Harpellales + Orphella & Kickxellales + Spiro-
myces.—The Harpellales includes two families: Har-
pellaceae, all with unbranched thalli and having hosts
that are lower dipterans, and Legeriomycetaceae, all
with branched thalli associated with various nonpre-
daceous larval aquatic insect hosts with one exception
in isopods (Lichtwardt 1986, Lichtwardt et al 1999,
2001, White 1999). Only the Harpellales possesses
asexual spores called trichospores (specialized, de-
ciduous, monosporous sporangia) that often bear
nonmotile appendages (F
IG
. 3A, B, D, E, F and I).
Four types (I–IV) of sexually produced biconical
zygospores are recognized based on orientation of the
zygospore on the zygosporophore (F
IG
. 3C, E, G and
H) but Valle and Santamaria (2005) documented
unique zygospores for Orphella (see this section,
below). Key morphological taxonomic characters
among the Harpellales include the size and shape of
the spores (F
IG
. 3A–I), the number of appendages
per spore, number of spores per branchlet (F
IG
. 3A–E
and H), shape or nature (with or without adhesive
exudate or mucilage) of the gut anchoring holdfast
and arthropod host type.
The Kickxellales was validated by Benjamin (1979)
who also suggested that the Harpellales—itself vali-
dated by Lichtwardt and Manier (1978)—be consid-
ered in the Zygomycetes, but the Harpellales has been
accepted more broadly as Trichomycetes (Benny et al
2001). Although seemingly rare members of Kickxel-
lales are fairly common on dung and they also can be
isolated from soil and insect cadavers (Benny et al
2001, Kurihara et al 2004). The Kickxellales have
F
IG
. 3.
Phase contrast photomicrographs of living (refractive) Trichomycetes illustrating sexual and asexual reproductive
structures. A–I. Harpellales from aquatic insect larvae. A. Orphella catalaunica. Sporulating heads with attached cylindrical
asexual spores. B–C. Harpellomyces eccentricus. B. Trichospores with their appendages visible within generative cells. C.
Zygospore (Type III) arising from conjugated cells. D. Smittium culisetae. Trichospores attached to fertile branchlets. E, F.
Capniomyces stellatus. E. Attached, biconical zygospore (Type II) and immature trichospores. F. Released trichospores with
multiple appendages. G. Furculomyces boomerangus. Released bent zygospore (Type II) with a short collar. H, I. Genistelloides
hibernus
. H. Biconical zygospores (Type I) with swollen zygosporophores (similar to suspensors in Zygomycetes) attached to
two conjugated branches. I. Released trichospore with two appendages. J. Asellariales from a marine isopod, Asellaria ligiae.
Thallus attached by a holdfast cell to the hindgut cuticle and releasing small cylindrical arthrospores. Bars: A–I 5 20 mm; J 5
100 mm.
878
M
YCOLOGIA
grown from eight (Benjamin 1961, 1963, 1979) to 11
genera with the recent addition of Myconymphaea,
Mycoe¨melia and Ramicandelaber (Kurihara et al 2001,
2004), although Ramicandelaber did not cluster with
other Kickxellales in the present study (see F
IG
. 1).
Like Harpellales all members of Kickxellales have
monosporous sporangia, usually borne on specialized
fertile branches (sporocladia) (Benjamin 1979, Benny
et al 2001, Kurihara et al 2004). Most Kickxellales
have asexual spores that collect in a fluid droplet,
except for Spiromyces and Spirodactylon, which are dry-
spored at maturity (Ingold 1978); zygospores are
hyaline and globose with two or three undifferentiat-
ed suspensors (Benjamin 1958, 1959, Benny et al
2001). Septa within Kickxellales are centrally perfo-
rate with biconvex plugs that persist in acid stains and
weak base (Kurihara et al 2001). In addition a sac-like,
labyrinthiform organelle or abscission vacuole is
located just below the sporangial septum in Kickxella
and Linderina (Benny et al 2001).
Several different relationships between the Kickxel-
lales and Harpellales have been proposed (Benjamin
1979, Moss 1979, Lichtwardt et al 2001) in part due to
the paucity of taxonomic characters (with unknown
character states in many instances) and the challenge
involved in working with unculturable microfungi
(about 80% of Harpellales) (White 2006). The
specialized, asexual trichospores of the Harpellales
are produced laterally from generative cells with
aseptate collars, and they are homologous to the
monosporous sporangia (borne on pseudophialides)
of the Kickxellales with the exception of Spiromyces
(Benjamin 1966, Moss and Young 1978). Zygospore
production, although different in the two orders
(biconical versus globose), also has long been used as
evidence for the close relationship of the two orders
(Benjamin 1979). Other possible homologies include
cell wall composition, septal ultrastructure, immuno-
logical affinities and sterol spectrum (Sangar et al
1972, Moss and Young 1978, Moss 1979, 1998, Benny
and White 2001).
Molecular phylogenetic studies began with Walk-
er’s (1984) 5S rRNA analyses; however this locus
lacked sufficient signal to infer natural relationships.
O’Donnell et al (1998) used 18S rRNA data with
additional support from morphological and physio-
logical characters to infer a sister group relationship
between the Harpellales and Kickxellales. Gottlieb
and Lichtwardt (2001) demonstrated a similar pattern
adding culturable members of Harpellales but also
suggested that Smittium was polyphyletic. White
(2006) used 18S and 28S rDNA to add various
unculturable taxa and discovered that the Harpellales
were monophyletic except for Orphella, a topology
(F
IG
. 1) repeated here using two Orphella species and
a broadened dataset (all harpellids in the tree are
represented in F
IG
. 3A–I). Orphella was thought to be
one of the most derived genera of the Harpellales
(Lichtwardt 1986) and species have distinctive
specialized reproductive cells (Santamaria and Girbal
1998, Valle and Santamaria 2005) that at maturity
protrude beyond the anus of their stonefly hosts
(White and Lichtwardt 2004). As predicted by White
(2002) the discovery of unusual coiled zygospores in
Orphella (Valle and Santamaria 2005), in contrast to
those found in both the Harpellales and Kickxellales,
is consistent with its unique position in the combined
rDNA phylogeny. Valle and Santamaria (2005)
suggested that Orphella derived directly from a kick-
xellid ancestor, presumably from one resembling
Ramicandelaber brevisporus (Kurihara et al 2004, see
below).
Although the Spiromyces clade was resolved as either
a sister of the Kickxellales or Harpellales in previous
molecular phylogenetic analyses (O’Donnell et al
1998, Tanabe et al 2000, White 2002, 2006),
Spiromyces was supported strongly as a sister of
Orphella + Kickxellales in the current study (F
IG
. 1).
The inconsistent placement of the Spiromyces clade
might reflect the need (i) for more sequence data,
(ii) to reconsider the genus or (iii) to consider the
effect of long-branch attraction (hereafter 5 LBA, see
next section).
Dimargaritales
+
Neozygites,
Zoopagales
and
uncertain lineages.—Genes of some taxa of the weakly
supported clade consisting of the Dimargaritales,
Zoopagales and several other fungi might have
undergone accelerated evolution and the cluster
might not be natural due to LBA (F
IG
. 1). LBA is
the phenomenon in which taxa that are evolving at
different rates are artificially attracted to each other
due to biases inherent in the method of tree
reconstruction (Bergsten 2005). Notably the unex-
pected inclusion of entomophthoralean Neozygites
with Dimargaritales on the longest branches observed
results in the paraphyly of Dimargaritales. Ramican-
delaber, originally classified within the Kickxellales
(Ogawa et al 2001) and the cockroach parasite,
Nephridiophaga, formerly considered a protozoan un-
til an affinity to the Zygomycota was suggested with
18S rDNA molecular analysis (Wylezich et al 2004),
appear as a basal members of the clade but with weak
support. The clade should be evaluated further with
more slowly evolving protein coding genes. The
monophyly of the Zoopagales is not supported
because of the uncertain grouping of Zoophagus
insidians with Harpellales + Kickxellales, consistent
with a study based on 18S rDNA (Tanabe et al 2000).
Given the diversity of ecology and morphology of
W
HITE ET AL
: P
HYLOGENY OF THE
Z
YGOMYCOTA
879