Fróðskaparrit - 01.01.1999, Page 181

HVUSSU DJÓRASLØG í FØROYSKUM FIRÐUM LAGA SEG EFTIR ALDUBROTUM
185
rina (L.) J.V. Lamour., Mastocarpus stella-
tus (Stackh. in With.) Guiry in Guiry et al.,
Palmaria palmata (L.) Kuntze, Pelvetia
canaliculata (L.) Decne. & Thur., Polysi-
phonia stricta (Dillwyn) Grev., Porphyra
umbilicalis (L.) J. Agardh; the lichen Ver-
rucaria mucosa Wahlenb.; and the inverte-
brates Littorina obtusata (L., 1758), Myti-
lus edulis L., 1758, Nucella lapillus (L.,
1758), Patella vulgata L., 1758, and Semi-
balanus balanoides (L., 1767). These
species were easy to recognise and quanti-
fy according to a defined abundance scale
(Table 1) and did not show bio-geographic
boundaries within the area.
The 146 stations were selected to cover
the main islands and the range of wave ex-
posure (Fig. 1A). Many stations were
reached from the sea by use of a Zodiac.
Areas with unstable boulders and stones, or
a tidal amplitude less than 0.40 m were not
included in the study. With small tidal am-
plitude, the atmospheric pressure has a re-
latively large effect on the water level,
thereby causing irregular exposure to air
and often prolonged desiccation. This may
have greater and more varied effects on the
structure of the biota of these shores than
wave action. For each station, a physical
exposure (PE) value was calculated by us-
ing the percentage of wind (W) stronger
than 15 m/sec and fetch to the nearest point
of land in each of 32 sectors. The fetch was
rated in three categories: local effect: 0.5 -
7.5 km (multiplier 1), fjord effect: 7.5 - 100
km (multiplier 10); ocean effect: >100 km
(multiplier 100) (Sjøtun et al., 1993). The
calculated values were transposed into a
scale from 1 (exposed) to 8 (sheltered) for
use as first exposure values (FEV) in Ex-
pon.
A station was defined as an area of the
coast with a length of 8 m and a vertical dis-
tribution from the lowest water level to the
upper limit of the 23 dominant species list-
ed above. The height of a station was di-
vided into equal intervals, each correspon-
ding to 1/10 of the mean tidal amplitude in
the area. The abundance of each species
was recorded at the interval where it had its
highest value, according to the simplified
method described and tested by Kruskopf
and Lein (1998). The abundance scales
used were originally defined by Crisp and
Southward (1958) and further modified by
Ballantine (1961), Dalby et al. (1978) and
Árrestad and Lein (1993) (Table 1). Also
recorded at each station were substrate,
slope, aspect of shore, tidal amplitude, tidal
current and topography using scales de-
fined in Bruntse et al. (in press). Rock
pools and backsides of boulders, etc. with
local shelter were avoided. The major part
of the fieldwork was carried out during July
- September 1995, May - November 1996
and May - October 1997.
Before running Expon, the data were sub-
jected to a correspondence analysis (CA)
and canonical correspondence analysis
(CCA) (Ter Braak, 1986) to assess the im-
portance of exposure (FEV) and the other
recorded environmental variables. Canoni-
cal correspondence analysis (CCA) is a
technique specifically designed to examine
relationships between species and environ-
mental factors (Ter Braak, 1986; Ter Braak
and Verdonschot, 1995). Following the