Árbók Hins íslenzka fornleifafélags - 01.01.1982, Qupperneq 54

58
ÁRBÓK FORNLEIFAFÉLAGSINS
slopes, provides an excellent natural laboratory within which to examine some of this evidence
together with the factors which have led to the present landscape variety.
The three long-abandoned, sites of Þuríðarstaöir, Þuriðarstaðir efri and Steinfinnsstaðir are
geomorphologically very similar. All three are severely eroded, with no trace remaining of the
former vegetation and soil cover, and it is obvious that the topography at the time the sites were
occupied was markedly different from that now in existence. In the vicinity of the Þuriðarstaðir
sites only the occasional patch of vegetation remains, and these are usually on the west slopes of
deep, steep-sided gullies. These vegetated slopes are now evolving through a combination of soil
creep and shallow earth slumping. The process of slumping appears to reflect the influence of the
grass cover on infiltration rates, allowing water to seep into the surface materials and not run
directly across the surface as happens on the bare slopes. Once the water has infiltrated it is
directed along the various loessic and tephra layers. The varying hydraulic conductivity of these
layers creates perched water tables and favourable conditions for shallow mass movements. Many
former earth slumps have re-vegetated, indicating that erosion is not necessarily progressive and
self-perpetuating. But the vegetated areas are the exception and the slopes are dominated by bare
tephra and loessic material.
Erosion is taking place in a variety of ways. Wind erosion is the most widespread and picks up
any particles loosened by frost, dessication and trampling by sheep. Iceland is a little unfortunate
in that its surface materials are amongst the most potentially erodible once the vegetation cover
has been destroyed. Ashwell (1963) has described, in general terms, the relationships between
wind directions and erosion, but in Þórsmörk, due to the varying influence of the nearby ice caps,
winds from many directions occur and move material locally from site to site. Some of this
material is deposited at the boundary of the vegetated and unvegetated areas but much is deposit-
ed in the numerous gullies. Some material is carried higher into the atmosphere and the dominant
direction of movement appears to be west to east, following the Markarfljót to the coast. Iron
pans and cemented tephra horizons temporarily halt wind erosion but these are soon broken up.
All three sites are characterised by a series of major gullies, some up to 6m deep, with numerous
smaller sub-parallel rills. During the summer months the gullies are generally dry but water see-
page is common from prominent impermeable tephra horizons. Water flow within the deposits is
high and a complex network of pipes, some up to 25 cm in diameter, occurs. The small rills
appear to be developing by the collapse of the larger pipes, much of the collapse being initiated by
sheep breaking through the surface. Piping, created by the eluviation of fine particles, is a wide-
spread phenomenon around the world and is often quoted as a likely cause of the rapid extensi-
ion of gully networks. Eluviation, resulting in the formation of pipes, seems to occur in soils
possessing initial weaknesses, caused by low bulk density, an unusual particle size distribution or
a structure which has been altered by chemical effects (Gilman and Newson, 1980). This may
apply to the Icelandic soils. Cracks in the soil caused by dessication may also lead to the initiation
of piping.
The larger gullies in Þórsmörk become infilled with windblown material during the summer
months. This material is eventually flushed out of the gully systems onto the valley sandur of the
Markarfljót to mix with the glaciofluvial sediments that have been so well described by
Haraldsson (1981).
Rofbarðs and gully exposures near all three sites give some indication of the processes of
erosion and deposition that have occurred in the historic period. A rofbarð near Steinfinnsstaðir
shows 12 cm of modern soil, with degraded birch and crowberry, resting on a 3 cm layer of ash
from the Hekla 1947 eruption. Three layers of loess, 19, 23 and 41 cm thick respectively, separa-
ted by two black ashes, one of which could be the 1918 eruption of Katla, rest on the very distinct
pale ash from the 1821 eruption of the Eyjafjallajökull volcano. A similar sequence, with slightly