Árbók Hins íslenzka fornleifafélags - 01.01.1983, Síða 64

68
ÁRBÓK FORNLEIFAFÉLAGSINS
SUMMARY
This article presents a pollen analytical study made in connection with archaeological
research at Herjólfsdalur in Heimaey, the biggest of the Westman Islands (off the south
coast of Iceland). The purpose of the study was to investigate the influence of the settle-
ment on the vegetation, and as the soil profiles in Heimaey may be dated by tephro-
chronological means, there is a possibility of relative dating. The dark part of the so-
called settlement layer (Vlla), most probably deposited about A.D. 900, is found in
Heimaey.
Two soil profiles have been analysed and pollen diagrams drawn (fig. 3, tables II and
III; for location see fig. 1). The profile named Herjólfsdalur situated some 100 m SSW
from the excavatcd area contained badly corroded pollen grains. Assuming that the cor-
rosion selectively acts upon various pollen types it makes all interpretation very difficult.
At the present time there are no peatlands in Heimaey, but south of Herjólfsdalur the
topographical name Torfmýri (Turf mire) still exists. Several sections at this site were
examined, most of them showed a discontinuous tephra stratigraphy (see Torfmýri B fig.
2). No peat was seen and the sediment was similar to the one at Herjólfsdalur. A profile
named Torfmýri A was analysed, however. The pollen preservation is somewhat better
than at Herjólfsdalur but still too bad for any definite conclusions about the former vege-
tation. The tephra layers K and tTl differ in chemical composition and a microscope
examination of tephra layers VHa and tT2 showed a marked difference. For these reasons,
bad pollen preservation and presumable hiatus in tephra stratigraphy, no attempt is made
to compare those two pollen diagrams.
The landnam phase in relative pollen diagrams from the lowland of Iceland is charac-
terized by an abrupt rise of the Gramineae curve and a fall of the Betula curve. There
are no indications of a Betula woodland in Heimaey, apparently the vegetation there al-
ways has been low and open (grassland communities). If that is true, one can’t expect
apophytes to react in the same way there as on the mainland. The appearance of pollen
grains from assumed synanthrops is discussed and so is every indication of cereal culti-
vation. All Gramineae pollen bigger than 40 /xm at Torfmýri were examined with Phase
contrast, most of them having Hordeum sculpturing type, thus belonging to Elymus
arenarius, Agropyron repens or Hordeum vulgare. The first maximum at L.P.A.Z. Tml
is considered to occur because of Elymus, as there is further indication of strand vege-
tation (Plantago maritima). The latter one at L.P.A.Z. Tm4 may be caused by some bar-
ley cultivation, as we don’t have any simultaneous rise in any strandplant curve.
In view of uncertainties in connection with the interpretation of the present data, it is
considered wise to refrain as yet from any definite conclusions based on the pollen ana-
lytical study.