The Botany of Iceland - 01.12.1945, Side 64

406
STEINDÓR STEINDÓRSSON
immediate neighbourhood of the valley, whence blown sand is constant-
ly carried across the valley; in this way the mýri is gradually filled up,
but at the same time the vegetation begins to change: the driest por-
tions are transformed into jaðar, which gradually passes into heath,
but at the same time a jaðar vegetation is found in the wettest parts.
This is the stage at which the vegetation of the valley is at present.
When certain parts have become so dry that the heath formation has
developed, the wind erosion begins in the driest places unless inhibiting
factors intervene. Thus the soil of the heath may be blown away and
the area be converted into a fell-field.
4. Calamagrostis neglecta-Ass. (Tab. VII. A-B, 7-8).
Calamagrostis is the dominant species in this association, though
Juncus balticus, also, is present in some quantity (cf. analysis VII, 7).
Of other conspicuous species mention should be made of Sagina nodosa.
The H percentage and the E percentage are fairly high. Analysis 7
is from Arnardalur, where the association is found in some few places,
while analysis 8 is from Landmannahellir, where the association is
distributed and is the commonest jaðar association. The surface is most-
ly level, and the physiognomy of the vegetation is very similar to that
of the Valllendi vegetation, but the soil is so damp that no actual
Valllendi vegetation can thrive there.
5. Calamagrostis neglecta-Ass., rich in Salix.
(Tab. VII. A-B, 9-11).
Calamagrostis neglecta is the dominant species in the association, but
in addition Salix herbacea is, as a rule, dominant also. Other species
abundantly represented are Equisetum arvense, E. variegatum, Carex
rigida, and Salix glauca. Analysis VII, 10 differs from the others in that
the latter species is very conspicuous, whereas S. herbacea is very sparse.
Analyses 1 o and 11 were taken in the immediate vicinity of each other,
while analysis 11 is from a somewhat damper locality, which may be
the cause of the difference. Analysis 9 was made at Landmannahellir,
analyses 10-11 at Kýlingar, where the association is fairly dense. Here
it forms a belt between the mýri series on the one hand and the shrub
heath or bare sand on the other; in the latter case it may be regarded
as a stage in the covering of the sand by vegetation. The surface is