Floral Construction and Pollination Biology in the Lamiaceae

Flowers are functional units for sexual reproduction. They balance several conflicting demands. These include offering pollen as a floral attractant vs. saving pollen for reproduction; availability of resources vs. the need for elaborate displays to attract pollinators; and ensuring the production of offspring by self-pollination vs. the need to increase genetic diversity by outcrossing. Each flower can thus be regarded as a species-specific compromise for managing sexual reproduction. Its ‘floral construction’ is defined as the syndrome of adaptive characters needed for pollen transfer that have evolved under phylogenetic, developmental and environmental constraints. The ‘bilabiate blossom’ characterizing the Lamiaceae is a construction for nototribic (dorsal) pollen deposition. It has evolved many times in parallel and illustrates that distantly related plants can approximate to similar functional solutions and that closely related plants can realise these solutions by diverse morphological means.

The five articles presented in this ‘Highlight’ section of Annals of Botany are based on talks given at the XVII International Botanical Congress, Vienna, Austria, 2005. They focus on the phenotypic diversity and adaptive significance of the bilabiate flower in the Lamiaceae. Starting from the general concept of a bilabiate blossom in the angiosperms (Westerkamp and Claßen-Bockhoff, 2007), three papers focus on Salvia and its characteristic lever mechanism. Walker and Sytsma (2007) reconstruct the multiple origin of the lever-like stamen construction, illustrating that even this particular feature has evolved in parallel. Reith et al. (2007) question the adaptive significance of the lever mechanism in Salvia pratensis, calculating its utility as a biomechanical tool for pollen dispensing. Wester and Claßen-Bockhoff (2007) illustrate the morphological diversification of the Salvia flower due to the shift from bee to bird pollination, in which relative proportions and positions play a considerable role. In the fifth paper, Rodríguez-Riaño and Dafni (2007) investigate the floral construction of different flower morphs in Teucrium capitatum and Origanum syriacum with respect to the breeding system and the evolution of separate sexes in plants.

Westerkamp and Claßen-Bockhoff (2007) widen the concept of the bilabiate flower and describe it as a general pattern. To become independent of morphological and taxonomic associations, the bilabiate blossom is characterized as a (reversible) one-way construction that in most cases offers nectar at its base. Ventrally, it is confined by a ‘floor’ and dorsally by a ‘roof’ covering the reproductive surface (i.e. pollen and stigma). In most cases, the visitor path is divided into the distal ‘pollination chamber’ and the proximal ‘alignment channel’. Usually, it is the narrow ‘alignment channel’ that defines the exact positioning of the visitor. Floor and roof are at such fixed a distance that legitimate visitors inevitably contact the reproductive structures with their dorsal side. Bilabiate flowers are thus nototribic by definition. They obviously evolved to protect pollen against pollen-collecting bees. Social bees have to nourish offspring and often maximize pollen harvest to such a degree that almost nothing is left for pollination. To reserve pollen for its original purpose, flowers hide pollen and place it on the insect in places out of the immediate reach of their legs. The most effective depositions are on the insect's back where access is impeded, especially in flight when grooming and reloading usually occur. The bilabiate flower construction evolved in many angiosperm groups in parallel, thus representing a convincing example for parallel evolution. Starting from the basic constrictions of the bilabiate construction, the authors summarize the main trends of diversification with emphasis on the Lamiales.

Parallel evolution is also found in Salvia, with about 1000 species. It is the largest group in the family and, moreover, characterized by the well-known staminal lever mechanism. Considering the derived stamen construction as a morphological support for the assumed monophyly of the genus, Himmelbaur and Stibal (1932–34) have long postulated parallel stamen transformations in the Old and New World. In the second article, Walker and Sytsma (2007) not only confirm this view but document multiple origins of the lever-like construction by referring stamen morphology to the phylogeny of the Mentheae. Based on a previous paper elucidating the polyphyly of Salvia (Walker et al., 2004), they distinguish nine different stamen morphologies in the three distinct lineages of Salvia. Starting from the bithecate stamen with only little connective widening, they repeatedly find a four-step progression from slight and significant elongation of the connective, to loss of fertility of the posterior thecae and fusion of the posterior branches of the connectives. In view of the multiple origin of the lever-like stamens, the authors discuss the underlying morphological homology and conclude that even complex constructions can be repeated in response to similar genetic canalizations and selective regimes.

As significant speciation only occurred in those lineages employing a staminal lever mechanism, the latter is postulated to be a key innovation forcing adaptive radiation (Claßen-Bockhoff et al., 2004). In the third paper, Reith et al. (2007) follow the hypothesis that pollen dispensing is one of the main functions of the staminal levers and that the forces arising between the pollen-bearing thecae and the bee's body play a significant role. To reconstruct stress and strains by qualitative and semi-quantitative analyses they study the interaction between flowers and bees (Apis mellifera). Their analyses illustrate that the spoon-shaped lower lever of S. pratensis has a small hole through which a bee introduces its proboscis into the corolla tube. It is concluded that (1) the spatial arrangement of the upper lever arm and the bee and (2) the physical properties of the upper lever arm such as elasticity and length are important for the forces involved in the process of pollen transfer. The entire flower proves to be a biomechanical construction in which all floral parts are synorganized, i.e. operate together.

The assumed utility of the staminal levers for pollen dispensing in bee-pollinated Salvia species raises the question of its functional significance in ornithophilous species. In the fourth article, Wester and Claßen-Bockhoff (2007) present a survey of the floral construction in 186 ornithophilous Salvia species. Their most surprising result is the identification of 63 species that transfer pollen without a functional lever mechanism, because it is either reduced or immobile because of spatial constrictions or stiff joints. Instead, the species have freely accessible pollen positioned a considerable distance from the nectar reward. The parallel reduction of the staminal lever mechanism comes about as a consequence of the adaptation to birds. Exclusion of bees is realised by reducing the landing place and by elongating and narrowing the pathway to nectar by diverse morphological means. Safe pollen deposition on the bird's feathered head is achieved by enlarging the distance between nectar and pollen. Within the bird pollinated Salvia species two lines of floral constructions have evolved. One maintains the lever mechanism and integrates it into the bird–flower syndrome and the other completely reorganizes the pollination mechanism, illustrating that even proved constructions may lose their significance in the course of evolution.

Variable proportions of floral structures are not only an adequate tool for adaptation to a certain pollinator guild but can directly influence the breeding system. In self-fertile individuals the relative position of the pollen sacs and the stigmatic tissues determine the degree of geitonogamy and xenogamy. In Lamiaceae, diverse mechanisms promote outcrossing, i.e. gynodioecy, dichogamy and herkogamy. In the fifth paper, Rodríguez-Riaño and Dafni (2007) study male–female relationships in hermaphrodite and male sterile flowers of the gynodioecious Teucrium capitatum and Origanum syriacum. Both species prove to be clearly protandrous, but differ in their modes of dichogamy (style movement in Teucrium capitatum, style elongation in Origanum syriacum). Reduction in size of floral characters is associated with male sterility in both species. This is also observed in intermediate plants with different degrees of pollen abortion. These are interpreted as the result of a delayed degeneration of one of the stages of microsporogenesis compared to the male sterile flowers. The study elucidates how floral characters, such as flower size and style length, can enhance the avoidance of pollen discounting and self-pollination.

In conclusion, this collection of papers illustrates how an analysis of the basic options of a given floral construction and its adaptive modifications can contribute to a deeper understanding of the ecological and morphogenetic constraints of flower evolution. By analysing functional traits and comparing phenotypic responses across many species beyond the level of monophyletic lineages, general evolutionary trends come to the fore that elucidate the origin of parallelisms.