Tribe Neottieae - The Pollination Biology of North American Orchids: North of Florida and Mexico - page 73

routinely liberated from the pollinia prior to or just after opening of the flower.
Based on these reports, occasional autogamy may result either from rotation of the
entire pollinium onto the stigma or from a lack of pollen coherence in older flowers
(Catling 1983 ). Talalaj and Brzosko ( 2008 ), however, note that pollen viability is
reduced in aging flowers.
Other European taxa of this genus are clearly facultatively to obligately autoga-
mous or even cleistogamous (e.g., Hagerup 1952 9OUNGAND2ENZ 1958 ; Sundermann
1975 ; Richards 1986 ; Claessens and Kleynen 1996 -EREDAAND0OTUCEK 1998 ).
Robatsch ( 1983 ), in fact, claimed that 60% of all species are autogamous. Most are
reported to produce very little nectar or floral scent and to share a breakdown of the
rostellum along with powdery pollen that falls directly onto the stigma. Some autog-
amous species are thought to have evolved independently within local populations
of E. helleborine (Richards 1982 ; Harris and Abbott 1997 %HLERSAND0EDERSEN
2000 ; Johnson and Edwards 2000 0EDERSENAND%HLERS 2000 ; Squirrell et al. 2001,
2002 ; Talalaj and Brzosko 2008 ). Facultative and obligate subspecies and varieties
are associated with dry habitats, such as coastal sand dunes of the Netherlands and
Denmark. Based on the levels of pollen removal, pollinator visitation may be lower
in dry than in mesic habitats (Ehlers et al. 2002 ) and could lead to selection for
autogamy (e.g., Baker 1955 ; Wyatt 1988 ).
Since the plants of E. helleborine may be multistemmed (e.g., Richards 1982 ;
Burns-Balogh et al. 1987 ; Muller 1988 ; Light and MacConaill 1998 ), between spike
as well as within spike geitonogamy is possible. Based on some reports, geitonog-
amy is the most probable mode of pollination and may be promoted by pollinator
behavior (Light 1994 ; Light and MacConaill 1998 ; Talalaj and Brzosko 2008 ). At
the same time, pollinators often carry the pollinia for extended periods, depositing
pollen slowly (Richards 1986 ), increasing pollen carryover and the potential for
cross-pollination. The extraction or deposition of pollinia does not affect the overall
ATTRACTIVENESSOFANINmORESCENCE0IPERAND7AITE 1988 ). Vectors bearing pollinia
may, therefore, visit previously pollinated flowers, and data on pollen germination
suggest the possibility of pollen-tube competition (Light 1994 ).
Both cross-pollination and geitonogamy are known to occur in Quebec (Mousley
1927 ; Light and MacConaill 1998 ), and cross-pollination is routinely reported in
European populations (e.g., Sundermann 1975 ; Richards 1982 ; Talalaj and Brzosko
2008 0IPERAND7AITE 1988 ) inferred a high frequency of cross-pollination
in England based on a marked temporal separation of pollen export and import.
A genetic structure consistent with Hardy-Weinberg expectations and random mat-
ing also suggests that outcrossing is predominant and selfing is rare or subject to
inbreeding depression within local populations in both North America and the Old
World (Hollingsworth and Dickson 1997 %HLERSAND0EDERSEN 2000 ; Squirrell et al.
2001 ). At the same time, significant differentiation is present among local popula-
tions, implying that pollen flow is largely restricted to flowers within populations
(Richards 1982 ; Hollingsworth and Dickson 1997 %HLERSAND0EDERSEN 2000 ;
Squirrell et al. 2001 ).