species, with dominance by this latter group potentially lasting for several
decades. Forest structural characteristics such as canopy height and basal area
may reach values similar to those of primary forest in as little as two decades.
Individuals that colonize the site quickly after its abandonment or assignment
to fallow dominate the site for decades, and succession unfolds because, to a
great extent, of the different life histories and degrees of shade tolerance of this
set of species. Successional processes similar to these also occur in tropical and
subtropical regions of Africa (de Namur and Guillaumet 1978; Donfack et al.
1995; King et al. 1997; Fuhr 1999) and Asia (Whitmore 1984; Riswan and
Kartawinata 1991). Botanical species, not tree functional groups, vary
between continents.
It is vitally important to go beyond basic general descriptions of successions
and characterize and understand the great diversity of successional processes
that are observed in reality. Tropical secondary successions occur as described in
the previous paragraph only under optimal conditions. These conditions
include little or no degradation of site conditions, a well-stocked seed bank,
seed trees located within dispersal range and with functional seed dispersal
processes, viable tree stumps and root systems for resprouting, and minimal
additional disturbance to the site after succession begins. Common sense sug-
gests that these conditions are most likely to be met on small areas, with light
or no agricultural use of the land, embedded in matrices that include large pro-
portions of primary or old secondary forest. It is unlikely that such conditions
will be met in most of the situations in which secondary succession will occur
in the tropics in the coming years. Successional processes on large areas such as
many abandoned pastures of the mainland neotropics are likely to be more
complex than in the optimal scenario and in successional communities of
slower development, lower diversity, and lower productivity. This is especially
likely when a large area is accompanied by a suite of site factors that represent
barriers to succession initiation (Janzen 1988; Nepstad et al. 1991; Finegan and
Delgado 2000). With respect to succession on shifting cultivation plots, it is
important to emphasize that these are small habitat patches created by a land
use system designed with the maintenance of site productivity for agriculture
as a principal goal. This scenario is favorable, in principle, for the development
of fallow vegetation—for secondary succession. On the other hand, these are
habitat patches that experience a high frequency of drastic disturbance (even
traditional fallow lengths should be considered short time periods in relation to
the recovery of many forest characteristics) embedded in a landscape in which
fallow vegetation of intermediate value as a seed source, probably is the most
important single land use. In seasonal environments, fallow vegetation is also
chronically prone to disturbance by uncontrolled fire (Smith et al. 2001), and
in general, long-term declines in site productivity caused by repeated cropping
cycles and ever-shorter fallows appear inevitable. All these factors, in different
ways, represent limitations on the development of secondary vegetation. Limi-
- The Biodiversity and Conservation Potential of Shifting Cultivation Landscapes 165