 1962. Functional AnatcMiiy of the Feeding Apparatus in Wa-
terfowl (Aves: Anatidae). S. Illinois Univ. Press, Carbondale.
Goudie, R. I. and C. D. Ankney. 1986. Body size, activity budgets, and diets of sea ducks

wintering in Newfoundland. Ecology 67:1475-1482.
Gould, J. 1841. Part III, Birds. The Zoology of the Voyage of H. M. S. Beagle, Under the

Command of Capuin Fitzroy, R. N., During the Years 1832 to 1836. Vol. 3. Smith, Elder

and Co., London.
Grandy, J. W. and J. A. Hagar. 1971. Analysing food habits of maritime Black Ducks. Trans.

Northeast Fish WUdl. Conf. 28:207-212.
Halse, S. A. 1983. Weight and particle size of grit in gizzards of Spur-winged Geese. Ostrich

54:180-182.
Halse, S. A. 1985. Diet and size of the digestive organs of Spur-winged Geese. Wildfowl

36:129-132.
Hanson, H. C. 1962. The dynamics of condition factors in Canada Geese and their relation to

seascMial stresses. Arctic Inst. North Amer. Tech. Pap. No. 12.
Harrison, LG. 1958. Skull pneumaticity in wildfowl in relation to their mode of life. WUd-

fowl 9:193-196.
Hartman, L. von. 1945. Zur Biologic der Wasser- und Ufervogel in Scharenmeer

Sudwestfinnlands. Acta Zool. Fennica 44.
Holmes, W. N. and J. G. Phillips. 1985. The avian salt gland. Biol. Rev. 60:213-256.
Housse, R. E. 1942. Les Oiseaux des Andes. Ann. Sci. Natur. Zool. (Ser. 11) 4:138-238.
Housse, R. E. 1945. Las Aves de Chile en su Clasificacion Modema. Su Vida y Costumbres.

Ediciones Univ. Chile, Santiago.
Housse, R. E. 1948. Les Oiseaux du ChUi. MassOT and Co., Paris.
Hubbard, R. 1907. Ornamental Waterfowl: A Practical Manual on the Acclimatizaticm of the

Swimming Birds, with References to Two Hundred Species. Walsall Press, London.
Humphrey, P. S. and B. C. Livezey. 1982a. Flightlessness in Flying Steamer-Ducks. Auk

99:368-372.
Humphrey, P. S. and B. C. Livezey. 1982b. Molts and plumages of Flying Steamer-Ducks

(Tachyeres patachonicus). Univ. Kansas Mus. Natur. Hist. Occas. Pap. No. 103:1-30.
Humphrey, P. S. and B. C. Livezey. 1985. Nest, eggs, and downy young of the White-headed

Flightless Steamer-Duck. In: P. A. Buckley, M. S. Foster, E. S. Morton, R. S. Ridgely, and

F. G. Buckley (eds.). Neotropical Omithology. Omithol. Monogr. No. 36, pp. 949-954.
Humphrey, P. S. and M. C. Thompson. 1981. A new species of steamer-duck (Tachyeres) from

Argentina. Univ. Kansas Mus. Nat. Hist. Occas. Pap. 95:1-12.
Humphrey, R S., D. Bridge, R W. Reynolds, and R. T Peterson. 1970. Birds of Isla Grande

(Tierra del Fuego). Smithsonian Inst., Washington, D.C.
James, F. C. and W. J. Boecklen. 1984. Interspecific morphological relationships and the

densities of birds. In: D. R. Strong, Jr., D. Simberloff, L. G. Abele, and A. B. Thistle (eds.).

Ecological Communities: Conceptual Issues and the Evidence, pp. 458-477. Princeton

Univ. Press, Princeton.

36 OCCASIONAL PAPERS MUSEUM OF NATURAL fflSTORY

Jenkin, P. M. 1957. The filter-feeding and food of flamingoes (Phoenicopteri). Philos. Trans.

Royal Soc. London (Sen B) 240:401-493.
Johnson, A. W. 1965. The Birds of Chile and Adjacent Regions of Argentina, Bolivia and Peru,

Vol. I. Piatt Establ. Graf. S. A., Buenos Aires.
Kehoe, F. P. and C. D. Ankney. 1985. Variation in digestive organ size among five species of

diving ducks {Aythya spp.). Canad. J. Zool. 63:2339-2342.
Kehoe, F. P. and V. G. Thomas. 1987. A comparison of interspecific differences in the

morphology of external and internal feeding apparatus among North American Anatidae.

Canad. J. Zool. 65:1818-1822.
Keith, L. B. 1961. A study of waterfowl ecology on small impoundments in southwestem

Alberta. Wildl. Monogr. 6:1-88.
Koehl, P S., T. C. Rothe, and D. V. Derksen. 1984. Winter food habits of Barrow's Goldeneye

in southeast Alaska. \n: D. N. Nettleship, G. A. Sanger and P. F. Springer (eds.). Marine

Birds: Their Feeding Ecology and Commercial Fisheries Relationships, pp. 1-5. Canad.

WUdl. Serv. Publ. CW66-65/1984.
Kiihnemann, O. 1970. Algunas consideraciones sobre los bosques de Macrocystis pyrifera.

Physis 29:273-296.
Lack, D. 1974. Evolution Illustrated by Waterfowl. BlackweU Sci. Publ., Osney Mead, Ox-
ford.
Livezey, B. C. 1986. A phylogenetic analysis of Recent anseriform genera using morphological

characters. Auk 103:737-754.
Livezey, B. C. and P. S. Humphrey. 1983. Mechanics of steaming in steamer-ducks. Auk

100:485-488.
Livezey, B. C. and P. S. Humphrey. 1984a. Diving behaviour of steamer-ducks Tachyeres spp.

Ibis 126:257-260.
Livezey, B. C. and P. S. Humphrey. 1984b. Sexual dimorphism in continental steamer-ducks.

Condor 86:368-377.
Livezey, B. C. and P. S. Humphrey. 1985. Territoriality and interspecific aggression in steamer-
ducks. Condor 87:154-157.
Livezey, B. C. and P. S. Humphrey. 1986. Flightlessness in steamer-ducks (Anatidae:

Tachyeres): its morphological bases and probable evoluticMi. Evolution 40:540-558.
Livezey, B. C, P. S. Humphrey, and M. C. Thompson. 1985. Notes on coastal birds of Puerto

Melo. Chubut, Argentina. Bull. Brit. Omilhol. Qub 105:17-21.
Lowe, P. R. 1934. On the evidence for the existence of two species of steamer-duck (Tachyeres),

and primary and secondary flightlessness in birds. Ibis 76:467-495.
Madsen, F. J. 1954. On the food habits of diving ducks in Denmark. Danish Rev. Game Biol.

2:157-256.
Markham, B. 1970. Reconocimiento faunistico del area de los fiordos Toro y Condor, Isla

Riesco, MagaUanes. Anal. Inst. Patagonia (Punta Arenas, Chile) 1:41-57.
Martin, A. C. and F. M. Uhler. 1939. Food of Game Ducks in the United Slates and Canada.

U. S. Dept. Agr. Tech. Bull. No. 634.
McGilvrey, F. B. 1967. Food habits of sea ducks from the north-eastern United States.

WUdfowl 18:142-145.
Meinertzhagen, R. 1954. Grit. BuU. Brit. Omithol. Qub 74:97-102.
Mester, H. and W. Prunte. 1966. Beobachtungen uber die Tauchdauer der ScheUente. Anthus

3:46-49.
Millais, J. G. 1913. British Diving Ducks. 2 vols. Lx)ngmans, Green and Co., London.
Monda, M. J. and J. T Ratti. 1988. Niche overlap and habitat use by sympatric duck broods

in eastern Washington. J. WQdl. Manage. 52:95-103.
Mudge, G. P. and D. S. Allen. 1 980. Wintering seaducks in the Moray and Dornoch firths, Scot-
land. WUdfowl 31:123-130.
Munro, J. A. 1939. Studies of waterfowl in British Columbia. Barrow Goldeneye, American

Goldeneye. Trans. Royal Canad. Inst. 22:259-318.
Murphy. R. C. 1936. Oceanic Birds of South America, Vol. 2. Amer. Mus. Natur. Hist., New

Yoik.

STEAMER-DUCK FEEDING ECOLOGY 37

Neter, J. and W. Wasserman. 1974. Applied Linear Statistical Methods. Richard D. Irwin,

Homewood.
NicoU, M. J. 1904. Ornithological journal of a voyage round the world in the 'Valhalla' (No-
vember 1902 to August 1903). Ibis 46:32-67.
Nicoll, M. J. 1908. Three Voyages of a Naturalist, Being an Account of Many Little-Known

Islands in Three Oceans Visited by the 'Valhalla.' R. Y. S. Witherby and Co., London.
Nilsson, L. 1969. Food consumption of diving ducks wintering chi the coast of Sweden in

relation to food resources. Oikos 20:128-135.
Nilsson, L. 1972. Habitat selection, food choice, and feeding habits of diving ducks in coastal

waters of south Sweden during the non-breeding season. Omis Scand. 3:55-78.
Norman, F. I. and R. S. Brown. 1985. Gizzard grit in some Austrahan waterfowl. Wildfowl

36:77-80.
Nudds, T. D. 1983. Niche dynamics and organization of waterfowl guilds in variable environ-
ments. Ecology 64:319-330.
Nudds, T D. and J. N. Bowlby. 1984. Predator-prey relationships in North American dabbling

ducks. Canad. J. Zool. 62:2002-2008.
Nudds, T D. and R. M. Kaminski. 1984. Sexual size dimorphism in relation to resource

partitioning in North American dabbUng ducks. Canad. J. Zool. 62:2009-2012.
Nudds, T. D., K. F Abraham, C. D. Ankney, and P. D. Tebbel. 1981. Are size gaps in dabbling-

and wading-bird arrays real? Amer. Natur. 118:549-553.
Nystrom, K. G. K. and O. Pehrsson. 1988. Salinity as a constraint affecting food and habitat

choice of mussel-feeding diving ducks. Ibis 130:94-1 10.
Oksanen, L., S. D. Fretwell, and O. Jarvinen. 1979. Interspecific aggression and the limiting

similarity of close competitors: the problem of size gaps in some community arrays. Amer.

Natur. 114:117-129.
Olney, P J. S. 1963. The food and feeding habits of Tufted Duck Xylhya fuligula. Ibis

105:55-62.
Olney, P. J. S. 1965. The food and feeding habits of Shelduck (Tadorna tadorna). Ibis

107:527-532.
Oustalet,E. 1891. MissicmScientifiqueduCapHom, 1882-1883. Vol. VI. Zoologie, Oiseaux.

Gauthier-Villars, Paris.
Owen, M. and G. Wilhams. 1976. Winter distribution and habitat requirements of Wigeon in

Britain. Wildfowl 27:83-90.
Pehrsson, O. 1976. Food and feeding grounds of the Goldeneye Bucephala clangula (L.) on

the Swedish west coast. Omis Scand. 7:91-1 12.
Pehrsson, O. 1984. Diving duck populatiwis in relation to their food supplies. In: P. R. Evans,

J. D. Goss-Custard, and W. G. Hale (eds.). Coastal Waders and Wildfowl in Winter, pp.

101-116. Cambridge Univ. Press, Cambridge.
Persson.L. 1985. Asynunetrical competition: are larger animals competitively superior? Amer.

Natur. 126:261-266.
Petersen, M. R. 1980. Observations of wing-feather moult and summer feeding ecology of

Steller's Eiders at Nelson Lagoon, Alaska. Wildfowl 31:99-106.
Petersen, M. R. 1981. Populations, feeding ecology and molt of Steller's Eiders. Condor

83:256-262.
Pettingill, O. S., Jr. 1965. Kelp Geese and FlighUess Steamer Ducks in the Falkland Islands.

Living Bird 4:65-78.
Phillips, J. C. 1925. A Natural History of the Ducks. Vol. 3. Houghton MiffUn Co., Boston.
Phillips,!. C. 1926. ANatural History of the Ducks. Vol.4. Longmans, Green and Co., London.
Player, P V. 1971. Food and feeding habits of the Common Eider at Seafield, Edinburgh, in

winter. Wildfowl 22:100-106.

P6ysa,H. 1983a. Morphology-mediated niche organization in a guild of dabbling ducks. Omis

Scand. 14:317-326.
Poysa, H. 1983b. Resource utilization pattern and guild structure in a waterfowl community

Oikos 40:295-307.
Poysa, H. 1985. Circumstantial evidence of foraging interference between two species of

38 OCCASIONAL PAPERS MUSEUM OF NATURAL fflSTORY

dabbling ducks. WUson BuU. 97:541-543.

Poysa, H. 1986. Foraging niche shifts in multispecies dabbling duck {Anas spp.) feeding
groups: harmftil and beneficial interactions between species. Omis. Scand. 17:333-346.

Raikow, R. J. 1970. The funaion and evolution of the supraorbital process in ducks. Wilson
BuU. 87:568-572.

Reynolds, P. 1987. Observations on the time budget and diving ecology of Long-tailed Ducks
in Eqalungmiut, Nunaat, West Greenland. WUdfowl 38:55-62.

Sanger, G. A. and R. D. Joies, Jr. 1984. Winter feeding ecology and trophic relationships of
Oldsquaws and White-winged Scoters on Kachemak Bay, Alaska. In: D. N. Nettleship, G.
A. Sanger, and P. F. Springer (eds.). Marine Birds: Their Feeding Ecology and Commercial
Fisheries Relationships, pp. 20-28. Canad. Wildl. Serv. Publ. CW66-65/1984.

Savard, J.-P. L. 1982. Intra- and inter-specific competition between Barrow's Goldeneye
(Bucephala islandica) and Bufflehead (Bucephala albeola). Canad. J. Zool. 60:3439-3446.

Savard, J.-P. L. 1984. Territorial behaviour of Common Goldeneye, Barrow's Goldeneye and
Bufflehead in areas of sympatry. Omis Scand. 15:211-216.

Schenkeveld, L. E. and R. C. Ydenberg. 1985. Synchronous diving by Surf Scoter flocks.
Canad. J. Zool. 63:2516-2519.

Schlatter, R. P. 1976. Aves observadas en un sector del Lago Rinihue, Provincia de Valdivia,
con alcances sobre su ecologia. Bol. Soc. Biol. Concepcion 50:133-143.

Schmidt-Nielsen, K. and Y. T Kim. 1964. The effect of salt intake on the size and function of
the salt gland of ducks. Auk 81:160-172.

Sclater, P. L. and O. Salvin. 1878. Reports on the collections of birds made during the voyage
of H. M. S. "Challenger." — No. IX. On the birds of Antarctic America. Proc. Zool. Soc.
London, pp. 431-438.

Scott, W. E. D. and R. B. Sharpe. 1912. Reports of the Princeton University expeditions to Pat-
agonia 1896-1899. Ornithology, Vol. 2, Part 3. Lancaster Press, Lancaster.

Siegfried, W. R. 1976. Segregation in feeding behaviour of four diving ducks in southem Mani-
toba. Canad. J. Zool. 54:730-736.

Sjoberg, K. 1 988. The flightless period of free-living male Teal Anas crecca in northern Swe-
den. Ibis 130:164-171.

SneU.R. R. 1985. Underwaterflightof Long-tailed Duck (Oldsquaw) C/aw^M/a/z^iewM/w. Ibis
127:267.

Stott,R.S. and D.P.Olson. 1973. Food-habitat relationship of sea ducks on the New Hampshire
coastline. Ecology 54:996-1007.

Strange, I. J. 1972. The Falkland Islands. David and Charles, Newtcm Abbot.

Thomas, G. J., M. Owen, and P. Richards. 1977. Grit in waterfowl at the Ouse Washes, Eng-
land. WUdfowl 28:136-138.

Toft, C. A., D. L. Trauger, and H. W. Murdy. 1982. Tests for species interactions: breeding phe-
nology and habitat use in subarctic ducks. Amer. Natur. 120:586-613.

Tome, M. W. and D. A. Wrubleski. 1988. Underwater foraging behavior of Canvasbacks,
Lesser Scaups, and Ruddy Ducks. Condor 90:168-172.

Tremblay, S. and R. Couture. 1986. MorjAologie bucco-linguale d'une guilde de canards bar-
boteurs. Canad. J. Zool. 64:2176-2180.

Trost, R. E. 1981. Dynamics of grit selection and retention in captive Mallards. J. WUdl.
Manage. 45:64-73.

VaUentin, R. 1901. Notes by a naturalist on his voyage to the Falklands and back, with remarks
on the fauna and flora of those islands. J. Royal Inst. Cornwall, pp. 339-364.

Vallentin,R. 1904. Notes on the Falkland Islands. Mem. and Proc. Manchester Lit. Philos. Soc.
48:23-45.

VaUentin, R. 1924. Zoology. In: V. R Boyson (ed.), The Falkland Islands, pp. 285-335.
Qarendon Press, Oxford.

Venegas C, C. and J. Jory H. 1979. Guia de Campo para las Aves de MagaUanes. Inst.
Patagonia, Punta Arenas, Chile.

Vermeer, K. 1982. Food and distribution of three Bucephala species in British Columbia
waters. Wildfowl 33:22-30.

STEAMER-DUCK FEEDING ECOLOGY 39

Venmeer, K. and N. Bourne. 1984. The White-winged Scoter diet in British Columbia waters:
resource partitioning with other scoters. In: D. N. Nettleship, G. A. Sanger, and P. E
Springer (eds.). Marine Birds: Their Feeding Ecology and Commercial Fisheries Relation-
ships, pp. 30-38. Canad. WUdl. Serv. Publ. CW66-65/1984.

WeUer, M. W. 1971. Waterfowl ecosystem studies on the Falkland Islands. Antarctic J. U. S.
6:107-108.

WeUer, M. W. 1972. Ecological studies of Falkland Islands' waterfowl. WQdfowl 23 :25^4.

WeUer, M. W. 1975. Habitat selection by waterfowl of Argentine Isla Grande. Wilson BuU.
87:83-90.

WeUer, M. W. 1976. Ecology and behaviour of steamer ducks. Wildfowl 27:45-53.

Weller, M. W. 1980. The Island Waterfowl. Iowa State Univ. Press. Ames.

White. D. H. and D. James. 1978. Differential use of fresh water environments by wintering
waterfowl of coastal Texas. Wilson BuU. 90:99-111.

Wiens.J.A. 1984. On undersunding a non-equUibrium world: myth and reality in community
patterns and processes. In: D. R. Strong, Jr.. D. Simberloff. L. G. Abele. and A. B. Thistle
(eds.). Ecological Conununities: Conceptual Issues and the Evidence, pp. 439^57.
Princeton Univ. Press. Princeton.

WUson. D. S. 1975. The adequacy of body size as a niche difference. Amer. Natur.
109:769-784.

Woods. R.W. 1975. The Birds of the FaUdand Islands. Compton Press, WUtshire.

Woods, R.W. 1982. FaUdand Island Birds. Anthony Nelson. Shropshire.

Young. D. A. and D. A. Boag. 1 982. Changes in physical conditicm of male MaUards {Anasplat-
yrhynchos) during moult. Canad. J. Zool. 60:3220-3226.

Zweers. G. A., A. F. C. Gerritsen. and P. J. Van Kranenburg-Voogd. 1977. Mechanics of Feeding
of the MaUard {Anas platyrhynchos L.; Aves. Anserlfonmes). Contr. Vert. Evol. 3, S. Karger,
Basel.

40 OCCASIONAL PAPERS MUSEUM OF NATURAL fflSTORY

Appendix I

Published data on food habits of steamer-ducks, by species. Num-
bers following taxa of food organisms refer to chronologically listed publica-
tions given in first footnote.' Genera of food organisms named in publications
follow the reference numbers. References under T. patachonicus which are
marked by asterisks pertain to birds on freshwater.

T. patachonicus

Molluscs (22, 24, 25, 34).

Gastropods. — snails (32* [Chilina], 36).

Bivalves.— mussels (21, 22, 24); clams (29, 36).
Arthropods

Crustaceans (14). — copepods (22, 24, 34); crabs (32* [Aegld], 34);
shrimp (22, 24).
Algae (25).
Vascular Plants (29).

T. brachypterus

Molluscs (7, 14).

Gastropods.— limpets (28, 29, 30, 34, 37); general (28, 34).

Bivalves (8, 34).— mussels (28, 29 [Mytilus], 30, 37).
Arthropods

Crustaceans. — isopods (29, 33); amphipods (29, 33); crabs (29 [Mumda],
30, 33 {Munida], 34); shrimp (28, 34, 37).
Bony Fish (29).
Algae (7 [Macrocystis], 29).
Vascular Plants (29).

'Sources: (l)Darwin (1839, 191); (2)Gould (1841, 191); (3)Cunninghani (1871a, 494);
(4)Sclater and Salvin (1878, 437); (5)Sharpe (1881, 13); (6)0ustalet (1891. 225); (7)Val-
lentin (1901, 350); (8)Vallemm (1904, 34); (9)Scott and Sharpe (1912, 496, 500); (lO)NicoU
(1904, 50; 1908, 172); (ll)Hubbard (1907. 217); (12)Blaauw (1916, 491); (13)Brooks
(1917, 156); (14)Bennett (1924. 280); (15)VaUentin (1924, 326); (16)Beck/Ide PhiUips
(1925, 295); (17)Cobb (1933, 82); (18)Reynolds in Lowe (1934, 479); (19)Murphy (1936,
962-963); (20)Housse (1942, 179 = 1948, 331); (21)Housse (1945. 92-93); (22)Goodall et
al. (1951. 163-164); (23)Delacour (1954. 269); (24)Johnson (1965. 197); (25)Humphrey et
al. (1970. 136); (26)Markham (1970. 45. 48); {21 )RcynoUs fide Humphrey et al. (1970, 131);
(28)Slrange (1972, 206); (29)Weller (1972, 34, 37); (30)Woods (1975, 121); (31)Daciuk
(1976, 28); (32)Schlatter (1976, 140-141, 143); (33)Weller (1976. 49); (34)Johnsgard
(1978. 135, 137-138); (35)Venegas and Jory (1979,73); (36)Weller (1980,27); (37)Woods
(1982. 50).

^Identified as T. patachonicus by Daciuk (1976).

'Dietary data in sources 13, 15, 16, and 17 were referred to T. brachypterus by Murphy (1936,
968).

STEAMER-DUCK FEEDING ECOLOGY 41

T. leucocephalus^

Molluscs

Gastropods (31).

Bivalves (31).
Arthropods

Cnistaceans. — copepods (31); isopods (31); amphipods (31).

T. pteneres

Molluscs

Gastropods (26, 35).— limpets (19 [Nacella], 34); snails (19 [Euthria]).

Bivalves. — clams (27, 33).
Arthropods

Cnistaceans (19, 22, 24, 34, 35).— crabs (12, 21, 33).
Bony Fish (22, 24).
Vascular Plants (29).

Tachyeres sp.^

Molluscs (1,2, 6, 11,15,20,23).

Gastropods. — limpets (13, 16).

Chitons (13).

Bivalves.— mussels (3 [Mytilus]), 5, 9, 10 [Mytilus], 13, 16, 17).
Arthropods

Crustaceans (6).— crabs (4, 9, 11, 12, 13, 17); shrimp (5, 13).

Insects (16).
Echinoderms (15 [Hemiaster]).
Bony Fish (16).

University of Kansas Publications

MUSEUM OF NATURAL HISTORY

The University of Kansas Publications, Museum of Natural
History, beginning with volume 1 in 1946, was discontinued with
volume 20 in 197 1 . Shorter research papers formerly published in
the above series are now published as Occasional Papers, Museum
of Natural History. The Miscellaneous Publications, Museum of
Natural History, began with number 1 in 1946. Longer research
papers are published in that series. Monographs of the Museum of
Natural History was initiated in 1970. Authors should contact the
managing editor regarding style and submission procedures before
manuscript submission. All manuscripts are subject to critical
review by intra- and extramural speciaUsts; final acceptance is at
the discretion of the publications committee. The Occasional
Papers and Miscellaneous PubUcations are produced by a Macin-
tosh computer, LaserWriter, and the software Microsoft® Word
and PageMaker®. Conversion to current format by E. O. Wiley.

Institutional libraries interested in exchanging pubUcations
may obtain the Occasional Papers and Miscellaneous Publications
by addressing the Exchange Librarian, University of Kansas Li-
brary, Lawrence, Kansas 66045. Individuals may purchase sepa-
rate numbers of all series. Prices may be obtained upon request ad-
dressed to Publications Secretary, Museum of Natural History,
University of Kansas, Lawrence, Kansas 66045 - 2454.

Editor: Robert D. Holt

Managing Editor: Joseph T. Collins

Design and Typesetting: E. O. Wiley, K. Wollter,

K. A. Shaw and J. T. Collins

PRINTED BY

UNIVERSFFY OF KANSAS PRINTING SERVICE

LAWRENCE. KANSAS

DATE DUE

■4fAR 3^1 1996

DEMCO. INC 38-2931

3 2044 093 361 707