capture fleeing insects, feeding exclusively on flower nectar, or foraging only
in clusters of suspended dead leaves. If critical habitats are missing or poorly
represented, the dependent species probably will disappear too.
Third, big fragments are proportionally less affected by edge effects, the
physical and biological changes associated with the abrupt boundary between
forests and adjoining modified habitats. Area and edge effects are difficult to
distinguish, and few studies have effectively done this, usually because samples
in small fragments are near edges, whereas those in large fragments are far
from edges, creating a strong correlation between edge and area predictors.
However, it is becoming clear that many population and community changes
in habitat fragments that were commonly attributed to area effects are in fact
the result of edge effects (Lovejoy et al. 1986; Laurance et al. 2002). Edge
effects are discussed in detail later in this chapter.
Finally, large fragments have lower rates of population extinction than
small fragments, especially for species that need large territories, are sensitive
to edge effects, are unable to cross even small clearings, or cannot tolerate con-
ditions in the surrounding modified habitats. An intriguing example is the
specialized ant-following birds of the neotropics, which accompany maraud-
ing swarms of army ants in order to capture fleeing insects. Each ant colony
raids extensive areas of up to 30 ha, and the birds’ home ranges must encom-
pass two or three colonies because every colony spends several weeks per
month in an inactive phase (Harper 1989). Because their ants need large areas
and because they must have access to several colonies, the specialized ant fol-
lowers are especially prone to extinction in small fragments (Harper 1989;
Stouffer and Bierregaard 1995b).
Despite these factors—and contrary to predictions of the island biogeog-
raphy theory (MacArthur and Wilson 1967)—larger habitat fragments do not
always support more species than smaller fragments. In certain taxonomic
groups, species richness can actually increase in fragments when there is an
influx of species from the surrounding modified habitats (Brown and Hutch-
ings 1997; Didham 1997b; Tocher et al. 1997) or when conditions near edges
become more favorable for a particular species or guild of species (Stouffer and
Bierregaard 1995a). Species that proliferate near edges or in the matrix can
include both nonforest species and those that were formerly limited to natu-
rally disturbed forest patches (Stouffer and Bierregaard 1995a; Brown and
Hutchings 1997).
Distance Effects
Interfragment distance can affect the movement of animals and plant propag-
ules between fragments, and even remarkably small clearings can become
impassable barriers for many rainforest organisms. In the Amazon, many ter-
restrial insectivorous birds have disappeared from forest fragments and failed
36 I. Conservation Biology and Landscape Ecology in the Tropics