Fróðskaparrit - 01.01.1999, Qupperneq 183

HVUSSU DJÓRASLØG í FØROYSKUM FIRÐUM LAGA SEG EFTIR ALDUBROTUM
187
portant when analysing species response to
a single environmental factor, such as wave
exposure, that this factor has the strongest
influence on the data. The data matrix,
therefore, was tested using CCA. This pre-
liminary test showed that wave exposure
(FEV) was closely related to axis 1, which
explained 21.0% of the species variance.
Cumulative percentage variance of axes 1-
4 was 24.4%, indicating that the other
recorded variables included in the analysis
were of little importance. Therefore, FEV
was the strongest of the physical variables
known to influence species abundance.
Further examination of the unexplained
variance (in correspondence analysis) did
not suggest that other unknown factors
were of major importance. Species re-
sponse to wave exposure and biological ex-
posure values (grade) (BEV) for each sta-
tion (Table 2) were then calculated using
Expon (Arrestad and Lein, 1993).
The abundance of 15 dominant species
showed a significant correlation (polyno-
mial regression, P<0.05) with wave expo-
sure (BEV) and response polynomials were
obtained (Fig. 2). The abundance value for
each of the 15 species at different biologi-
cal exposure values (grade) (BEV), the lev-
el of significance, and the R2 value (coeffi-
cient of determination) are shown in Table
3. Porphyra umbilicalis, Polysiphonia
stricta, Fucus distichus ssp. anceps and
Cladophora rupestris were stable in their
response to wave exposure as shown by the
high R2 values. Ascophyllum nodosum and
Verrucaria mucosa have low R2 values
(Table 3). These species have a patchy oc-
currence and were missing at several sta-
tions where they were expected to occur.
Eight of the 23 dominant species examined
did not show an unequivocal response to
changes in wave exposure. These were not
included in the final Expon calculation, but
were later plotted against the biological ex-
posure value (BEV) for each station (Fig.
3). Laminaria digitata, Palmaria palmata,
and Patella vulgata had a more or less even
distribution throughout the exposure range,
whereas Fucus spiralis and F. evanescens
were only found on the moderately exposed
to sheltered coasts, and L. saccharina and
F. vesiculosus only on sheltered coasts.
Even though some of these species seemed
to respond to change in wave exposure, the
species response polynomials were not sig-
nificant probably because of a patchy dis-
tribution of these species and/or the influ-
ence of some physical factors or biological
interactions not accounted for in this study.
Abundance of species in relation to
height on the shore is shown in Fig. 4, A-I.
The MLWS and MHWS are indicated on
the diagrams for comparison. The identi-
fied species, with a frequency of occur-
rence of at least 33% in any one of the three
exposure groups, are listed in Table 4.
The species predominantly found on ex-
posed coasts were Aglaothamnion seposi-
tum, Alaria esculenta, Himanthalia elon-
gata, Polysiphonia stricta and Porphyra
umbilicalis. Most species found at these lo-
calities extended high above MHWS (Fig.
4, A-D). Fucus distichus ssp. anceps was
not recorded at any of the profiles given in
Fig. 4, but was found elsewhere at exposed
localities (Fig. 2) in association with P. um-
bilicalis. Lomentaria articulata occurred