in the species diversity of organisms with a coarse-grained perception of habi-
tat, such as diurnal raptors, and undoubtedly has an important influence on
species-level diversity in general. Trends in landscape structure and diversity
that intuition tells us probably are detrimental to biodiversity and probably are
associated with the trend toward shorter fallows would include greater domi-
nance of the landscape by younger fallow habitats and increases in the mean
areas of patches of anthropogenic vegetation. Specific aims for management to
counter these trends might be to maintain or increase the area of older fallow
vegetation and to maintain a high degree of interspersedness of different patch
types. More detailed quantitative analyses of shifting cultivation landscapes
than those available should provide further pointers to technically desirable
management objectives (Metzger 2002).
The landscape-scale consequences of adopting improved (i.e., planted) or
managed fallows would depend on how this change affects the relative area
and spatial configuration of natural fallow vegetation and other patch types in
the landscape, which is impossible to assess at present. It is clear that
improved, planted fallows, which are short (often less than 3 years) and often
feature a single planted species as a major component (Szott et al. 1999), in
general are of lower diversity at the patch scale than natural fallows.
Finally, by analogy with ecological principles related to fragmented com-
munities (Laurance et al. 2000; Metzger 2000), factors such as patch size and
shape, the type of community or communities bordering a given patch, and
distances to similar patch types must also influence biodiversity in any given
patch. There appears to be no published information on this aspect of biodi-
versity and its dynamics in shifting cultivation landscapes.
Community-Level Management
Box 8.2 contains a summary of our review of aspects of the biodiversity of the
communities that make up shifting cultivation landscapes, on which we base
the following suggestions for biodiversity management at the level of individ-
ual communities or patch types within such landscapes. Box 8.2 makes clear
that fallow vegetation and crop fields are anthropogenic communities whose
characteristics are largely shaped by drastic, high-frequency disturbances. As
such, they are inhospitable to forest-dependent plant species, and whether
forest-dependent vertebrates are observed in them is likely to depend at least
partly on the presence of older secondary or mature forest in the landscape.
The anthropogenic nature and hostility to patch-scale biodiversity of forest
species probably is even more marked in planted fallows.
Analysis and management at the scale of the individual patch arguably are
most important from the point of view of plants because mobile animals and
birds range over a variety of patches within the landscape. However, aspects of
vegetation composition and structure of fallow and secondary forest could be
184 III. The Biodiversity of Agroforestry Systems