Fróðskaparrit - 31.12.2000, Side 117

TITTLEIKIN AV VANLIGASTU FØROYSKU DýRA- OG TARASLØGUNUM
A KLETTASTRONDI MUN TIL YMISK UMHVØRVISVIðURSKIFTI
121
tude larger than 0.4 m implied that at low
wave exposure levels another factor was of
importance. This was not reflected in the
CCA plot, and the axis was only weakly
correlated with any environmental variable
(negatively with the exposure index and
positively with the substrate index). The in-
terp'retation of this axis may only be specu-
lated. It may have reflected a different as-
pect of exposure, for instance a different
time aspect, than the first axis. From the
species plot, it seemed that the axis may
have reflected life-history pattems, with
temporally stable, long-lived species such
as Ascophyllum nodosum and Corallina of-
ficinalis at the top of the diagram, and tem-
porally more variable and potentially op-
portunistic species such as Fucus evane-
scens at the bottom of the diagram. This
might be connected to the particular histo-
ries of the sites, but needs further exploring.
Biotic factors
The results indicated that direct effects of
grazing by Littorina or Patella, or preda-
tion by Nucella could not explain any large
part of the variation in the ordination. How-
ever, as already pointed out, such interac-
tions may surely have occurred, as known
from experimental studies elsewhere (see
e.g. Hartnoll and Hawkins, 1985; Chap-
man, 1995). but their effects were not pro-
nounced with the scale and sampling
method used. Indirect effects of grazing
could not be ruled out as a possible expla-
nation for some of the variation. If so, these
effects caused a similar species pattern to
that caused by wave exposure. It was not
possible to say to what extent competition
between species influenced the observed
patterns. The degrees of distributional over-
laps along the gradient were indicated by
the distances between the species centroids
in the DCA plot (Fig. 3) as well as by com-
parisons of the plots of species abundances
versus site scores on the first DCA axis
(Fig. 4). For instance, the centroids of Fu-
cus distichus ssp. anceps and Ascophyllum
nodosum were distanced far apart in the
DCA plot, suggesting little distributional
overlap, which was further confirmed by
the plots of their abundances along the first
DCA axis, from which it seemed that the
two species did not occur together at all.
Such pattems might be induced both by ex-
trinsic factors such as wave exposure as
well as by interactions between the species,
or perhaps most likely, by a combination of
both. Experimental studies are needed to
discern the causative factors for the distrib-
utions.
Species abundance curves
Considering the discussion of the factors
influencing the first DCA axis for sites with
tidal amplitude larger than 0.4 m, it seems
reasonable to interpret the plots of species
abundance versus site scores on the axis as
the species responses to wave exposure.
The three groups identified were, thus, (1)
species that increase in abundance with in-
creasing exposure, (2) species with abun-
dance optimums at intermediate exposure
or with no clear response to exposure, and
(3) species that increase in abundance with
decreasing exposure.