It’s clear through several studies that certain wasps much prefer orchids that are at a certain height – or more specifically fall within a certain height range. Most of the time this is not a big deal, but it obviously affects the selection and evolution of the flowers themeselves. Here’s an interesting article describing the influences of the pollenating insects on the flowers themselves:
Our field experiments clearly showed that when given a choice, the pollinators of Chiloglottis trilabra consistently preferred to visit flowers at 15 cm and 22 cm high more than flowers presented above or below this height range. The results suggest that mate search and subsequently visits to this sexually deceptive orchid, is highly stereotypic and focused on a narrow vertical flight zone compared with the much broader foraging flight zone. Female thynnines in general are known to call for mates from the ground or low vegetation (Ridsdill Smith 1970a,b; Alcock 1981, Alcock and Gwynne 1987; Peakall 1990), and the response of patrolling males may correspond with this pattern. In the case of Neozeloboria proxima, the range of heights that females call from is unknown, but given our results, we predict that the typical calling height is between 15 cm and 22 cm.
Because the thynnine pollinators of C. trilabra preferred low but not the lowest flowers they have the potential to exert strong stabilizing selection for floral height. However, several conditions must be met for selection to operate. First, variation in floral height must be heritable. Indirect evidence suggests that floral height may be genetically controlled. (1) Despite very different preflowering seasonal conditions including drought in 1991 and extensive rain in 1992, the floral height of C. trilabra was similar between years. (2) Within this genus of 22 species floral height in many species is relatively invariant. For example, species within the Chiloglottis gunnii complex typically have flowers less than 5 cm above the ground, whereas species belonging to the Chiloglottis reflexa group, such as C. trilabra, have flowers in the range 5 cm-15 cm (Jones 1988, 1991). Second, for any pollination system, variation in the phenotype can result only in selection if there is a difference in pollination success and seed set among phenotypes. In this sexual deceptive case, although we found a significant preference for flowers in the 15 cm-22 cm range, we observed some flower visits and pseudocopulation at all heights. Consequently, given sufficient pollinators all flowers could be visited, and if pollen removal and deposition were equal among heights, there would be no fitness differential through either male or female function. Pollinator preferences can drive stabilizing selection only when height is both heritable and pollen removal or deposition is limited.
The experimental results suggest an optimal floral height between 15 cm and 22 cm, however, the distribution of floral heights for the species ranges from 5 cm-15 cm. Discrimination between 15-cm versus 8-cm flowers was particularly great relative to the small difference between the two heights. For example, the comparison of means via analysis of variance revealed the mean proportion of visits for 15 cm versus 8 cm was significantly greater than for 15 cm versus 30 cm but equivalent to that for the 15-cm versus 50-cm treatment. The natural range of floral heights thus appears to correspond with the region of strongest discrimination by the pollinator.
Excerpt: Handel, Steven N., and Rod Peakall. “Pollinators discriminate among floral heights of a sexually deceptive orchid: implications for selection.” Evolution 47.6 (1993):