Fróðskaparrit - 01.01.2004, Blaðsíða 223

HÆDDARÚTBREIÐSLA AV PLANTUSAMFELØGUM I FØROYUM
221
Based on the results of the cluster analyses,
the number of significantly different com-
munities was determined. A cut level was
found that gave the maximum number of
communities that were statistically inde-
pendent (p<0.01), using a t-test based on
scores from axis 1 and axis 2 in the De-
trended Correspondence Analysis (DCA)
(Table 7). Syntaxonomic sorting of tables
was performed to determine the vegetation
types using the methods of Braun-Blanquet
(1932); Mueller-Dombois and Ellenberg
(1974); and Westhoff and van der Maarel
(1978). The synoptic tables present taxa
with their constancy class in each commu-
nity, categorized in Roman numerals frorn I
toV: V = 81-100%; IV = 61-80%; 111=41-
60%; II = 21-40%; I = 1-20%. The abun-
dances are given as a percentage for each
species in the communities. Only taxa with
a constancy class > II in at least one com-
munity are shown in the Tables. The abun-
dance and the constancy class of species in
the meso-plots were used to determine the
communities. The indicator species used to
name the plant communities were species
with constancy class values between V and
III and high abundances in the community.
Variation between vegetation and the
environment was analysed using DCA.
Following an ordination with all samples
together, the main vegetation types were
analysed separately in order to see more
clearly the relationship between a speciftc
part of the mountain and its communities.
The altitudinal zonation of the plant
communities was determined by testing for
a significant difference (t-test) in altitude
between pairs of communities and then
combining those that were not significantly
different in altitude into zones.
We used the 25th percentile and the 75lh
percentile in a box plot to define lower and
upper boundaries for the zones, respective-
ly (Fig. 3). The boundary between the tem-
perate and the low alpine zone was found
to be at about 200 m a.s.l. The boundary
between the low alpine and the alpine zone
was found to be at about 400 m a.s.l.
Results
At the chosen cut level, the cluster analy-
sis resulted in twelve communities (Fig.
2), ten of which were significantly differ-
ent based on a t-test of the DCA scores.
The remaining two communities, Bistorta
vivipara-Festuca vivipara community and
Festuca vivipara-Agrostis capillaris com-
munity, are similar to previously described
communities (Bocher, 1937;Hansen, 1967;
Hobbs and Averis, 1991) and are, therefore,
kept as separate plant communities.
From these twelve communities, it was
possible to define three significantly dif-
ferent altitudinal zones (p< 0.01) (Fig. 3):
the temperate vegetation zone (up to 200
m a.s.l.); the low alpine vegetation zone
(200-400 m a.s.l.); and the alpine vegeta-
tion zone (above 400 m a.s.l.). The three
defíned altitudinal zones were found to be
consistent with the grouping ol' the twelve
communities into four main vegetation
types. Definitions of the four main vegeta-
tion types were based upon species of high
constancy shared among communities. In
this way, vegetation in the three altitudinal
zones was grouped into the four main veg-