speed with which soil fertility is recovered decrease, and the time needed to
control certain weeds in the patch increases. Increases in fallow length may be
observed under some circumstances (Smith et al. 2001). Among the costs
associated with longer fallow lengths is the need to achieve a greater area of
land under the shifting cultivation cycle to maintain a given quantity of agri-
cultural produce and the attendant possibility that this greater land area must
be obtained by the labor-intensive cutting of more primary forest. A critical
moment arrives when it is no longer possible to cut more primary forest, so
that land area under cultivation can be increased—as a response to declining
crop yields or an increasing population to be supported from the land—only
by shortening fallow length. The shortening of fallows is much more common
than their lengthening. For example, Thiele (1993) reports that traditional fal-
low lengths in seasonal environments of lowland Bolivia were 6–12 years but
by the 1980s had declined markedly, with a modal length of 4 years. Fallow
lengths had declined to a range of 3–10 years in the 1990s in northeastern
India from a traditional value of 60 years (Ramakrishnan 1992; Shankar
Raman 1996). Traditional fallow lengths in the Taï forest region of the Côte
d’Ivoire were 14–30 years and had been reduced to 6–10 years by the end of
the twentieth century (de Rouw 1995).
Shortening fallow lengths can exacerbate the decline of crop yields and
increase weed problems and therefore labor needs during cultivation (de Rouw
1995; Roder and Maniphone 1998; Smith et al. 1997, 2001). For example,
short fallows in the Côte d’Ivoire led to a 72 percent increase in weed biomass
during cropping periods (Becker and Johnson 2001). A logical response to
these tendencies, from both farmers and the research and development com-
munity, has been to seek alternative technologies. Permanent agriculture—as
sought by the government of Laos to replace shifting cultivation, which is con-
sidered to be environmentally undesirable (Fujisaka 1991)—is one alternative
(see also Thiele 1993). Improved (planted) fallows are another main focus of
attention (Fujisaka 1991; Buckles and Triomphe 1999; Szott et al. 1999).
Improved fallows are intended to fulfill the agroecological function of the fal-
low in short time periods (perhaps 3 years or even less) and may themselves
bring increased labor needs (Szott et al. 1999), although examples of their
spontaneous development and wide adoption by farmers are documented
(Buckles and Triomphe 1999).
To conclude, the relationship between fallow length and the floral and
faunal characteristics of the community is one of the most important ele-
ments in the potential of shifting cultivation landscapes for biodiversity con-
servation, as discussed later in this chapter. The tendency toward shorter fal-
lows and the development of techniques for fallow “improvement” represent
severe limitations on this potential. In particular, improved fallows seem to
be of little value for forest biodiversity conservation because of their often
monospecific composition and short duration and will not be considered fur-
160 III. The Biodiversity of Agroforestry Systems