THE AUK
A QUARTE1LLY JOURNAL OF
ORNITHOLOGY
VOL. 64 JULY, 1947 No. 3
THE TROPICAL AVIFAUNA OF THE UPPER MAGDALENA
VALLEY, COLOMBIA
BY ALDEN H. MILLER
Plates 12-13
T• arid tropical area of the upper Magdalena Valley of Colombia
extends from Honda, Tolima, south through the departments of
Tolima and Huila to about latitude 2 ø N., a distance of 375 kilometers
(see Chapman, Bull. Amer. Mus. Nat. Hist., 36:84 if., pl. 26, 1917).
This large district is unforested and its xerophilous vegetation reflects
a greater extreme of aridtry than does that of the floor of the Cauca
Valley which lies to the westward on the other side of the central
Andes. Along the main water courses in the upper Magdalena basin
are strips of woodland with tracts of savanna adjacent to them, but in
large measure the area is covered with open thorn scrub, sometimes
forming a chaparral, yet often broken or scattered. Clumps of cacti
are frequent and there are areas of short-grass plains and of barren,
clay or gravel-surf?ced badlands.
This arid basin is bounded on the east, sough and west by the
eastern and central divisions of the Andes on which the Subtropical
and Temperate zones appear. Only at Andalucia in southeastern
Huila is there a moderately low gap leading eastward through the
encircling mountains. This gap is stated by Chapman to be 7000 feet
in elevation; it lies in the Subtropical Zone. The Tropical Zone occurs
below about 5000 feet. To the north the upper Magdalena Valley
adjoins the more humid middle section of the Magdalena drainage
which, by reason of its climate and forests, forms a varyingly effective
ecologic barrier between the biota of the arid upper region and that of
the moderately arid Caribbean coast.
Chapman's report (op. cit.) on the bird life of Colombia dealt much
more fully with the fauna of the arid tropical Cauca district than with
351
THE AUK, VOL. 64
PLATE 12
i
(Upper) EDGE Ore PLAINS 10 KZLOMETERS SOUTH Ot• VILLA¾IEJA, HUILA, CO-
LOMBIA, SHOWING EROSION EXPOSURES AND PATCHES O1' SCRUB COVER.
(Lower) SLOUG•I AT VILLAVIEIA, SHOWING WOODLAND BORDERS INCLUDING
•AMBOOS AND •ANANASo
359. MI•,•,•I•, The Tropical Avifauna of Colombia [•u•k•
that of the upper Magdalena. He stated (p. 124) "that our work has
been done about the borders of the upper Magdalena district. Of the
fauna of the floor of the valley, I feel that we have still much to learn."
More specifically, Chapman's parties worked at Honda, in a narrow
part of the valley, where the arid tropical fauna is not typically de-
veloped, and at Chicoral, near Giradot. Even at Giradot the full
measure of aridity is not encountered. In travelling south from this
point, it may be seen that woodlands become more restricted, sa-
vannas more open and the grass shorter; scrub, open plains and bad-
lands predominate. Chapman's parties also worked at San Agustfn
at the head of the valley at 5000 feet in the upper edge of the 2•ropical
Zone, and below Andalucia at 3000 feet near the margin of the tropical
area. Neither of these southern stations may be depended upon to
afford an adequate representation of the fauna of the valley floor. A
very few native-made skins were available to Chapman from Purifi-
caci6n, 50 kilometers south of Giradot.
In January and l•'ebruary of 1945 I had opportunity to collect and
observe birds intensively at Villavieja, Huila, at an elevation of 435
meters (1427 feet), 30 kilometers north of Neiva. 2•his is in the middle
of the arid tropical district and represents its most extreme manifes-
tation. 2•hrough the kindness of Dr. R. A. Stirton, I shared his
quarters at Villavieja, where he was engaged in collecting fossils in
the late Miocene beds of the near-by badlands; Stirton also in many
other ways aided and encouraged my work in Colombia. I am further
indebted to the Servicio Geo16gica of Colombia, to the 2•ropical Oil
Company, and particularly to Dr. J. Wyatt Durham of that com-
pany, who generously gave various critical assistance to Dr. Stirton
and to me. I was able to work uninterruptedly at Villavieja from
January 12 to l•'ebruary 11, spending half of each day hunting and
observing in the field. A total of 380 birds was taken; also there were
six specimens taken by Stirton in near-by areas in Huila and Tolima.
•'he specimens represent 123 species. Additionally seven species were
identified adequately by sight so that 130 species may be recorded
from Villavieja. In the summer of 1946 identifications were reviewed
and critical specimens were compared directly with Chapman's and'
other collections at the American Museum of Natural History. I am
much indebted to John •'. Zimmer for the use of the Museum's facili-
ties and for aid in taxonomic problems relating to South American
birds.
A variety of habitats was accessible at Villavieja. About five
kilometers northeast along Quebrada La Venta lay the large area of
barren land where some of the best fossil exposures occurred. The
Vol. 64']
1947 / •V[II•I•I•R, The Tropical Avifauna of Colombia
ground surface here was chiefly blue or brown clay, but in some areas
the surface was covered with a residue of Pleistocene cobbles and
gravel. There was little or no grass. Small mimosa bushes grew
scatteringly in the intermittent water courses. South and east of
town, six to ten kilometers, a short-grass plain could be reached. On
this were several small, shallow playas, some with borders of tall grass,
thorny brush and cactus (Plate 13, lower figure). The thorn scrub
nearer town varied in height from three to 18 feet. Much of it could
be traversed easily, but occasionally tracts of an acre or more proved
impenetrable. Grass two feet in height was intermixed with the scrub.
Closer to the river were bottomlands with open savannas with trees
15 to 30 feet tall, 50 to 150 feet apart, and with waist-high grass,
except where it was overgrazed. In the Laja Valley, three kilometers
northeast, a tributary of the Magdalena, there was a merging of thorn
scrub and streamside woodland. The latter was often 60 feet in
height and usually only 300 feet wide (Plate 13, upper figure). The
woodland along the Magdalena River itself was somewhat less xero-
philous and even taller than that of the Laja Valley. Tall bamboo
clumps were frequent. Probably originally these woods were in
places two kilometers across, but they were rather thoroughly broken
up by pasture lots and small plantations of corn and bananas so that
they usually consisted of narrow borders of trees. Just south of
Villavieja was a slough two kilometers in length, the lower course of a
small tributary, which had little flow and was overhung by trees. The
water was choked with floating vegetation except for a narrow central
channel (Plate 12, lower figure). 2•his situation was particularly
favorable for the naturalist, for the slough was attractive to marsh
birds and by wading in it the species of the woodland borders were
readily accessible. The Magdalena River was swift-flowing, with loose
rock bottom in many places and with steep, muddy banks.
In the accounts of species which follow, taxonomic and distributional
matters are included where new information has come to light, but
primarily attention is devoted to observations on habitat and to
fragmentary records of behavior, items which are particularly lacking
for many South American birds. Data on sexual activity and weight
are for similar reasons included. 2•he breeding cycles of birds in
equatorial regions are but poorly known. Specimens were taken with-
in five kilometers of Villavieja and in 1945 unless otherwise noted.
Some spedmens saved as alcoholics or as skeletons were not weighed,
and alcoholics were not sexed. Age has been determined from the
condition of the skull unless otherwise stated.
[Auk
354 MILLER, The Tropical Avifauna of Colombia tJulv
A nhinga anhinga leucogaster (Vieillot).--Encountered but once, when one was taken
as it was perched on a post standing in a muddy section of the slough where there
were broad borders of water vegetation. The bird was fully exposed, as there was no
foliage screen. When first seen, it was standing with wings outspread, possibly in
preparation for flight. I follow Wetmore (Proc. U.S. Nat. Mus., 93: 232-233, 1943)
with respect to the racial taxonomy of this species.
c•, February 3, wing 338 min., tail 264, eulmen 97.
Ardea cocoi Linnaeus.--One was seen flying along a wooded section of the Mag-
dalena River on January 24.
Butorides striatus striatus (Linnaeus).--Of regular occurrence along river courses
and sloughs and about ponds on the plateau. The adherence of these herons to
foliage-screened borders, the character of their notes and their behavior in every
way suggested the closely related Green Heron, Butorides virescens, of North and
Central America.
c•, January 31, 163 gm., testis 7 mm., iris and face yellow; 9, first-year, February
2, I0 km. S. Villavieja, 135 gm.
Theristicus caudatus (Boddaert).--These ibises proved to be fairly common on the
plains and in the badlands where they worked along muddy stream banks, about
playa lakes, in dry short grass, and well out on the stony slopes and mesas. In feeding
they picked objects from the surface and did no probing or stone-turning. Often they
were associated in pairs. Once two birds alighted in the top of a twenty-five foot tree
after they were flushed from the ground. At times they flew high overhead, calling,
often at such heights that their markings could not be made out. Chapman did not
record this species in the Magdalena Valley.
c•, February 1, bare skin of face black, iris and feet red; 9, same data; immature
c•, 10 km. S. Villavieja, February 8, iris dark, feet pink.
Arias discors Linnaeus.--A flock of about 50 frequented two shallow ponds on the
plateau 10 kilometers south of Villavieja. These ponds were each about two acres in
extent, well grown with water vegetation and bordered with some low thorn trees;
they nowhere were more than two feet deep. Few male teals in nuptial plumage
were noted.
9, February 2.
Coragyps atratus (Bechstein).--Common about towns and dwellings and along the
river.
Cathartes aura (Linnaens).--Seen often in the open scrub country, soaring about
the mesas and escarpments. In a small side valley, on January 18, an adult flushed
from a nest when I fired into a tangle of trees and thorny bushes. The clump of
vegetation was about 75 feet across, and near its middle was a boulder five feet high.
Beside the rock, partly protected above by an overhang, was a single vulture egg on
bare earth from which leaves had been cleared. The nest was clean, but immediately
flies settled on the egg. Fifty yards away the adult sat on a low bluff, craning its
neck. On January 23 an adult was covering the nest closely and had to be pushed off.
The single egg had hatched and the young one gave a loud hiss. I judged it to be
only about a day old; half egg shells were in the nest. Flies immediately swarmed
on the young bird's head. The parent reigned injury, beating through the brush 15
feet away. On February 4 the adult flushed before I entered the thicket. The
young was then the size of an adult Screech Owl and sat upright on its haunches,
hissing.
tteterospi$ias meridionalis meridionalis (Latham).--These hawks, which behave and
on the wing look like buteonines, especially rough-legs, were encountered regularly