and Cox 1987, Hobbs 1992, Simberloff et al. 1992).
Since June 1993, we have spent over 1000 hours on
local farms investigating whether birds use planted
and natural remnant windbreaks as corridors. We
mist-netted birds in windbreaks, pastures, and forest
fragments, recorded the net in which each bird was
captured and the direction from which it entered the
net, and marked and released them. The sampled
windbreaks were short and narrow (50 x 5 m) and
were connected on both ends to a forest fragment,
connected on one end only, or totally unconnected.
Forest fragments of various degrees of isolation and
pastures were also sampled.
Natural remnant windbreaks were used by at least
68 species of birds; 52 species were found in planted
windbreaks composed of only two introduced tree
species: cypress (Cupressus lusitanical, Cupressaceae)
and tubu (Montanoa dumicola, Asteraceae), neither
of which produce resources such as fruit. Five
neotropical migrants occurred in planted windbreaks,
as well as 14 others in natural windbreaks and forest
fragments. Insect-eating birds comprised almost one-
half of the species captured. The large amount of edge
of windbreaks may make the areas good places to for-
age. Birds that feed on fruit and insects, and hum-
mingbirds (which eat nectar and insects) were the next
two largest guilds of birds captured, followed by fruit-
and seed-eaters, despite the fact that trees in planted
windbreaks did not have fruit, nectar, or seeds at the
time of sampling. In unconnected windbreaks, frugi-
vores and seed-eaters comprised a much smaller per-
centage of the total number of birds captured than in
the connected windbreaks.
However, there were more forest edge species
using connected windbreaks than unconnected wind-
breaks. Moreover, we found more forest-dwelling
species in connected windbreaks than in unconnected
windbreaks. The total number of species captured and
the hourly capture rate were significantly higher in
windbreaks connected to forest fragments than in
those without connections. There was also a positive
correlation between hourly capture rate and distance
to the nearest forest fragments.
Windbreaks appeared to serve as corridors for
many bird species. The daily capture rates should
have dropped to nearly zero after 4-5 consecutive
days of mist-netting if only birds residing in the
area were using the windbreaks (Karr 1981). How-
ever, in connected windbreaks, the daily capture
rate increased after 3 consecutive days of mist-net-
ting, whereas the daily capture rate in unconnected
windbreaks declined monotonically. Many birds
were captured while using windbreaks to return
to forest fragments, rather than crossing open ground.
Several forest-dwelling species were also captured
in windbreaks other than the one in which theyFigure 12.11. Windbreaks planted in the San Luis valley below Monteverde. Photograph by Paul
Butler.449 Conservation Biology