development of the cabruca system. As Alger and Caldas (1992) note, farmers
on large estates tended to plant cocoa under native tree shade (i.e., in the
cabruca system) and to invest in only part of their holdings, leaving more
under forest, whereas small farmers were more likely to clearcut the forest on
a larger part of their property and use bananas and other planted shade instead
of native trees for their cocoa crop. Planting cocoa in the cabruca system took
smaller investments per unit area than the clearcut system and was more
amenable to the minimal management system practiced by absentee owners of
large estates (Hill 1999). The cabruca system also gave them the flexibility to
hire workers and intensify plantation management when the cocoa prices were
high and to fire the workers when the prices fell. Under the sociopolitical and
legal conditions of the late twentieth century, with improved workers’ rights,
and since the outbreak of the witches’ broom fungus, which necessitates con-
tinuous efforts to keep the disease in check, these former advantages (from the
perspective of estate owners) of the cabruca system have to some extent been
lost.
In the late 1980s and 1990s, low cocoa prices coincided with the spread of
the witches’ broom disease in Bahia, encouraging some planters to (illegally)
sell timber from their residual forests that had been reserved for future planta-
tions and the cocoa agroforests themselves to compensate for low cocoa rev-
enues. Moreover, some cocoa agroforests were abandoned, but others were
converted into pasture (Alger and Chaldas 1992; Trevizan 1996; Johns 1999).
In a November 1996 survey, 30 farmers declared an average loss of 70 percent
of their labor force and a similar drop in production. All planters mentioned
tree felling and timber selling by neighbors, and several planters anticipated a
large-scale conversion of cocoa farms into other land uses, including pasture.
In 2003, however, the picture looks different. According to Brazilian
agronomists and private cocoa pod counters and forecasters, the rate of tree
felling and transformation of cabruca cocoa into pastures in the mid- to late
1990s, though higher than before, did not exceed 10 percent in the last 10
years (P. Petersen, pers. comm., 2003). Most of it occurred in drier parts of the
region where the cocoa trees had also been affected by an increased frequency
of droughts and inconsistency of rainfalls since 1982 (Carzola et al. 1995).
These climatic events appear to have increased the sensibility of the Bahian
cocoa planters to the ecological functions of shade trees, especially the reten-
tion of soil moisture and the microclimatic protection of the cocoa trees. In
the mid-1990s, a study highlighted the importance of ecological functions of
the shade canopy in the farmers’ perception: protection from the sun and con-
servation of soil moisture and soil fertility (Johns 1999). In the same period,
many farmers accused the extension service of having misled them in the
1980s by promoting shade removal in old farms and the establishment of new
farms with little or no shade and expressed their worries about the climate
change, especially more frequent drought, some giving drought the same
126 II. The Ecological Economics of Agroforestry