explanation for between-patch type differences in total species richness; for
example, the secondary vegetation in their study lacked foliage above 20 feet
(6.1 m) and therefore lacked the species that use this component of the habi-
tat in other patch types, such as primary forest. Other authors demonstrate
differences in vegetation vertical structure between fallows of different ages
and point out their likely relationship to changes in the species richness of ver-
tebrate assemblages (Blankespoor 1991; Medellín and Equiha 1998; Shankar
Raman et al. 1998). In general, woody secondary vegetation tends to have a
more uniform structure than mature forests in terms of canopy height and the
absence of treefall gaps. Nevertheless, some of the other ways in which habi-
tat structural factors are linked to vertebrate community characteristics may be
subtle. For example, Terborgh and Weske (1969) found that variation in
foliage height profiles was insufficient to explain all the observed variation in
bird community composition and diversity at their site, so they invoked and
justified an additional set of special habitat quality factors. Examples of such
factors are vine tangles, absent, along with their associated birds, from the
second-growth site studied by these authors.
Habitat structure may also influence the characteristics of vertebrate com-
munities at the landscape scale. Although fallow vegetation is less structurally
diverse than primary forest at the stand level, the landscapes with shifting cul-
tivation activity in Anderson’s (2001) study area were morestructurally diverse
than those with only primary forest. Three raptor species of the landscapes
modified by human activity are common in anthropogenic open habitats of
Honduras, so that the greater structural diversity of these landscapes presum-
ably underlies the greater richness, diversity, and abundance of raptors
observed in them, in ways linked to the habitat preferences, foraging tactics,
and preferred prey of the different bird species (Anderson 2001).
Overall, the information available justifies the simple conclusion that ele-
ments of the forest fauna that use habitat structural elements or types that are
absent or uncommon in areas influenced by shifting cultivation probably will
be less abundant in those areas than in forest. Conversely, habitat features
associated with shifting cultivation will bring species adapted to those fea-
tures into the community. These relationships will operate at different spatial
scales depending on the characteristics of the vertebrate guild or species
involved.
It is highly likely that spatial and temporal patterns in the availability of
food influence the characteristics of vertebrate communities in shifting culti-
vation landscapes. Cowlishaw and Dunbar (2000) emphasize that Musanga
cecropioides,a pioneer tree abundant in African shifting cultivation landscapes,
produces fruit attractive to primates over much of the year and so contributes
to the use of agricultural habitats by these animals. However, habitat use pat-
terns by primates may vary over the year in relation to the fruiting phenolo-
gies of the different communities of the landscape (Fimbel 1994, cited by
178 III. The Biodiversity of Agroforestry Systems