Environmental Engineering Reference
In-Depth Information
when one includes the use of residues for fuel. Even in afl uent nations, data on
average burn fractions indicate much higher regional rates both for i eld and for
orchard crops (Jenkins, Turn, and Williams. 1992). Yevich and Logan (2003) esti-
mated that in the mid-1980s, 400 Mt of crop residues were burned in i elds, ranging
from just 1% in China to 70% in Indonesia. In deforested regions most of the
residues ended up as household fuel: as recently as the late 1980s roughly 75% of
China's crop residues, including more than two-thirds of all cereal straws, were
burned in cooking stoves (Smil 1993). The modernization of fuel supply has reduced
that demand throughout Asia, and the use of wheat straw as fuel is minimal in
Europe and nonexistent in North America. The demand for bedding straw has also
decreased with modern milk and beef production arrangements, but postharvest
straw burning in i elds remains common around the world.
I put the overall share of recycling during the late 1990s at no less than 60% of
the total residual phytomass in high-income countries and at 40% of the total in
low-income countries, resulting in the most likely global mean of just below 50%
(Smil 1999b). In addition, a signii cant share of residues used in ruminant feeding
and bedding is returned to i elds as recycled manure. This means that close to 2 Gt
of dry-matter phytomass in recycled residues and in undigested feed and used
bedding belongs to a special category. This phytomass either is not removed from
i elds or is rather promptly returned to them, and hence it should not be added to
the human appropriation total—or it should be counted only with qualii cations:
once it becomes available for other heterotrophs, its recycling serves to replenish
organic matter in agricultural soils and hence to enhance the production of
croplands, not of natural, unmanaged ecosystems. There is no obvious unequivocal
solution to this accounting problem.
Even greater challenges arise with the inclusion of phytomass consumed by
domesticated animals either grazing on pastures created by the conversion of natural
ecosystems or as they share natural grasslands with wild herbivores, a situation
common in Africa and parts of Asia and South America. A theoretical maximum
can be found easily by assuming that all annual production of all permanent
meadows and pastures (3.35 Gha, according to the FAO) is consumed by domesti-
cated herbivores to be converted to meat and milk: that was, indeed, the method
used by Vitousek et al. (1986) and by Rojstaczer, Sterling, and Moore (2001). Even
with a conservative productivity average of 5 t/ha, that would amount to nearly 17
Gt/year.
This is an obviously unrealistic assumption: even in those pastures that are
heavily overgrazed some phytomass remains, and in well-managed, fully stocked
