Mashau, Aluoneswi Caroline2024-01-292024-01-292024https://hdl.handle.net/10539/37452A thesis submitted in fulfilment of the requirements for the degree of Doctor of Philosophy to the Faculty of Science, School of Animal, Plant and Environmental Sciences, University of the Witwatersrand, Johannesburg, 2023Geographic range size is the outcome of both evolutionary and ecological processes. Therefore both historical contingencies, and the ecological characteristics (traits) of particular species, interact to result in observed distribution patterns. These distribution patterns are also changing – expanding as species invade into new ecosystems, and shrinking as species are filtered from ecosystems due to climate change or changing land management. Understanding current distributions and range sizes is therefore important for helping explain biogeographic patterns and processes, for informing conservation action, the management of invasive plants, and interventions to adapt to climate change. The grass family (Poaceae) covers approximately 31– 43% of the land surface globally, and started to spread during the Miocene period (approximately 8–20 Million years ago) to achieve its current dominance. This would have occurred through rapid range expansion as well as speciation and has resulted in some species with almost cosmopolitan (global) distributions, as well as rare grass species found in only a few localities. This study aims to understand the drivers of range size and dispersal traits of grasses in Africa with the purpose of quantifying differences between clades and functional types, and determine the floral traits that likely influence dispersal modes. In Chapter 1 I compiled general introduction of the whole thesis including background, rationale, aims and objectives. In Chapter 2 I aimed to understand the geographical distribution of grasses in sub-Saharan Africa with reference to key plant traits thought to affect range size in this family (Poaceae). Specifically, to test hypotheses on the importance of plant height and lifespan in determining range size and invasion potential in the context of their evolutionary history. The range sizes of 757 grass species native to southern Africa were estimated for the sub-Saharan African region from geo-referenced herbarium records using the alpha hull function. Phylogenetic generalised least squares models and linear mixed effects models were fitted to test whether grass range size was related to plant height and lifespan. Tribe-level relationships between range size and plant height were assessed with linear models. For species introduced to other continents, generalised linear mixed effects models were fitted to test whether invasiveness was related to native range size, plant height and lifespan. Differences in native range size 4 among species in four invasion-related categories were assessed with linear mixed effects models. Geographic range sizes were larger for taller grass species and for species with shorter lifespans. The relationship between plant height and range size varies widely among tribes, with some environmentally-restricted tribes not showing significant responses to plant height. Grasses with larger native range sizes and shorter lifespans are more likely to become invasive after being introduced to other continents. Grass species introduced to other continents have larger native range sizes than those that have not, and native range size increases along the introducednaturalised-invasive continuum. The increased dispersal opportunities of annual-biannual grasses appears to have a greater positive effect on range size than do the longer generation times of perennial grasses. Grass height has and continues to be an important driver of grass biogeography, with implications for understanding the spread of certain grass tribes over the Miocene. Factors that promote large native range sizes are also likely to increase the probability of a species becoming invasive. Grass floral structures vary greatly but we have very little understanding of their functional significance. Due to the varied dispersal mechanisms shown by grasses, certain syndromes of floral traits would likely be associated with particular strategies for dispersal, and consequently, different environments. In particular, effective seed maturation and dispersal in fire-prone tall grasslands would require different floral trait syndromes than in short, frequently grazed ecosystems. In Chapter 3 I quantified floral traits of nearly 200 Poaceae species from savanna and grassland ecosystems in southern Africa and explored how their floral structures co-vary and correlate with other functional characteristics such as grass height. Using field information on the dominance disturbance regime of 163 of these grass species it was tested whether certain floral traits are more associated with fire vs grazing. Non-metric multi-dimensional scaling (NMDS) was used to illustrate how floral traits covaried among grass species, and to group them into syndromes based on these traits. Analysis of variance (ANOVA) was used to test whether certain floral trait syndromes were more associated with fire vs grazing. I identified four clear floral trait syndromes separated largely by awn length and the presence of hooks/prickles or bristles. Longawned species were more likely to be found in frequently burned environments and were also usually taller than species without awns. Grazer-dominated systems appear to select for two different floral trait syndromes. The study has improved our ecological and taxonomic 5 understanding of how floral traits differ among the range of tribes in one family across African countries. It can help in understanding dispersal limitations in grasses and predicting which species are likely to flourish in particular grassland habitats. The grass family (Poaceae), despite having only emerged and spread in the last 50 million years, is cosmopolitan, and many species have large, almost cosmopolitan distributions. Lineage age and dispersal ability are two factors thought to explain the variation of range size and grasses show a wide range of floral structures and heights associated with different dispersal strategies. In Chapter 4 I aimed to assess how dispersal syndrome (inferred from floral structures and other functional traits) and evolutionary history affect range size in the grass subtribe Eleusininae – a tropical grass clade with variation in floral structures. Global location records for 97 grass species of 29 Eleusininae genera were used to quantify range size, and linear models were used to test the relationship between range size and interaction between plant height, and lemma awn state (absent/present), caryopsis length (mm) and genus age. Taller grass species with awned lemmas were found to have a larger range size, and this supports my hypothesis (developed in Chapter 2) that the importance of grass height in driving range size depends on the dispersal syndrome. It was found that there is no relationship between genus age versus floral and functional traits used in this analysis. The study can help to explain some of the differences in biogeographic history between different lineages and also determine dispersal syndromes. In Chapter 5 I compiled general discuss or overview of the study, including geographical distribution of the southern African grasses, grass clades co-vary according to their floral traits, conservation and management implications, limitations of this study and needs for future research and conclusion.enPoaceaeEcological processesGeographic range sizeThesis