A farm. survey based. on the method of the nitrogen apparent balance at the
farm scale quantified the nitrogen inputs and outputs due to the farm comm
ercial transactions. A total of 555 farms, partly situated in France and pa
rtly in north-west Italy, was analyzed. This method considers the farm. as
a "black box" and also takes the nitrogen fixation into account. However, n
itrogen input due to atmospheric deposition, or nitrogen losses due to vola
tilisation, denitrification or leaching are disregarded: this is the reason
why it is called "apparent balance". The study quantified the excess of ni
trogen and its most: important sources of variation in ii important, types
of farm production systems including stockless fanning, crop-less intensive
rearing, forage crop-livestock integrated farming systems, mixed farming w
ith livestock. The surplus of N was smaller in the stockless fanning system
(50 kg N ha(-1) y(-1), on average), intermediate in dairy cow or beef bree
ding farms (from 30 to 250 kg ha(-1)), and higher in intensive stocking sys
tems where forage was not produced in the farm (from 250 to 500 kg-ha(-1),
or even more). in general, where the apparent N surplus was small (< 200 kg
ha(-1)) the excess of N was mainly caused by the purchase of mineral ferti
lizers, while in farms with a high surplus (>200 kg h(-1)) the N surplus wa
s highly dependent on N input of purchased animal feedstuffs. Two groups of
farming systems were clearly identified: one was oriented to crop producti
on and was characterized by a favourable N efficiency or N conversion (72%)
; the other one was oriented to livestock production, with very variable N
fluxes, but little outputs, and was characterized by lower values (22-44%).
The very large variability in N surplus and in indicators (N efficiency, N
conversion and N losses) suggests that in many situations it should be pos
sible to reduce the environmental impact of a farm without drastic changes
in the production system.