were originally captured, showing that birds use
windbreaks to move from one area to another. Fur-
thermore, there were large differences in species
diversity between isolated and connected frag-
ments; isolated forest fragments had approximately
one-half as many species as connected forest fragments.
Thus, connected windbreaks are used as corridors be-
tween forest fragments by many bird species in Monte-
verde. Linking habitats with windbreaks may help
maintain biodiversity by enabling birds to use for-
est fragments that would otherwise have been left as
isolated remnants.WINDBREAKS AS HABITATS FOR TREES
Celia A; Harveyindbreaks are common features in temper-
ate and tropical agroecosystems (Caborn
1965, Forman and Baudry 1984, Budowski
1987). Their primary function is to protect farms from
wind damage or soil erosion, but they also increase
the amount of forest in the landscape and provide
habitat for forest species (Caborn 1965). In temperate
regions, windbreaks and hedges serve as important
habitats or corridors for a variety of species (birds:
Yahner 1982, Haas 1995; see Nielsen and DeRosier,
"Windbreaks"; mammals: Yahner 1983, Bennett 1980;
insects: Lewis 1970, Pasek 1988; see Williamson and
Darling, "La Ventana"; plants: Corbit 1995), especially
in areas where forest habitat is scarce (Forman and
Baudry 1984, Forman 1995). Species composition,
structure, age, width, and proximity of windbreaks to
the forest can influence the abundance and diversity
of forest species that use them (Forman and Baudry
1984, Johnson and Beck 1988; Bennett 1990). Little
is known about the value of windbreaks as habitat
for tropical forest species. If agricultural windbreaks
provide essential habitat, the planting of windbreaks
could offer a pragmatic approach to the conservation
of forest species within agroecosystems.
In the Monteverde region, more than 180 km of
windbreaks have been planted on dairy farms as part
of the Monteverde Conservation League's reforesta-
tion program (see Chap. 10, Conservation Institu-
tions). These windbreaks are excellent study systems
because the majority of windbreaks are of known age
and similar in species composition and design. The
objectives of my study were to determine whether
windbreaks in Monteverde provided habitat for for-
est trees and to examine how the location of wind-
breaks influences the abundance and species richness
of trees within them. I compared tree abundance and
species richness in windbreaks connected to forests
to those of windbreaks that were not connected, and
between windbreaks with or without remnant trees.
In 1995, I surveyed understory vegetation in 51
windbreaks on 20 farms in Monteverde and the
nearby communities of Cerro Piano, Santa Elena, Los
Llanos, La Cruz, and Cafiitas. All the windbreaks had
been planted in pastures 5-6 years earlier and con-
tained 2-4 rows of trees (primarily Montonoa guate-
malensis, Croton niveus, Cupressus lucitanica, and
Casuarina equisetifolia). They were all fenced to
protect them from cattle grazing. Eight pairs of wind-
breaks were established; one windbreak of each pair
was connected to forest and the other was not con-
nected. There were also 10 pairs in which one wind-
break contained a remnant tree and the other did not.
The remainder of the windbreaks were chosen ran-
domly. To survey tree seedlings, I divided wind-
breaks into 20 m segments and randomly placed 10
plots (1.5 x 1 m) within each segment. Within each
plot, I counted and identified all woody plants
greater than 4 cm in height. Identifications were veri-
fied by W. Haber.
My survey indicated that Monteverde windbreaks
provided habitat for at least 83 forest tree species. Al-
though the most common species (Viburnum costari-
canum and Trichilia havenensis) were pioneer spe-
cies, 50 of the 83 species were primary forest species.
Most of the forest tree species occurred infrequently
within the windbreaks; 12% of the species were rep-
resented by only a single individual. Such low fre-
quencies may reflect the relatively short time for seed
dispersal and colonization to occur. Alternatively, the
low frequencies could indicate that the physical con-
ditions of windbreaks (higher temperatures, higher
light levels, and lower humidity compared to forests)
are unfavorable for forest tree species.
Windbreaks that were connected to forests had sig-
nificantly more forest tree species than windbreaks
that were not connected (paired rank sign test, P =
.016). Connected windbreaks also had significantly
higher forest tree seedling densities than isolated450 Conservation BiologyW