New (and old) aspects of the island syndrome in plants on New Zealand’s outlying islands

For reasons not fully understood, plant communities on islands differ predictably from mainland ones. For example, plants with herbaceous relatives on the mainland are often woody on islands (Zizka et al., 2022). Island flowers are often white and inconspicuous (Whittaker et al., 2023). This suite of consistent differences between island and mainland species has been termed the island syndrome. The island syndrome can arise from processes of biased colonization (i.e. differences in dispersal and/or establishment) and/or in situ evolution (i.e. evolution on the island, after colonization) (Carlquist, 1974; Barrett, 1996; Ottaviani et al., 2020). The island syndrome is also shaped by abiotic (i.e. climatic and geographic variables) and biotic (i.e. co-occurring island organisms) factors (Burns, 2019). This thesis investigates insular changes in a range of plant functional traits and tries to disentangle the relative roles of biased colonization and in situ evolution, as well as abiotic and biotic factors, in shaping the island syndrome in plants. First (Chapter 2), I explored flower size changes in plants that colonized islands from the New Zealand and Australian mainland. I asked whether pollination mode causes flower size to evolve differently after island colonization, leading animal-pollinated species to follow the island rule (an evolutionary trend wherein small organisms become large and large organisms become small after island colonization). Animal-pollinated flowers followed the island rule, but wind-pollinated flowers did not, instead increasing consistently after colonization. Next (Chapter 3), I explored the relative roles of biotic and abiotic factors in influencing island colonization and establishment. I asked whether Baker’s rule, predicting that the incidence of dioecy among early island colonizers be lower than that of the mainland source pool, can result from trait correlation rather than insularity. The incidence of dioecy was found unrelated to insularity but regulated by dioecy-related traits and climate. Third (Chapter 4), I explored the relative roles of colonization biases and in situ evolution in shaping the island syndrome in plants. I asked whether trait convergence among island Coprosma can result from a combination of in situ evolution and colonization biases. Leaf size was found to be favoured during island establishment, while both leaf size and stature consistently increased on islands as a result of in situ evolution Finally (Chapter 5), I shed some light on the long-standing loss of divarication hypothesis. I asked whether the divaricate habit was actually lost on islands, or whether the trait evolved in the mainland sister taxon after island colonization, thus never appearing on islands. In most cases, island endemics had never been divaricate. Still, when present on islands, divarication was consistently lost. These findings challenge the pairwise comparison method, a long-standing analytical method of island biogeography. Overall, this thesis reveals novel patterns in plant biology and helps disentangle the distinct roles of evolution and colonization biases, as well as biotic and abiotic factors, in shaping plant traits and communities on islands. It also illustrates how investigating changes in plant traits can help identify the evolutionary mechanisms driving island life, and plant life in general.