180
andrias, 19
(2012)
in width due to shrinkage of the included spores
(a similar effect is induced by shortly heating the
slide). Living asci shrink to a similar rate (10-15 %
in length and 15-20 % in width) when killed by
KOH or heat.
The effect is provoked by an increase in water
content and volume of the living spores as soon
as the ascus turgor is released. In the spore clus-
ter the spores are rather strongly dehydrated.
When the asci loose turgor, the spores fill the
complete ascus by keeping the elastic ascus
wall in a state of tension (Fig. 6d-e). As a conse-
quence, such dead asci are scarcely smaller in
size than living asci (Figs 5f, 6c). This peculiarity
of D. triangulispora might be due to the saccate
shape of the asci which are not much longer than
the pars sporifera.
Despite the observed real variation in ascus size,
the asci appear to be always 64-spored, with rare
exceptions of single asci with perhaps only 32
spores or intermediate spore numbers. The varia-
tion in length is partly explained by the occasional
presence of a more or less pronounced stalk.
The first collection from Folkend near Fischbach
on Ilex consisted of only two apothecia. This spec-
imen differed in the asci which were noted much
shorter (*21 × 9 µm, with a pars sporifera of only
20 × 6 µm). Ascospore size was not evaluated;
also spore number was not noted but might well
have been only 32. This site was revisited in 2011,
but again only a few apothecia could be detected.
Here the asci were distinctly longer (†23-27 × 7.5-
10 µm, containing living spores).They were clearly
more than 32-spored and matched the specimens
on the other hosts in every respect.
Ascospores
.The living ascospores are quite con-
sistent in size when measured outside the asci.
Their shape is somewhat variable, but strongly
depends on the direction of view: the character-
istic triangular shape is only seen in profile view,
whereas in dorsal view they look ± deltoid and es-
pecially in oblique view more ellipsoid (Figs 1.1l,
1.2d upper right). Shrinkage of the spores when
killed by KOH lies in the range of 5-20 % in both
length and width.
In turgescent asci the angular spore shape per-
mits a very dense packing within the spore cluster.
This is best seen in asci that lost turgor prior to
spore release: the living spores completely fill the
ascus and fit together like a honeycomb.
Anamorph
. In one collection of D. triangulispora
a few ascospores were seen to show yeast-like
germination, i.e. to bud off conidia at one end,
apparently by phialidic conidiogenesis (Fig. 1.4).
This phenomenon seems to be rare because it
was not observed in any of the other collections
studied.
When the ascospores were shot on agar, they
formed dense heaps. Germination did either not
occur at all, or spores germinated only 7 days af-
ter shooting, especially when the heaps were sep-
arated with a glass rod. Often, germinated spores
very soon stopped growing, but in one culture
(on CMA1:4) they formed a very slow-growing
mycelium, with up to 110 µm long hyphae within
ca. 25 days after germination, i.e., ca. 30 µm per
week.When transferred one week later to another
plate, the hyaline mycelium formed a dense mat
that hardly increased in diameter within the next
month. Simultaneously, several dozens of ochre-
to red-brown conidiomata developed (Fig. 6f-g).
These somewhat resemble the ascomata in both
size and shape, but open by a transversal slit.
The abundantly produced phialoconidia resemble
those formed on the ascospores but are longer
and partly also wider.
The conidia of Deltopyxis resemble those obtained
by W
eber
(2002) in pure culture of Tromeropsis
microtheca, the ascospores of which produced a
likewise slowly growing mycelium on agar (MEA).
However, the small, ± cylindrical, straight to me-
dium curved conidia of T. microtheca emerged
from little pegs on integrated conidiogenous cells
(conidiogenesis holoblastic). Also larger conidia
were observed which germinated yeast-like to
form small conidia.
Ecology
. The sites where Deltopyxis triangulispo
ra was collected are open woodlands or hedges,
particularly at ± S-exposed slopes but also in ar-
eas with indistinct inclination. The geology was
generally more or less calcareous or basic, and
the vegetation usually thermophilous. The inhab-
ited substrates are dead branches with a usually
initial to optimal, rarely final stage of wood decay,
attached to living or dead shrubs or small trees
one or a few meters above ground, sometimes
also broken but hanging on other branches. Quite
an undisturbed vegetation over many years is a
prerequisite for the detection of many of these
desiccation-tolerant ascomycetes that are con-
fined to xeric branches. Most of the branches on
which D. triangulispora occured were previously
infected by Vuilleminia which was, however, no
more viable when this and other discomycetes
appeared on their basidiomata or on the bark
around. Such branches are usually entirely corti-
cated, but with the periderm replaced by the Vuil
leminia on one side (Figs. 2c, 3a). The branches
