Invited Research Paper 
93 
The first task Gerald Smith and I first undertook was to sort out the 
relationships of the three endemic southeastern rain-lilies. For several decades 
some botanists, including Dr. Flory and I, thought that Z. treatiae might be an 
ecological variant of Z. atamasca. The two have been traditionally distinguished 
by their leaves and it was thought that their flowers were not distinguishable. 
Varietal status was formalized by Alan Meerow: Z. atamasca var. treatiae (S. 
Wats.) Meerow (1985). 
Dr. Smith thought that the flowers of Z. atamasca and Z. treatiae were 
different, and was adamant that we do detailed studies of them (Flagg and Smith, 
2001). To shorten a very long story, we took a lot of measurements of the flower 
parts of all three taxa (Flagg and Smith, 2008a). The flowers of the three species 
are distinguished by characteristics related, by and large, to the degree of fusion 
of the perianth into a tube (Fig. 11). In summary, the length of the perianth tube 
in Z. atamasca is less than 1/4 the total perianth length, in Z. treatiae it is at least 
1/4 the total perianth length, and in Z. simpsonii it is usually at least 1/3 the total 
perianth length (Flagg and Smith, 2008a). 
Figure 11. Floral diagrams of Zephyranthes species endemic to southeastern 
USA (Flagg and Smith, Castanea 2008a). 
The morphometries are illustrated in an old-fashion pictorialized scatter 
diagram (Fig. 12) and a Principal Component Analysis (PCA) (Fig. 13) of flower 
part lengths. These revealed three distinct species, and indicated that the 
variation between Z. atamasca and Z. treatiae is greater than the variation 
between Z. treatiae and Z. simpsonii. Publication of DNA fingerprinting 
Cladograms (Spurrier and Smith, 2011) and ongoing research by Spurrier, Smith 
and Flagg will cast additional light on these relationships. 
