Ecologists and Forest Conservation 477populations of tree species reportedly vary, from
rapid losses (e.g., Leighet al. 1993) to long-term
persistence (Thomas 2004) in even small frag-
ments. And although edge effects undoubtedly
also occur in large fragments in Brazil (Nepstad
et al. 1999, Laurance 2000, 2004, Lauranceet al.
2002, Silva Diaset al. 2002), Phillipset al. (2006)
failed to find the expected extensive edge effects in
Amazonian Peru. Furthermore, there is little evi-
dence that forest fragments of more than 10,000
ha lose a large portion of species if hunting is
controlled and wildfires are avoided. Nevertheless,
results of fragmentation research are often used to
justify pleas for parks much larger (> 106 ha) than
any studied forest fragments (e.g., Terborghet al.
1997, Peres 2005, Tabarelliet al. 2004, Tabarelli
and Gascon 2005). Although the “Single Large
Or Several Small” (SLOSS) reserves debate has
been described by some critics as a purely theoret-
ical exercise (e.g., Simberloff 1997), the “larger is
better” argument seems to prevail in some conser-
vation and scientific circles (e.g., Laurance 2005,
Peres 2005). Ecologists could hel pby studying
diversity maintenance in larger fragments, but
their research is likely to have more impact if it
were focused on improving matrix management
(Kupferet al. 2006).
Similarly, despite substantial reservations
about its relevance to conservation (Hanski 1997,
Freckleton and Watkinson 2002), metapopula-
tion theory has been invoked to support cam-
paigns to connect forest fragments with habitat
corridors (e.g., Laurance and Bierregaard 1997).
Corridors make sense and sometimes work as
predicted (Beier and Noss 1998, Haddadet al.
2003), but their likely biodiversity benefits should
not be exaggerated (Harrison and Bruna 1999),
especially when they are long (Simberloffet al.
1992) and the matrix surrounding fragments is
already favorable for many species (e.g., Gascon
et al. 1999). Despite these possible limitations, the
theory has drawn attention to the value of small
habitat patches for species survival, to the issue
of species movement, and to the dynamic aspects
of species maintenance. Research on increasing
the retention of biodiversity in multifunctional
landscapes consisting of mosaics of forest frag-
ments (small to large, connected to isolated),
production forests, plantations, and agricultural
lands is urgently required (Zuidema and Sayer
2003).
A body of ecological theory that has not yet
been fully utilized in tropical forest conservation
concerns the capacity of disturbed ecosystems
to return to pre-intervention states, the roles
of diversity in this resilience, and the likelihood
that further disturbance will precipitate dramatic
changes in ecosystem structure and composi-
tion (e.g., Holling 2001; see http://www.resalliance.org).
Resilience theory is broad and flexible enough
to include human-induced as well as other sorts
of perturbations, and accepts that responses to
stresses and disturbances are often non-linear
and sometimes completely unexpected. Such sur-
prises are sometimes manifest in “phase shifts”
(sensuFolkeet al. 2004), such as when tropical
forest is degraded to the point that it becomes
savanna (e.g., Oyama and Nobre 2003). Unfor-
tunately, models based on resilience theory are
unavoidably only as good as the input data,
which must include social as well as biophysi-
cal variables. For example, failure to recognize
the importance of historical land-use practices
and inappropriate extrapolations of ecological evi-
dence resulted in mistaken impressions of the role
of local people in forest destruction in both West
and East Africa (Fairhead and Leach 1996, Bas-
sett and Zuéli 2003). Given the complexity of
tropical forests and the diversity of issues rel-
evant to their management and conservation,
the humbling occurrence of surprising responses
to planned and unexpected interventions is and
will remain common. In light of this uncer-
tainty, a diversity of locally tailored approaches
to conservation is warranted, with adjustments
as justified from experience and frequent mon-
itoring (Figure 28.1; Bormann and Kiester
2004).CONSERVATION SOLUTIONS
VARY IN SIZE AND LAND-USE
INTENSITIES
While we might like to think of tropical forests
as extensive and uninhabited wilderness areas,
most do not fit this description (e.g., Denevan
1992, van Gemerdenet al. 2003, Williset al.