Diversity and evolution of the floral morphospace

Evolutionary biology witnessed during the past decades the development of genomics at the expense of morphological studies. However, phenotypes (and not nucleotide sequences) are the biological components, which interact directly with the environment. Recently, the scientific community has again begun to realize the importance of morphology in fields ranging from evolution and systematics to ecology, where morphological concepts and hypotheses are often referred to or build on. This renewed interest in morphology together with the availability of modern analytical methods and the exciting results of genomic disciplines, opens new opportunities for the study of the evolution of morphological traits. It is in this scientific context that we propose to study the evolution of morphological shapes from a phylogenetic perspective. This project has a strong mathematical component, as we seek innovative synthetic and analytical representations of shape evolution. The starting point for this project stems from the fact that not all the theoretically possible architectures have been explored during the evolution of species. Existing morphologies are restricted to a limited set because of developmental, functional, and phylogenetic constraints. We propose here to compare the occupied portion of the morphological space (or morphospace) through geological time and among non-sister taxa of angiosperms (flowering plants). More precisely, we will study the floral morphospace of the two morphologically diverse orders Laurales (magnoliids) and Ericales (asterids). The study of the evolution of morphospace filling through geological time will be based on morpho-anatomical observations of flowers of extant and fossil taxa, ancestral character state reconstructions, and dated molecular phylogenies. The morphospace approach provides a mathematical conceptualization which is a prerequisite to make previously unknown patterns of shape evolution emerge and be detectable. Hence, new questions about physical and biological rules, convergence and contingency at play during the evolution of natural shapes can be addressed. This project is highly innovative, not only because it will bring new insights into the evolution of morphospace filling in non-sister taxa of angiosperms, but most of all because it is intrinsically transdisciplinary, integrating knowledge and methods from fields as different (but complementary) as paleobotany, modern morphology, molecular phylogenetics, mathematics, and theoretical biology.

The past two decades have seen rapid and tremendously successful developments of molecular techniques and genomic approaches in evolutionary biology. At the same time, morphology, i.e. the description and analyses is of organismal form, seemed to have little to contribute to modern research. However, more recently, the scientific community has again begun to realize the importance of morphology in fields ranging from evolution and systematics to ecology, where morphological concepts are often referred to or built on. This renewed interest together with modern analytical tools opened new opportunities for the study of the evolution of morphological traits. In this project, we focused on the morphology and evolution of flowers using a morphospace approach. Morphospaces are mathematical spaces describing and relating the morphology of organisms. This method allowed us to quantify the distribution of floral morphological diversity across angiosperms. In addition, we reviewed the theoretical background and modern statistical tools applied in morphospace analyses. At the core of the project is an extensive floral morphological data set that we compiled for the order Ericales. Next to many economically important taxa such as tea, kiwi, persimmon, and ebony, Ericales also include several horticulturally important groups such as primroses and rhododendrons. The data set comprises 381 species representing 275 genera and all 22 families of the order. Our data set also contains data for several well-preserved fossil flowers from the Cretaceous. We scored 37 floral traits and each of the currently 12,503 data entries is linked explicitly to the reference from which the information is derived. This approach ensures the reproducibility of our analyses and - likewise important - will allow us and other researchers to use our data set for other types of analyses in the future. The data set described above allowed us, for the first time, to quantify detailed floral morphological diversity in a major group of angiosperms. We show that major lineages of Ericales occupy different portions of the floral morphospace in a mosaic pattern. Our most exciting finding is that morphological diversity is unevenly distributed among different floral organs. We show that the male part of the flower, i.e. the stamens, account for significantly higher levels of morphological diversity than either female or sterile organs. This is surprising as it is the female part of the flower, which is the morphologically and functionally most complex and, therefore, has generally been considered to contribute more to total floral diversity than the other organ categories. Very few studies have so far used morphospaces to investigate the evolution of broad-scale patterns of plant morphological diversity. With this project, we have shown that these types of analyses are highly rewarding and that they have the potential to provide novel insights into the evolution and diversity of plants even in times of ever more powerful genomic approaches.