Subtribe Spiranthinae - The Pollination Biology of North American Orchids: North of Florida and Mexico - page 33

The absence of a rostellum and the occurrence of autogamy are unknown among
other North American species of Spiranthes (Catling 1980b, 1983b ). However,
among European and Asiatic members, autogamy is present in northern Danish
populations of S. spiralis (L.) Chev. (Hagerup 1952 ), at least some Australian popu-
lations of S. sinensis (Pers.) Ames (Darwin 1862 ; Fritzgerald 1876 ; Ridley 1888 )
and Hong Kong populations of S. hongkongensis Hu and Barr (Hu 1977 ). The col-
umn also lacks a rostellum in some populations of S. sinensis (Ridley 1888 ) and a
viscidium in S. hongkongensis (Hu 1977 ; Sun 1996 ), and the pollinia come into
direct contact with the stigma in both.
Another type of mixed mating system based on varying levels of selfing and
outcrossing is found in the rare orchid S. diluvialis at sites in Colorado and Utah
(Sipes and Tepedino 1995 ). Experimental self-pollination, geitonogamous pollina-
tion, and cross-pollination produced no statistically significant differences in fruit
and seed set (Sipes 1995 ; Sipes and Tepedino 1995 ). Thus, no indication of self-
incompatibility or inbreeding depression is present in this species, and although
outcrossing is favored by protandry, acropetally, and pollinator foraging habits (see
below), male and female phases overlap and facilitated selfing and geitonogamy
occur. Tests for autogamy and agamospermy led to no fruit production, and no poly-
embryonic seeds were found in any of the populations examined (Sheviak 1984 ;
Sipes and Tepedino 1995 ). As in other nonautogamous members of Spiranthes
(Catling 1982 ), autogamy is precluded by the structure of the column which pre-
vents contact between the pollinium and stigma in the absence of a pollinator
(Fig. 2.4 ) (Sipes and Tepedino 1995 ).
Various factors contribute to the maintenance of species integrity in taxa with
overlapping distributions (Sheviak and Brown 2002 ) . Phenology is highly variable,
but may be of some significance in the isolation of particular species. For example,
in southwestern Ontario, S. ochroleuca and S. magnicamporum bloom simultane-
ously after S. cernua has passed its peak (Catling and Brown 1983 ). Both can cross
with S. cernua but, like most members of the S. cernua complex, are more or less
genetically isolated from one another (Sheviak 1982 ). They tend to be spatially
isolated by soil pH as well, S. ochroleuca preferring acidic conditions and S. magni-
camporum alkaline (Catling and Brown 1983 ). In addition to a degree of phenological
isolation, S. cernua is also partly isolated from S. magnicamporum and S. ochroleuca
by a preference for relatively moist sites (Catling and Brown 1983 ). Differences in
ploidy level may also restrict introgressive hybridization between S. cernua with
2 n = 60 chromosomes and S. ochroleuca or S. magnicamporum with 2 n = 30 (Sheviak
1976 ; Catling 1980b ; Sheviak and Catling 1980 ). Occasional crosses do occur, how-
ever, and it is likely that hybridization between S. cernua and related diploids accounts
for a large part of the variability observed in this species (Sheviak 1982 ; Catling
1990 ; Johnson 2006 ).
Sheviak ( 1973, 1982 ) suggested that pollinator discrimination might further iso-
late S. cernua and S. magnicamporum . Although this hypothesis has not been cor-
roborated by subsequent studies (Sheviak 1976, 1982 ; Catling 1983c ) and both
species share at least one common pollinator in eastern North America (Tables 2.3
and 2.4 ), the flowers and inflorescences of these plants differ in appearance, and