Jökull - 01.12.1990, Side 81

pollen and macrofossil analysis the investigation is
supported by six radiocarbon dates (Hel-558 — Hel-
563, Vasari and Vasari 1990). Biostratigraphically
pollen zone c at Hafratjöm resembles that of KHM-
2 and KHM-3 at Krosshólsmýri. Pollen zone b at
Hafratjöm with very high frequencies of Juniperus
and Salix does not have a counterpart at the Kross-
hólsmýri sequence, and vice versa concerning the first
pollen zone at Krosshólsmýri, i.e. KHM-1. Inaccor-
dance with radiocarbon dates Vasari and Vasari (1990)
place zone c in the Chronozones Late Boreal, Early
Atlantic and Middle Atlantic. At the same time the
comparable biozones at Krosshólsmýri, i.e. KHM-2
and KHM-3, have a somewhat longer duration back-
wards in time, spanning a part of the Late Preboreal,
Boreal, Early Atlantic and the beginning of the Middle
Atlantic Chronozone.
At Hafratjöm the first Betula maximum is dated
to 8040 ± 150 B.P., which is consistent with KHM-
2b at Krosshólsmýri. Betula alba fruits appear in the
sediment for the first time somewhat later, yet still
ln the Early Atlantic Chronozone (Vasari and Vasari,
1990). Thus the very first Betula appearance at Hafra-
tjöm may be due to Betula nana, as evidenced by the
macrofossils (Vasari and Vasari 1990). Yet a long
distance transport can not be excluded, as at the time
°f the first Betula maximum at Hafratjöm the Betula
pollen frequencies are already passing 50 % in Skaga-
fjörður(the author’s unpublishedpollen diagram from
the lake Vatnskotsvatn, Hallsdóttir, 1979).
inferred climatic change in the
early atlantic chronozone
Probably due to delayed immigration of Betula
pubescens in westem North Iceland, the vegetational
succession of the westemmost locality Hafratjöm is
standing somewhat by itself, with Salix and Juniperus
being the main pollen producers at the beginning of the
Late Boreal Chronozone. At the other localities Ju-
niperus increases together with Betula, but often more
rapidly (Vatnskotsvatn and Ytri-Bægisá). Juniperus
Pollen were not yet identified at the time of analysis
°f the sequence at Moldhaugar, i.e. in the fifties (pers.
comm. Þorleifur Einarsson; cf. Einarsson, 1961 and
Þórarinsson 1955).
A successional pattern like this is well known from
outside Iceland, e.g. in the British Isles (Pennington,
1974; Bennett, et al., 1990), where it characterizes the
change from tundra to forested landscape. As the pi-
oneer tree, the birch, immigrates, it soon shades out
the light demanding juniper. This has been shown in
Southwest Iceland (Vasari, 1972; 1973, Hallsdóttir,
1987) as well as in North Iceland. The latter re-
gion shows some peculiarity, as just before the disap-
pearance of juniper something happens, which makes
juniper recover its former place as the formerly ex-
panding birch woodland contracts (LPAZ KHM-2c).
As far as can be inferred from the present material,
the unpublished pollen diagram from Vatnskotsvatn
(Hallsdóttir, 1979) supported by six radiocarbon dat-
ings (Hakansson, 1980), and the already published
data(i.e. Moldhaugar: Einarsson, 1961; Ytri-Bægisá:
Bartley, 1973; Hafratjörn: Vasari and Vasari, 1990),
these changes seem to be synchronous at all the sites
concemed, falling within the Early Atlantic Chrono-
zone.
The reason for this must be looked for amongst
the climatic parameters affecting the water regime of
the soils occupied by birch trees and maybe the pro-
tecting snow cover in winter time. Badly drained and
water soaked soils are as a habitat not suitable for
birch woodland formation, but favourable for Salix
and in certain cases for juniper too. An increase in
humidity, increased precipitationand orlowertemper-
atures, resulting in more wetness in the environment,
are well known factors contributing to expansion of
mires, which is consistent with the rising Cyperaceae
curve at Krosshólsmýri. Also evidence for a rising
water level in the former lake Krosshólsmýri can be
inferred from the sudden expansion of Isoetes echi-
nospora on the lake bottom of that time. The climatic
change responsible for these changes in environmental
factors may possibly be ascribed to changes in ocean
currents at the north coast. Perhaps LPAZ KHM-2c
denotes a period, where a branch of the Gulf Stream
became stronger and reached further east along the
north coast. This would result in more precipitation
at the north coast, as well as higher temperatures and
less snow during the winter. Wet and unstable winters
JÖKULL, No. 40, 1990 77