Phylogenetic studies are providing powerful new insights into the evolution of complex traits. Metal hyperaccumulation is an unusual and complex physiological trait found in about 500 plant species and is associated with an exceptionally high degree of tolerance of metalliferous soils. Alyssum L. (Brassicaceae) is the largest known hyperaccumulator genus, comprising approximately 188 species distributed throughout the Mediterranean region and south-west Asia. Approximately one-quarter of these are largely restricted to areas of serpentine soils and have the ability to accumulate nickel to high concentrations in shoot tissue. This genus provides a good example in which to study the origins of a complex physiological trait, but its phylogeny is currently poorly understood. To produce a well-resolved phylogenetic tree to investigate the number and timing of origins of nickel hyperaccumulation within Alyssum, DNA sequences were generated for four chloroplast regions (matK, rps16–trnK, trnD–T and trnL–F) from 170 of 255 species in the tribe Alysseae. Additional sequencing was carried out for the chloroplast genes ndhF and rbcL and the nuclear gene PHYA. A Bayesian analysis employing a relaxed uncorrelated lognormal molecular clock and multiple fossil-age calibration points was carried out to reconstruct a time-calibrated phylogeny of this tribe using appropriate outgroups. Optimization of the nickel hyperaccumulation trait onto the resulting phylogenetic tree suggests that nickel hyperaccumulation arose twice in the Alysseae in the late Miocene/early Pliocene: 3.3–8.3 Mya in Alyssum and 6.3–8.8 Mya in Bornmuellera. The single origin in Alyssum is strongly associated with a significant acceleration in net species diversification rate, suggesting the ability to hyperaccumulate nickel could have provided a key evolutionary innovation facilitating rapid range expansion and subsequent species diversification. The scattered distribution of nickel hyperaccumulators across small island-like patches of serpentine soil suggests that allopatric speciation may have driven rapid diversification in this clade.

This thesis addresses the methodological challenges of determining the variability of large herbivore populations through time and their impact on European vegetation. Large herbivores are at the heart of conservation policy. However, opinions widely diverge on whether we should aim for fewer herbivores and managed populations or, on the contrary as advocated by the rewilding movement, more herbivores and self-regulating populations acting as ecosystem engineers. This controversy has roots in a debate regarding the nature of ecosystems before the prevalence of human activities. Baseline ecosystems are either described as continuous forest cover with passive large herbivores, or, in contrast, as mosaics with patchy forest cover driven <i>inter alia</i> by bison, aurochs and horses, now rare or extinct in Europe. The main obstacle in moving this debate forward is a poor understanding of large-herbivore densities in the past. I analysed modern pollen and spore assemblages from known environmental settings to improve palaeoecological interpretation of fossil assemblages dating from the pre-human (baseline) period. The sites investigated are the rewilded grasslands of the Oostvaardersplassen (The Netherlands), the mosaic habitats of The New Forest (UK) and the old-growth closedcanopy forest of Białowieża (Poland). I demonstrate that the common practice of interpreting pollen percentages fails to estimate past forest cover in situations with natural grazing. As an explanation, I suggest that pollen productivity fluctuates with biotic factors such as herbivory and canopy shading. As a result, new insights into the baseline debate require additional lines of evidence. In this thesis, I develop an existing methodology to reconstruct past herbivore presence using fossil dung fungal spores. I synthesise current knowledge of this method with an emphasis on spore identification and, finally, I demonstrate that dung fungal spore abundance in lake sediments can be translated into large herbivore numbers. The evidence presented in this thesis contributes to the debate on re-wilding and addresses a fundamental challenge of nature conservation in the human-dominated landscapes of Europe.

The presence of exotic plants both in Africa and in Asia has long attracted the attention of scholars who have attempted to understand the human activities linked to them. Archaeological and ethnographic evidence for the reconstruction of these activities is often very limited, but indirect methods such as the study of DNA have become useful tools in building models of early human dispersal. Among the plants that were carried across the Indian Ocean, sometimes known as the “tropical food kit”, the staple crop taro (Colocasia esculenta (L.) Schott) has continued to be the subject of ongoing research. The use of this crop in antiquity is well documented by discoveries of ancient taro starch granules found on archaeological artefacts from sites in Southeast Asia and the Pacific islands, making it one of the oldest plants consumed by people. However, less is known about the use of taro in Africa and the Mediterranean region, where it is found both in the wild and under cultivation - often representing a staple crop in Sub-Saharan Africa. In this doctoral thesis, genetic analysis was performed on modern samples of taro collected from Africa and other regions of the Indian Ocean, using four molecular markers. Two main clusters have been identified, and within this main sub-division four populations of taro have been detected in Africa. By integrating the genetic results with historical and linguistic research, and extensive ethnobotanical fieldwork in Africa, two of these populations are proposed to represent early translocations, with modern distribution patterns suggesting diverse dispersal routes at different times. These results open up a new scenario in which the “tropical food kit” is finally unpacked, with important historical implications for each of the crops contained within it.

Rab GTPases are major regulatory proteins of vesicle traffic and thus responsible for membrane identity, vesicle targeting and vesicle fusion. The angiosperm Rab GTPase family is grouped into eight clades (Rab-A to Rab-H) that are broadly conserved in animals and yeasts. It has been proposed that the Rab-A clade has diversified in land plants giving rise to six plant- specific structural subclasses, Rab-A1 to Rab-A6. Previous work suggests that the Arabidopsis Rab-A2 and Rab-A3 proteins define a novel endosomal compartment that lies on a pathway between the Golgi and the plasma membrane. In dividing cells, the Rab-A2/A3 compartment is implicated in biosynthetic traffic to the cell plate but it is unclear what traffics through this compartment in non-dividing cells. In this project, I investigated a range of membrane trafficking pathways in Arabidopsis thaliana. These were probed for dependency on RAB-A2a function, using the dominant negative approach combined with fluorescent marker technology. The data presented in this thesis suggests that RAB-A2a acts on a protein recycling pathway that is used by PIN2:GFP.

Little is known about the composition of tropical epiphytic communities and the influence of environmental variables on community composition. In this thesis I quantify the diversity and biomass of bromeliads, and other vascular epiphytes along an altitudinal transect on the eastern slope of the southeast Peruvian Andes and then look for species’ adaptations related to patterns of diversity and biomass. I compare patterns with those of woody species. Bromeliad species, like tree species, were found to form ecological zones related to climate. The lowest altitude ecological zone (below 1250 m) is the lowland rainforest (LRF), which has the warmest climate and highest evapotranspiration. In LRF vascular epiphytes are less prominent than other ecological zones, with the lowest bromeliad species richness and lowest vascular epiphyte biomass. However, low water-availability gives rise to most variable shoot morphology of bromeliads. The tropical montane forest (TMF), between 1250 m and 2250 m, is intermediate in climate between the LRF and the tropical montane cloud forest (TCF). The TMF has the highest α-diversity, but species richness is lower than the TCF. The shoot morphology of bromeliads is intermediate between TCF and LRF. The highest altitude ecological zone with forest is the TCF (above 2250 m). The TCF has the highest bromeliad species richness, and lowest diversity of shoot forms. The low diversity of shoot forms represents the need for a large phytotelm (water-impounding shoot) to intercept and store precipitation. The TCF has the highest vascular epiphyte biomass, although the biomass is variable as a consequence of the natural disturbance caused by landslides. Along the transect bromeliad species with CAM photosynthesis are only present in the LRF. Terrestrial bromeliad distribution records covering the Neotropics show CAM photosynthesis is more prevalent in drier environments showing that CAM photosynthesis is primarily an adaptation to drought. Epiphytic bromeliads, pre-adapted to a water-stressed environment show no differences in presence along rainfall gradients, but species with CAM photosynthesis occupy warmer environments.

A key question in developmental biology is how form is generated. The model species Arabidopsis thaliana produces simple leaves with marginal outgrowths termed serrations. Serration development in A. thaliana requires both the transcription factor CUP-SHAPED COTYLEDON2 (CUC2) and the auxin efflux facilitator PIN-FORMED1 (PIN1), which regulates polar auxin transport by forming convergence points (Hay et al., 2006; Nikovics et al., 2006; Scarpella et al., 2006). In Chapter 3, I investigate how CUC2, PIN1 and auxin interact to control serration development. I demonstrate that CUC2 promotes PIN1 convergence point and auxin activity foci formation along the margin of the leaf, whilst high auxin activity represses CUC2 expression. Furthermore, interspersed peaks of CUC2 and auxin activity pattern serration development along the proximo-distal axis of the leaf. Thus, auxin, PIN1 and CUC2 form a negative feedback loop that patterns serration development. CUC genes and PIN1 are required for leaflet development in Cardamine hirsuta (Barkoulas et al., 2008; Blein et al., 2008), a close relative of A. thaliana that produces compound leaves subdivided into units termed leaflets. However, it is unclear how CUC and PIN1 interact to control leaflet development. In Chapter 4, I demonstrate that similar to A. thaliana, CUC genes promote PIN1 convergence point and auxin activity foci formation at the C. hirsuta leaf margin, whilst high auxin activity represses CUC2 expression. These genetic interactions likely create interspersed peaks of CUC2 and auxin activity that pattern leaflet development. Thus, the same negative feedback loop between CUC, PIN1 and auxin patterns both leaflet development in C. hirsuta and serration development in A. thaliana. KNOTTED1-LIKE HOMEOBOX (KNOX) genes are expressed in C. hirsuta leaves, and interact with ChCUC and PIN1 in positive and negative feedback loops, respectively, to control leaflet development (Barkoulas et al., 2008; Blein et al., 2008). KNOX genes are not expressed in A. thaliana leaves, but deeply lobed margins reminiscent of leaflets develop in association with ectopic KNOX expression in leaves (Chuck et al., 1996; Hay et al., 2006). However, it is unclear whether regulatory interactions of PIN1, CUC and KNOX which occur in C. hirsuta leaflets are employed during KNOX-induced lobe development in A. thaliana. In Chapter 5, I demonstrate that CUC2 and polar auxin transport are required for ectopic KNOX expression. Conversely, I show that KNOX misexpression up-regulates CUC2 expression in A. thaliana leaves. Thus, interactions between KNOX, CUC and PIN1 that occur in leaflet development in C. hirsuta also occur in association with KNOX-induced lobe development in A. thaliana. In addition to investigating the regulatory interactions between known components of leaf development pathways, I sought to identify novel genes that mediate CUC2-dependent serration development in A. thaliana. In Chapter 6, I identify a suppressor of the smooth margin phenotype of cuc2 leaves that partially restores PIN1 localisation in the absence of functional CUC2. Finally, in the General Discussion I evaluate how interlinking feedback loops between CUC, KNOX and auxin pattern serration and leaflet development. I then discuss why interlinking feedback loops may have been deployed to control outgrowths in both plant and animal systems.

Tomatoes are the fourth most valuable commodity in agriculture after rice, wheat and soybeans globally with 151 million tonnes of fruit being produced in 2012. The tomato fruit is also a model system for fleshy fruit development. During ethylene-regulated fruit ripening there are complex changes in fruit chemical composition due to degradation and synthesis of a number of soluble and volatile metabolites. Ultimately, these changes control the composition of the ripe fruit and dictate its flavour and texture. It is known that ripening can proceed when mature green fruit are removed from the plant (and indeed this is standard commercial practice) but the extent to which metabolic changes are sustained when fruit are ripened in this way has yet to be established. A modelling approach such as constraints-based modelling can provide system-level insights into the workings of the complex tomato metabolic network during ripening. The first aim of this thesis was therefore to construct a genome-scale metabolic network model for tomato and to use this model to explore metabolic network flux distributions during the transitions between the stages of fruit ripening. The flux distributions predicted provided insight into the production and usage of energy and reductants, into routes for climacteric CO₂ release, and the metabolic routes underlying metabolite conversions during ripening. The second aim of this thesis was to use the model to explore metabolic engineering strategies for increased production of lycopene in tomato fruit. The model predictions showed that rearrangement of dominant metabolic fluxes were required to cope with the increased demand for reductants at high lycopene accumulation, which came at a cost of a lower accumulation of other secondary metabolites. Overall the thesis provides an approach to connect underlying metabolic mechanisms to the known metabolic processes that happen during ripening.

Germination rates and germination fractions of seeds can be predicted well by the hydrothermal time (HTT) model. Its four parameters hydrothermal time, minimum soil temperature, minimum soil moisture, and variation of minimum soil moisture, however, must be determined by lengthy germination experiments at combinations of several levels of soil temperature and moisture. For some applications of the HTT model it is more important to have approximate estimates for many species rather than exact values for only a few species. We suggest that minimum temperature and variation of minimum moisture can be estimated from literature data and expert knowledge. This allows to derive hydrothermal time and minimum moisture from existing data from germination experiments with one level of temperature and moisture. We applied our approach to a germination experiment comparing germination fractions of wild annual species along an aridity gradient in Israel. Using this simplified approach we estimated hydrothermal time and minimum moisture of 36 species. Comparison with exact data for three species shows that our method is a simple but effective method for obtaining parameters for the HTT model. Hydrothermal time and minimum moisture supposedly indicate climate related germination strategies. We tested whether these two parameters varied with the climate at the site where the seeds had been collected. We found no consistent variation with climate across species, suggesting that variation is more strongly controlled by site-specific factors. / Keimungsgeschwindigkeit und Anteil gekeimter Samen lassen sich gut mit dem Hydrothermalzeit-Modell bestimmen. Dessen vier Parameter Hydrothermalzeit, Mindesttemperatur, Mindestbodenfeuchte und Streuung der Mindestbodenfeuchte müssen jedoch durch aufwendige Keimungsversuche bei Kombinationen von mehreren Temperatur- und Feuchtigkeitsstufen bestimmt werden. Für manche Anwendungen des Hydrothermalzeit-Modells sind aber ungefähre Werte für viele Arten wichtiger als genaue Werte für wenige Arten. Wenn die Mindesttemperatur und die Streuung der Mindestfeuchte aus Veröffentlichungen und Expertenwissen geschätzt würde, können die Hydrothermalzeit und Mindestbodenfeuchte aus vorhandenen Daten von Keimungsversuchen mit nur einer Temperatur- und Feuchtigkeitsstufe berechnet werden. Wir haben unseren Ansatz auf einen Keimungsversuch zum Vergleich der Keimungsquote wilder einjähriger Arten entlang eines Trockenheitsgradienten in Israel angewendet. Mit diesem Ansatz bestimmten wir die Hydrothermalzeit und Mindestfeuchtigkeit von 36 Arten. Der Vergleich mit genauen Werten für drei Arten zeigt, dass mit unserem Ansatz Hydrothermalzeit-Parameter einfach und effektiv bestimmt werden können. Hydrothermalzeit und Mindestfeuchtigkeit sollten auch bestimmte klimabedingte Keimungsstrategien anzeigen. Deshalb testeten wir, ob diese zwei Parameter mit dem Klima am Ursprungsort der Samen zusammenhängen. Wir fanden jedoch keinen für alle Arten übereinstimmenden Zusammenhang, so dass die Unterschiede vermutlich stärker durch standörtliche als durch klimatische Ursachen hervorgerufen werden.

Japan is one of the world’s 34 biodiversity hotspots, according to Conservation International (CI). The methods used by various organisations to define priorities differ, however, and all have weaknesses when trying to identify hotspots at finer resolutions. The goal of this thesis is to investigate how biodiversity hotspots in Japan could be revealed and mapped in order to encapsulate conservation elements of biodiversity in practical ways and at various scales. Bioquality is a term that emphasises the concentration within a community of elements of biodiversity with high conservation value. It evaluates the global rarity and taxonomic distinctiveness of plant species or infra-specific taxa using four Star categories. At a plant community level, the Genetic Heat Index (GHI), which is a standardised global range size rarity score, is calculated using weighted Star statuses of species in the community. Bioquality hotspots are assessed here for the first time for the flora and vegetation in Japan – and for temperate Asia – by categorising the Japanese flora into Stars and by applying GHI to survey data and literature-based sources. Keys to Stars are developed for the Japanese flora, with adjustments for variability in species geographic range size information and for taxonomic relatedness. A Flora of Japan (FOJ) database was compiled as a BRAHMS database, containing 8,262 accepted names (30,656 taxon names in total, including synonyms) in 258 families – the first full database of Japanese vascular plants. A total of 7,145 taxa are assigned Stars; from the rarest to the widespread class, there are 884 Black, 756 Gold, 833 Blue, and 4,672 Green Star taxa, confirming that Japan as a whole contains a high proportion of globally rare taxa (23% taxa in Black or Gold). A protocol for calibrating the weight of Stars based on species geographic range is developed based on fine-resolution distribution maps within Japan and coarse–resolution Taxonomic Database Working Group (TDWG) code information. The protocol optimises calculation for temperate regions. The first ever bioquality hotspot maps of Japan are produced using two independent data sources on species distribution at national level: 1) 50 botanical prefectures using 4,830 species from the FOJ database; 2) 1,418 Horikawa ‘geoquadrats’ (0.1° latitude by 0.15° longitude grid) maps covering 829 species. The Ryukyu Islands and Ogasawara Islands are identified as bioquality hotspots, and high mountain ranges in mainland Japan are predicted to contain areas potentially high in GHI; the spatial patterns of GHI are generally concordant between maps of different resolutions. These findings highlight that bioquality assessment can be applied meaningfully at various spatial resolutions. Using field sampling data and existing literature, three study sites are further investigated on a local level: 1) the satochi-satoyama landscape, the current national priority area for biodiversity conservation; 2) various vegetation types of Okinawa-jima Island, the Ryukyu Islands; and 3) the Utaki sacred groves within the predicted hotspot of the Ryukyu Islands. The Ryukyu Islands are confirmed to contain bioquality hotspots within many individual sites, while there was generally low GHI across the satochi-satoyama landscape. The field study outcomes, together with a gap analysis of the existing coverage of protected areas, highlight three important points that are directly relevant to national biodiversity conservation planning: 1) the Ryukyu Islands urgently need newly designated protected areas; 2) the satochi-satoyama landscape conservation should redirect its focus on cultural benefits to the public; 3) the existing protected areas, particularly on mountain areas, need re-evaluation in terms of upgrading their status in light of the bioquality assessment.

Tomato is an important food crop and a model for fleshy fruit development. The process of fruit ripening involves changes in chemical composition and in particular the accumulation of sugars, organic, amino acids and carotenes. The research described in this thesis aimed to identify key regulatory aspects associated with the accumulation of the major acids in tomato fruit by analysis of introgression lines resulting from a cross between a cultivated variety, Solanum lycopersicum, and a wild progenitor species, Solanum pennellii. Line 2-5 showed increases in citrate, malate, aspartate and glutamate in fruit grown under greenhouse conditions. The genetic differences between line 2-5, its overlapping lines, sub-introgression lines and the recurrent parent were used to link the metabolite phenotypes to smaller chromosomal regions. This analysis suggested multiple epistatic loci control fruit metabolite accumulation. Investigation of the biochemical differences between line 2-5 and the recurrent parent revealed that organic and amino acid accumulation did not dependent upon increased TCA cycle capacity. Regulation at the metabolic level was identified for citrate accumulation with changes in cytosolic aconitase in line 2-5. As these metabolites accumulate in the vacuole, tonoplast transport was investigated. Correlation of ATPase-dependent malate influx with altered malate content suggested malate tonoplast transport plays a role in malate accumulation and highlights the importance of vacuolar storage and transport in the regulation of organic and amino acid accumulation.