bon is essentially zero. This would rank agroforestry
with other mitigation strategies, such as increased
building energy efficiency and vehicle efficiency, that
yield net benefits (NAS 1991). The profitability of
agroforestry is very important for its widespread adoption;
short-term cash profits are more attractive to
farmers than are long-term environmental benefits
(Hosier 1989).
It is risky to attempt to estimate the total amount of
carbon that could be stored by agroforestry because
current estimates of land available for conversion to
agroforestry are uncertain. The most realistic estimate
may be the 160 million ha in the tropics derived by
Trexler & Haugen (1991). They included economic,
social, and political factors that affect land availability
in addition to its biological suitability. Their analysis
distributed the total 160 million ha roughly equally
between the 3 tropical zones (tropical Africa, Asia and
America), but it did not distinguish between ecozones.
A very rough approximation of the potential range for
total carbon storage is possible by multiplying their
total estimate of available land by each of the carbon
storage estimates in Table l . The result is between ca
1.5 billion t C and 8.0 billion t C. Improved and refined
estimates of land availability are required to reduce
the uncertainties in these values.
Estimates of the total carbon conservation potential
of agroforestry resulting from reduced deforestation
are unavailable. Limited evidence currently available
indicates that the potential is real but impossible to
quantify. Additional efforts should be directed to
answering this question because reducing deforestation
might have a larger impact on the global carbon
cycle than direct carbon storage by agroforestry
practices.