Agriculture Reference
In-Depth Information
uncertainty. Lack of consistency in the system
boundary used in various LCAs of livestock prod-
ucts makes it difficult to compare across studies
and makes interpretation of results uncertain.
As the largest proportion of the GHG emis-
sions associated with producing meat and milk
occurs up to the product leaving the farm, most
LCAs of livestock products use a system bound-
ary that ends at the farm gate, i.e. 'cradle to farm-
gate'. For example, a LCA for lamb produced in
New Zealand and consumed in the UK reported
that 80% of the GHG emissions were from the
cradle to farm-gate stage with only 20% from
manufacturing, retailing and transportation
activities beyond the farm gate (Ledgard
et al
.,
2011). Similarly, in a study of global dairy pro-
duction, it was estimated that cradle to farm-gate
emissions contribute on average 93% of total
GHG emissions (Gerber
et al
., 2010; 73-83% in
developed countries). Because the greatest pro-
portion of the GHG emissions is associated with
the cradle to farm-gate stage, it can be advanta-
geous to produce livestock in areas with a natu-
ral resource advantage and then transport the
final product to markets in less advantaged areas.
In contrast, grains for human consumption tend
to be more efficiently produced close to the
intended market because transportation and
activities beyond the farm gate represent a much
greater proportion of the total C-footprint.
wherein the environmental burden is assigned
to the co-product based on a parallel production
system (e.g. for a dairy-beef system, the burden
would be assigned to the meat based on a paral-
lel traditional beef production system, which
assumes that meat from dairy production dis-
places an equal amount of beef production);
(ii) allocation on the basis of a physical relation-
ship between the products (sometimes also
called biological allocation); and (iii) allocation
on the basis of economic value. There is consid-
erable disagreement amongst experts on the best
way of handling allocation, with valid argu-
ments in favour of and against each method.
System expansion is often considered the
best approach (Weidema, 2001; Cederberg and
Stadig, 2003; Weidema and Schmidt, 2010),
but this implies that there is an alternative
means of producing the product and that the
environmental burden of that production sys-
tem is known. System expansion generally
results in the lowest GHG intensity of milk
production, compared with other methods of
allocation (Flysjö
et al
., 2011a). However, sys-
tem expansion can be difficult to apply because
LCA for substitute products may not exist or an
alternative production system may not exist.
Thus, in most LCAs of livestock production, both
economic and physical allocation has been
applied. With economic allocation, the environ-
mental burden is allocated to co-products based
on their respective price, usually averaged over
some period of time (e.g. Kristensen
et al
., 2011;
McGeough
et al
., 2012). A physical allocation
can be applied by partitioning the feed energy
consumed to the requirements for milk and meat
production. For example, a LCA of Swedish milk
production used physical allocation based on
energy partitioning, which resulted in an 85%
allocation to milk and 15% to meat (Cederberg
and Stadig, 2003). The International Dairy
Federation (IDF, 2010) also recommends a phys-
ical allocation of environmental burden to milk
and meat, as follows:
Allocation factor (AF) = 1 − 5.7717 × R,
where R = Mmeat/Mmilk, and Mmeat is the sum
of live weight of all animals sold (including calves
and culled cows) and Mmilk is the sum of milk
sold (corrected for true protein and fat content).
A default value of 0.025 kg meat kg
−1
milk to
yield an allocation of 14.4% to meat and 85.6%
Co-product Allocation
Livestock production systems can, in some
cases, generate more than one product, which
raises the issue of how environmental burdens
should be allocated across the co-products. For
example, dairying produces milk as the primary
product, but substantial quantities of meat are
also produced from cull animals and surplus
calves. To deal with this complex issue, the ISO
(2006) standard recommends that whenever
possible, allocation of environmental burden be
avoided meaning that environmental impact
should be assessed for each co-product sepa-
rately. That is rarely possible with livestock
production systems because the production of
the co-products is mutually dependent. In that
case, several options are available for dealing
with allocation of the environmental burden to
co-products. These include: (i) system expansion,
