- Formalized reproduction of an expert-based phytosociological
classification -
605
Results
Using the Cocktail method, we defined several spe-
cies groups, of which 12 were used in the formal defini-
tions of associations (Table 1). Some of these groups
have a rather narrow ecological range and they more or
less correspond to the groups of diagnostic species for
associations or alliances within the class Mulgedio-
Aconitetea (e.g. Aconitum plicatum Group, Betula
carpatica Group, Bupleurum vapincense Group, and
Laserpitium archangelica Group). Some other groups
have a wide ecological range, with an optimum beyond
the subalpine tall-forb vegetation, e.g. the Mercurialis
perennis Group in beech forests and the Trientalis
europaea Group in spruce forests.
Φ values for the
species of the groups in Table 1 are proportional to the
group quality, higher values indicating that the species
of the group have a stronger co-occurrence tendency.
Of 16 associations of the expert-based classifica-
tion (Kočí 2001; App. 1), 14 were reproduced by the
Cocktail definitions (Table 2). In contrast with the
original classification, association 3, Sileno-Calama-
grostietum was merged with association 2, Crepido-
Calamagrostietum, because the former was a species-
poor variant of the latter. Association 9, Piceo-Salicetum
was abandoned since it was a rare community known
only from two sites and mainly defined by its structure
rather than by floristic composition.
When the Cocktail definitions of associations were
applied to the original data set of 718 relevés used by
Kočí (2001), 376 relevés were assigned to the associa-
tions. Of these 376 relevés, 15 were simultaneously
assigned to two associations; these relevés were subse-
quently assigned to one of the two suggested associa-
tions by similarity calculations. Assignments by the
Cocktail definitions were in good agreement with the
original expert-based classification (Table 3), except
for the associations with few or poor diagnostic species
(4, Bistorto-Deschampsietum, 5, Violo-Deschamps-
ietum, and 12,
Trollio-Geranietum). Some relevés origi-
nally classified to association 8, Pado-Sorbetum were
assigned to the structurally and floristically similar
association 7, Salici-Betuletum by the Cocktail defini-
tions.
This good agreement was even improved, particu-
larly for associations 4 and 5, after similarity-based
assignment of the so far unclassified relevés to the
constancy columns consisting of the relevés that had
been assigned by the Cocktail definitions (Table 4).
Cocktail definition of the combined association 2-3
mainly captured the relevés of association 2, Crepido-
Calamagrostietum, which had been defined in the ex-
pert-based classification by the presence of several spe-
cialist species. Relevés of the expert-based association
3, Sileno-Calamagrostietum, which had a similar floristic
composition as association 2 but lacked the specialist
species, were satisfactorily assigned to the combined
association 2-3 only by similarity calculations.
Table 1. Species groups used for the Cocktail classification of subalpine tall-forb vegetation of the Czech Republic. Species of each
group are ranked by decreasing fidelity to the group, calculated as the
phi coefficient in a data set of 21 794 relevés of all vegetation
types of the Czech Republic. The range of
Φ values (multiplied by 100) between the first and the last species is given, higher values
indicating more distinct groups. The number of relevés in which the group occurs is the number of relevés containing at least half of
the group’s species.
Species group
Φ
Φ
Φ
Φ
Φ ×
×
×
×
×
100
No. of relevés
Species (ranked
by decreasing
Φ value)
Aconitum plicatum
64 - 38
47
Aconitum plicatum,
Viola biflora,
Epilobium alpestre, Carduus personata
Betula carpatica
70 - 36
14
Salix silesiaca,
Ribes petraeum, Betula carpatica, Prunus padus ssp.
borealis
Bupleurum vapincense
67 - 43
5
Thymus pulcherrimus ssp.
sudeticus, Pleurospermum austriacum,
Bupleurum longifolium
ssp. vapincense, Thesium alpinum
Cardamine amara
55 - 34
300
Cardamine amara, Chrysosplenium alternifolium, Chaerophyllum hirsutum,
Crepis paludosa,
Chrysosplenium oppositifolium
Geranium sylvaticum
65 - 35
215
Geranium sylvaticum,
Cardaminopsis halleri,
Phyteuma spicatum,
Silene dioica, Crepis
mollis, Cirsium heterophyllum
Laserpitium
67 - 28
6
Laserpitium archangelica,
Delphinium elatum, Campanula latifolia,
Stachys alpina,
archangelica
Scrophularia scopolii, Aconitum lycoctonum ssp.
lycoctonum
Ligusticum mutellina
63 - 57
25
Campanula barbata, Ligusticum mutellina, Viola lutea ssp.
sudetica, Avenula planiculmis
Mercurialis perennis
78 - 71
1688
Mercurialis perennis, Galeobdolon luteum s. lat.,
Galium odoratum, Dryopteris filix-mas
Petasites albus
62 - 45
312
Petasites albus, Stellaria nemorum, Cicerbita alpina, Thalictrum aquilegiifolium
Trientalis europaea
71 - 60
410
Trientalis europaea, Homogyne alpina, Calamagrostis villosa
Vaccinium myrtillus
79 - 55
1739
Vaccinium myrtillus, Avenella flexuosa, Polytrichum formosum, Dicranum scoparium
Veratrum lobelianum
57 - 32
108
Rumex arifolius,
Adenostyles alliariae, Athyrium distentifolium,
Veratrum album ssp
.
lobelianum,
Gentiana asclepiadea, Ranunculus platanifolius