Jökull - 01.12.1990, Blaðsíða 71

During the Early Preboreal Chronozone mountain
and valley glaciers in the tributary valleys advanced
and reached out to the main valley of Flateyjardalur
and left there terminal moraines and fluvioglacial de-
posits (Norðdahl, 1990; 1991). One of these is the
moraine in front of the twin tributary valley Brettings-
staðadalir, which belongs to the Langhóll glacier ad-
vance (Norðdahl, 1979; 1991). In the basin that
formed behind these terminal landforms (Fig. 2) or-
ganic sediments accumulated, at first diatom-rich gyt-
tja, but later on peat.
The present paper deals with a pollen analytical
study on the lower part of these organic strata, thus
representing a continuation of the environmental his-
tory described by Norðdahl (1979; 1983; 1990; 1991).
His contribution includes investigations on the Weich-
selian glaciation in central North Iceland, terminating
at the Langhóll-Ljósavatn glacier advance in early Pre-
boreal time in the Fnjóskadalur Sequence, the strati-
graphical column for central North Iceland (Norðdahl,
1983; 1990). The present study is dealing with the dis-
continuous vegetation cover some 9650 B.P. up to the
first signs of a woodland formation 7000 B.P.
field work and methods
Coring in Krosshólsmýri, a former small lake,
but now overgrown by mire vegetation, was carried
°ut by Hreggviður Norðdahl and Guðmundur Hjalta-
son in summer 1977. The geographical location is
66o05'55"N and 17°54'20/,W, altitude approximately
25 m a.s.l. (Fig. 2). The coring was done with a peat
sampler of a Russian type of 5 cm diameter and 50 cm
length (c/. Jowsey, 1966).
The sediment description and subsampling for
pollen and radiocarbon analysis was done in the lab-
oratory and only for that part of the sequence which
was taken for analysis. The characterization system
°f Troels-Smith (1955) was used.
Another sediment sequence from an earlier coring
at almost the same locality was described in the field
and subsequently published in Norðdahl 1979 (p. 10,
site 2). The lowest part of that sequence was ignited.
The loss on ignition was about 3% in the deepest sam-
Ple rising up to 20-26% 50 cm higher up in the core
(Norðdahl, pers. comm.).
POLLEN ANALYSIS
Quantitative pollen samples were processed by
standard methods, including warm HF treatment (e.g.
altemative A in Berglund and Ralska-Jasiewiczowa,
1986). Stockmarr’s Lycopodium spore tablets were
used to determine the pollen concentration (Stock-
marr, 1971). The dosage was one tablet (12500±100
spores) per cm3 of fresh sediment. Whole slides were
counted, to avoid the effects of any nonrandom dis-
tribution of pollen and spores on the slide. Two to
four slides had to be counted to get reliable number of
pollen.
The total number of terrestrial pollen grains
counted and included in JjA ranges from 160 to
500 per sample. The pollen-percentage diagram is
constructed using ^TA for the main part but ^ ,4 ±
B ... K is used for the taxa excluded from /jA, i.e.
spores of pteridiophytes [B] and Sphagnum [E], pollen
of aquatic plants [C], telmatophytes [D], unidentified
and exotic plants [G] (i.e. trees, which have not lived
in Iceland since the Tertiary and or Lower Quater-
nary Period). Counts of the algae Pediastrum [K] are
given in the same way. The PC program POLLDATA
by H. J. B. Birks (1989) of the Botanical Institute in
Bergen was used for construction of the diagrams.
I4C DATINGS
Five cm of the diatom-gyttja were taken from the
core at two levels, 555-560 and 522,5-527,5 cm for
radiocarbon dating. These two samples were submit-
ted to the Radiocarbon Dating Laboratory at the De-
partment of Quatemary Geology in Lund in Sweden
(Norðdahl, 1979). One of the samples was treated with
HCl. Both had to be diluted with CO2 from anthracite
to allow dating. The <513C value of the samples was
measured and a normalization due to isotopic fraction-
ation performed. The calculation of the radiocarbon
age is based on a half-life of 5568 years for 14C. The
results are given as the number of years before 1950
and are not corrected for the secular 14C/12C varia-
tions. The dates are published in the official dating
list of the laboratory (Hakansson, 1979; 1983).
When comparing the present results with datings
from the period before 1970, some precaution must be
JÖKULL,No. 40, 1990 69