Content uploaded by Malle Leht
Author content
All content in this area was uploaded by Malle Leht
Content may be subject to copyright.
Ann. Bot. Fennici 41: 53–61 ISSN 0003-3847
Helsinki 27 February 2004 © Finnish Zoological and Botanical Publishing Board 2004
Variation of Potentilla sect. Potentilla (Rosaceae) in
Estonia and neighbouring countries
Malle Leht1 & Jaanus Paal2
1) Institute of Zoology and Botany, Estonian Agricultural University, Riia Str. 181, Tartu 51014,
Estonia (e-mail: malle@zbi.ee)
2) Institute of Botany and Ecology, University of Tartu, Lai Str. 40, Tartu 51005, Estonia (e-mail:
jpaal@ut.ee)
Received 8 Jan. 2003, revised version received 16 May 2003, accepted 8 June 2003
Leht, M. & Paal, J. 2004: Variation of Potentilla sect. Potentilla (Rosaceae) in Estonia and neigh-
bouring countries. — Ann. Bot. Fennici 41: 53–61.
One-hundred and eighty specimens of Potentilla erecta, 58 of P. reptans, 49 of P.
anglica and 19 of P. ¥ italica were studied phenetically using 19 macromorphological
characters. As P. erecta and P. reptans are quite common in Estonia, while P. anglica is
absent from that country, the occurrence of P. ¥ italica in Estonia points to its possible
origin from hybridization of the former two species. Estonian P. ¥ italica produces no
seeds and is hexaploid (2n = 42). Potentilla ¥ italica and P. anglica appeared, from
the statistical point of view, morphologically well separable from each other as well
as from their putative parents. In comparison with the varieties of P. erecta, var. erecta
and var. strictissima, which do not cluster into separate groups, it is reasonable to treat
P. ¥ italica, which clearly forms clusters of its own, on the same level as P. reptans and
P. anglica, i.e., to recognize it as a morphologically stable nothospecies.
Keywords: Potentilla anglica, Potentilla erecta, Potentilla ¥ italica, Potentilla
reptans, phenetics, taxonomy
Introduction
The group Tormentillae (sect. Potentilla), as
described by Wolf (1908), comprises eight spe-
cies and some hybrid taxa, of which Potentilla
reptans, P. erecta, P. anglica, and one of their
hybrids, P. ¥ italica are found in the Baltic
states. These three species are related so that P.
anglica is an allo-octoploid (2n = 56) originat-
ing from the hybridization of P. erecta (2n =
28) and P. reptans (2n = 28) (Matfield & Ellis
1972, Ietswaart & Kliphuis 1985). Hybrids of P.
anglica ¥ P. erecta (P. ¥ suberecta), P. erecta ¥ P.
reptans (P. ¥ italica) and P. anglica ¥ P. reptans
(P. ¥ mixta) have been found in the nature (Ball
et al. 1968, Ietswaart & Kliphuis 1985), and they
have also been produced experimentally (Czapik
1968, Matfield et al. 1970, Matfield 1972, Mat-
field & Ellis 1972).
Experimental hybridizations have shown that
hybrids very similar to Potentilla ¥ mixta can be
produced in two ways: either by occasional suc-
cessful pollination of P. reptans by P. anglica,
or by hybridization between autopolyploid P.
54 Leht & Paal • ANN. BOT. FENNICI Vol. 41
reptans and tetraploid P. erecta. Thus, it is pos-
sible that natural hexaploid hybrids include
plants whose origin is different, but which are
morphologically and cytologically indistinguish-
able (Matfield et al. 1970, Czapik 1975, Matfield
& Walters 1975). In floral lists even the name P.
¥ mixta s. lato with an explanation (P. erecta or
P. anglica ¥ P. reptans) is used (Staffordshire
Biodiversity Action Plan 2001, Flora of Northern
Ireland 2000–2001). We refer to P. erecta ¥ P.
reptans as P. ¥ italica.
A typical Potentilla ¥ mixta has 42 chromo-
somes as does also P. ¥ italica; they both are
often sterile but reproduce abundantly by run-
ners. In their genetic constitution, four genomes
of P. reptans and two genomes of P. erecta are
involved; therefore, in many characters of P. ¥
mixta s. lato, domination of P. reptans is recog-
nizable (Matfield et al. 1970, Czapik 1975). An
extremely poor seed set combined with highly
effective vegetative reproduction classifies P.
¥ mixta and P. ¥ italica as vegetative apomicts
(Czapik 1975).
Potentilla erecta is very common in Estonia
and quite common in Latvia and Lithuania on
moderately moist to moist mineral soils and
on peat. Potentilla reptans thrives in meadows
and wooded meadows, on grasslands, sea-
shores, roadsides and in other open habitats in
the western regions of Estonia as well as in all
parts of Latvia and Lithuania (Leht et al. 1996).
Potentilla anglica is absent from Estonia and
Lithuania and is known only from one locality
in Latvia near Riga on a sandy pine forest edge,
where it was refound in 1988 after a one-hun-
dred-year interval (Leht 1989). In 1997 several
specimens among Estonian herbarium material,
dated from the 1930s and 1940s, were identified
as P. ¥ italica. They all originated from neigh-
bouring localities on the shores of Lake Võrts-
järv. However, in 1997, P. ¥ italica plants were
found in only one of these localities, while the
other sites were overgrown with shrubs.
Phenotypically, Potentilla erecta is a vari-
able taxon, which has been dealt with in various
ways: Hegi (1923) cited 19 taxa of different
ranks and various taxonomic significance that
can be joined under the name P. erecta. Its vari-
ation in Estonia was studied by Leht and Paal
(1998b) and will not be discussed in detail in the
current paper.
Morphological variation is said to be com-
paratively limited in Potentilla reptans, moder-
ate in P. anglica, and fairly large in the hybrids
Table 1. Morphological characters used in analysis. Characters 7–19 were measured in triplicate and average
values used for calculations. Characters 7–13, 16, and 17 were measured on successive cauline leaves in the
central part of the shoot.
No. Denotation Character
01 NOD Number of nodes under the first branch
02 BRCH Number of branches on the shoot
03 HU Hairiness of the upper side of the leaflet (1 = glabrous, 2 = sparsely hairy, 3 = densely hairy)
04 HL Hairiness of the lower side of the leaflet (1 = glabrous, 2 = sparsely hairy, 3 = densely hairy)
05 RL Rosette leaves (1 = over 5, 2 = up to 5, 3 = absent)
06 FLWS Number of flowers
07 LFL Length of the central leaflet (mm)
08 LFW Width of the central leaflet (mm)
09 TEETH Number of teeth of the central leaflet
10 STPL Length of the stipule (mm)
11 STPW Width of the stipule (mm)
12 TOL Length of the central tooth of the central leaflet (mm)
13 TOW Width of the central tooth of the central leaflet (mm)
14 SEPL Length of the sepal (mm)
15 SEPW Width of the sepal (mm)
16 LFN Leaflet number
17 PETIOL Length of the petiole (mm)
18 LPET Length of the petal (mm)
19 WPET Width of the petal (mm)
ANN. BOT. FENNICI Vol. 41 • Variation of Potentilla sect. Potentilla 55
of these species (Ietswaart & Kliphuis 1985).
During the cultivation of P. ¥ mixta, Czapik
(1968) noted that it showed remarkable con-
stancy only in respect of its sterility, while the
other characters varied largely, which is prob-
ably one of the reasons why P. ¥ italica and P. ¥
mixta are often overlooked in the nature.
The questions that we address in our study
are:
1. How variable phenetically are P. reptans, P.
anglica, and P. × italica?
2. What is the chromosome number of Estonian
P. × italica?
3. Is P. × italica a morphologically stable taxon?
Material and methods
A total of 180 specimens of Potentilla erecta,
58 of P. reptans, 49 of P. anglica and 19 of P.
¥ italica were studied phenetically using 19
macromorphological characters (Table 1). The
Estonian material of P. erecta and P. reptans
analysed was mostly collected in 1988 and 1996;
herbarium specimens from the Herbarium of the
Institute of Zoology and Botany (TAA) were
also used. The Finnish and Swedish material of
P. anglica originates from the Herbarium of the
Botanical Museum of the Finnish Museum of
Natural History (H). The Estonian material of
P. ¥ italica was collected in 1997; herbarium
specimens from TAA were used as well, and four
specimens were obtained from Prof. R. Czapik
from Kraków (KRA). The material collected
from Estonia is preserved in TAA.
The characters (Table 1) were measured on
air dried herbarium material using a binocular
microscope MBS-2 or a ruler. To reduce the
effects of individual variability, characters 7–15,
18 and 19 were measured in triplicate, and the
corresponding average values were used in fur-
ther calculations. Characters 7–13 were meas-
ured on successive cauline leaves in the central
part of the stem.
Among the material of Potentilla erecta,
using Richards’s (1973) descriptions, 101 speci-
mens were identified as P. erecta var. strictissima
and 45 as P. erecta var. erecta; 34 specimens
appeared intermediate (Leht & Paal 1998b). In
the current study, two of Richards’s characters
used for identification, division depth of the
stipule and length of the dentate part of the leaf-
let, were not included, as they do not vary in P.
reptans, P. ¥ italica and P. anglica.
Chromosomes
Chromosomes were counted from root tips taken
from young shoots of Potentilla ¥ italica which
started to root in water. Root tips were pre-
treated with 8-hydroxyquinoline for three hours
for accumulation of metaphases, then hydrolysed
and stained with concentrated HCl and 1% aceto-
orcein (1:9) for two hours or longer at room tem-
perature, and squashed in 45% acetic acid.
The material for chromosome counts was
collected in 2001.
Data processing
Clustering of the standardized data was carried
out with the program package SYN-TAX 2000
(Podani 2001). The method of the minimal
incremental sum of squares, or Ward’s method
(Podani 2000), with the Manhattan-metric as the
proximity measure was employed.
For evaluation of the clusters’ distinctness,
the probability of the a-criterion (Duda & Hart
1976), termed the coefficient of indistinctness
(CI; Paal 1987), was used as in Leht and Paal
(1998a, 1998c). In addition, for elucidating
mutual relationships between the clusters
and the main directions of their variation in
the multivariate character space, the adja-
cency matrix was calculated. Adjacency was
expressed as the percentage of specimens in
the considered cluster for which the centroid
of the cluster to be compared is the closest
in the character space (Paal & Kolodyazhnyi
1983).
Discriminant analysis of the STATISTICA
package (StaSoft Inc.) was used to determine the
set of variables yielding the best discrimination
between the species and the established clusters.
Canonical discriminant analysis was carried out
to produce a scatterplot for the first two discrimi-
nant functions (canonical roots).
56 Leht & Paal • ANN. BOT. FENNICI Vol. 41
Results
Morphological variation
All four conventionally established species-clus-
ters are mutually highly distinct (CI = 0.0) from
the statistical point of view. The within-group
variation of Potentilla erecta and P. reptans
specimens is directed towards P. anglica and P. ¥
italica clusters, while for 79.6% of the P. anglica
specimens, the centroid of the P. erecta cluster
appears to be the closest (Table 2). Of the P. ¥
italica specimens, 52.6% show certain affinity
with the cluster of P. reptans; to a lesser extent,
their variation is directed also towards the cen-
troids of P. erecta and P. anglica.
On the scatterplot of the canonical roots, where
grouping was established according to the spe-
cies-clusters (Fig. 1), the specimens of Potentilla
erecta form a well-separated and compact group.
The clusters of P. anglica and P. reptans are more
diffuse and partly overlapping with the cluster of
P. ¥ italica located between them. Due to reduced
dimensions, the scatterplot does not reflect the
relationships between the clusters as explicitly as
does Table 1. Yet the mutually directed variation
of the clusters of P. erecta and P. anglica is quite
obvious, while variation of P. reptans specimens is
directed mainly towards the P. ¥ italica cluster.
The dendrogram showing the results of clus-
ter analysis (Fig. 2) is clearly split into two clus-
ters on the highest level, the first (I1) comprising
Potentilla erecta and the second (I2) consisting
of specimens of P. reptans, P. anglica, and P. ¥
italica. Both clusters are further divided again
into two large clusters. In clusters II1 and II2,
specimens of P. erecta var. erecta and P. erecta
var. strictissima stand intermixed both with each
other and with intermediate specimens (speci-
mens that were not identified as var. erecta or
var. strictissima, see Leht & Paal 1998b). Cluster
II3 comprises P. reptans and three specimens of
P. ¥ italica; cluster II4 includes specimens of P.
anglica and P. ¥ italica.
On lower dissimilarity levels, within rather
short distances, the phenogram can be split into
nine (level III) and finally into 15 subclusters
(level IV). When divided into subclusters of level
III, the clusters of Potentilla erecta (III1–III5)
remain mixed clusters consisting of specimens
of var. erecta and var. strictissima, as is also the
case with subclusters formed on level IV.
The cluster comprising specimens of Poten-
tilla reptans (II3) is divided into two subclusters
(III6 and III7), while cluster III6 also includes
three specimens of P. ¥ italica. Cluster II4 is split
into subcluster III8, joining exclusively speci-
mens of P. anglica and subcluster III9, compris-
ing only P. ¥ italica. All nine subclusters estab-
lished on level III are reliably distinct, as their
coefficients of indistinctness are close to zero
in all cases. The material of P. anglica collected
from Finland and the material originating from
Sweden belong mostly to separate subclusters on
level IV. But since the subclusters of level IV are
small and usually indistinct, only the subclusters
of level III will be further analysed.
The subclusters of Potentilla erecta are
mostly adjacent to each other in the charac-
ter space, except for subcluster III2 where for
32.4% of specimens the nearest neighbour is
the centroid of P. anglica (III8) (Table 3). Two
other subclusters of P. erecta are varying to some
extent towards P. italica: the centroid of sub-
cluster III9 is the most adjacent for 13.3% of the
specimens of subcluster III4 and for 6.7% of the
specimens of subcluster III5.
The within-group variation of the first Poten-
tilla reptans subcluster (III6) is directed mainly
towards P. anglica (III8, 66.7%) and P. ¥ italica
(III9, 20.4% ), while the other subcluster of P.
reptans (III7) is the nearest neighbour for only
13.0% of the specimens. For all specimens of
the smaller subcluster of P. reptans (III7), the
closest neighbour is the centroid of P. ¥ italica
(III9). Specimens of P. anglica (III8) recognise
two subclusters of P. erecta as the most adjacent
(III1, 22.5% and III2, 71.4%). The subcluster of
P. ¥ italica (III9) varies in many directions; varia-
Table 2. Adjacency matrix of conventionally estimated
species-clusters.
Cluster Cluster compared
analysed
P. erecta P. reptans P. anglica P. ¥ italica
P. erecta ¥ – 56.7 42.8
P. reptans – ¥ 43.1 56.9
P. anglica 79.6 16.3 ¥ –
P. ¥ italica 31.6 52.6 15.8 ¥
ANN. BOT. FENNICI Vol. 41 • Variation of Potentilla sect. Potentilla 57
tion is more pronounced towards the subclusters
of P. reptans (III6, 25.0% and III7, 43.8%), and to
a lesser extent also towards four subclusters of P.
erecta (Table 2).
The scatterplot of canonical roots of nine
subclusters (Fig. 3) is in a good concordance
with the scatterplot of species (Fig. 1) as well
as with the adjacency matrix of the subclusters
(Table 3). The first four subclusters of Potentilla
erecta form an almost totally overlapping cloud,
only the specimens of subcluster III5 reveal a
wider variation and some overlapping with the
Fig. 1. Scatt erp lot of
specimens by two first
canonical roots (axes).
Groups correspond to the
conventionally established
species; ellipses represent
the prediction interval of
the respective clusters in
which a single new obser-
vation can be expected to
fall with 95% probability.
Fig. 2. Phenogram of clus-
tering of standardized mor-
phometric data of Poten-
tilla erecta, P. reptans, P.
¥ italica and P. anglica
specimens according to
the minimal incremental
sum of squares algorithm
and Manhattan-metric.
58 Leht & Paal • ANN. BOT. FENNICI Vol. 41
cluster of P. ¥ italica. A very large overlapping is
also characteristic of the subclusters of P. reptans
(III6 and III7). The cluster including P. ¥ italica
specimens has again a transitional position
between all other subclusters.
Characters
The means of the characters of phenotypically
similar Potentilla reptans, P. anglica and P. ¥ ital-
ica appeared according to multivariate ANOVA
analysis all significantly different (Table 4).
In distinguishing conventionally identified
species, the most important character is number
of leaflets followed by number of rosette leaves,
length of the sepal, width of the stipule and width
of the central tooth.
The characters important in separation
of the subclusters of level III are not exactly
the same as in discrimination of the species;
now the most important character, number of
leaflets, is followed by number of flowers and
length of the petiole, slightly less important are
the number of rosette leaves and width of the
sepal.
Table 3. Adjacency matrix of level III clusters in Fig. 2.
Cluster Cluster compared
analysed
III1 III2 III3 III4 III5 III6 III7 III8 III9
III1 ¥ 51.7 43.3 – – – – – –
III2 67.6 ¥ – – – – – 32.4 –
III3 73.0 – ¥ 8.1 16.2 – – – –
III4 30.0 – 56.7 ¥ – – – – 13.3
III5 – – 86.7 6.7 ¥ – – – 6.7
III6 – – – – – ¥ 13.0 66.7 20.4
III7 – – – – – – ¥ – 100.0
III8 22.5 71.4 – – – – – ¥ –
III9 6.3 – 12.5 6.3 6.3 25.0 43.8 – ¥
Fig. 3. Scatterplot of speci-
mens by two first canonical
roots (axes); groups cor-
respond to the clusters of
level III on the phenogram.
ANN. BOT. FENNICI Vol. 41 • Variation of Potentilla sect. Potentilla 59
Chromosomes
The chromosome number counted for P. ¥ italica
was 2n = 42.
Discussion
As Potentilla erecta and P. reptans are quite
common in Estonia, while P. anglica is absent,
the occurrence of P. ¥ italica here points to its
possible origin from the hybridization between
P. erecta and P. reptans. The relatedness of these
species is also confirmed by the evident pheno-
typical variation of the specimens of P. ¥ italica
towards P. reptans and P. erecta (Table 2), while
much less in the direction of P. anglica. How-
ever, a certain morphological resemblance of P.
¥ italica with P. anglica can be concluded from
the fact that on the phenogram P. ¥ italica and P.
anglica belong on the second level of the same
cluster (II4).
According to Czapik (1975), Potentilla ¥
italica has four genomes of P. reptans and two
genomes of P. erecta, which could explain the
stronger association between P. ¥ italica and
P. reptans compared with its other parent. P. ¥
italica produces no seeds in Estonia and repro-
duction takes place only vegetatively by runners.
As P. erecta and P. reptans were not found in the
nearest vicinity of the only Estonian population
of P. ¥ italica known at present, we regard it here
as a vegetatively persistent, not repeatedly born
nothospecies rather than an accidental hybrid.
According to Matfield et al. (1970), P. ¥ italica
is sometimes, but not always, isolated from any
other member of the section Potentilla.
Our data show that Potentilla anglica and
P. ¥ italica, the taxa of hybrid origin, are mor-
phologically well separable from each other and
from their putative parents from the statistical
point of view. Although P. ¥ italica is known to
be morphologically highly variable, the means of
the morphological characters used in the study
differ from those of P. anglica as well as from
those of P. reptans, which is genetically very
close to P. ¥ italica (Czapik 1975), and agree
well with the morphological data obtained from
the Dutch material (Ietswaart & Kliphuis 1985).
Ietswaart and Kliphuis (1985) stressed that the
exact identification of section Potentilla (Tor-
mentillae) specimens is sometimes possible only
Table 4. Mean ± standard error of the species characters. P = significance level according to univariate ANOVA,
other notations as in Table 1.
Character Species P
P. erecta P. reptans P. anglica P. ¥ italica
NOD 2.3 ± 0.1 6.8 ± 0.3 3.3 ± 0.1 2.4 ± 0.3 < 0.001
BRCH 2.0 ± 0.0 1.1 ± 0.0 1.7 ± 0.1 2.3 ± 0.2 < 0.001
HU 2.0 ± 0.0 1.4 ± 0.1 1.8 ± 0.1 1.6 ± 0.1 < 0.001
HL 2.0 ± 0.0 2.1 ± 0.1 1.9 ± 0.1 2.1 ± 0.1 0.043
RL 3.0 ± 0.0 1.8 ± 0.1 1.9 ± 0.1 2.2 ± 0.1 < 0.001
FLWS 10.4 ± 0.6 3.9 ± 0.2 5.9 ± 0.7 11.4 ± 1.6 < 0.001
LFL 23.3 ± 0.4 25.9 ± 1.2 16.6 ± 0.7 30.3 ± 2.7 < 0.001
LFW 7.6 ± 0.2 11.0 ± 0.5 7.6 ± 0.4 12.2 ± 1.5 < 0.001
TEETH 9.7 ± 0.5 16.2 ± 0.4 8.9 ± 0.2 15.1 ± 0.6 < 0.001
STPL 12.4 ± 0.2 8.2 ± 0.3 8.3 ± 0.3 13.1 ± 1.3 < 0.001
STPW 9.4 ± 0.2 3.1 ± 0.1 2.3 ± 0.1 4.7 ± 0.4 < 0.001
TOL 2.6 ± 0.1 1.6 ± 0.1 3.2 ± 0.1 3.5 ± 0.3 < 0.001
TOW 1.2 ± 0.0 1.3 ± 0.1 1.7 ± 0.1 2.1 ± 0.1 < 0.001
SEPL 3.1 ± 0.0 8.0 ± 0.2 4.2 ± 0.1 5.2 ± 0.2 < 0.001
SEPW 1.5 ± 0.0 3.4 ± 0.1 2.3 ± 0.0 3.0 ± 0.1 < 0.001
LFN 3.0 ± 0.0 5.2 ± 0.1 3.1 ± 0.0 4.6 ± 0.1 < 0.001
PET 0.1 ± 0.0 5.0 ± 0.4 1.5 ± 0.1 3.8 ± 0.5 < 0.001
LPET 3.8 ± 0.1 8.5 ± 0.2 5.8 ± 0.1 7.1 ± 0.3 < 0.001
WPET 3.5 ± 0.1 7.1 ± 0.1 5.7 ± 0.1 6.7 ± 0.2 < 0.001
60 Leht & Paal • ANN. BOT. FENNICI Vol. 41
with the aid of the data of chromosome numbers
and fertility degree. Hence, as Estonian P. ¥
italica produces no seeds and is hexaploid (2n
= 42), the studied plants definitely belong to this
nothospecies.
The variation of Potentilla reptans is not
very wide, as its subclusters of level III did not
appear separated on the scatterplot (Figs. 2 and
3), but were situated within each other, while the
subclusters of level IV were mutually indistinct.
Morphological differences in Potentilla
anglica are not important either (being even
smaller than in P. reptans), as the subclusters of
level IV appeared indistinct and revealed no sub-
groups on the ordination scheme; still, specimens
originating from different regions were clustered
into separate subclusters. Hence, the existence of
geographic morphotypes of P. anglica is possible
and needs further consideration.
The specimens identified as Potentilla erecta
var. erecta and P. erecta var. strictissima (some
authors have considered these taxa subspecies
or even species) did not form clusters of their
own either when the intraspecific variation of
P. erecta was analysed (Leht & Paal 1998b) or
when their variation was studied in compari-
son with P. reptans, P. anglica and P. ¥ italica
(Fig. 1). At the same time, P. reptans, P. anglica
and P. ¥ italica formed clusters consisting of
only one taxon. As the subtaxa of P. erecta did
not cluster out even as small indistinct groups on
level IV, it confirms once more that these subtaxa
do not deserve the rank of subspecies but should
be referred to as varieties.
Hence, as in the analysis Potentilla ¥ italica
behaved in the same way as the “good species”
P. reptans and the stabilized hybrid species P.
anglica, which each formed a separate cluster,
it is certainly justified to recognize it at the same
taxonomic level as P. reptans and P. anglica, i.e.,
as a morphologically stable nothospecies.
Acknowledgements
The work was supported by ESF grants No. 4833 and No.
0180552. The investigation of the material in Helsinki was
financed by CIMO. We are very thankful to Prof. R. Czapik for
the donation of the herbarium material and for checking the iden-
tifications of the Estonian specimens. We are very grateful to Mrs.
Ester Jaigma for revising the English text of the manuscript.
References
Ball, P. W., Pawlovski, B. & Walters, S. M. 1968: Potentilla
L. — In: Tutin, T. G., Heywood, V. H., Burges, N. A.,
Moore, D. M., Valentine, D. H., Walters, S. M. & Webb,
D. A. (eds.), Flora Europaea 2: 36–47. Cambridge Univ.
Press, Cambridge.
Czapik, R. 1968: Karyological studies on Potentilla reptans
L. and P. mixta Nolte. — Acta Biol. Cracov. Ser. Bot. 11:
187–197.
Czapik, R. 1975: Apomixis in a sterile hybrid species of
Potentilla. — In: Walters, S. M. (ed.), European floristic
and taxonomic studies: 38–47. E. W. Classey, London.
Duda, R. & Hart, P. [Duda, R. & Hart, P.] 1976: [Patterns
classification and scene analysis]. — Mir, Moscow. [In
Russian].
Flora of Northern Ireland 2000–2001. — Available on the
web at http://
www.ulstermuseum.org.uk/flora/
Hegi, G. 1923: Potentilla. — In: Hegi, G. Illustrierte Flora
von Mittel-Europa 4(2): 809–892. Carl Hansen Verlag,
München.
Ietswaart, J. H. & Kliphuis, E. 1985: Hybrids between Dutch
Tormentillae (Potentilla, Rosaceae). — Acta Bot. Neerl.
34: 157–170.
Leht, M. 1989: Otsigem maranaid. — Estonian Nature 3: 150.
Leht, M., Eglïte, Z. & Lapelë, M. 1996: Potentilla L. — In:
Kuusk, V., Tabaka, L. & Jankevièienë, R. (eds.), Flora of
the Baltic countries. Compendium of vascular plants 2:
68–79. Eesti Loodusfoto, Tartu.
Leht, M. & Paal, J. 1998a: Potentilla L. sect. Rivales Wolf
and related taxa in the Baltic states. — Folia Geobot.
33: 171–186.
Leht, M. & Paal, J. 1998b: Variation of Potentilla erecta
(Rosaceae) in Estonia. — Ann. Bot. Fennici 35: 11–19.
Leht, M. & Paal, J. 1998c: Variation of Potentilla sect.
Aureae (Rosaceae) in the Baltic states. — Nordic J. Bot.
18: 339–351.
Matfield, B. 1972: Potentilla reptans L. — Identification of
its hybrids. — Watsonia 9: 137–139.
Matfield, B., Jones, J. & Ellis, J. R. 1970: Natural and experi-
mental hybridization in Potentilla. — New Phytol. 69:
171–186.
Matfield, B. & Ellis, J. R. 1972: The allopolyploid origin and
genomic constitution of Potentilla anglica. — Heredity
29: 315–327.
Matfield, B. & Walters, S. M. 1975: Potentilla L. — In:
Stace, C. A. (ed.), Hybridization and the flora of the Brit-
ish Isles: 206–209. Academic Press, London.
Paal, J. 1987: Taxonomic continuum, some problems and
methods for its quantitative analysis. — In: Laasimer,
L. & Kull, T. (eds.), The plant cover of the Estonian
SSR. Flora, vegetation and ecology: 108–122. Valgus,
Tallinn.
Paal, J. L. & Kolodyazhnyi, S. F. [Paal‘, Ä. L. &
Kolodäånyj, S. F.] 1983: [Quantitative methods
for analysing transitions between vegetation syntaxa].
— Bot. Zh. 68: 1467–1474. [In Russian].
Podani, J. 2000: Introduction to the exploration of multivari-
ate biological data. — Backhuys Publ., Leiden.
Podani, J. 2001: SYN-TAX 2000 computer program for data
ANN. BOT. FENNICI Vol. 41 • Variation of Potentilla sect. Potentilla 61
analysis in ecology and systematics for WINDOWS 95,
98 & NT. User’s manual. — Scientia Publ., Budapest.
Richards, A. J. 1973: An upland race of Potentilla erecta (L.)
Räusch. in the British Isles. — Watsonia 9: 301–317.
Staffordshire Biodiversity Action Plan 2001, 2 ed. — Aval-
able on the web at http://
www.sbap.org.uk/audit/.
Wolf, T. 1908: Monographie der Gattung Potentilla L.
— Biblioth. Bot. 71: 1–714.
This article is also available in pdf format at http://www.sekj.org/AnnBot.html
