lower costs) of a technique can more than offset this effect by providing an
incentive to cultivate a larger total area. And because kudzu saves labor, the
farmers will have the capacity to do so.
One should be very cautious in drawing general conclusions about
improved fallows based on a single study. First, kudzu is not a tree, although
introducing it in some cases can be seen as a technological change in an agro-
forestry context. Second, improved tree fallows might be more labor intensive
than kudzu fallows; clearing certain planted tree fallows takes more work than
clearing a spontaneous fallow (see Franzel 1999 and other articles in the same
issue). The general lesson is that the impact on natural forests depends on the
characteristics of the agroforestry practice in question, in combination with
farmer and market characteristics of the area, as well as government policies.
Agroforestry could reduce the pressure on forest if it reduces ecological and
economic risks for the farmer. Agriculture, particularly rain-fed tropical agri-
culture, is a risky business, and risk considerations are central in farmers’
decision-making processes. Farmers may overexploit natural resources as
(short-term) insurance against yield and price risk, thereby ensuring that their
income, even in a bad year, is above their subsistence needs. Coxhead et al.
(2001) found evidence of this in their study from northern Mindanao, in the
Philippines. Risk-reducing technologies therefore should enable farmers to
convert less forest to agriculture.
Another aim of agroforestry is to make the system more resilient to ecolog-
ical shocks such as dry years or pest outbreaks, which will not affect all species
in a mixed system or not to the same extent (Schroth et al. 2000; see also
Chapter 16, this volume). The multi-output nature of the system is also insur-
ance against fluctuating market prices; all eggs are not put in the same basket.
On this account, mixed tree crop systems (multistrata agroforestry), compet-
ing with monoculture systems, should be favorably placed to reduce the pres-
sure on natural forests.
Not all systems reduce risk, however. Introducing cash tree crops (e.g., fruit
trees) at the expense of crops for domestic use (e.g., cassava) or cattle exposes
the farmer to higher market risks. Moreover, when a number of farmers simul-
taneously adopt new tree crops, the increased supply puts downward pressure
on prices and could easily make them unprofitable. Furthermore, we need to
consider the sustainability of the technology and how suitable it is for cultiva-
tion on recently cleared forests. Ruf and Schroth (Chapter 6, this volume)
point out that cocoa is particularly suitable for recently cleared forestland and
that cocoa farmers enjoy a “forest rent” (i.e., a higher profit compared with
planting on previously cultivated land) in recently converted forests. Suitability
also refers to the necessary infrastructure. Perishable products, such as fruits,
necessitate proximity to markets and good infrastructure (something normally
not found at the forest frontier), whereas nonperishable products such as rub-
ber and cattle are less dependent on regular transportation.
98 II. The Ecological Economics of Agroforestry