Fróðskaparrit - 31.12.2000, Qupperneq 109

TITTLEIKIN AV VANLIGASTU FØROYSKU DýRA- OG TARASLØGUNUM | 13
A KLETTASTROND I MUN TIL YMISK UMHVØRVISVIðURSKIFTI
Asco
Clad ru ■ ■ Li obt La dig Nucel* Verr * Corall ■
M asto • Patel * Pol str ■
Pelv • Seraiba ■ Palra" Agl sep ■ Porph • dis an ■
• F ves Myt ed • Alaria " H irael *
Fspir •
Fevan •
-1.0 +S.0
Fig. 3. Detrended Correspondence Analysis (DCA)for
rocky shore communities in the Faroe Islands: dia-
gram ofspecies and ordination axes 1 (horizontal) and
2 (vertical), explaining 52% ofthe inertia (= weighted
variance) in species abundances. The eigenvalues of
axes 1 and 2 are 0.40 and 0.07, respectively. The data
comprise 159 sites with tidal amplitude larger than 0.4
m. Species names are given infull in Table 1.
caused a similar species response as the en-
vironmental variables.
In the Detrended Correspondence
Analysisis (DCA), wherein the axes were
not constrained by the environmental vari-
ables, the first axis explained 45% of the
species variation and the second axis 7%
(eigenvalues 0.40 and 0.07). The sequence
of the species along the first axis (Fig. 3) re-
sembled their sequence along the first axes
in the CCA and in the partial CA plots (rs =
-0.99 and -0.71, respectively, p<0.01 for
both). This suggested the presence of one
strong trend in the data, which was reflect-
ed in the first axis in each of the three plots.
It also implied imperfect estimations of the
underlying factor(s) by the environmental
variables, or possibly, an unknown factor
causing a similar species pattern. The high
correlations between some of the environ-
mental variables and the first CCA axis as
well as the first DCA axis (Table 4) sug-
gested, however, that these variables were
in fact important to explain the observed
pattern. The DCA may provide the best
representation of the major trend in the data
since it does not depend on imperfect vari-
ables and since it seems to reflect the same
major trend as found by the CCA.
The species sequence along the second
DCA axis was correlated with the species
sequences along the second and the third
axes in the partial CA (rs = 0.43 and 0.63,
0.01<p<0.05 and p<0.01, respectively). It
was not significantly correlated with
species sequences along any of the CCA
axes. The site scores along the axis were
weakly correlated with the wave exposure
index (negatively, correlation coefficient
-0.28) and the substrate index (positively,
correlation coefficient 0.24) (Table 4).
Biotic fcictors
Biotic factors such as grazing, predation
and competition are known to play impor-
tant roles in intertidal communities. This
study has been designed to focus on extrin-
sic, physical factors. Other sampling de-
signs or experimental methods would have
been better suited to study the effects of the
biotic factors. However, to see to what ex-
tent the abundances of the predator Nucella
lapillus and the grazers Patella vulgata and
Littorina obtusata could explain the other