Meeting Abstract
Plants produce huge numbers of pollen grains as part of their reproductive cycle. Pollen grains are tough and widely dispersed and they also fossilise readily in a wide variety of environments. Consequently, pollen grains are widely used as an empirical record of plant evolution. In this talk I begin by outlining a few examples where pollen grains have provided data on plant biodiversity and then focus on the tropics, which harbour extraordinarily high levels of flowering plant diversity. This diversity is expressed in the morphology of flowering plant pollen grains, which are by characterised by striking variety in their shape and patterning. I approach this vast morphological diversity from the perspective of enumerative combinatorics. This involves generating pollen morphotypes by algorithmically combining character states and enumerating the results of these combinations to establish a raw un-ordinated morphospace for flowering plant pollen. I map the pollen grains of 1,008 species of neotropical angiosperms onto this morphospace, which facilitates the comparison of forms that exist in nature and forms that do not. This mapping highlights that neotropical rainforests represent an enormous reservoir of morphological diversity, but whether the patterns of morphospace occupation can be interpreted in terms of an adaptive landscape is unclear. The large number of discrete character combinations that arises from the morphological complexity of pollen grains presents a data visualisation challenge, and I have experimented with ideas ranging from multipartite graphs, to “imagespaces” in which each pixel represents a unique combination of discrete characters. The approaches I discuss here have been developed to quantify the diversity of plant life as recorded by pollen grains, but could be employed in any situation where some aspect of organismal form can be captured by discrete characters.
