MAP Technical Reports Series No. 106 UNEP

- 63 -
estimates are about half of the UNEP figures. Besides, Béthoux budget estimates remain partial
in any case neglecting retention by sedimentation of both, nitrogen and phosphorus, and nitrogen
losses by denitrification. However, it might be that nitrogen retention and denitrification may
approximately be balanced by the aeolian contribution. If our figures are correct as to the order
of magnitude, aeolian nitrogen load could make out some 15 to 25% of the total.
Table 18
Areal Nitrogen and Phosphorus Load in g/m 2 .year;
Selected Examples*
Nitrogen Phosphorus **
Baltic proper (210,000 km 2 ): 4.3 0.3 (0.1-0.2)
North Sea (500,000 km 2 ):
Atlantic exchagne load +
atmospheric contribution
River + direct discharges
Inner Oslo Fjord (200 km 2 ): 17.0 3.0
Laholm Bay (300 km 2 ): 24.0 1.4
Japan:
Tokyo Bay (1,400 km 2 )
Ise Bay (170 km 2 )
Seto Sea (22,000 km 2 )
Adriatic Sea: Emilia-Romagna
Coast (2,000 km 2 )
(ortho-phosphate P)
4.2
1.9
89.2
43.4
8.2
0.9
0.2
10.7 (0.5-1.0)
5.8 (0.5-1.0)
0.8 (0.15-0.3)
70.0 7.8 (0.4-0.8)
3.6
Mediterranean Sea
(2,505,000 km 2 )
land-based + aolian input 1.5$0.5 0.15$0.05 (#0.1)
* Data elaborated from GESAMP, 1988
** Values in parentheses indicate loading tolerance range that would apply to lakes
of corresponding morphometry and hydrology
A similar equalisation cannot be postulated for phosphorus as long as the aeolian
contribution remains unknown. Also, it does not seem that as a whole the discrepancies could
be explained by concentration increases due to increased nutrient discharges from land-based
sources, as Mediterranean deep water concentrations remained practically constant over the
last two decades (cf. data reported by Coste, 1969; McGill, 1969; Krom et al., 1991). It is more
likely that a) the UNEP phosphorus load estimates have been overestimated, a conclusion that
derives from an unacceptably low N/P average ratio (6.4 by atoms = 2.9 by weight), which is
in contrast to our data; b) the return flow balance of 77,500 t/y is an underestimate. Taking our
upper load estimate from land-based sources of 250,000 t/y, then