Pollen and pollinators in legumes
Kew scientist Hannah Banks, a member of the Comparative Plant and Fungal Biology team, speculates on whether we can use pollen structure to predict which species are vulnerable or resilient to environmental change.
The huge diversity of legume pollen
Scientists at Kew have been studying pollen structure for many years, and trying to unravel how and why certain structures have evolved. In this blog, I will focus on how structures in pollen of caesalpinioid legumes might be adapted to aid their dispersal by particular insects, mammals or even by wind.
The legume family is one of the most spectacular examples of evolutionary success in the flowering plants. Members of this family are found in all terrestrial habitats from the equatorial tropics to the edges of dry and cold deserts, dominating large parts of African and Australian savannas, plus some South American and African forests.
The family Leguminosae exhibits wide-ranging ecological success and is hugely diverse in both structural and physiological traits. Legumes are also one of the most economically important families, providing many important protein-rich food crops (for example peas, beans, lentils and peanuts), animal fodder (for example alfalfa, clover and vetch), and wood for building shelter and burning as fuel. In addition, the family is agriculturally important because many species have nitrogen-fixing bacteria associated with the roots, providing a vital source of this important macronutrient. Recent systematic studies have changed our understanding of legume classification, but my current focus is on the traditional subfamily Caesalpinioideae (about 171 genera, 2,250 species).
Pollen morphology is diverse in legumes. Some features appear to have evolved repeatedly in the family, possibly associated with factors such as available pollinators, the need to avoid excessive desiccation, facilitation of pollen germination, and changes to the shape of the pollen wall due to dehydration and re-hydration. To explore how the diversity of pollen structures may have evolved, I have mapped the pollen data onto the family tree of legumes, which was mostly generated using phylogenetic data from DNA studies.
Bird and bat pollination
Most legumes are pollinated by insects such as bees producing Acacia or clover honey. Pollination by birds (ornithophily) and bats (chiropterophily) is relatively uncommon in flowering plants, though perhaps records are sparse as it is difficult to determine the pollinators of remote or inaccessible species. However, both ornithophily and chiropterophily occur in Caesalpinioideae. At least 13 caesalpinioid genera include species that are bird pollinated, and 11 genera include species that are bat pollinated. It appears that ornithophily and chiropterophily are often associated with round lumps (termed verrucae or gemmae), or rope-like structures (striae) covering the surface of the pollen (Fig.1). These structures might increase the surface area of the grain, which when released from the mature anthers is covered in sticky substances that help it stick to the pollinator.
Flowers that attract birds and bats are relatively large and sturdy and therefore demand considerable energy requirements from the plant. However, the benefits are substantial because these large pollinators can carry relatively large pollen loads and travel great distances. The potential benefits of bird and bat pollination are important in habitats where insect activity is limited by harsh climatic conditions, or where the distances between individual trees might be great. Some studies have demonstrated a correlation between pollen grain size and pollinator body size, and perhaps also with the size of the flower, particularly the length of the style.
Pollen of most species is dispersed as individual grains that have separated from each other during development. However, pollen of some angiosperms is clumped together into groups by various means, perhaps facilitating efficient dispersal. Some pollen grains have strong threads that link them together. In the legume family, pollen-connecting threads occur in a few species from only four genera: Arapatiella (Fig. 2), Caesalpinia, Delonix, and Jacqueshuberia, which are all associated with bird or bat pollination. Indeed, Jacqueshuberia is the only caesalpinioid genus that is exclusively bat pollinated.
In some species, pollen grains do not separate during development, but mature pollen is released permanently connected in groups of four (known as tetrads). In the same way that pollen-connecting threads could facilitate more efficient pollination, this type of pollen aggregation could have similar benefits. Permanent tetrads are the most common form of pollen aggregation in flowering plants, but the selective advantages of this feature remain elusive. Some reports suggest that permanent tetrads occur in species that are visited relatively infrequently by pollinators. Tetrads have apparently evolved independently four times in caesalpinioid legumes, in a few species of Bauhinia, Diptychandra, Afzelia, and one species of Dinizia (Fig. 3). Tetrads are even more common in mimosoid legumes, but are entirely absent from papilionoid legumes.
Wind pollination in caesalpinioid legumes
In two closely related caesalpinioid legumes, Colophospermum and Hardwickia, pollen is dispersed not by insects or vertebrates but by the wind. Wind pollination might be expected to expose the pollen to dehydration, which might help to explain the characteristic pollen structures of these two species. The openings (apertures) are reduced to very small pores, like those of grasses, potentially reducing the likelihood of desiccation by water loss. Many pores are scattered over the surface of the grain, which could increase the possibility of the small pores coming into contact with the stigma; this could facilitate germination of the pollen tube so that fertilisation can occur. These pollen grains are relatively light and thin-walled, which would make them more readily carried by air currents.
Future perspectives in legume pollen studies
Much remains to be learned about pollination in legumes, particularly for those species growing in remote and inaccessible places. For many species, the pollinators are unknown but the pollen structures resemble those of species where the pollinators are known. The question arises whether one can predict the pollinator based on pollen structure. To do this, it is vital to distinguish between structural similarities arising from a shared evolutionary history and those from convergent evolution due to functional constraints.
We do not know how many species rely on specific pollinators or specialised groups of pollinators and would therefore be at risk if those pollinators were lost. Climate change, deforestation, and other ecosystem changes are likely to threaten species with highly specialised, co-adapted pollination syndromes. To overcome this lack of data, we are currently investigating whether we can use data on pollen morphology to predict the level of pollinator specialisation. If this is possible, we may then be able to predict those species most at risk of extinction due to loss of pollinators.
Further, if we can understand the function of particular pollen structures, for example those that maintain viability in dry conditions, this may help predict which species will be vulnerable or resilient should rainfall patterns change. Because legumes are not only one of the largest, but also one of the most economically important flowering plant families, understanding how and why their evolution has been so successful in conquering their huge spectrum of habitats could contribute to our understanding of how they will continue to perform in future environments.