etable matter as well as seeds, they 
might easily have shifted from a forest 
to a shrub environment. In the puna, 
the endemic thick-billed siskin 
(, Spinus crassirostris) appears to be 
restricted to some very high altitude 
woodlands found within the grassland 
and scrub and may have derived from 
the smaller-billed hooded siskin (5. 
magellanicus), a species with a wide 
range and broad habitat tolerance. 
The buntings, which include the yel- 
low-finches, may have moved in from 
open areas farther south, perhaps the 
grassy pampas or the Patagonian 
steppes, rather than the forested slopes 
below the newly formed paramo and 
puna zones. The arid woodlands of 
the montane basin, however, are al- 
most as likely an original habitat for 
these stout-billed finches. As for the 
doves, they probably came from lower 
altitudes, and originally perhaps from 
North and Central America. 
What about the small, brown 
ground-dwelling miners, other mem- 
bers of the diverse family Furnariidae, 
and the ground tyrants of the genus 
Muscisaxicolal These birds subsist on 
a diet that consists primarily of in- 
sects. Since none of the high Andean 
furnariids appear related to Central 
American stocks, they would seem to 
have come from farther south or south- 
east in South America, either the cam- 
pos (grasslands) of Brazil, the pampas, 
or the Patagonian steppes. But they 
could also, in part, have evolved lo- 
cally in the central Andes, from stocks 
living in cloud forests down the slopes. 
The landscape probably progressed 
slowly enough from savanna or wood- 
land to arid grassland or shrubland 
for such an ecological and evolution- 
ary transition to be possible. 
The Muscisaxicola ground tyrants 
are an extraordinary group of birds 
with unusual habits — for flycatchers. 
They live away from trees and hop 
and run along the ground on their 
long, thin legs. The proto-Mwsmax- 
icola , having “discovered” the ground 
niche, then diversified into a large 
number of species by allopatric spe- 
ciation, a good example of adaptive 
radiation. Several of the species are 
exclusively high Andean. Did they 
originally come from stocks already 
living in open spaces elsewhere, such 
as Patagonia? Perhaps, but not nec- 
essarily. While eight or so species of 
ground tyrants do live in open habitats 
along the Andean axis, from moun- 
taintops in Colombia to southernmost 
South America, one species lives along 
rivers at the forested base of the An- 
des. The possibility thus exists that 
the ground tyrants evolved from low- 
land ancestors living in openings in 
forested habitats. The ecological pref- 
erences of the closely related chat ty- 
The twenty-foot-tall puyas 
(relatives of the pineapple ) growing 
at 13,000 feet in Peru, below, and 
the yellow-finch resting on a 
cushion plant in Bolivia, right, are 
characteristic of puna landscapes. 
rants (genus Ochthoeca) may shed 
light on the origin of the high Andean 
ground tyrants. Chat tyrants encom- 
pass all ecological intermediates, from 
cloud-forest species to arid-scrub spe- 
cies living high up in the puna , without 
much morphological or behavioral 
change. One species, the brown- 
backed chat tyrant ( Ochthoeca fumi- 
color), even shows intraspecific vari- 
ation in habitat preferences: in the 
Venezuelan Andes, for instance, I 
have observed these birds in dense 
montane woodlands at about 11,000 
feet and in increasingly more open 
habitats all the way up to 14,500 feet 
in desert paramo. 
If the chat tyrants have invaded 
the geologically young paramo and 
puna habitats from lower-altitude 
A Greensmith, Ardea London 
