Jökull - 01.12.1990, Qupperneq 77

(dwarf-)shrub B. nana has been met in two ways.
Firstly by measuring the largest diameter of all unbro-
ken Betula pollen grains (cf. Usinger, 1978) and sec-
ondly by taking into account the morphological char-
acteristics of the Betula pollen grains (cf'. Berglund,
1966). The latter method was applied to samples 11,
12, 14, 16-18, 20 and 21. Pollen of Betula nana type
do not exceed 2,3 % of the terrestrial pollen sum in
these samples. In sample 6 the first pollen grain ap-
peared which without doubt can be referred to as of a
tree birch type.
The results of the size measurements are given in
absolute number of pollen per score (size class) in a
size frequency diagram in Fig. 5. The scores range
from 8 to 17 ocular lines corresponding to 16,8 jim-
35,7 /im. Size frequency curves were drawn for each
of the uppermost three samples separately, but below
sample 21 they were grouped in accordance with the
Local Pollen Assemblage Zones and Subzones.
Investigations on size distribution of recent Be-
tula pollen grains from Iceland have so far not been
carried out. Such reference material is needed for a
more comprehensive statistical analysis, cf. Prentice
(1981). Therefore the curves presented from Flat-
eyjardalur tell us little more than that there has been a
succession, from a vegetation assemblage producing
small and few pollen grains, probably B. nana, to an
assemblage producing both larger and more numer-
ous pollen grains, most likely an assemblage with B.
pubescens. Investigationsoutsidelceland have shown
that pollen of tree birch is larger than that of B. nana
(cf. Berglund and Digerfeldt 1970). ForBritishBetula
species Birks (1968) gives the size range 14,3-23,4
pm for B. nana and 18,2-28,6 pm for B. pubescens.
It is obvious that in the first group in Flateyjardalur
there is not the same population as in the uppermost
three samples. The distribution pattem of this group
is rather weak, probably due to the few grains present
in the lowest LPAZ. None of the following six curves
are showing any tendency of bimodal distribution, as
would be the case if both B. nana and B. tortuosa
were represented (Birks 1968). On the other hand
the size distribution of the next subzone, KHM-2a
(Fig. 5), shows a positively skewed distribution, in-
dicating an assemblage where B. nana is the most
8 9 10 11 12 13 14 15 16 ocular lines
h---------1------T---------1---------1-----
15 20 25 30 35 nm
Figure 5. Size frequency plots for Betula pollen
grains in the Krosshólsmýri sediment sequence. —
Stœrðardreifing Betula frjókorna (birkilfjalldrapi) í
Krosshólsmýrarsniðinu.
prominent species of the two in the vegetation, con-
trary to the three uppermost samples showing a neg-
atively skewed distribution, indicating an assemblage
of both B. nana and the tree birch type B. pubescens
with the latter type more prominent.
JÖKULL.No. 40, 1990 73