( 1998 ) to his new order, theSalilagenidiales
(Table 3.4), along with a number of other
marine genera such asAtkinsiellaand marine
Lagenidium spp. The type species,
Haliphthoros milfordensis, was described by
Vishniac as an endoparasite of crustacean
eggs and oyster drill, and all members of the
order are parasites of molluscs, crustaceans, or
their eggs (Diggles 2001 ; Hatai 2012 ; Hatai et al.
1980 , 1992 ). They are the only group of early
diverging genera to produce irregularly
branched eucarpic mycelial thalli (Fig.3.7s, t,
v) that can be cultured on artificial media. Large
segments of the thalli convert into zoosporan-
gia (Fig.3.7u, w, x), in which peripheral zoo-
spore initials are defined by a large central
vacuole, as described in the Saprolegniales
(Beakes 1994 ). All genera form very long
narrow hypha-like exit tubes (Fig. 3.7u–x)
(Hatai 2012 ), from which the zoospores escape
in uniseriate fashion, reminiscent of sporula-
tion in Aphanomyces. In LSU rRNA trees
(Hatai 2012 ) all haliphthoralean species formed
a monophyletic clade, which diverged just
before the peronosporomycete/saprolegniomy-
cete split (Hatai 2012 ; Maurosa et al. 2009 ).
However, in mitochondrial cox2 gene trees
there is no good statistical support for
separating or combining the Haliphthorales
and Olpidiopsidales clades, and greater taxon
and gene sampling is required before these
basal orders can be properly resolved
(Sekimoto 2008 ; Sekimoto et al. 2007 ).
F. Saprolegniomycetes
The class “Saprolegniomycetes”(Figs.3.8,3.9,
3.10, and3.11) replaces the Saprolegniomyceti-
dae subclass (Dick 2001) and broadly encom-
passes those species generally known aswater
moulds. Most have eucarpic mycelial thalli
(Figs.3.8d, fand 3.9b) and can usually be
cultured. They are mainly freshwater sapro-
trophs (Willoughby 1962 ) or opportunist
necrotrophic pathogens (Dick 1976 ; Dick
2001a; Sparrow 1960 ). Molecular studies (Hud-
speth et al. 2000 ;Le ́clerc et al. 2000 ; Petersen and
Rosendahl 2000 ; Riethmu ̈ller et al. 1999 ) have
shown that this class forms a well-supported
monophyletic clade (Fig.3.5), although there is
still some uncertainty about the placement of a
small number of basal genera. Saprolegniomy-
cetes are able to synthesize sterols using both
lanosterol and cycloartenol pathways and can
utilize ammonium and organic sulphur (Dick
2001a; Gleason 1976 ). They mostly formsep-
tum-delimited zoosporangiaand can produce
two morphologically distinct generations of
asexual zoospores (diplanetic), each followed
by a walled cyst stage (Fig.3.11q) (Dick1973b,
1990 ,2001a,b; Sparrow 1960 ). Oospheres are
formed as a result ofcentrifugal cleavagewith-
out the formation of a peripheral periplasmic
layer (Beakes 1981 ). We take a conservative
approach to their classification, recognizing
three orders (Fig.3.6, Table3.5), theAtkinsiel-
lales, theLeptomitales,and theSaprolegniales.
There is no molecular support (Figs.3.5and3.6)
for retaining either the ordersSalilagenidiales
or Sclerosporales introduced by Dick
(Table3.4) and placed in his Saprolegniomyce-
tidae subclass, and both are excluded from our
revised classification (Table3.5).- Atkinsiellales
We propose a new order, the “Atkinsiellales”,
for the usually earliest diverging clade in Sapro-
legniomycete phylogenetic trees (Fig. 3.6)
(Cook et al. 2001 ; Sekimoto et al. 2007 ). This
order includes the monotypic familyAtkinsiel-
laceaefor the crustacean parasite Atkinsiella
dubia, which Dick ( 1998 ,2001a) placed in the
family Haliphthoraceae in the order Salilagen-
idiales (Table3.4). However,cox2-based trees
revealed that this species was not associated
with Haliphthoros but formed a clade that
diverged before the Leptomitales (Cook et al.
2001 ; Sekimoto et al. 2007 ). A second mono-
typic family, the Crypticolaceae, is also
included in this order. This family contains
two holocarpic species that are parasites of
mosquito larvae. The first,Crypticola entomo-
phaga(Dick 1998 ), had originally been placed
in the genusAtkinsiellaby Martin ( 1977 ), and
the second, Crypticola clavulifera, was
described by Frances et al.( 1989 ). Their place-
ment in this order is supported by unpublished58 G.W. Beakes et al.