Spore size and shape in taxonomy

Erast Parmasto erast at park.tartu.ee
Mon Mar 4 05:53:30 EST 1996

On 21 Feb 1996, R.N. Weinstein wrote:
> Is there a consensus that spore SIZE, within a range, and spore SHAPE is
> enough to be the last word in species identification? This seems to be the
   No good taxonomy is possible when using only one character. Not even
on species level.	 
   Spore size, volume (transformed to the diameter of an equivoluminous
sphere) and shape (length/width quotient) are, as a rule, normally or 
lognormally variable characters. At species level, the variation coeffi-
cient CV is usually 5-10, rarely 3-13. Accordingly, if you have TWO specimens
with mean spore length difference one and a half times (e g, one with mean
spore length 10 micrometers and another with mean = 15), there is not more
than 90 % probability that they belong to different species. 90 % signi-
ficance level is statistically a quite low one, however.
   To use spore size data for distinguishing species, you must have a
good enough picture on infraspecific variation of this character. For
this, not less than some 7-10 specimens randomly collected in various 
places and in various time must have been studied, and not less than 
21 (better: 30 or 50) randomly taken spores measured.
   Together with my wife Ilmi I published a small book on spore
variability in Hymenomycetes, with some examples on Oomycetes, Erysiphales
and Uredinales. There is also a chapter on use of spore size in taxonomy.
The book has been rarely cited; however, its content may be unpleasant
for many taxonomists. We demonstrated that one COLLECTION of a new
species is not enough to describe variation of any character (incl.
spore size) in the new SPECIES. Alas! About 70-75 % of the new
species of fungi have been based on only one collection. In other words:
in 70 % cases the description of a new species does not give any
information on the limits of variation of this species. It is much
more interesting for a taxonomist to search for, find and describe
new and new and new species instead of spending time studying the
variation of a species already known.
   In another paper we described the variation of spore statistics on
order and family levels in the Hymenomycetes, and tried to characterize
these taxa using spore data. Usually the variation found was hyper-
lognormal: there are always more species with extremely long, wide, 
voluminous spores than may be expected. - It was published in the 
"Mycologia Helvetica", a journal dealing mainly with macromycetes.
	 1) E. & I. Parmasto. Variation in basidiospores in the Hymenomycetes
and its significance to their taxonomy. - Bibliotheca Mycologica 115.
Berlin & Stuttgart, 1987. 168 p.
   2) E. & I. Parmasto. Size and shape of basidiospores in the Hymenomy-
cetes. - Mycologia Helvetica 5 (1): 47-78. 1992.

   Erast Parmasto
	 erast at park.tartu.ee
	 Institute of Zoology & Botany, Tartu, Estonia

More information about the Mycology mailing list