Content uploaded by Thom W Kuyper
Author content
All content in this area was uploaded by Thom W Kuyper
Content may be subject to copyright.
A COMPARISON OF THE APPLICATION
OF A BIOLOGICAL AND PHENETIC SPECIES CONCEPT
IN THE HEBELOMA CRUSTULINIFORME COMPLEX
WITHIN A PHYLOGENETIC FRAMEWORK
DUUR K. AANEN
1 & THOMAS W. KUYPER
2
A method is presented to derive an operational phenetic species concept for the Hebe-
loma crustuliniforme complex in northwestern Europe. The complex was found to
consist of at least 22 biological species (intercompatibility groups; ICGs). Almost
none of these biological species could be recognised unambiguously by morpholog-
ical criteria. It is therefore necessary to base a phenetic species concept on combina-
tions of biological species. However, such species delimitation must be performed
within an explicitly phylogenetic context. It is crucial therefore to have a reliable
estimate of the phylogeny of 22 biological species in that complex. Based on two
nuclear sequences, we present a best estimate of the phylogeny of biological species
within the complex. Using this phylogeny, on the basis of strict monophyly only two
species can be morphologically recognised among 22 biological species. Relaxing the
criterion of monophyly and allowing paraphyletic groupings of biological species as
phenetic species would result in the recognition of three phenetic species. A tree, with
the ve ICGs of the previously dened morphospecies H. crustuliniforme (1, 2, 3,
4 and 5) constrained as a monophyletic group, can not be rejected. This constrained
tree, together with the relaxed criterioon that allows for paraphyletic groupings of
biological species, leads to the recognition of four phenetic species, viz. H. crustu-
liniforme, H. helodes, H. incarnatulum and H. velutipes. These phenetic species are
described and a key is provided. Other taxon names are briey discussed. The very
limited ability to translate a biological species concept into an operational phenetic
species concept is explained by the lack of qualitative characters and the plasticity
of quantitative characters. Recency of common evolutionary history is also a major
factor. Intercompatibility tests and DNA based phylogenies indicate that most bio-
logical species are very closely related and hence provide support for the claim that
correspondence between a biological species concept and a phenetic species concept
in the H. crustuliniforme complex is not likely to be forthcoming. In an Appendix
morphological descriptions are provided of the 22 ICGs.
Among the various genera of the Agaricales the genus Hebeloma (Fr.) Kumm. has often
been regarded as taxonomically difcult. The status of a number of described species
is uncertain, and taxonomic controversies abound. This somewhat frustrating situation
has been eloquently described by Favre (1960): “Il nʼest pas de genre où la taxonomie
des espèces soit plus embrouillée. Cʼest un véritable chaos. Même pour les espèces
les plus répandues le désaccord règne entre les mycologues. Placé dans la nécessité de
1) Department of Population Ecology, University of Copenhagen, Universitetsparken 15, DK-2100
Copenhagen, Denmark; dkaanen@zi.ku.dk
2) Subdepartment of Soil Quality, Wageningen University, P. O. Box 8005, 6700 EC Wageningen,
The Netherlands; thom.kuyper@wur.nl
P E R S O O N I A
Volume 18, Part 3, 285 – 316 (2004)
286 PERSOONIA — Vol. 18, Part 3, 2004 Aanen & Kuyper: Biological and phenetic species concept in Hebeloma crustuliniforme 287
parler des hébélomes (…), je me trouve dans le plus grand embarras.” Within the genus
the complex of taxa around H. crustuliniforme (Bull.) Quél. has been particularly prob-
lematic. For the purpose of this paper that species complex is operationally described
as follows: Very small to large mushrooms. Pileus viscid, ranging from white to dark
reddish brown, but often with a paler margin. Cortina absent. Lamellae clay-brown,
often exuding drops of water (ʻlamellae weepingʼ). Stipe shorter to (much) longer
than pileus diameter, white brillose, with a pruinose to occulose apex. Spore print
clay-brown. Spores ornamented. Cheilocystidia cylindrical to clavate to (sub)capitate,
thin-walled, but sometimes with thickened wall in the median or upper part, hyaline.
Pleurocystidia absent. Ectomycorrhizal with a very wide range of trees, under a large
range of ecological conditions.
In the framework of a study on species and speciation in the H. crustuliniforme com-
plex (Aanen & Kuyper, 1999; Aanen et al., 2000a, 2001) the question about an optimal
taxonomy was also addressed. However, an optimal taxonomy can only be dened in
relation to species concepts (Levin, 2000). Discussions on species concepts are as old
as the taxonomic practice. Even a classication of the various kinds of species con-
cepts is liable to heated debate, because the terms for the various concepts have as much
an ideological as an explanatory function. For the purpose of this paper we recognise two
classes of species concepts, viz. the mechanistic and non-mechanistic or historical spe-
cies concepts. Mechanistic species concepts, often taken together under the denominator
biological species concept, are based on the various processes and mechanisms by which
species originate (speciate) or by which they cohere. Mechanisms of speciation relate to
the origin of reproductive isolation related to the origin of genetic divergence (incompat-
ibility rst or divergence rst; Aanen et al., 2000b), and mechanisms of cohesion refer
to genetic and ecological mechanisms that allow interbreeding within that species and
simultaneously prevent breeding with members of different species. Among the non-
mechanistic species concepts, which consider pattern instead of process, two concepts
have attracted much attention, viz. a phylogenetic concept, which emphasises mono-
phyly of all the members of the species, and a phenetic concept, which emphasises
morphology.
Each of the three species concepts has been applied to the H. crustuliniforme complex.
Aanen & Kuyper (1999) studied intercompatibility groups (ICGs) and arrived at the
conclusion that at least 20 different ICGs (ʻbiological speciesʼ) could be recognised.
However, a subsequent analysis of a subset of these isolates showed that intercompat-
ibility was not a qualitative character (ʻall or nothingʼ), but that degrees of compatibility
could be recognised. One isolate turned even out to be fully compatible with members
of 2 ICGs, which were otherwise incompatible (Aanen et al., 2000b). Apparently, inter-
compatibility may be a plesiomorphous character and the mechanistic species concept
does not always allow to determine unambiguously to which species a certain fungus
belongs. The ICGs were subsequently subjected to a phylogenetic analysis, based on
sequence data of the internal transcribed spacers (Aanen et al., 2000a). This phyloge-
netic analysis also showed that ICGs did not always meet the criterion of monophyly.
The phenetic species concept (ʻmorphological species conceptʼ) has been applied by
different authors (e. g. Bruchet, 1970; Vesterholt, 1995) but, as Favreʼs words testify,
consensus was hardly reached.
286 PERSOONIA — Vol. 18, Part 3, 2004 Aanen & Kuyper: Biological and phenetic species concept in Hebeloma crustuliniforme 287
In this paper we attempt to reconcile the various species concepts, or, if the concepts
are fundamentally incompatible, at least to arrive at an operational taxonomy that is
consistent with genetics, morphology and phylogeny. We compare a mechanistic (ʻbio-
logicalʼ) and phenetic (ʻmorphologicalʼ) species concept within an explicit phylogenetic
framework. Our ultimate aim is the phenetic species concept as this is the only concept
useful for the general user of a Flora (Kuyper, 1988). The criterion for consistency is
intended to put constraints on the phenetic species concept. First, no intercompatible col-
lections should be classied as different species. Second, interincompatible collections
should be regarded as belonging to the same phenetic species, if there are no morpho-
logical criteria by which these biological species can be separated. Third, combina-
tions of biological species as phenetic species should only be accepted if the group of
biological species forms a monophyletic entity. However, even these conditions can
conict. For example, if one non-basal ICG can be morphologically separated from a
monophyletic group of ICGs, the remainder of that group automatically turns into a
paraphyletic group.
The strict criterion of monophyly of species has therefore been challenged. We
recognise that biological species can form paraphyletic groups. De Queiroz & Dono-
ghue (1988, 1990) have stressed that interbreeding units need not necessarily be
monophyletic. As speciation often involves the splitting off of marginal and / or local
populations (Levin, 1993; Rieseberg & Brouillet, 1994), a consequence is that after
such a speciation event the parent species has become paraphyletic. If the rules of the
cladistic game prevent recognition of such paraphyletic groupings as species, we must
accept a new mechanism of speciation, described by Templeton (1998) as speciation by
remote control. Aanen et al. (2001) noted that ICG 17 (H. velutipes) was paraphyletic,
as it contained two ITS types that belonged to different clades. Paraphyly of biological
species has also been observed in the genus Pleurotus (Vilgalys & Sun, 1994). In that
genus biological species were monophyletic within a continent, but paraphyletic when
investigated over various continents. The same pattern likely has occurred in ICG 17.
The ITS polymorphism occurred in Europe and in North America. A plausible scenario
for this polymorphism in the face of concerted evolution is divergence in allopatry,
followed by bilateral migration to both continents. This could have occurred with the
introduction of plantation forest trees, such as Pseudotsuga menziesii in Europe, where
the phenetic species H. velutipes is regularly found (Aanen et al., 2001).
If, however, one accepts paraphyletic taxa at the level of biological species, one may
wonder why combinations of biological species, forming phenetic species, could not also
form paraphyletic entities. We therefore also considered the consequences of relaxing
the criterion of strict monophyly and decided to recognise paraphyletic phenetic species
as well. Alternatively, we could accept such paraphyletic taxa on infraspecic level.
However, we considered polyphyletic entities unacceptable as phenetic species. To test
for monophyly of groups of ICGs, which have formerly been recognised as morphospe-
cies, the most parsimonious tree(s) were compared with constrained trees in which such
morphospecies (H. crustuliniforme, H. lutense, H. pusillum) were monophyletic.
A crucial step in our approach is to have a reliable estimate of the phylogeny of the
biological species. For the H. crustuliniforme complex as a whole, we have estimated
phylogenetic relationships based on ITS sequences (Aanen et al., 2000a). Taxonomic
288 PERSOONIA — Vol. 18, Part 3, 2004 Aanen & Kuyper: Biological and phenetic species concept in Hebeloma crustuliniforme 289
resolution based on ITS sequences turned out to be insufcient for a group of nine ICGs
belonging to different morphospecies such as H. crustuliniforme, H. leucosarx and
H. pusillum). For that group the Intergenic Spacer (IGS) was also studied (Aanen et
al., 2000b).
Different data sets that have the same evolutionary history are expected to converge
onto the true species phylogeny of the group under study, if analysed using appropriate
phylogenetic methods (Mes, 1995). In principle, such data sets can be combined. Kluge
(1989) proposed that phylogenetic analysis should always be performed using all the
available evidence (the ʻtotal evidenceʼ approach). In this approach, all of the independ-
ent characters available to the systematist should be combined and then analysed using
parsimony. However, others have argued against this approach (e. g. Lutzoni & Vilgalys,
1995). Miyamoto & Fitch (1995) argued that phylogenetic trees should be estimated
separately from each data set and the different estimates should be compared using
taxonomic congruence. Under this separate analysis approach, each partition represents
an independent estimate of the tree, and these different estimates can be judged for
congruence. It is often argued that congruence among different data partitions provides
some of the strongest evidence that a particular phylogenetic estimate is accurate (Hillis
et al., 1996). A compromise between the ʻtotal evidenceʼ approach and the ʻseparate
analysisʼ approach is the ʻconditional combinationʼ approach (Huelsenbeck et al., 1996a)
as advocated by Bull et al. (1993) and De Queiroz (1993). Under this approach, data sets
are statistically tested for homogeneity. Heterogeneous data sets are those that result in
signicantly different estimates of phylogeny when analysed separately and these data
sets can not be combined. If the test result is non-signicant, i. e. the data sets do not
result in signicantly different estimates of the phylogeny, then these data sets should
be combined (Huelsenbeck et al., 1996a). As an alternative to combining the data sets,
the resulting trees can be combined (Mes, 1995; Sanderson et al., 1998). A ʻsupertreeʼ
is an estimate of a phylogeny assembled from sets of smaller estimates (source trees)
sharing at least some taxa (Sanderson et al., 1998).
To the morphological characters studied belong those traditionally used in Hebeloma
taxonomy (Bruchet, 1970; Vesterholt, 1995). Since many of the characters used are
quantitative, we did not reconstruct phylogenies based on these characters. Instead,
we i) reconstructed organismal phylogenies based on molecular data; and ii) tried to
dene morphologically recognisable monophyletic entities. Using the best estimate
of the phylogenetic relationships of ICGs within the H. crustuliniforme complex, we
addressed the following questions:
1. How many morphological taxa, consisting of (strictly) monophyletic groups of ICGs
(biological species) can be recognised in this complex?
2. How would relaxing the criterion of monophyly and allowing paraphyletic groupings
of ICGs affect the number of phenetic species that can be recognised?
3. How would relaxing the criterion of monophyly and allowing groupings of ICGs in
previously recognised morphospecies that can not statistically be rejected against
the most parsimonious tree(s), affect the number of phenetic species that can be
recognised?
4. What is the phylogenetic quality of some previously recognised morphospecies such
as H. alpinum, H. lutense or H. pusillum?
288 PERSOONIA — Vol. 18, Part 3, 2004 Aanen & Kuyper: Biological and phenetic species concept in Hebeloma crustuliniforme 289
MATERIALS AND METHODS
MATERIAL
Sexual intercompatibility was tested for 110 collections (Aanen & Kuyper, 1999).
This analysis led to the recognition of at least 20 intercompatibility groups (ICGs). Two
collections (isolates 9692 and 9694) were not compatible with any of the other collec-
tions. However, since these collections have neither shown compatibility in intracol-
lection pairings nor in intercollection pairings, the possibility that these two collections
were ʻincompetentʼ could not be excluded (R. Petersen, pers. comm.). Therefore, it was
not warranted to give a formal status as ICG to these two collections. However, assum-
ing that the two collections are competent, they represent two other ICGs. We therefore
consider them as representants of two further biological species: ICG 13 and ICG 22.
The macroscopical characters were determined for all 110 collections, the microscopy
for 78 collections.
MORPHOLOGICAL DESCRIPTIONS OF ICGS AND
MORPHOLOGICAL CHARACTERS USED
Each ICG was described morphologically and the range of character states was describ-
ed for each ICG. The morphological characters used are listed in Table I. Many of these
characters are quantitative. Full descriptions of the ICGs are given in the Appendix.
PHYLOGENETIC ANALYSIS AND COMBINING DATA SETS
For clade I (the H. velutipes clade), we had only the ITS data to reconstruct an organ-
ismal phylogeny. For clade IIa, we had different data sets. The rst estimate of the
phylogeny was based on ITS sequences (Aanen et al., 2000a). Clade IIa, except for the
two ICGs of H. pusillum, was studied in detail using different sequences: the nuclear
IGS and a mitochondrial intron (Aanen et al., 2000b). Here we include IGS sequences
of the two ICGs of the morphospecies H. pusillum, ICG 7 and 9. We performed a new
parsimony analysis with the inclusion of those two additional taxa. For the details of
the parsimony analysis we refer to Aanen et al. (2000a). Gaps were coded according to
Hibbett et al. (1995) for all data sets. The reason that we used gap coding for the ITS
data here but not in Aanen et al. (2000a) is that the analysis here was limited to a group
of closely related taxa, the alignment of which was straightforward, whereas the align-
ment with the extended data set was more ambiguous.
Sixteen collections were common for the two nuclear data sets, the ITS and IGS
sequences (ICG 1: 9503, 9618, 9621, 9673; ICG 2: 9570, 9627; ICG 3: 9680; ICG 4:
9602; ICG 5: 9581; ICG 7: 9654; ICG 8: 9538; ICG 9: 9509; ICG 14: 9566; ICG 15:
9624; ICG 20: 9688; ICG 21: 9650). Eleven collections were common for all data sets
(ICG 1: 9503, 9618, 9621, 9673; ICG 2: 9570, 9627; ICG 3: 9680; ICG 4: 9602; ICG 14:
9566; ICG 20: 9688; ICG 8: 9538). The Partition Homogeneity Test (Farris et al., 1995;
Huelsenbeck et al., 1996b; implemented in PAUP*) was used (with 1000 replicates) to
determine whether the different data sets were in conict. In this test, the observed sites
from all genes for each individual are pooled and resampled without replacement to
give an articial data set in which sites have been swapped randomly among loci. Many
290 PERSOONIA — Vol. 18, Part 3, 2004 Aanen & Kuyper: Biological and phenetic species concept in Hebeloma crustuliniforme 291
Table I. Morphological characters used to describe ICGs.
Macroscopical: pileus: diameter
colour
shape
presence of hygrophanous spots
lamellae: number
shape
weeping
stipe: length
width
presence of bulb
presence of pendent marrow strand
covering
general habit
smell
Microscopical: spores: length
width
Q (ratio l / w)
dextrinoidy (scale D0 – D4, see Vesterholt, 1995)
shape
perispore loosening (scale P0 – P3, see Vesterholt,
1995)
ornamentation (scale O0 – O4, see Vesterholt,
1995)
cheilocystidia: length
width at median part
width at apex
Q (width apex / width median part)
shape
wall thickness
presence of apical bifurcations
Host tree genera
such articial data sets are produced. MP trees are then made for each newly sampled
partition in each articial data set. If the data sets have the same evolutionary history,
the sums of the lengths of the gene trees for the observed and resampled data should
be similar, but if they have different evolutionary histories, the sums of the tree lengths
should be longer than that for the actual data, because of extra homoplasy in the data
(Geiser et al., 1998).
Furthermore, some alternative topologies were tested. Three species traditionally
considered to be ʻgoodʼ morphospecies are H. crustuliniforme sensu stricto, H. pusillum
and H. lutense. We rst did a parsimony analysis with the constraint that the biological
species of which the morphospecies H. crustuliniforme consisted (ICGs 1, 2, 3, 4 and 5)
formed a monophyletic group. Secondly, an analysis was performed with the morphospe-
cies H. pusillum as a monophyletic group (ICGs 6, 7, 8 and 12). Thirdly, a constrained
analysis was performed with the morphospecies H. leucosarx as a monophyletic group
(ICG 14 and 15). The constrained trees found were compared with the unconstrained
trees using the Kishino-Hasegawa (1989) test and Templetonʼs (1983) nonparametric
test as implemented in PAUP*.
290 PERSOONIA — Vol. 18, Part 3, 2004 Aanen & Kuyper: Biological and phenetic species concept in Hebeloma crustuliniforme 291
STRATEGY TO ARRIVE AT AN OPERATIONAL TAXONOMY
The rst species concept that was tested as an operational species concept was the
biological species concept. We considered the morphology of biological species and
tested if different biological species could be recognised by morphological criteria. We
recognised the possibility that biological species could represent paraphyletic taxa. For
eight ICGs we have included more than one strain in the phylogenetic analysis: ICGs 1,
2, 3, 4, 9, 11, 17 and 21. For those ICGs we tested the hypothesis that strains of a single
ICG form a monophyletic group.
The second species concept that was tested as an operational species concept was
based on combinations of biological species within a phylogenetic framework. On the
basis of an estimate of the phylogenetic relationships within Hebeloma, we tested for
every sister group whether both sister taxa could be morphologically separated. The
morphological descriptions of the ICGs were used to do this. As an initial help, we used
a set of 13 morphological characters, divided into discrete classes (Table II). Sister taxa
were separated if they showed no overlap in at least one of these characters. If both of
them could indeed be unambiguously demarcated, they were (at least provisionally) ac-
cepted as valid phenetic species. The process was then repeated at the next higher level
till all sister group relations had been dealt with. If sister taxa could not be recognised
separately as phenetic species, both sister taxa were lumped and the morphological
variability for the composite species was assessed. Again the process was repeated till
all sister group relations had been dealt with. We introduced an additional criterion for
recognition, viz. that morphological relationships between such provisional morphotaxa
could be upheld across hierarchical levels. Essentially, in this approach the two sister
groups A and B, even when sufciently different to be kept apart by standard taxonomic
practice, were lumped when clade C, the sister group of AB, could not be treated as
separate from either group A or B.
In cases where the consensus cladogram did not yield sister group relationships but
showed unresolved polytomies, each taxon in a polytomy was compared with every
other taxon. Inevitably, this could result in a complex pattern of relationships within the
polytomy where some taxa could be unambiguously separated from each other whereas
some other ones could not. Again, the criterion of consistency across levels was used.
A more relaxed version of this procedure was tested as well. In this version, para-
phyletic taxa were recognised, viz. when the sister groups A and B could be separated,
but clade C could only be separated from A, but not from B, we recognised the mono-
phyletic A, and the paraphyletic (B, C).
RESULTS
PHYLOGENY OF ICGs
For the 16 collections for which both ITS and IGS sequences were determined, we
determined whether these data sets were in conict using the partition homogeneity test.
The actual summed tree length of 171 was equal to or longer than 65.6% of the articial
data sets, indicating that the gene trees did not have signicantly different topologies
(Fig. 1a). Therefore, we combined the ITS and IGS data sets to reconstruct a nuclear
292 PERSOONIA — Vol. 18, Part 3, 2004 Aanen & Kuyper: Biological and phenetic species concept in Hebeloma crustuliniforme 293
phylogeny. The sister taxon of clade II, H. sarcophyllum was used as the outgroup. Doing
a parsimony analysis with gaps coded according to Hibbett et al. (1995), four trees were
found of length 237 (c. i. = 0.86; excluding uninformative characters, l = 113, c.i. = 0.69),
the strict consensus of which is depicted in Fig. 2. For the 11 collections for which all
three sequences were determined, we also did the partition homogeneity test. The actual
summed tree length of 134 was smaller than 99.9% of the articial data sets, indicating
that the gene trees did have signicantly different topologies (Fig. 1b). Therefore, we
conclude that the mitochondrial and nuclear phylogenies can not be combined. Aanen
et al. (2000b) showed that the incongruence between the nuclear and mitochondrial
tree was mainly due to ICG 1, which had a different position in both phylogenies. As
a possible cause we proposed a hybridisation with different mitochondrial and nuclear
contributions. Here we use the nuclear phylogeny, but we consider the consequences
of other positions of ICG 1.
RECOGNIZABILITY AND MONOPHYLY OF ICGS
In the appendix morphological descriptions are given of 20 ICGs and two putative
ICGs (ICGs 13 and 22). Hebeloma incarnatulum is the single ICG that can be separated
from all the other taxa of the H. crustuliniforme complex by the shape of its cylindrical
to very narrowly clavate cheilocystidia. All other species of this complex have clavate
to (sub)capitate cheilocystidia. Of the ICGs represented by more than one collection,
ICGs 1, 2, 9 and 21 were monophyletic, and the two strains of ICG 11 had identical
ITS sequences but did not form a monophyletic group. The partially compatible ICGs
3 and 4 did not form monophyletic groups, together they constituted a monophyletic
group, however. Strains of ICG 17 did not form a monophyletic group. Two ITS types
were found within this ICG that belonged to two different clades.
Fig.1. Partition homogeneity test results.
292 PERSOONIA — Vol. 18, Part 3, 2004 Aanen & Kuyper: Biological and phenetic species concept in Hebeloma crustuliniforme 293
MONOPHYLETIC RECOGNISABLE ENTITIES
Since most of the ICGs could not be uniquely characterised, we tested if we could
recognise monophyletic combinations of ICGs. For every sister group we tested whether
both sister taxa could be morphologically separated. The consecutive steps for combin-
ing ICGs into monophyletic units are illustrated in Fig. 3 and described in Table III. Some
of the characters of the morphological descriptions of the ICGs and of combinations of
sister groups are presented in Table II.
To illustrate the procedure, we discuss some examples. At the lowest taxonomic
level some sister groups could not be separated and were combined and others could be
separated and were, at least provisionally, maintained (Fig. 3 and Table III). ICG 1 and
14 could be separated on the basis of general habit and pileus colour and were therefore
maintained at this point. The same was the case for ICG 5 and 7 (stipe-pileus ratio and
pileus colour). However, in subsequent steps, those taxa could not be maintained any
longer, because ICG 1 could not be separated from 5. ICGs 10 and 15 could be separated
on the basis of pileus colour, spore form, and general habit (stipe-pileus ratio). These
taxa were therefore maintained at this point. However, the sister group of the pair [10,
15], ICG 20, could not be separated from 10, although it could be separated from 15.
Therefore, these three taxa were lumped to g.
This analysis ultimately led to the recognition of 2 morphologically recognisable
monophyletic groups, one consisting of three ICGs (clade I), and one consisting of 19
ICGs (clade II). If paraphyletic species would be recognised, ICG 18 (H. incarnatulum)
could be recognised as an additional monophyletic morphospecies, with the two ICGs
of H. velutipes forming a paraphyletic phenetic species. The acceptance of paraphyletic
taxa (e. g. the pair [1, 5] or [10, 20]) did not have any inuence on the nal number of
species recognised in clade II. Only the moment of combining ICGs was postponed in
some cases if paraphyletic entities were (temporarily) accepted.
CONSTRAINED ANALYSIS
The main difference between the nuclear and mitochondrial phylogenies was the posi-
tion of ICG 1. In the mitochondrial tree, ICG 1 belonged to a clade together with ICG
2, 3 and 4 (and probably 5, see hypothesised gain and loss of different introns in Aanen
et al., 2000b). To test the hypothesis that ICG 1, 2, 3, 4 and 5 formed a monophyletic
group in the nuclear tree as well, we did a parsimony analysis on the nuclear data set
with the constraint that H. crustuliniforme was monophyletic. A total of 16 trees were
found of length = 241, which could not be rejected in favour of the unconstrained trees
(Kishino-Hasegawa test: p ≥ 0.10, Templetonʼs test: p ≥ 0.22). Performing the sister
group analysis on the strict consensus tree of those 16 constrained trees did not give other
conclusions than the analysis of the unconstrained trees. However, under the relaxed
version of the sister group analysis, viz. if we recognised paraphyletic taxa as well, the
5 ICGs of H. crustuliniforme could now be recognised as a monophyletic group versus
the paraphyletic rest of clade II.
A parsimony analysis on this data set with the constraint that the four ICGs of
H. pusillum formed a monophyletic group (ICGs 6, 7, 8 and 12) gave 583 trees of length
(text continued on p. 298)
294 PERSOONIA — Vol. 18, Part 3, 2004 Aanen & Kuyper: Biological and phenetic species concept in Hebeloma crustuliniforme 295
Taxon Pileusa pileus stipe stipe Qd stipe spore spore Qg spore Perispore spore cheilo– cheilo– optional
diametera colour
b width c covering
e length
f dextrinoidyh looseningi smoothnessj cystidia widthk cystidia Qav1 charactersm
ICG 1 2–3 1 2 –3 1 1 1–2 2– 3 0–1 0 –1 1–2 1– 2 4 0
ICG 2 2–3 1– 2 3 1 –2 (–3) 1 1– 2 2–3 0 –1 0 1– 3 1– 2 3– 4 0
ICG 3 2–3 1 3 1 –2 1 2 –3 1– 3 0– 1 0– 1 1– 3 2 3– 4 0
ICG 4 2–3 1 2 –3 1 1 2–3 2– 3 0–1 0 1–2 2 3– 4 0
ICG 5 2–3 1 3 1 –2(– 3) 1 2 3 0 –2 0 1– 3 1– 2 3– 4 0
ICG 6 1 3 1 3 2 3 3 0–1 0 2 1 4 0
ICG 7 1 3 1 3 2 1 –2 2 –3 0 –2 0 2– 3 1–2 4 0
ICG 8 1 3 1 3 2 2 –3 3 0– 2 0–1 1– 3 1– 2 4 0
ICG 9 1–3 2– 3 1 –2 2 –3 2 1 –2 2–3 0 –1 0 –1 2– 3 1 3–4 0
ICG 10 2 –3 1 –2 2 2 2 1 –2 2 –3 0– 1 0 1– 3 1 2 0
ICG 11 2 –3 1 –2 2–3 1 –2 2 1 –2 2 –3 0 –1 0 2– 3 1 –2 2– 4 0
ICG 12 1 –2 2 –3 1–2 2–3 2 1 –2 2–3 0 –2 0 2– 3 1 –2 4 0
ICG 13 2 1 –2 2 2 2 1 2 0–1 0 2– 3 1 2– 3 0
ICG 14 1 –2 3 1–2 1 –2 2 2 –3 2 –3 0 –1 0 –1 2 –3 1 –3 4 0
ICG 15 1 –2 3 1–3 1 –2 2 3 3 0 –2 0 –1 1 –3 1 –3 3 –4 0
ICG 16 2 –3 (1–)2 2 –3 1– 3 2 1 –2 1 –2 2 –4 0– 1 2 –3 3 2– 3 1
ICG 17 2 –3 (1–)2 2 –3 1– 3 2 1 –2 1 2– 4 0– 1 1 –3 2–3 2 –3 1
ICG 18 3 (1–)2 2 3 3 1 1 3 –4 0 2– 3 3 1 1
ICG 19 2 2 1– 2 3 2 1 2 0–1 0 –1 2–3 2 2 0
ICG 20 2 –3 2 3 1 –2 2 2 2 0 –2 0 1– 3 1–2 3 0
ICG 21 2 –3 2 2–3 2 –3 2 1 –2 2 –3 0 –1 0 2– 3 1 –2 2– 4 0
ICG 22 2 –3 2 2–3 1 –2 2 1 –2 2 –3 0 –1 0 1– 2 1 3 0
A (1,5) 2 –3 1 2– 3 1–2 1 1 –2 2–3 0 –2 0 –1 1– 3 1 –2 3– 4 0
B (3,4) 2 –3 1 2 –3 1– 2 1 2 –3 1 –3 0 –1 0– 1 1 –3 2 3– 4 0
C (8,9) 1 –3 2 –3 1 –2 2– 3 2 1 –3 2 –3 0 –2 0–1 1– 3 1 –2 3–4 0
D (13,22) 2– 3 1– 2 2 –3 1 –2 2 1–2 2 –3 0–1 0 1 –3 1 2 –3 0
E (16,17,18) 2 –3 2 2– 3 1–3 2 –3 1– 2 1 –2 2 –4 0– 1 1 –3 2 –3 1 –3 1
F (1,5,7,14) 1–3 1– 3 1 –3 1 –3 1 –2 1 –3 2 –3 0–2 0 –1 1–3 1– 3 3–4 0
G (2,a) 2 –3 1 –2 2 –3 1– 2 1 1 –3 1 –3 0 –1 0–1 1– 3 1 –2 3–4 0
H (10,15,20) 1 –3 1 –3 1–3 1 –2 2 1 –3 2 –3 0– 2 0 –1 1– 3 1 –3 2– 4 0
I (e,f) 1 –3 1 –3 1 –3 1– 3 1 –2 1 –3 1 –3 0–2 0 –1 1–3 1–3 2 –4 0
J (c,g,11,19) 1 –3 1 –3 1–3 1 –3 2 1 –3 2 –3 0 –2 0 –1 1 –3 1 –3 2– 4 0
K (h,b) 1 –3 1–3 1 –3 1 –3 1– 2 1– 3 1–3 0 –2 0 –1 1– 3 1 –3 2– 4 0
L (i,12) 1– 3 1– 3 1–3 1–3 2 1 –3 2 –3 0–2 0 –1 1–3 1– 3 2–4 0
M (k,j) 1 –3 1 –3 1 –3 1– 3 1 –2 1 –3 1–3 0 –2 0 –1 1 –3 1 –3 2 –4 0
N (6,l) 1–3 1– 3 1–3 1 –3 1–2 1 –3 1 –3 0– 2 0 –1 1– 3 1 –3 2– 4 0
O (m,21) 1 –3 1–3 1 –3 1 –3 1– 2 1– 3 1–3 0 –2 0 –1 1– 3 1 –3 2– 4 0
294 PERSOONIA — Vol. 18, Part 3, 2004 Aanen & Kuyper: Biological and phenetic species concept in Hebeloma crustuliniforme 295
←
Table II. Summary of 13 morphological characters for 22 ICGs and combinations of ICGs. — a =
Maximum for a collection – 1: ≤ 30, 2: 30 – 50, 3: ≥ 50 (in mm); b = 1: pale, 2: pale with dark centre,
3: dark; c = maximum for a collection – 1: ≤ 4, 2: 4 – 8, 3: ≥ 8 (in mm); d = 1: ≤ 7, 2: 7–10, 3: ≥ 10;
e = 1: coarsely occose, 2: occose, 3: occulose; f = 1: ≤ 11, 2: 11–12, 3: ≥ 12 (in μm); g = 1: < 1.7,
2: 1.7–1.8, ≥ 1.8; h = grouped 0– 4 (Vesterholt, 1995); i = grouped 0–3 (Vesterholt, 1995); j = grouped
0–4 (Vesterholt, 1995); k = just below the apex – 1: < 4.5, 2: 4.5 – 4.8, 3: > 4.8 (in μm); l = 1: < 1.2,
2: 1.2 –1.7, 3: 1.7– 2, 4: ≥ 2; m = 0: less than four of the following six (optional) character states:
bulbous stipe, pendant marrow strand, forked apex of cheilocystidia, hygrophanous spots on pileus,
dextrinoid spores (≥ 3), Q spores ≤ 1.7; 1: at least four of these six (optional) character states.
1st level
– compare 1, 14 – can be separated (general habit, stipe-pileus
ratio, pileus colour) – maintain 1 and 14.
– compare 5, 7 – can be separated (general habit, stipe-pileus
ratio) – maintain 5 and 7.
– compare 3, 4 – can not be separated – combine to a.
– compare 8, 9 – can not be separated – combine to b.
– compare 10, 15 – can be separated (general habit, stipe-pileus
ratio) – maintain 10 and 15.
– compare 13, 22 – can not be separated – combine to c.
– compare 16, 17, 18 in all combinations – 16 and 17 can not be separated – combine to
d.*
2nd level
– compare 1, 14, 5 and 7 in all combinations – 1 and 5 can not be separated – combine to e.
– compare a, 2 – can not be separated – combine to f.
– compare 10, 20 and 15, 20 – 10, 20 can not be separated – combine 10, 15
and 20 to g.
3rd level
– compare e and f – can not be separated – combine to h.
– compare g, 19, 11 and c in all combinations – none of them can be separated – combine to i
4th level
– compare h, b – can not be separated – combine to j.
– compare i, 12 – can not be separated – combine to k.
5th level
– compare j, k – can not be separated – combine to l.
6th level
– compare l, 6 – can not be separated – combine to m.
7th level
– compare m, 21 – can not be separated – combine to n.
8th level
– compare d, n – can be separated – maintain d and n.
Table III. Consecutive steps for combining ICGs into monophyletic units.
* ICG 18 can be separated from all ICGs because of the shape of its cheilocystidia. However, under
the constraint of strict monophyly of recognisable groups, ICG 18 can not be maintained, since
16 and 17 can not be separated. Moreover, strains of ICG 17 itself form a paraphyletic group.
296 PERSOONIA — Vol. 18, Part 3, 2004 Aanen & Kuyper: Biological and phenetic species concept in Hebeloma crustuliniforme 297
Fig. 2. Phylogenetic relationships in the Hebeloma crustuliniforme complex based on nuclear ribos-
omal ITS sequences. For clade IIa and strains 9624, 9688 and 9650 IGS sequences were determined
as well and these sequences were also used in this phylogenetic analysis. The two clades I and II
were analysed separately, but are placed in the same gure. Indicated are bootstrap values and decay
indices (preceded by d).
296 PERSOONIA — Vol. 18, Part 3, 2004 Aanen & Kuyper: Biological and phenetic species concept in Hebeloma crustuliniforme 297
Fig. 3. Consecutive steps for combining ICGs into monophyletic units. If two sister groups can not be
separated, this is indicated with *, if two sister groups can be separated, they are combined and a letter
is given to the provisional taxon. These steps are repeated at the next level. This analysis ultimately
leads to the separation of two morphologically recognisable monophyletic groups, one consisting of
three ICGs (clade I), and one consisting of 19 ICGs (clade II).
298 PERSOONIA — Vol. 18, Part 3, 2004 Aanen & Kuyper: Biological and phenetic species concept in Hebeloma crustuliniforme 299
= 245, that could be rejected in favour of the unconstrained trees (Kishino-Hasegawa
test: p < 0.05, Templetonʼs non parametric test: p < 0.1). We conclude that the ICGs of the
morphospecies H. pusillum do not form a monophyletic group and that H. pusillum can
not be maintained as a valid phenetic species. A parsimony analysis on the data set with
the constraint that the two ICGs of H. lutense formed a monophyletic group (ICG 14 and
15) gave 60 trees of length = 255. These trees could be rejected in favour of the uncon-
strained trees (Kishino-Hasegawa test and Templetonʼs non parametric test: p < 0.01
in both cases). We conclude that the ICGs of the morphospecies H. lutense do not form
a monophyletic group and that H. lutense can not be maintained as a valid species.
DISCUSSION
Most biological species could not be recognized using morphological characters. Only
ICG 18 could be separated unambiguously from the remaining taxa. ICG 18 is accepted
as the phenetic species H. incarnatulum (Smith, 1984). The remaining ICGs could not
be uniquely characterised, implying that cryptic biological species are the rule within
the H. crustuliniforme complex. Similar observations have been made in other species
groups such as the Corticiaceae (Hallenberg, 1991) and the genera Paxillus (Fries,
1985) and Laccaria (Mueller, 1991; Mueller & Gardes, 1991), although in some cases
morphological differences could be found between ICGs. The biological species of
the H. crustuliniforme complex are also not meaningful ecological entities. Most col-
lections (70%) of H. crustuliniforme and H. helodes were made with members of the
Salicaceae as the ectomycorrhizal host tree while most collections of H. velutipes (93%)
were made with other trees as ectomycorrhizal hosts. As the rst group consisted of 19
biological species, with very short branch length in the molecular phylogeny, we conclude
that extensive speciation took place after a host switch. Extensive speciation after host
switches is apparently not uncommon. The same phenomenon has been reported for Suil-
lus (Kretzer et al., 1996) and Leccinum (H. den Bakker, pers. comm.). It is also known to
occur in several species complexes in the genus Lactarius (three spp. of the L. obscuratus
complex associated with Alnus; four spp. of the L. torminosus complex associated with
Betula (Molina et al., 1992)). Interestingly, extensive speciation after host switches to the
Salicaceae has occurred at least in three clades in Hebeloma, viz. the H. crustuliniforme /
H. helodes clade, the H. mesophaeum complex (Vesterholt, 1989) and the H. sacchariolens
complex (Gröger & Zschieschang, 1981). One would be tempted to speculate whether
there is a causal relationship between the host switch to Salicaceae and fungal speciation.
Salicaceae belong to the monophyletic order Malphigiales. All families in this order form
arbuscular mycorrhiza, except Salicaceae that form predominantly ectomycorrhiza. The
most parsimonious explanation for this pattern of host tree colonisation is that Salicaceae
have been colonised by ectomycorrhizal fungi separately and probably relatively recently.
The colonisation of this new and empty niche may have been the factor that favoured rapid
speciation. Rapid speciation in the genus Hebeloma may also be related to a recent eco-
logical switch from saprotrophy to the mycorrhizal symbiosis. Both the genus Hebeloma
and its sister group Alnicola (Aanen et al., 2000a; Peintner et al., 2001) contain species
that live as ectomycorrhizal symbionts and species that have (maintain) a saprotrophic life
style, e. g. on old re places. The sister group of that clade may be the saprotrophic genus
Agrocybe (Aanen et al., 2000a; Moncalvo et al., 2000).
298 PERSOONIA — Vol. 18, Part 3, 2004 Aanen & Kuyper: Biological and phenetic species concept in Hebeloma crustuliniforme 299
Since most of the biological species in this species complex could not be recog-
nized, we tested if monophyletic (or paraphyletic, but not polyphyletic) combinations
of biological species could be recognized by morphological criteria. Crucial in this ap-
proach was to have a reliable estimate of the phylogeny of the biological species. The
two nuclear data sets were shown to be not signicantly different and were therefore
combined. The mitochondrial based phylogeny, however, was shown to be signicantly
different. We used the nuclear phylogeny as our best guess for the phylogeny of the
ICGs, but we considered the mitochondrial based tree as well. If paraphyletic taxa are
recognised, three taxa can be recognised. Under the constraint that the ve ICGs of
H. crustuliniforme (ICGs 1, 2, 3, 4 and 5) form a monophyletic group, trees are found
that can not be rejected against the most parsimonious tree. The idea that the ve ICGs
of H. crustuliniforme form a monophyletic group is in agreement with the mitochon-
drial tree.
After evaluating the pros and cons of a phenetic species concept that allows inclusion
of paraphyletic groupings of biological species and groupings of biological species that
can not be rejected as belonging to a monophyletic group, we decided to accept four
phenetic species in the H. crustuliniforme complex. They are keyed out as follows.
KEY TO FOUR RECOGNISED SPECIES OF THE H. CRUSTULINIFORME COMPLEX
1a. Spores distinctly, often rather strongly dextrinoid (D2 – D4), ellipsoid to oblong
(Qav ≤ 1.7); cheilocystidia cylindrical to cylindrico-clavate, but sometimes with
bid apex; stipe usually distinctly bulbous, occulose; generally associated with
Pinaceae, Betulaceae, Carpinaceae and Fagaceae . . . . . . . . . . . . . . . . . . . . . . 2
b. Spores not to weakly dextrinoid (D0 – D1(– D2)), oblong to fusiform (Qav ≥ 1.7);
cheilocystidia clavate to subcapitate, never with bid apex; stipe usually cylin-
drical to clavate to subbulbous, often (coarsely) occose; generally associated
with Salicaceae . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
2a. Cheilocystidia cylindrico-clavate, in upper part on average more than 6.0 µm broad
(6.2 –10.2 µm), and Qav = 1.2 – 2.0 . . . . . . . . . . . . . . . . . . . . . . . . . . . 1. H. velutipes
b. Cheilocystidia cylindrical, in upper part on average less than 6.0 µm broad (5.6 µm),
and Qav = 1.1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2. H. incarnatulum
3a. Stipe 2 –10.5 mm broad, minutely occulose to (sub)occose, distinctly darkening
from base upwards; pileus 13 –75 mm, usually with straight margin when young,
yellowish to red-brown, often distinctly paler towards margin and then ± bicolorous;
cheilocystidia usually (sub)capitate . . . . . . . . . . . . . . . . . . . . . . . . . . . 3. H. helodes
b. Stipe 6 –14 mm broad, coarsely occose, white, not or hardly darkening from base
upwards; pileus 35 –170 mm, with involute margin when young, whitish to yellow-
ish, ± unicolorous, cheilocystidia clavate to slightly subcapitate
4. H. crustuliniforme
General note
This paper does not (and could not) aim at a full taxonomic revision of all taxa in this
complex. It could not do so, because the biological species concept can not be applied
to type collections (unless there exist ex-type cultures). Therefore our nomenclator will
inevitably be incomplete. We only list major names that have been used recently and
300 PERSOONIA — Vol. 18, Part 3, 2004 Aanen & Kuyper: Biological and phenetic species concept in Hebeloma crustuliniforme 301
comment upon the biological (what is the relationship between biological species and
morphological taxa?), phylogenetic (are morphospecies that are commonly mentioned,
mono-, para- of polyphyletic?) and morphological (how well are different morpho-
species in common use separated from each other?) quality of species names in com-
mon use.
Hebeloma velutipes Bruchet
Hebeloma velutipes Bruchet, Bull. mens. Soc. linn. Lyon (Suppl.) 39 (1970) 127.
Hebeloma bulbiferum Maire, Publ. Inst. bot. Barcelona 10 (3) (1937) 108. — Hebeloma crustu-
liniforme var. bulbiferum (Maire) J. Favre, Ergebn. wiss. Unters. schweiz. NatParks, NF 6 (1960) 488
(invalid, basionym not cited).
Hebeloma bulbosum Romagn., Sydowia 36 (1983) 263, non H. bulbosum Fayod 1893. — Hebeloma
favrei Romagn. & Quadr., Doc. Mycol. 14 (56) (1985, ʻ1984ʼ) 31.
Misapplied. Hebeloma crustuliniforme sensu auct.; Hebeloma longicaudum sensu J. E. Lange, Fl.
agar. dan. 3 (1938) 95; sensu P. J. Keizer & Arnolds, Persoonia 16 (1995) 92; sensu Bruchet, Bull. mens.
Soc. linn. Lyon (Suppl.) (1970) 77 (see notes); Hebeloma leucosarx sensu Vesterh., Svampe 25 (1992)
16 ; Symb. bot. Upsal. 30 (3) (1995) 136 (see notes).
Pileus to 32 –78 mm, convex to applanate, without umbo to rather distinctly umbonate,
dry, slightly to distinctly viscid, sometimes seemingly hygrophanous with irregular spots,
in centre red-brown, (pale) yellow-brown to pale ochraceous yellow (Mu. 5 YR 4 – 5 / 3,
10 YR 4 – 6 / 4, 5 – 6 / 6, 2.5 Y 7– 8 / 2 – 4, 10 YR 7– 8 / 4 – 6), uniformly coloured (especially
in paler specimens) to ± distinctly paler outwards and at margin sometimes even whitish.
Lamellae, L = 40 –70, l = 3 –7, thin, (very) crowded, rather broadly to narrowly adnate,
to 8 mm, not ventricose to subventricose, ochraceous buff to brownish ochraceous (Mu.
10 YR 7/ 2 – 3 to 6 / 3 – 4); edge mbriate, whitish; weeping (but sometimes not distinctly
so). Stipe to 34 –120 × 5 –10 mm, Q = 5.3 –12, shorter to longer than diameter of pileus,
usually ± distinctly bulbous (to 20 mm), sometimes (sub)clavate or even equal, stulose,
with pendent narrow strand, sometimes solid, whitish, discolouring to brownish on dam-
age from base upwards, (sub)occulose to subocculose, especially in upper part. Context
thick, rm, white. Smell raphanoid.
Spores (9.5–)10.0–13.0 × 6.0–7.5 µm, on average 10.4–11.9 × 6.3–7.2 µm, Q = 1.5 –
1.8(–1.9), Qav = 1.57–1.80, weakly to distinctly dextrinoid (D2 – D4), regular to subamyg-
daliform, exceptionally sublimoniform; perispore not or very slightly loosening (P0 – P1);
almost smooth or slightly to rather distinctly verruculose (O1– O3). Cheilocystidia (36 –)
40 – 87(–106) × 4 –7(– 8) × 6 –13 µm, on average 45.5 –72.2 × 4.5 – 6.3 × 6.2 –10.2 µm, Q
= (1.0 –)1.2 – 2.2(– 2.8), Qav = 1.2 – 2.0, straight to exuose, subcylindrical to subclavate,
usually not distinctly enlarged apically, but exceptionally tending to subspathuliform or
subcapitate, exceptionally also subcylindrical and not swollen towards apex, sometimes
slightly swollen in basal part and then slenderly subutriform, thin-walled to slightly thick-
walled, sometimes bid in apical part in varying frequency (absent to fairly common, and
then apex to 19 µm broad).
Habitat — Associated with various deciduous and coniferous trees (Betula, Fagus,
Quercus, Carpinus, Corylus, Picea, Pinus), only very exceptionally in the vicinity of
Salix.
300 PERSOONIA — Vol. 18, Part 3, 2004 Aanen & Kuyper: Biological and phenetic species concept in Hebeloma crustuliniforme 301
Notes — 1. Hebeloma velutipes is well recognized by relatively broad, dextrinoid
spores and clavate, non-capitate cheilocystidia. Many collections also show a number
of the following characters: pileus with hygrophanous spots; stipe distinctly bulbous
(H. bulbiferum Maire); stipe hollow with pendent marrow strand; part of the cheilo-
cystidia with bid apex. The taxon is accepted as a paraphyletic phenetic species, con-
sisting of ICG 16 and 17. Hebeloma leucosarx sensu Vesterholt and H. longicaudum
sensu Keizer & Arnolds are identical (ITS-RFLP patterns of both taxa studied, D. K.
Aanen, unpubl. obs.).
2. In the literature on ectomycorrhizal fungi the name H. crustuliniforme is very
repeatedly encountered. It is likely that many, if not most, of these cultures actually
refer to H. velutipes.
3. Hebeloma longicaudum has originally been characterised by a pale pileus and
a long stipe. However, the case of H. pusillum serves as a warning with regard to the
taxonomic value of habit characters. Descriptions of several pale-coloured ICGs also
indicate that habit can be very variable. The name is therefore considered a nomen ambi-
guum.
4. Hebeloma leucosarx was described by Orton (1960) as a species with relatively
slender spores, distinctly capitate cheilocystidia and associated with Salix. On the
basis of that description it has been considered a member of the H. helodes complex
by Dutch mycologists. Vesterholt (1995), however, noted that the holotype had dis-
tinctly dextrinoid spores and non-capitate cheilocystidia, which makes this collection
a member of the H. velutipes clade. Vesterholt also suggested that H. velutipes might
be identical. We have not studied the type. Considering the divergent interpretations
of the name H. leucosarx, the name is not accepted in our paper; instead we con-
tinue to use the name H. velutipes. Hebeloma leucosarx sensu auct. neerl. belongs to
H. helodes (q. v.).
5. Hebeloma fragilipes Romagn. was dened on the basis of the shape and median
wall thickening of the cheilocystidia. The microscopical characters mentioned by
Vesterholt (1992, 1995) on the basis of a large number of collections from a number
of European countries, suggest both elements of H. velutipes (spores distinctly dextri-
noid, cheilocystidia that are not much swollen apically) and H. helodes (spores index-
trinoid, oblong to fusiform). Slightly thick-walled cheilocystidia have been observed in
both ICG 16 and 17 (H. velutipes) and in various ICGs of the H. helodes complex (also
mentioned by Vesterholt); our collections with a somewhat thickened cheilocystidial
wall (in median or apical part) were completely interfertile with specimens with thin-
walled cystidia, so we think that at least some doubt exists whether this morphospecies
could be maintained. No collections have been made by us that exactly t Vesterholtʼs
descriptions, so we refrain from a conclusion about its taxonomic status.
Hebeloma incarnatulum A. H. Sm.
Hebeloma incarnatulum A. H. Sm., Sydowia 37 (1984) 280.
Hebeloma bryogenes Vesterholt, Windahlia 20 (1993) 55.
302 PERSOONIA — Vol. 18, Part 3, 2004 Aanen & Kuyper: Biological and phenetic species concept in Hebeloma crustuliniforme 303
Pileus to 60 mm, convex to almost applanate, with a low broad umbo, very viscid,
uniformly yellow-brown (Mu. 10 YR 7– 8 / 4 – 6). Lamellae, L = 55, l = 1– 3, thin, nor-
mally crowded, to 5 mm, not ventricose, broadly adnate, ochraceous (10 YR 7/ 2 – 3); edge
mbriate, whitish; weeping. Stipe to 110 × 7 mm, Q = 15.7, longer than diameter of pileus,
bulbous (to 20 mm), stulose with pendent marrow strand, white, nely occulose. Con-
text thin, rm, white. Smell raphanoid.
Spores (10.0 –)10.5 –11.5(–12.0) × (6.0 –)6.5 –7.0 µm, on average 10.9 × 6.5 µm, Q =
1.6 –1.7(–1.8), Qav = 1.67, distinctly dextrinoid (D3 – D4), regular to subamygdaliform,
not sublimoniform; perispore not loosening (P0); distinctly verruculose (O2 – O3).
Cheilocystidia (45 –)46 – 59(–72) × (4 –)5 – 6 × 5 – 6(–7) µm, on average 54.5 × 5.0 × 5.6
µm, Q = 1.0 –1.2(–1.3), Qav = 1.1, cylindrical, partly somewhat inated in basal part and
then subventricose-slenderly utriform, near apex not or hardly inated, not clavate, thin-
walled.
Habitat — Associated with Pinus among living Sphagnum.
Note — Only a single collection was studied of ICG 18. The differences with
H. velutipes are rather subtle (narrower cheilocystidia). Possibly H. incarnatulum has
also a slightly different ecology (natural moist forests with Sphagnum).
Hebeloma helodes J. Favre
Hebeloma helodes J. Favre, Beitr. Krypt.-. Schweiz 10 (3) (1948) 214.
Hebeloma hiemale Bres., Fung. trident. 2 (1892) 52.
Hebeloma pusillum J. E. Lange, Fl. agar. dan. 5 (1940) iv.
Hebeloma cavipes Huijsman, Persoonia 2 (1961) 97.
Hebeloma lutense Romagn., Bull. trimest. Soc. mycol. Fr. 81 (1965) 342.
Hebeloma oculatum Bruchet, Bull. mens. Soc. linn. Lyon 39 (Suppl.) (1970) 126.
Hebeloma pusillum var. longisporum Bruchet, Bull. mens. Soc. linn. Lyon 39 (Suppl.) (1970)
126.
Misapplied. Hebeloma leucosarx sensu auct. Neerl. (see notes).
Excluded. Hebeloma helodes sensu Keizer & Arnolds, Persoonia 16 (1995) 88 (= H. velutipes).
Pileus 13 –75 mm, plano-convex to applanate, nally even slightly depressed, with or
without umbo, margin sometimes subinvolute, viscid to rather dry, subshiny, two-coloured
and in centre reddish ochraceous to (dark) reddish brown (Mu. 2.5 – 5 YR 3 / 2, 5 YR
3 – 4 / 4, 5 / 6, 7.5 YR 4 / 2 – 4, 5 / 4, 5 / 6, 10 YR 6 –7/ 6, 5 / 4 – 6, 3 – 4 / 3), outwards slightly to
distinctly paler, at margin slightly paler to whitish or rather uniformly coloured and paler,
pale yellow-brown or pale yellow (10 YR 7– 8 / 3, 2.5 Y 8 / 2 – 4, 2.5 Y 7/ 8). Lamellae, L =
25 –70, l = 1– 5(–7), thin, normally crowded, sometimes very crowded, to 6 mm broad,
rather narrow to subventricose, broadly to narrowly adnate or emarginate, pale brown to
(greyish) ochraceous brown (10 YR 4 / 4, 5 / 4, 6 / 3 – 4, 7/ 2 – 3) edge mbriate, whitish;
weeping, sometimes only indistinctly. Stipe 18 – 90 × 2 –10.5 mm, Q = 2.4 – 25, shorter
to (much) longer than diameter of pileus, equal to slightly swollen, sometimes clavate to
subbulbous (to 10 mm), solid, with age stulose, whitish above, somewhat to distinctly
darkening from base upwards, yellow-brown to brown in lower part, especially with age,
(minutely) occulose or even (sub)occose, especially in upper part, in lower part more
brillose, sometimes occulose over whole length. Context thin to thick in larger spec-
imens, whitish to (pale) brownish buff. Smell raphanoid, sometimes weakly so.
302 PERSOONIA — Vol. 18, Part 3, 2004 Aanen & Kuyper: Biological and phenetic species concept in Hebeloma crustuliniforme 303
Spores 8.5 –17.0 × 5.0 – 8.0(– 9.0) µm, on average 9.0 –14.9 × 5.0 –7.2 µm, Q = (1.5 –)
1.6 – 2.1(– 2.2), Qav = 1.64 – 2.03, not to weakly dextrinoid (D0 – D2), subamygdaliform, a
few tending to sublimoniform; perispore not or slightly loosening (P0 – P1); almost smooth,
weakly ornamented to distinctly verruculose (O1– O3(– O4)). Cheilocystidia (34 –)36 –
78(– 80) × 3 – 6 × (5 –)6 –16(–17) µm, on average 41.0 – 64.0 × 3.7– 5.1 × 5.9 –13.3 µm,
Q = 1.2 – 3.8(– 4.0), Qav = 1.4 – 3.0, straight, but in some collections exuose, (sub)cylindri-
co-(sub)clavate to cylindrico-subspathuliform or cylindrico-subcapitate, sometimes rather
conspicuously so, but sometimes only hardly swollen apically, thin-walled or with slightly
to distinctly thickened yellowish or brownish wall in upper part, especially in subcapitate
cheilocystidia, exceptionally slightly thick-walled in median part or throughout.
Habitat — Usually associated with Salix, sometimes with other deciduous trees (Popu-
lus, Quercus, Betula, Fagus, Tilia), exceptionally with conifers (Picea, Pinus).
Notes — 1. Hebeloma helodes was originally described as a taxon very close to
H. pusillum (same habit, same size, and same colour). Subsequent authors have
gradually enlarged this circumscription (H. helodes sensu Vesterholt, 1995, is paler) or
misinterpreted the name (H. helodes sensu Keizer & Arnolds). Bruchet (1970) did not
treat H. helodes. On the basis of our cladogram H. helodes is accepted as the name for
a paraphyletic grouping, as a sister group to H. crustuliniforme, which is considered
a separate phenetic species. Species circumscription of H. helodes is still quite broad,
which is consistent with the relatively high amount of molecular variation (compared
to H. crustuliniforme). Intermediates between H. helodes and H. crustuliniforme
could possibly occur (cf. H. cavipes, note 3). Considering the wide circumscription of
H. helodes, it becomes inevitable that H. pusillum and H. lutense have to be included;
this conclusion is not surprising, considering the enlarged description of H. helodes by
Vesterholt. As a consequence, variation in morphology and microscopical characters
is substantial. The phenetic species is comprised of ICGs 6, 7, 8, 9, 10, 11, 12, 13, 14,
15, 19, 20, 21, 22.
2. The name H. helodes is not the oldest name available. We decided to use it as a
name that is less liable to confusion over its application and circumscription than e. g.
H. pusillum or H. hiemale.
3. Hebeloma cavipes was accepted by Vesterholt (1995) as a valid species, only
known from the type collection in an open vegetation under Cistus. Vesterholt noted
that in its microscopical characters it was very similar to both H. crustuliniforme and
H. helodes, and these three taxa could only be kept apart on the basis of size, hollow
stipe (also a supposed characteristic of H. alpinum, see under H. crustuliniforme),
and cheilocystidia that are often widened in the basal part. As the species is known
from one collection only, it is likely therefore that nding additional collections of
H. cavipes could either collapse the distinction with H. crustuliniforme (if somewhat
larger specimens were found) or with H. helodes (if smaller specimens were found).
Vesterholt further suggested that H. cavipes could be conspecic with H. lutense (note
6), a species that is usually regarded as having a darker pileus and for that reason was
not treated in Vesterholt (1995). A culture collection of H. cavipes had the same ITS
sequence as collections of H. hiemale, H. lutense (ICG 15), and H. helodes (ICG 10).
The taxon is therefore accepted as a synonym of H. helodes.
304 PERSOONIA — Vol. 18, Part 3, 2004 Aanen & Kuyper: Biological and phenetic species concept in Hebeloma crustuliniforme 305
4. Hebeloma hiemale was not treated by Vesterholt. According to Bruchet (1970) it
is a xerophilous species, usually associated with Cistus in the Mediterranean region,
although he also cited a note from Huijsman that the species is common in the Dutch
dunes (which raises suspicion that it is very close to H. lutense, see note 6). On the basis
of the species description, it seems to come very close to H. helodes. The ITS sequence
of a culture collection of H. hiemale was identical to those of H. cavipes, H. lutense
(ICG 15) and H. helodes (ICG 10). The taxon is therefore accepted as a synonym of
H. helodes.
5. Hebeloma leucosarx was described by Orton (1960) as a species with relatively
slender spores, distinctly capitate cheilocystidia and associated with Salix. On the basis
of that description it has been considered a member of the H. helodes complex by Dutch
mycologists. Vesterholt (1995), however, noted that the holotype had distinctly dextri-
noid spores and non-capitate cheilocystidia, which makes this collection a member of
the H. velutipes clade. Vesterholt also suggested that H. velutipes might be identical.
We have not studied the type. Considering the divergent interpretations of the name
H. leucosarx, the name is not accepted.
6. In the Netherlands H. lutense (= H. leucosarx sensu auct. neerl.) is recognized
mainly by habit (relatively stout specimens, compared to H. pusillum) and habitat (usu-
ally associated with Salix repens often in early successional, relatively open sites). The
rst character may reect an adaptation to the relatively open structure (and hence drier
microclimate) of S. repens vegetation. The species also differs from several interpreta-
tions of H. helodes by darker colours. Collections that t the description of H. lutense
belong to two different ICGs. ITS sequence of ICG 15 is similar to that of H. cavipes,
H. hiemale, and H. helodes (ICG 10). A constrained tree in which both ICGs with the
characteristics of H. lutense were combined, had to be rejected. Morphological vari-
ability within some of the ICGs also suggests that H. lutense can neither be separated
from H. helodes nor from H. pusillum. Vesterholt (1995) suggested that H. lutense could
be a synonym of H. cavipes (note 2).
7. The name H. pusillum is used for small and slender specimens with a dark pileus,
associated with Salix. The species is somewhat variable in spore size. This species also
got a slightly enlarged circumscription, e. g. by Bruchet (1970; where it almost certainly
includes H. helodes) and Phillips (1981; which also seems to t better into the concept of
H. helodes). H. pusillum consists of four different ICGs. A tree, constrained to make
Hebeloma pusillum a monophyletic entity, must statistically be rejected against the most
parsimonious trees, indicating that the dening characters of H. pusillum (slender habit
with small basidiocarps) have likely arisen repeatedly.
8. Hebeloma fragilipes Romagn. was dened on the basis of the shape and median
wall thickening of the cheilocystidia. The microscopical characters mentioned by
Vesterholt (1992, 1995) on the basis of a large number of collections from a number of
European countries, suggest both elements of H. velutipes (spores distinctly dextrinoid,
cheilocystidia that are not much swollen apically) and H. helodes (spores indextrinoid,
oblong to fusiform). Slightly thick-walled cheilocystidia have been observed in both ICG
16 and 17 (H. velutipes) and in various ICGs of the H. helodes clade (also mentioned
by Vesterholt); our collections with a somewhat thickened cheilocystidial wall (in
median or apical part) were completely interfertile with specimens with thin-walled
cystidia, so we think that at least some doubt exists whether this morphospecies could
304 PERSOONIA — Vol. 18, Part 3, 2004 Aanen & Kuyper: Biological and phenetic species concept in Hebeloma crustuliniforme 305
be maintained. No collections have been made by us that exactly t Vesterholtʼs descrip-
tions, so we refrain from a conclusion about its taxonomic status.
Hebeloma crustuliniforme (Bull.) Quél.
Agaricus crustuliniformis Bull., Herb. Fr. (1787) pl. 308; Hebeloma crustuliniforme (Bull.) Quél.
in Mém. Soc. Emul. Montbéliard, sér. II, 5 (1872) 128.
Hebeloma populinum Romagn., Bull. trimest. Soc. mycol. Fr. 81 (1965) 326.
Hebeloma crustuliniforme var. alpinum J. Favre, Ergebn. wiss. Unters. schweiz. NatParks, NF 5
(1955) 121; Hebeloma alpinum (J. Favre) Bruchet, Bull. mens. Soc. linn. Lyon 39 (Suppl.) (1970) 68.
Hebeloma ochroalbidum Bohus, Annls hist.-nat. Mus. nat. hung. 64 (1972) 71.
Hebeloma crustuliniforme var. tiliae Bresinsky, Z. Mykol. 53 (1987) 294.
Excluded. Hebeloma crustuliniforme sensu auct. (= H. velutipes).
Pileus 35 –170 mm, convex to applanate, without umbo or with indistinct umbo, mar-
gin (sub)involute, (slightly) viscid when moist, rather pale, in centre pale yellow to pale
yellow-brown (Mu. 10 YR 7– 8 / 3; 2.5 Y 6 –7– 8 / 2 – 4–6), but sometimes more brownish
(10 YR 5 – 6 / 4 – 6 to 4 / 4), paler towards outer part or rather uniformly pale ochraceous
yellow, at margin whitish or white. Lamellae, L = 45 –100, l = 1– 3 – 5, thin, (very) crowded,
to 8 mm, subventricose, narrowly adnate to emarginate, ochraceous to pale grey-brown
(10 YR 7/ 3 – 6 / 3 – 4); edge mbriate, whitish; weeping. Stipe 23 –115 × 6 –14 mm,
Q = 2.1–11.5, usually shorter to longer than diameter of pileus, at base somewhat clavate
(to 16 mm) to (almost) equal, solid, stulose with age, white, coarsely occose, especially
in upper part. Context thick in pileus, rm, white. Smell raphanoid, sometimes mixed with
a sweetish component.
Spores (9.5 –)10.0 –13.0(–14.0) × (5.0 –)5.5 –7.5 µm, on average 10.3 –12.5
× 5.8 –7.1 µm, Q = 1.6 – 2.1, Qav = 1.68 –1.92, not dextrinoid, sometimes indis-
tinctly dextrinoid (D0 – D1(– D2)), regular to subamygdaliform, not or exceptionally
tending to sublimoniform; perispore not (or very slightly) loosening (P0 –P1); almost
smooth to distinctly verruculose (O1– O2(– O3)). Cheilocystidia (36 –)43 –77(– 90) ×
(3 –)4 – 6 × (6 –)7–12(–14) µm, on average 49.8 – 66.4 × 3.8 – 4.8 × 7.6 –10.3 µm, Q =
(1.2 –)1.4 – 2.8(– 3.0), Qav = 1.7– 2.2, (slenderly) (sub)cylindrico-(sub)clavate, exception-
ally cylindrical, gradually broadened towards apex, but only a (small) minority tending
to subcapitate and then apical part more distinctly enlarged, thin-walled, exceptionally
slightly thick-walled in upper part and slightly refringent.
Habitat — Mainly associated with Salix and Populus, but also with Tilia, Betula, Cory-
lus, Quercus, or Dryas and Helianthemum. Occurring from the lowland to the alpine
region.
Notes — 1. Hebeloma crustuliniforme as presently conceived by Vesterholt (1995) is
maintained as an autonomous species, after the criteria of monophyly and parsimony are
relaxed to include paraphyletic groupings of ICGs and less parsimonious trees that can-
not be rejected against the most parsimonious trees. Moreover, the mitochondrial tree
supported the monophyly of ICGs 1, 2, 3 and 4 of H. crustuliniforme and did not exclude
the possibility that ICG 5 belonged to it as well. It consists of ICG 1, 2, 3, 4 and 5.
2. In the literature on ectomycorrhizal fungi the name H. crustuliniforme is very
repeatedly encountered. It is likely that many, if not most, of these cultures actually
refer to H. velutipes.
306 PERSOONIA — Vol. 18, Part 3, 2004 Aanen & Kuyper: Biological and phenetic species concept in Hebeloma crustuliniforme 307
3. The application of the name Agaricus fastibilis Pers. :Fr. was extensively discussed
by Kuyper & Vesterholt (1990). They concluded that the name, as originally conceived,
referred to a taxon very close to H. crustuliniforme. However, another interpretation
has been widespread in which H. fastibile is a cortinate species, also recently known as
H. mesophaeum var. crassipes (Vesterholt, 1989).
4. Hebeloma alpinum was originally described by Favre (1960) as a variety of
H. crustuliniforme. Favre invoked a number of morphological characters to delimit this
taxon (small habit, broader lamellae, hollow stipe) but wondered whether these characters
might just reect adaptations to the microclimatic conditions in the alpine zone. However,
as Favre did not observe specimens that were transitional between the alpine variant and
the typical H. crustuliniforme (actually, he did not even observe typical H. crustuliniforme
in the upper subalpine zone), he considered these differences to be genetically xed and
hence worthy of recognition on varietal status. Bruchet (1970) elevated the taxon to spe-
cies rank and gave an enlarged description of it (much larger size than Favreʼs taxon, taste
less bitter), whereas Vesterholt (1995), who also accepted species status, used a partly
different set of characters than Favre to keep this taxon apart (spores weakly dextrinoid,
in H. crustuliniforme indextrinoid or nearly so). Our alpine collections from this group
showed substantial variation in size. They belong to three different ICGs, two of which
contained both alpine and lowland collections (ICGs 1 and 2), and one of which (ICG 4)
was partially compatible with a lowland ICG (ICG 3). Partial intercompatibility between
these four ICGs furthermore strongly suggest that H. alpinum can not be maintained any
longer. Moreover, in a phylogeny between populations of these ICGs, the alpine popula-
tions did not form a distinct monophyletic group (Aanen et al., 2000b).
A CONSENSUS TAXONOMY FOR THE H. CRUSTULINIFORME COMPLEX?
Most of the characters used, traditionally and also in this study, in Hebeloma taxonomy,
are quantitative. Different types of characters can be recognised. First, there are characters
that are absolutely discriminating between taxa, one taxon always has state A, whereas
the other taxon always has state B. Such characters are rare in this group of ICGs. In fact
the only character of this type is the shape of the cheilocystidia, with a unique state for
H. incarnatulum. A second type of characters are characters that have a rare unique state.
The presence of such a state is informative and can be decisive to place a specimen in
a certain taxon, whereas its absence is uninformative. Examples of such a character are
cheilocystidia with a bid apex, or the presence of hygrophanous spots on the pileus. Both
characters have only been found in H. velutipes and never in H. helodes or H. crustulini-
forme. However, in H. velutipes these character states do not occur constantly. Therefore,
only the presence of this state is informative, whereas its absence is not. Some combinations
of characters are unique for certain species groups and can be used as well. The combi-
nation of characters, coded in Table III, is such a set. This combination of characters, of
which none is necessary, is jointly sufcient to assign a collection to the phenetic species
H. velutipes. The individual characters that comprise a set, however, do not necessarily
uniquely characterise phenetic species. Relatively broad spores, or slender, non-capitate
cheilocystidia occur in H. helodes and H. crustuliniforme, but the combination of these
306 PERSOONIA — Vol. 18, Part 3, 2004 Aanen & Kuyper: Biological and phenetic species concept in Hebeloma crustuliniforme 307
characters has never been encountered in this clade. The last type of characters we recog-
nise are statistical characters. For example, 74% of the collections of H. helodes and
H. crustuliniforme have been found with Salicaceae, whereas only 7% of the collections
of H. velutipes. The latter species usually occurs in association with members of Pinaceae,
Fagaceae and Betulaceae. Such characters can never be decisive themselves but can give
additional support for doubtful collections.
The lack of reliable qualitative characters may seem surprising, considering the
number of species described in that group. The almost complete similarity in micro-
scopical characters between the different intercompatibility groups is consistent with a
scenario of a slow evolution (or even stasis) of morphological characters. However, as
the molecular data show a high sequence similarity and consequently short branch length
between biological species, especially in the H. crustuliniforme / H. helodes clade, it is
more likely that the members of the group diverged only recently, with ITS sequences
evolving at slower rates than compatibility characteristics and micro-morphological
characteristics evolving at an even lower rate, while some macroscopical characters
are highly plastic.
A scenario of rapid speciation would at rst sight contradict the existence of well
recognised morphospecies such as H. alpinum, H. pusillum or H. lutense. However,
both latter taxa had to be rejected because phylogenetic trees, in which these taxa were
constrained to form a monophyletic group, performed signicantly worse. It is more
likely, that these morphospecies are recognised on the basis of plastic characters such as
habit (length / width ratio of stipe; ratio of pileus diameter and stipe length) and that these
characters reect more habitat conditions than genetically xed characters. A similar
explanation may be true for H. alpinum, where again differences in habit (small pileus,
short and thick stipe) could be more plastic than normally assumed. The high plasticity
and variability in macroscopical characters and the relative uniformity in microscopical
characteristics may ultimately be the explanation for our failure to recognise more than
four phenetic species with a minimal phylogenetic quality in this taxon complex.
The comparison of a biological and morphological species concept within a phylo-
genetic framework indicated that these various concepts can not be reconciled to pro-
duce an unambiguous, unique solution to the species problem in the H. crustuliniforme
complex. Even under the assumptions (which not all mycologists would accept!) that
i) the phylogeny estimate is sufciently accurate and ii) the value of morphological
characters has been exhaustively studied, the number of species ultimately depends
on the rules of the game (acceptability of paraphyletic groupings, acceptability of less
parsimonious trees that can not be rejected in favour of the most parsimonious tree). It
would therefore be necessary to seek consensus about these rules, so that these incom-
patible demands on taxonomy (species should really exist, be recognisable and have
minimal phylogenetic quality) can be sorted out and an acceptable solution can be found
for this complex. Between the Scylla of morphology (which would have forced us to
accept taxa that must statistically be rejected because of lack of phylogenetic quality)
and the Charybdis of the biological species concept (which would have forced us to
produce a taxonomy that can not be applied in daily practice), the recognition of four
species seems a workable alternative.
308 PERSOONIA — Vol. 18, Part 3, 2004 Aanen & Kuyper: Biological and phenetic species concept in Hebeloma crustuliniforme 309
ACKNOWLEDGEMENTS
Most of the work reported here was carried out while both authors were at the Biological Station
at Wijster. The investigations were supported by the Netherlands Organisation for Scientic Research
(NWO). We thank Eef Arnolds, Béatrice Senn-Irlet, Ivano Brunner, Frank Graf, Wim van der Sluijs,
and various members of the Netherlands Mycological Society (NMV) during the three collecting
seasons. We acknowledge stimulating discussions with Teun Boekhout, Rolf Hoekstra, Jan Vesterholt,
Greg Mueller, and Rytas Vilgalys about the way in which the results of an experimental taxonomic
investigation should be translated into a workable taxonomy. Mariska Oude Elferink provided support
in the lab.
REFERENCES
Aanen, D. K & T.W. Kuyper. 1999. Intercompatibility tests in the Hebeloma crustuliniforme complex
in northwestern Europe. Mycologia 91: 783 –795.
Aanen, D. K., T.W. Kuyper, T. Boekhout & R. F. Hoekstra. 2000a. Phylogenetic relationships in the
genus Hebeloma based on ITS1 and 2 sequences, with special emphasis on the Hebeloma crustu-
liniforme complex. Mycologia 92: 269 – 281.
Aanen, D. K., T.W. Kuyper, T. H. M. Mes & R. F. Hoekstra. 2000b. The evolution of reproductive iso-
lation in the ectomycorrhizal Hebeloma crustuliniforme aggregate (Basidiomycetes) in northwest-
ern Europe: a phylogenetic approach. Evolution 54: 1192 –1206.
Aanen, D. K., T.W. Kuyper & R. F. Hoekstra. 2001. A widely distributed ITS polymorphism within a bio-
logical species of the ectomycorrhizal fungus Hebeloma velutipes. Mycol. Res. 105: 284 – 290.
Bruchet, G. 1970. Contribution à lʼétude du genre Hebeloma (Fr.) Kummer. Bull. mens. Soc. linn.
Lyon 39 (Suppl.): 3 –132.
Bull, J. J., J. P. Huelsenbeck, C.W. Cunningham, D. L. Swofford & P. J. Waddell. 1993. Partitioning
and combining data in phylogenetic analysis. Syst. Biol. 42: 384 – 387.
De Queiroz, A. 1993. For consensus (sometimes). Syst. Biol. 42: 368 –3 72.
De Queiroz, K. & M. J. Donoghue. 1988. Phylogenetic systematics and the species problem. Cladistics
4: 317– 338.
De Queiroz, K. & M. J. Donoghue. 1990. Phylogenetic systematics and species revisited. Cladistics
6: 83 – 90.
Farris, J. S., M. Kallersjo, A. G. Kluge & C. Bult. 1995. Testing signicance of incongruence. Cla-
distics 10: 315 – 319.
Favre, J. 1960. Catalogue descriptif des champignons supérieurs de la zone subalpine du Parc National
Suisse. Ergebn. wiss. Unters. schweiz. NatParks 6 (NF): 321– 610.
Fries, N. 1985. Intersterility groups in Paxillus involutus. Mycotaxon 24: 403 – 409.
Geiser, D. M., J. I. Pitt & J.W. Taylor. 1998. Cryptic speciation and recombination in the aatoxin-
producing fungus Aspergillus avus. Proc. Natl. Acad. Sci. USA 95: 388 – 393.
Gröger, F. & G. Zschieschang. 1981. Hebeloma-Arten mit sacchariolens-Geruch. Z. Mykol. 47: 195 – 210.
Hallenberg, N. 1991. Speciation and distribution in Corticiaceae (Basidiomycetes). Pl. Syst. Evol.
177: 93 –110.
Hibbett, D. S., Y. Fukusama-Nakai, A. Tsuneda & M. J. Donoghue. 1995. Phylogenetic diversity in
shiitake inferred from nuclear ribosomal DNA sequences. Mycologia 87: 618 – 638.
Hillis, D. M., B. K. Mable & C. Moritz. 1996. Applications of molecular systematics: the state of the
eld and a look at the future. In: D. M. Hillis, C. Moritz & B. K. Mable (eds), Molecular systemat-
ics, 2nd ed.: 515 – 543. Sinauer Associates, Sunderland Mass.
Huelsenbeck, J. P., J. J. Bull & C.W. Cunningham. 1996a. Combining data in phylogenetic analysis.
Trends Ecol. Evol. 11: 152 –158.
Huelsenbeck, J. P., D. M. Hillis & R. Nielsen. 1996b. A likelihood-ratio test of monophyly. Syst. Biol.
45: 546 – 558.
Kishino, H. & M. Hasegawa. 1989. Evaluation of the maximum likelihood estimates of the evolu-
tionary tree topologies from sequence data, and the branching order in Homoidea. J. mol. Evol.
29: 170 –179.
308 PERSOONIA — Vol. 18, Part 3, 2004 Aanen & Kuyper: Biological and phenetic species concept in Hebeloma crustuliniforme 309
Kluge, A. G. 1989. A concern for evidence and a phylogenetic hypothesis among Epicrates (Boidae,
Serpentes). Syst. Zool. 38: 7– 25.
Kretzer, A., Y. Li, T. Szaro & T. D. Bruns. 1996. Internal transcribed spacer sequences from 38 rec-
ognized species of Suillus sensu lato: phylogenetic and taxonomic consequences. Mycologia 88:
776 –785.
Kuyper, T.W. 1988. Specic and infraspecic delimitation. In: C. Bas, T.W. Kuyper, M. E. Noordeloos
& E. C. Vellinga (eds.), Flora agaricina neerlandica 1. Balkema, Rotterdam.
Kuyper, T.W & J. Vesterholt. 1990. The typication of Agaricus fastibilis Pers.: Fr., the type species
of the genus Hebeloma (Fr.) Kumm. Persoonia 14: 189 –192.
Levin, D. A. 1993. Local speciation in plants: the rule not the exception. Syst. Bot. 18: 197– 208.
Levin, D. A. 2000. The origin, expansion, and demise of plant species. Oxford University Press,
Oxford.
Lutzoni, F. & R. Vilgalys. 1995. Integration of morphological and molecular data sets in estimating
fungal phylogenies. Can. J. Bot. 73 (Suppl.): S649 – S659.
Mes, T. H. M. 1995. Origin and evolution of Macaronesian Sempervivoideae (Crassulaceae). PhD
thesis, University of Utrecht.
Miyamoto, M. M. & W. M. Fitch. 1995. Testing species phylogenies and phylogenetic methods with
congruence. Syst. Biol. 44: 64 –76.
Molina, R. H., H. Massicotte & J. M. Trappe. 1992. Specicity phenomena in mycorrhizal symbioses:
community-ecological consequences and practical implications. In: M. F. Allen (ed.), Mycorrhizal
functioning, an integrative plant-fungal process: 347– 423.
Moncalvo, J. M., F. M. Lutzoni, S. A. Rehner, J. Johnson & R. Vilgalys. 2000. Phylogenetic relation-
ships of agaric fungi based on nuclear large subunit ribosomal DNA sequences. Syst. Biol. 49:
278 – 305.
Mueller, G. M. 1991. Laccaria laccata complex in North America and Sweden: intercollection pairing
and morphometric analyses. Mycologia 83: 578 – 594.
Mueller, G. M. & M. Gardes. 1991. Intra- and interspecic relations within Laccaria bicolor sensu
lato. Mycol. Res. 95: 592 – 601.
Orton, P. D. 1960. New check list of British agarics and boleti. 3. Notes on genera and species in the
list. Trans. Br. mycol. Soc. 43: 159 – 439.
Peintner, U., N. L. Bougher, M. A. Castellano, J. M. Moncalvo, M. M. Moser, J. M. Trappe & R. Vilga-
lys. 2001. Multiple origins of sequestrate fungi related to Cortinarius (Cortinariaceae). Amer. J.
Bot. 88: 2168 – 2179.
Phillips, R. 1981. Mushrooms and other fungi of Great Britain and Europe.
Rieseberg, L. H. & L. Brouillet. 1994. Are many plant species paraphyletic? Taxon 43: 21– 32.
Sanderson, M. J., A. Purvis & C. Henze. 1998. Phylogenetic supertrees: assembling the tree of life.
Trends Ecol. Evol. 13: 105 –109.
Smith, A. H. 1984. Studies of species of Hebeloma (Fr.) Kummer from the Great Lakes region of
North America I. Sydowia 37: 271– 283.
Templeton, A. R. 1983. Phylogenetic inference from restriction endonuclease cleavage site maps with
particular reference to the evolution of humans and the apes. Evolution 32: 221– 244.
Templeton, A. R. 1998. Species and speciation: geography, population structure, ecology, and gene
trees. In: D. J. Howard & S. H. Berlocher (eds.), Endless forms – species and speciation: 32 – 43.
Oxford University Press, Oxford.
Vesterholt, J. 1989. A revision of Hebeloma sect. Indusiata in the Nordic countries. Nord. J. Bot. 9:
289 – 319.
Vesterholt, J. 1992. Almindelig Tåreblad (Hebeloma crustuliniforme) og dens almindelige dobbelt-
gængere. Svampe 25: 15 – 22.
Vesterholt, J. 1995. Hebeloma crustuliniforme and related taxa – notes on some characters of taxonomic
importance. Symb. bot. Upsal. 30 (3): 129 –137.
Vilgalys, R. & B. L. Sun. 1994. Ancient and recent patterns of geographic speciation in the oyster
mushroom Pleurotus revealed by phylogenetic analysis of ribosomal DNA sequences. Proc. natl
Acad. Sci. USA 91: 4599 – 4603.
310 PERSOONIA — Vol. 18, Part 3, 2004 Aanen & Kuyper: Biological and phenetic species concept in Hebeloma crustuliniforme 311
APPENDIX
ICG 1
Pileus to 40 – 85 mm, convex to plano-convex, without umbo, margin involute, (slightly) vis-
cid when moist, rather pale, in centre pale yellow to pale yellow-brown (Mu. 10 YR 7– 8 / 3, 2.5
Y 6 –7/ 4), paler towards outer part, at margin whitish or white. Lamellae, L = 55 – 60, l = 1– 3(– 5),
thin, crowded, to 6.5 mm, subventricose, narrowly adnate to emarginate, ochraceous (10 YR
7/ 3); edge fimbriate, whitish; weeping. Stipe 35 – 50 × 7–14 mm, Q = 2.6 – 5.7 usually shorter
than diameter of pileus, but sometimes equal to or slightly longer than diameter of pileus, at base
somewhat clavate (to 16 mm) to almost equal, solid, but sometimes fistulose with age, white,
coarsely occose, especially in upper part. Context thick in pileus, rm, white. Smell raphanoid.
Spores (9.5 –)10.0 –13.0(–14.0) × (5.5 –)6.0 –7.0 µm, on average 10.5 –11.6 × 6.0 – 6.4 µm, Q =
1.7– 2.1, Qav = 1.74 –1.92, not dextrinoid (D0 – D1), regular to subamygdaliform, not or exception-
ally tending to sublimoniform; perispore not (or very slightly) loosening (P0 – P1); almost smooth to
indistinctly verruculose (O1– O2(– O3)). Cheilocystidia (41–)44 –70(– 80) × 4 – 6 × (7–)8 –12(–13)
µm, on average 56.1– 61.3 × 4.3 – 4.7 × 8.9 –10.3 µm, Q = (1.4 –)1.8 – 2.8(– 3.0), Qav = 2.0 – 2.2,
(sub)cylindrico – (sub)clavate, gradually broadened towards apex, but only a (small) minority tending to
subcapitate and then apical part more distinctly enlarged, thin-walled.
Associated with Salix (4 ×), Betula (1 ×) or Dryas (1 ×).
ICG 2
Pileus to 36 –170 mm, convex to applanate, without umbo or with indistinct umbo, margin
(sub)involute, slightly viscid when moist, in centre yellowish (Mu. 10 YR 7/ 3, 10 YR – 2.5 Y 7– 8 / 4 – 6),
but sometimes more brownish (10 YR 5 / 6 – 4 / 4), paler outwards, at margin whitish. Lamellae, L = 55 –70,
l = 3 – 5, thin, crowded, to 7 mm, subventricose, (very) narrowly adnate to emarginate, ochraceous brown
(10 YR 6 / 4); edge mbriate, whitish; weeping. Stipe to 25 –100 × 8 –13 mm, Q = 2.1–11.1 shorter to
longer than diameter of pileus, slightly swollen to clavate (to 16 mm), but sometimes equal, solid or
stulose, white, coarsely occose in upper part. Context thick, rm, white. Smell raphanoid.
Spores (9.5 –)10.0 –12.0(–12.5) × (5.0 –)5.5 – 6.5 µm, on average 10.3 –11.2 × 5.8 – 6.1 µm, Q =
1.6 –1.9(– 2.0), Qav = 1.76 –1.87, not dextrinoid (D0 – D1(– D2)), regular to subamygdaliform, not or
very exceptionally tending to sublimoniform; perispore not loosening (P0); nely to distinctly verruculo-
se (O1– O3). Cheilocystidia (36 –)45 –77(– 83) × (3 –)4 – 5(– 6) × 7–10(–14) µm, on average 50.3 – 66.4
× 3.8 – 4.5 × 7.6 – 8.6 µm, Q = (1.5 –)1.6 – 2.3(– 2.8), Qav = 1.7– 2.0, (slenderly) cylindrico-(sub)clavate,
very exceptionally tending to clavate-subcapitate with enlarged apical part, thin-walled, exceptionally
slightly thick-walled in upper part, colourless.
Associated with Salix (8 ×), Dryas & Helianthemum (1 ×) or Corylus (1 ×).
ICG 3
Pileus to 45 – 58 mm, plano-convex to applanate, without umbo to indistinctly umbonate, margin
involute when young, slightly viscid when moist, pale ochraceous yellow ( Mu. 2.5 Y 7– 8 / 2 – 4),
uniformly coloured or paler outwards and at margin whitish. Lamellae, L = 55 – 65, l = 3 – 5, thin, crowded,
to 8 mm, subventricose, narrowly adnate to emarginate, pale ochraceous (10 YR 7/ 3); edge mbriate,
whitish; weeping. Stipe to 45 – 60 × 8 –11 mm, Q = 4.4 – 9.8, usually longer than diameter of pileus, equal
to subclavate (13 mm), solid, with age becoming stulose, white, coarsely occose, especially in upper
part. Context thick in pileus, rm, white. Smell raphanoid.
Spores (10.5 –)11.0 –13.0 × 6.0 –7.5 µm, on average 11.3 –12.5 × 6.3 –7.1 µm, Q = 1.6 –1.9, Qav =
1.68 –1.81, not dextrinoid (D0 – D1), regular to (sub)amygdaliform, not to partly tending to sublimoniform,
perispore not to slightly loosening (P0 – P1); nely to distinctly verruculose. Cheilocystidia (40 –)43 –
74(– 90) × 4 – 6 × 7–11 µm, on average 49.8 – 60.6 × 4.5 – 4.8 × 8.3 – 9.4 µm, Q = (1.3 –)1.5 – 2.5, Qav =
1.8 – 2.0, (sub)cylindrico-(sub)clavate to (sub)clavate, broadened towards apex, exceptionally to partly
subclavate-subcapitate with enlarged apex, thin-walled or slightly thick-walled in upper part, colour-
less.
Associated with Populus (4 ×), Salix (2 ×) and Tilia (1 ×).
310 PERSOONIA — Vol. 18, Part 3, 2004 Aanen & Kuyper: Biological and phenetic species concept in Hebeloma crustuliniforme 311
ICG 4
Pileus to 42 –70 mm plano-convex to applanate, without umbo, margin involute, slightly viscid when
moist, pale yellow in centre (Mu. 2.5 Y 7– 8 / 2 – 4), but sometimes darker, to ochraceous or brownish
ochraceous (10 YR 5 – 6 / 4 – 6, 10 YR 6 / 4), outwards paler and at margin whitish. Lamellae, L = 45 –70,
l = 3 – 5, thin, crowded, narrowly adnate; ochraceous; edge mbriate, whitish; weeping. Stipe to 23 – 35
× 6 –12 mm, Q = 2.3 – 4.2 shorter than diameter of pileus, clavate (to 14 mm), white, coarsely occose.
Context thick, rm, white. Smell raphanoid, sometimes mixed with sweetish component.
Spores (10.5 –)11.0 –13.0(–13.5) × 6.0 –7.0(–7.5) µm, on average 11.3 –12.4 × 6.4 –7.0 µm, Q = (1.6 –)
1.7–1.9, Qav = 1.72 –1.83, not dextrinoid (D0 – D1), regular to subamygdaliform, not to partly tending
to sublimoniform; perispore not loosening (P0); slightly verruculose to almost smooth (O1– O2(– O3)).
Cheilocystidia (49 –)50 –70(–75) × 4 – 5 × (7–)8 –11 µm, on average 56.4 – 65.3 × 4.5 – 4.6 × 8.0 – 9.6
µm, Q = 1.6 – 2.5, Qav = 1.8 – 2.1, (sub)cylindrico-(sub)clavate, somewhat broadened towards apex, not
or exceptionally tending to (sub)capitate, thin-walled.
Associated with Salix (8 ×) and Dryas (3 ×).
ICG 5
Pileus to 45 – 60 mm, convex to applanate, without umbo or with an indistinct umbo, margin sometimes
involute, viscid when moist, in centre pale yellowish (Mu. 2.5 Y 7– 8 / 2 – 4) to ochraceous (10 YR 5/4 – 6),
outwards paler and at margin whitish to white. Lamellae, L = 80 –100, thin, (very) crowded, to 6 mm
broad, sometimes subventricose, narrowly adnate to slightly emarginate, ochraceous to pale grey-brown
(10 YR 6 / 3); edge mbriate, whitish; weeping. Stipe to 40 –115 × 10 –11.5 mm, Q = 3.6 –11.5 shorter to
longer than diameter of pileus, equal to subclavate (16 mm), solid to stulose, white, coarsely occose,
especially in upper part. Context rm, white. Smell raphanoid.
Spores (10.5 –)11.0 –12.5 × 6.0 –7.0 µm, on average 11.5 –11.9 × 6.1– 6.4 µm, Q = 1.7– 2.0, Qav =
1.81–1.90, not to indistinctly dextrinoid (D0 – D2), subamygdaliform, not tending to sublimoniform; peris-
pore not loosening (P0), rather weakly to ± distinctly verruculose (O1– O3). Cheilocystidia 40 –71(– 89) ×
(3 –)4 – 5 × (6 –)7–10(–11) µm, on average 58.7– 62.5 × 4.1– 4.6 × 7.9 – 8.5 µm, Q = (1.2 –)1.4 – 2.3(– 2.5),
Qav = 1.7– 2.1, cylindrico-clavate to subclavate, but sometimes almost cylindrical, a minority tending
to clavate-subcapitate, but in one collection not swollen at apex at all, thin-walled, colourless or with
slightly refringent wall.
Associated with Tilia (2 ×) or Quercus (2 ×), sometimes mixed with Corylus.
ICG 6
Pileus to 13 – 23 mm, plano-convex to applanate, with or without umbo, margin sometimes subinvo-
lute, viscid, subshiny, in centre reddish ochraceous to red-brown (Mu. 5 YR 5 / 6, 7.5 YR 5 / 4), outwards
slightly to distinctly paler, at margin slightly paler to whitish. Lamellae, L = 25 – 35, l = 1– 3(– 5), thin,
normally crowded, to 2 mm broad, narrowly adnate, ochraceous brown; edge mbriate, whitish; weeping.
Stipe to 30 – 33 × 2.5 – 3 mm, Q = 11–12 longer than diameter of pileus, equal, whitish above, somewhat
darkening downwards, especially with age, slightly occulose, especially in upper part. Context thin.
Smell weak, raphanoid.
Spores 12.0 –14.5(–15.0) × 6.0 –7.5 µm, on average 12.6 –13.5 × 6.5 –7.0 µm, Q = (1.7–)1.8 – 2.0(– 2.1),
Qav = 1.92 – 1.93, not dextrinoid (D0 –D1), subamygdaliform, not tending to sublimoniform; perispore not
loosening (P0); moderately weakly ornamented (O2). Cheilocystidia (50 –)53 –74(–75) × 4 –5 × (8 –)
9 –16(–17) µm, on average 60.2 – 64.0 × 4.1– 4.4 × 9.3 –13.3 µm, Q = 2.0 – 3.8(– 4.0), Qav = 2.3 – 3.0, cylin-
drico-subclavate to cylindrico-subspathuliform or cylindrico-subcapitate, sometimes rather conspicuously
so, thin-walled or with slightly thickened yellowish wall in upper part, especially in subcapitate cheilocystidia.
Associated with Salix (2 ×).
ICG 7
Pileus to 21– 31 mm, plano-convex to applanate, with or without umbo, viscid, in centre (dark) brown
(Mu. 7.5 YR 4 / 4 – 5 / 6), towards margin paler. Lamellae, L = 30 – 40, l = 3, thin, normally crowded,
emarginate, pale brown; edge mbriate, whitish; weeping. Stipe to 35 – 40 × 3.5 – 4 mm, Q = 10 equal to
longer than diameter of pileus, equal to slightly swollen, stulose, ochraceous, occulose. Context thin.
Smell raphanoid.
312 PERSOONIA — Vol. 18, Part 3, 2004 Aanen & Kuyper: Biological and phenetic species concept in Hebeloma crustuliniforme 313
Spores 10.0 –12.0(–12.5) × (5.5 –)6.0 –7.0 µm, on average 10.9 –11.3 × 6.1– 6.3 µm, Q = 1.6 –1.9(– 2.1),
Qav = 1.72 –1.84, not to weakly dextrinoid (D0 – D2), subamygdaliform, exceptionally tending to sub-
limoniform, perispore not or very slightly loosening (P0(– P1)); (sub)distinctly verruculose (O2 – O3).
Cheilocystidia 48 –79 × 4 – 5 × (6 –)7–12(–13) µm, on average 58.0 – 59.9 × 4.2 – 4.8 × 8.3 –10.3 µm,
Q = (1.4 –)1.5 – 2.5(– 2.6), Qav = 2.0 – 2.1, subcylindrico-(sub)clavate, partly more tending to (sub)capitate,
thin-walled or very slightly thick-walled in upper part.
Associated with Salix (2 ×).
ICG 8
Pileus to 18 – 25 mm, plano-convex, usually (sub)umbonate, but sometimes without umbo, viscid,
two-coloured, in centre red–brown (Mu. 5 YR 3 – 4 / 4), outwards paler, at margin pale yellow-brown
(Mu. 10 YR 7– 8 / 4) to whitish. Lamellae, L = 25 – 35, l = 1– 3, thin, normally crowded, to 3 mm broad,
rather narrow, narrowly adnate to emarginate, brownish ochraceous (10 YR 7/ 3); edge mbriate, whitish;
weeping. Stipe to 25 – 58 × 2 – 3 mm, Q = 10 – 25, equal to much longer than diameter of pileus, equal
to subclavate, soon stulose, initially whitish, on damage discolouring to (yellow-)brown from base
upwards, at apex (minutely) occulose, downwards slightly brillose. Context thin, rm, whitish to pale
brownish buff. Smell raphanoid.
Spores (10.5 –)11.0 –14.0(–15.0) × 5.5 –7.0(–7.5) µm, on average 11.8 –13.3 × 5.9 – 6.8 µm, Q =
(1.8 –)1.9 – 2.0(– 2.1), Qav = 1.93 –1.99, not to weakly dextrinoid (D0 – D2), subamygdaliform, none to
a few tending to sublimoniform; perispore not or slightly loosening (P0 – P1); weakly to distinctly ver-
ruculose (O1– O3). Cheilocystidia (36 –)39 –73(– 80) × 4 – 5(– 6) × (7–)8 –15 µm, on average 44.1– 60.9 ×
4.2 – 4.5 × 9.3 –12.2 µm, Q = (1.6 –)2.0 – 3.5(– 3.8), Qav = 2.1– 2.9, cylindrico-(sub)clavate, partly tending
to subspathuliform or subcapitate, thin-walled or with slightly thickened yellowish wall in apical part.
Associated with Salix (6 ×).
ICG 9
Pileus to 20 – 66 mm, plano-convex to applanate, usually not or hardly umbonate but sometimes more
distinctly umbonate, viscid, sometimes only slightly so, in centre orange ochraceous to reddish brown
(Mu. 10 YR 6 –7 / 6, 5 / 4 – 6), outwards paler, sometimes rather contrasting with centre of pileus and then
± bicoloured, at margin whitish to white. Lamellae, L = 30 – 45, thin, normally crowded, broadly to nar-
rowly adnate or emarginate, to 6 mm, subventricose, ochraceous (10 YR 6 / 3); edge mbriate, whitish;
(distinctly) weeping. Stipe to 26 – 90 × 3 –7.5 mm, Q = 6.4 –14, longer than diameter of pileus, equal, not
clavate or bulbous, solid but sometimes becoming stulose, white, occulose over whole length. Context
thick, rm, white to brownish. Smell raphanoid.
Spores (9.0 –)10.0 –12.0(–12.5) × 5.0 – 6.5 µm, on average 10.2 –11.4 × 5.5 – 6.1 µm, Q = 1.7– 2.0, Qav =
1.79 –1.88, not dextrinoid (D0 – D1(– D2)), regular to subamygdaliform, a few tending to sublimoniform;
peri-spore not or very slightly loosening (P0 – P1), (moderately) distinctly verruculose ((O1–)O2 – O3).
Cheilocystidia (39 –)40 –74 × 3 – 5 × 6 –12 µm, on average 50.5 – 58.6 × 3.7– 4.1 × 6.7– 9.1 µm, Q =
1.5 – 2.5(– 3.0), Qav = 1.8 – 2.2, often (conspicuously) exuose but sometimes straight, cylindrico-subcla-
vate, towards apex partly more (sub)spathuliform or subcapitate, but sometimes not or hardly broadened
towards apex, thin-walled.
Associated with Salix (5 ×).
ICG 10
Pileus to 50 mm, applanate to slightly depressed, without umbo, viscid, pale yellow (Mu. 2.5 Y 8 / 2 – 4),
outwards slightly paler. Lamellae, L = 63, l = 3, thin, (very) crowded, emarginate, ochraceous; edge
mbriate, whitish; weeping. Stipe to 75 × 8 mm, Q = 9.3, longer than diameter of pileus, equal, stulose,
whitish, indistinctly occulose. Context white. Smell raphanoid.
Spores (10.0 –)10.5 –12.5(–14.5) × (5.5 –)6.0 – 6.5(–7.0) µm, on average 10.9 –12.0 × 6.2 – 6.4 µm,
Q = 1.6 – 2.0(– 2.1), Qav = 1.72 –1.94, not dextrinoid (D0 – D1), subamygdaliform, not to partly tending to
sublimoniform; perispore not loosening (P0); almost smooth, slightly to moderately verruculose (O1– O3).
Cheilocystidia 36 – 50(– 52) × 4 – 5 × (5 –)6 –7(– 8) µm, on average 41.0 – 46.3 × 4.0 – 4.2 × 6.4 – 6.6 µm,
312 PERSOONIA — Vol. 18, Part 3, 2004 Aanen & Kuyper: Biological and phenetic species concept in Hebeloma crustuliniforme 313
Q = (1.3 –)1.4 –1.8, Qav = 1.5 –1.7, cylindrico-subclavate, only slightly swollen towards apex, not tending
to subcapitate or subspathuliform, in general rather small and narrow, thin-walled.
Associated with Salix (1 ×) and Betula (1 ×).
ICG 11
Pileus to 37–75 mm, plano-convex to applanate, without or with low broad umbo, (slightly) viscid
when moist, in centre yellowish to pale yellow-brown (Mu. 10 YR 7– 8 / 3, 6 –7/ 4), outwards paler, at
margin whitish to white. Lamellae, L = 50 – 60, l = 3, thin, normally crowded, to 4 mm, not or hardly ven-
tricose, narrowly adnate or emarginate, greyish ochraceous (10 YR 7/ 2); edge mbriate, whitish; weep-
ing. Stipe to 50 – 53 × 5 –10 mm, Q = 5 –10, shorter to longer than diameter of pileus, clavate to ± bulbous
(10 mm), solid to substulose, white, occulose. Context thick, rm, white to brownish. Smell rapha-
noid.
Spores (9.5 –)10.0 –12.0(–12.5) × 5.5 –7.0 µm, on average 10.7–11.1 × 5.9 – 6.4 µm, Q = (1.6 –)
1.7–1.9, Qav = 1.74 –1.82, not dextrinoid (D0 – D1), subamygdaliform, not to exceptionally tending to
sublimoniform; perispore not loosening (P0); rather distinctly verruculose (O2 – O3). Cheilocystidia
(39 –)41– 63(–76) × 4 – 5 × (5 –)6 –12 µm, on average 52.3 – 53.7 × 4.2 – 4.5 × 6.1– 9.7 µm, Q = 1.2 – 2.5,
Qav = 1.4 – 2.2, cylindrical to (sub)clavate, not or only a minority tending to subspathuliform or subcapi-
tate, thin-walled.
Associated with Salix (2 ×) or at forest edge with various trees (Fagus, Picea).
ICG 12
Pileus to 25 – 38 mm, convex to applanate, without or with low umbo, margin sometimes involute,
distinctly viscid, in centre pale yellow, ochraceous, (dark) yellow-brown to red-brown (Mu. 10 YR 7/ 4 – 6,
5 – 6 / 6, 4 – 5 / 4, 7.5 YR 4 / 4), outwards paler, at margin whitish. Lamellae, L = 40 – 50, l = 3 – 5, thin, crowd-
ed, narrowly adnate to emarginate, pale ochraceous; edge mbriate, whitish; weeping. Stipe to 53 – 65 ×
3 –7 mm, Q = 9.3 –17.7, shorter to longer than diameter of pileus, equal to slightly clavate, stulose or
solid, white, occulose. Context white. Smell raphanoid.
Spores 8.5 –12.0 × 5.0 – 6.5(–7.0) µm, on average 9.0 –11.2 × 5.0 – 5.8 µm, Q = (1.5 –)1.6 – 2.0, Qav
= 1.74 –1.96, not to weakly dextrinoid (D0 – D2), regular to subamygdaliform; not to partly tending to
sublimoniform; perispore not loosening (P0); verruculose, sometimes rather coarsely so (O2 – O3(– O4)).
Cheilocystidia (37–)40 – 63(– 68) × 4 – 5 × (7–)8 –13 µm, on average 44.9 – 55.5 × 4.2 – 4.5 × 9.4 –10.8
µm, Q = (1.4 –)1.8 – 3.0(– 3.3), Qav = 2.1– 2.6, (sub)clavate, usually (distinctly) swollen towards apex
and sometimes tending to subcapitate, a minority remaining subcylindrico-subclavate, thin-walled or
very slightly thick-walled, especially in apical part in ± subcapitate cheilocystidia, exceptionally slightly
thick-walled throughout.
Associated with Salix (5 ×) or Populus (1 ×).
ICG 13
Pileus to 25 – 37 mm, plano-convex to applanate, slightly umbonate, viscid, in centre pale yellow (Mu.
2.5 Y 6 –7/ 4), paler outwards, at margin white. Lamellae, L = 55, l = 3 – 5, thin, crowded, emarginate, pale
ochraceous; edge mbriate, whitish; weeping. Stipe to 45 × 6 mm, Q = 7.5, slightly longer than diameter
of pileus, subclavate, solid, white, (sub)occose. Context white. Smell raphanoid.
Spores (9.0 –)9.5 –10.0(–11.0) × (5.0 –)5.5 – 6.0(– 6.5) µm, on average 9.9 × 5.7 µm, Q = (1.6 –)1.7–
1.8(–1.9), Qav = 1.74, not dextrinoid (D0 – D1(– D2)), regular to subamygdaliform, not tending to subli-
moniform; perispore not loosening (P0); verruculose (O2 – O3). Cheilocystidia 38 – 56 × 4 – 5 × (5 –)6(–7)
µm, on average 45.0 × 4.3 × 5.9 µm, Q = 1.2 –1.5(–1.8), Qav = 1.4, cylindrical to somewhat subclavate,
only slightly broadened apically, not tending to subspathuliform or subcapitate, sometimes even more
subutriform and slightly broadened in lower part, thin-walled.
Associated with Populus (1 ×).
ICG 14
Pileus to 22 – 35 mm, plano-convex to applanate, without umbo, not or hardly viscid, usually bicol-
oured, in centre (dark) red-brown to ochraceous brown (Mu. 7.5 YR 4 / 2, 4 – 5 / 4, 10 YR 5 – 6 / 4), at margin
paler, pale brown to whitish (10 YR 6 –7/ 4, 8 / 3 or paler). Lamellae, L = 30 – 40, l = 1– 3, thin, normally
314 PERSOONIA — Vol. 18, Part 3, 2004 Aanen & Kuyper: Biological and phenetic species concept in Hebeloma crustuliniforme 315
crowded, to 4 mm, subventricose, almost free to narrowly adnate or emarginate, ochraceous brown (10 YR
5 / 4); edge mbriate, whitish; not distinctly weeping. Stipe to 18 – 32 × 4 –7.5 mm, Q = 2.6 – 8, usually
shorter than but sometimes equal to diameter of pileus, equal or slightly bulbillose, solid to stulose,
white, discolouring to brown with age or on damage from base upwards, occulose to suboccose.
Context thick, rm, white to brownish. Smell raphanoid.
Spores (11.0 –)11.5 –14.0(–15.0) × 6.5 – 8.0(– 8.5) µm, on average 12.0 –13.1 × 6.9 –7.2 µm, Q =
1.6 –1.9(– 2.0), Qav = 1.72 –1.84, not dextrinoid (D0 – D1(– D2)), regular to subamygdaliform, not to partly
tending to sublimoniform; perispore not to very indistinctly loosening (P0(– P1)); moderately weakly to
distinctly verruculose ((O1–)O2 – O3). Cheilocystidia 39 –75 × (3 –)4 – 6 × (7–)8 –12(–14) µm, on aver-
age 49.9 – 62.7 × 4.1– 5.1 × 9.3 –11.2 µm, Q = (1.3 –)1.6 – 3.5(– 3.7), Qav = 2.1– 2.6, usually clavate to
(indistinctly) (sub)capitate or more (sub)spathuliform, a minority more cylindrico-(sub)clavate and hardly
broadened towards apex, sometimes somewhat broadened in middle part and then cylindrico-subutriform,
thin-walled or slightly thick-walled with brownish wall in apical part in subcapitate cheilocystidia.
Associated with Salix (4 ×).
ICG 15
Pileus to 19 – 45 mm, plano-convex to applanate, without or with rather indistinct umbo, dry to
slightly viscid, in centre dark red-brown to orange brown (Mu. 2.5 – 5 YR 3 / 2, 5 –7.5 YR 4 – 5 / 4 – 6, 10
YR 3 – 4 / 3), outwards almost concolorous to paler. Lamellae, L = 35 – 55, l = 1– 3, thin, crowded, to 4.5
mm, subventricose, narrowly adnate to emarginate, brown, ochraceous brown to greyish brown (10 YR
4 / 4, 5 – 6 / 4, 6 –7/ 3 – 4); edge mbriate, whitish; weeping, but sometimes not (distinctly) weeping. Stipe
to 23 – 45 × 3 –10.5 mm, Q = 2.4 – 9, shorter than to equal to diameter of pileus (exceptionally somewhat
longer than diameter of pileus), equal, solid to stulose, white, discolouring to yellow-brown on ageing
or damage from base upwards, at apex occulose. Context thick, rm, white. Smell raphanoid.
Spores 12.5 –17.0 × (6.5 –)7.0 – 8.0(– 9.0), on average 13.7–14.9 × 6.9 –7.6 µm, Q = (1.7–)
1.8 – 2.1(– 2.2), Qav = 1.83 – 2.03, not to weakly dextrinoid (D0 – D2), (sub)amygdaliform, partly tend-
ing to sublimoniform; perispore not or slightly loosening (P0 – P1); slightly to distinctly verruculose
(O1– O3). Cheilocystidia (34 –)41– 63(– 65) × 4 – 6 × (6 –)7–13(–14) µm, on average 50.4 – 52.0 ×
4.4 – 4.9 × 8.3 –10.7 µm, Q = 1.3 – 2.8(– 3.0), Qav = 1.9 – 2.3, subcylindrical to clavate, partly more tend-
ing to subcapitate, partly somewhat swollen in middle part and subcylindrical-subutriform, thin-walled
or with a slightly thickened yellowish wall in apical part, especially in subcapitate cheilocystidia, in one
collection with slightly thickened wall halfway.
Associated with Salix (4 ×), Populus (1 ×) or Pinus (1 ×).
ICG 16
Pileus to 34 – 65 mm, convex to applanate, without or with rather distinct umbo, slightly viscid, in
centre red-brown, yellow-brown to ochraceous (Mu. 5 YR 4 – 5 / 3, 10 YR 4 – 6 / 4, 5 – 6 / 6), uniformly
coloured (especially in paler specimens) to ± distinctly paler outwards and at margin sometimes even
whitish. Lamellae, L = 40 – 65, l = 3 –7, thin, (very) crowded, rather broadly to narrowly adnate, to 6 mm,
not ventricose to subventricose, ochraceous buff to brownish ochraceous (10 YR 7/ 2 – 3 to 6 / 3 – 4); edge
mbriate, whitish; weeping (but sometimes not distinctly so). Stipe to 34 – 60 × 5 – 9 mm, Q = 6.7–10.5,
usually ± distinctly bulbous, sometimes (sub)clavate, stulose, with pendent marrow strand, whitish,
(sub)occulose to suboccose. Context thick, rm, white. Smell raphanoid.
Spores (9.5 –)10.0 –12.5(–13.0) × 6.0 –7.0 µm, on average 10.5 –11.8 × 6.4 –6.6 µm, Q = 1.5 –1.8(–1.9),
Qav = 1.62 –1.80, weakly to distinctly dextrinoid (D2 – D4), regular to subamygdaliform, exceptionally
sublimoniform; perispore not or very slightly loosening (P0 – P1); slightly to rather distinctly verruculose
(O2 – O3). Cheilocystidia (40 –)47– 87(–106) × (4 –)5 – 6(– 8) × 6 – 9(–12) µm, on average 55.2 –72.2 ×
4.9 – 5.7 × 6.7–10.2 µm, Q = 1.2 –1.8(– 2.4), Qav = 1.3 – 2.0, subcylindrical to subclavate, usually not
distinctly enlarged apically, but exceptionally tending to subspathuliform, sometimes slightly swollen
in basal part and then slenderly subutriform, thin-walled to very slightly thick-walled.
Associated with various deciduous trees in mixed forest (Betula, Fagus, Quercus, Carpinus, Cory-
lus).
314 PERSOONIA — Vol. 18, Part 3, 2004 Aanen & Kuyper: Biological and phenetic species concept in Hebeloma crustuliniforme 315
ICG 17
Pileus to 32 –78 mm, convex to applanate, without umbo to ± distinctly umbonate, very viscid to almost
dry, sometimes seemingly hygrophanous with irregular spots, in centre usually varying between pale ochr-
aceous yellow to pale yellow-brown (Mu. 2.5 Y 7– 8 / 2 – 4, 10 YR 7– 8/4 – 6), sometimes more ochraceous
brown (7.5 –10 YR 5 – 6 / 4), uniformly coloured (especially in paler specimens) to distinctly paler outwards
and then whitish at margin. Lamellae, L = 45 –70, l = 1– 3 –7, thin, (very) crowded, to 8 mm, subventri-
cose, rather broadly to narrowly adnate, ochraceous brownish (10 YR 6 –7/ 3 – 4); edge mbriate, whitish;
weeping. Stipe to 40 –120 × 5 –10 mm, Q = 5.3 –12, shorter to longer than diameter of pileus, usually
distinctly bulbous (to 20 mm), but sometimes only subclavate or even equal, usually stulose with
pendent marrow strand but sometimes solid, white, discolouring to brownish on damage from base up-
wards, minutely occulose to suboccose, especially in upper part. Context thin, rm, white. Smell rapha-
noid.
Spores (9.5 –)10.0 –13.0 × 6.0 –7.5 µm, on average 10.4 –11.9 × 6.3 –7.2 µm, Q = 1.5 –1.7(–1.8), Qav
= 1.57–1.69, weakly to distinctly dextrinoid (D2–D4), regular to subamygdaliform, sometimes tending
to sublimoniform; perispore not loosening (P0(– P1)); almost smooth to distinctly verruculose (O1– O3).
Cheilocystidia (36 –)40 – 81(– 83) × 4 –7(– 8) × 6 –13 µm, on average 45.5 – 66.0 × 4.5 – 6.3 × 6.2 – 9.3
µm, Q = (1.0 –)1.2 – 2.2(– 2.8), Qav = 1.2 – 2.0, straight to exuose, usually subcylindrico-subclavate, only
slightly broadened towards apex (but in two collections more distinctly broadened and even tending to
subspathuliform or subcapitate), a few more subcylindrical and hardly swollen towards apex, some-
times slightly swollen in basal part and then slenderly subutriform, thin-walled to slightly thick-walled,
sometimes bid in apical part in varying frequency (absent to fairly common, and then apex to 19 µm
broad).
Associated with various conifers (Pinus, 6 ×; Picea, 4 ×) and deciduous trees (Betula, 5 ×, Quer-
cus, 2 ×, Fagus, 1 ×, Carpinus, 2 ×); in two collections vicinity of Salix also noted.
ICG 18
Pileus to 60 mm, convex to almost applanate, with a low broad umbo, very viscid, uniformly yellow-
brown (Mu. 10 YR 7– 8 / 4 – 6). Lamellae, L = 55, l = 1– 3, thin, normally crowded, to 5 mm, not ventricose,
broadly adnate, ochraceous (10 YR 7/ 2 – 3); edge mbriate, whitish; weeping. Stipe to 110 × 7 mm, Q =
15.7, longer than diameter of pileus, bulbous (to 20 mm), stulose with pendent marrow strand, white,
nely occulose. Context thin, rm, white. Smell raphanoid.
Spores (10.0 –)10.5 –11.5(–12.0) × (6.0 –)6.5 –7.0 µm, on average 10.9 × 6.5 µm, Q = 1.6 –1.7(–1.8),
Qav = 1.67, distinctly dextrinoid (D3 – D4), regular to subamygdaliform, not sublimoniform; perispore not
loosening (P0); distinctly verruculose (O2 – O3). Cheilocystidia (45 –)46 – 59(–72) × (4 –)5 – 6 × 5 – 6(–7)
µm, on average 54.5 × 5.0 × 5.6 µm, Q = 1.0 –1.2(–1.3), Qav = 1.1, cylindrical, partly somewhat inated
in basal part and then subventricose-slenderly utriform, near apex not or hardly inated, not clavate,
thin-walled.
Associated with Pinus among living Sphagnum.
ICG 19
Pileus to 35 – 49 mm, applanate, only indistinctly umbonate, viscid, ochraceous yellow-brown (Mu. 10
YR 6 / 6) in centre, outwards paler. Lamellae, L = 55, l = 3, thin, normally crowded, to 5.5 mm, subventri-
cose, narrowly adnate, ochraceous brown (10 YR 6 / 4); edge mbriate, whitish; weeping. Stipe 42 –75 ×
4 –7 mm, Q = 10 –12.5, longer than diameter of pileus, equal, not bulbous, white, occulose in upper part.
Context thick, rm, white. Smell raphanoid.
Spores 10.5 –11.0 × 6.0 – 6.5 µm, on average 10.7 × 6.2 µm, Q = (1.6 –)1.7–1.8, Qav = 1.74, not
dextrinoid (D0 – D1), subamygdaliform, not tending to sublimoniform; perispore not or hardly loosen-
ing (P0(– P1)); moderately coarsely verruculose (O2 – O3). Cheilocystidia (38 –)39 – 55(– 57) × 4 – 5 ×
(5 –)6 –7(– 8) µm, on average 46.2 × 4.6 × 6.4 µm, Q = (1.2 –)1.6(–1.8), Qav = 1.4, subcylindrical-subcla-
vate, exceptionally more distinctly clavate, partly somewhat swollen below middle part and then tending
to slenderly subutriform, thin-walled.
Associated with Quercus (1 ×).
316 PERSOONIA — Vol. 18, Part 3, 2004
ICG 20
Pileus to 48 – 60 mm, plano-convex to applanate, with or without umbo, viscid, pale brownish yellow
(Mu. 10 YR – 2.5Y 6 – 8 / 4), outwards paler, at margin whitish. Lamellae, L = 60 –70, l = 1–7, thin, normally
crowded, to 5 mm, not ventricose, narrowly adnate to emarginate, pale ochraceous grey; edge mbriate,
whitish; probably weeping. Stipe to 60 – 62 × 8 – 9 mm, Q = 6.7–7.8, equal to diameter of pileus, equal
to slightly clavate, whitish, solid, occulose in upper part. Context thick, rm, white. Smell raphanoid.
Spores (10.5 –)11.0 –12.0(–12.5) × 6.0 –7.0(–7.5) µm, on average 11.3 –11.4 × 6.5 – 6.8 µm, Q =
1.6 –1.8, Qav = 1.64 –1.76, not to weakly dextrinoid (D0 – D2), subamygdaliform, not sublimoniform;
perispore not loosening (P0); very slightly to distinctly verruculose (O1– O3). Cheilocystidia (44 –)45 –
72(–74) × 4 – 5 × (6 –)7– 9 µm, on average 55.0 – 59.7 × 4.3 – 4.5 × 7.4 –7.5 µm, Q = 1.4 – 2.0(– 2.3), Qav
= 1.7, subcylindrico-subclavate to somewhat more distinctly clavate, partly even tending to somewhat
subcapitate, but partly somewhat swollen in lower part and then tending to slenderly utriform, thin-
walled or sometimes distinctly thick-walled in upper part, especially in subcapitate cheilocystidia.
Associated with Quercus (2 ×).
ICG 21
Pileus to 46 – 50 mm, plano-convex to applanate, slightly umbonate, slightly viscid, in centre brown-
ish yellow (Mu. 10 YR 6 –7/ 4 – 6), outwards paler, at margin whitish. Lamellae, L = 40 – 50, l = 3, thin,
normally crowded, not ventricose, emarginate, ochraceous; edge mbriate, whitish; weeping. Stipe to
70 –75 × 6.5 – 9 mm, Q = 8.3 –10.8, longer than diameter of pileus, equal to slightly swollen, stulose,
whitish, occulose. Context thick, rm, white. Smell raphanoid.
Spores 9.5 –12.5 × 5.5 – 6.5(–7.0) µm, on average 10.4 –11.6 × 5.8 – 6.3 µm, Q = 1.7– 2.0, Qav =
1.75 –1.88, not dextrinoid (D0 –D1), subamygdaliform, not to weakly sublimoniform; perispore not loos-
ening (P0); verruculose (O2 – O3). Cheilocystidia (42 –)45 – 68(–73) × 4 – 5 × 6 –11(–15) µm, on average
48.9 – 57.3 × 4.3 – 4.6 × 7.3 – 9.8 µm, Q = (1.2–)1.4 – 2.5(– 3.0), Qav = 1.6 – 2.3, cylindrico-subclavate,
usually only slightly broadened apically to more distinctly subspathuliform or subcapitate, a minority
tending to subcylindrical-subclavate, thin-walled, but sometimes with slightly thickened wall in middle
part.
Associated with Betula (1 ×) and Tilia (1 ×).
ICG 22
Pileus to 35 –70 mm, convex, without umbo, viscid, pale ochraceous yellow(Mu. 2.5 Y 7/ 8), more
or less uniformly coloured, only at margin somewhat paler. Lamellae, L = 55, l = 3 –7, thin, normally
crowded, emarginate, ochraceous; edge mbriate, whitish; weeping. Stipe 20 – 45 × 3.5 –10 mm, Q =
4.5 –7.5, shorter than diameter of pileus, at base slightly swollen, white, subocculose. Context thick,
rm, white. Smell raphanoid.
Spores (11.0 –)11.5 –12.0(–12.5) × 6.0 – 6.5 µm, on average 11.7 × 6.2 µm, Q = 1.8 – 2.0, Qav = 1.88,
not dextrinoid (D0 – D1), (sub)amygdaliform, partly tending to sublimoniform; perispore not or very
slightly loosening (P0); almost smooth to slightly verruculose (O1– 02). Cheilocystidia (43 –)47– 67(–70)
× (3 –)4(– 5) × 6 – 8(– 9) µm, on average 57.5 × 3.9 × 7.2 µm, Q = (1.5 –)1.6 – 2.0(– 2.3), Qav = 1.8, cylin-
drico-(sub)clavate, at apex slightly to distinctly broadened but not or hardly tending to (sub)capitate or
(sub)spathuliform, thin-walled, colourless.
Associated with Betula (1 ×).
