Jökull - 01.12.2007, Blaðsíða 42

McPeek et al.
Figure 1. (Left) Satellite image of the 1783 AD Laki flow and the 934 AD Eldgjá flow. (Right) Satellite image
of the 1554 AD Hekla flow and the 1300 AD Hekla flow. Maps were adapted from Jóhannesson et al. (1990)
and Jóhannesson and Sæmundsson 1998. – Gervitunglamynd af Eystra gosbeltinu og Heklusvæðinu. Hraun frá
Skaftáreldum 1783, Heklu 1300 og 1554 og Eldgjá 934 eru teiknuð eftir jarðfræðikortunum frá 1990 og 1998.
1996). An estuarine sample, SRM1646a, was used
as a standard. The resulting three fractions are occ-
luded, mineral, and organic P. The occluded fraction
was analyzed using the Leeman Labs P950 Inducti-
vely Coupled Plasma-Atomic Emission Spectrometer
(ICP-AES). The mineral and organic fractions were
analyzed on the Shimadzu-UVPC spectrophotometer.
The resulting concentrations of P fractions were mul-
tiplied by the density of the soil and the depth of the
sample trench to obtain a total amount of each fraction
of P in the soil.
RESULTS AND DISCUSSION
At each sample site, the concentration of organic C
was determined as a function of soil depth. On the ol-
dest flow of 934 AD, this was a straightforward exer-
cise as it had a deep soil profile. Soil organic C was
the highest in the top soil, but was progressively lower
in deeper layers of soil (Figure 2). Our results seem to
be inconsistent with earlier Icelandic studies that de-
monstrated soil organic C to be lower in top than in
deeper soil where erosion is severe (Oskarsson et al.,
2004). The different patterns are presumably caused
by geographical variations, which can affect C content
and distribution in Icelandic Andosols (Oskarsson et
al., 2004). The pattern of soil organic C distribution
40 JÖKULL No. 57