This thesis presents two studies where natural and sexual selection have interacted to evolve sexual behaviours. The thesis uses mathematical modelling to understand how these forces have caused each behaviour to evolve. This is useful because the results allow for reflection on the potential role of sexual selection in adaptation of these species to a changing environment. The first study is of early male arrival to spring breeding grounds in migratory avian species, this is termed protandry. The study explores the main hypotheses for avian protandry and then tests the susceptibility of each hypothesis to changing environment. The second study is of convenience polyandry in species where there is conflict over mating rate. Females have multiple strategies to avoid harassive males but strategies vary in cost and success rate; she must balance her strategy use to minimise her fitness depreciation. The study identifies the main factors that cause convenience polyandry to evolve and paves the way for future studies to investigate if sexual selection over resistance strategy provides these species a future advantage in adaptation to a changing environment.

With maritime industry carrying out nearly 90% of the volume of global trade, the algorithms and solutions to provide quality of services in maritime transportation are of great importance to both academia and the industry. This research investigates an optimization problem using evolutionary algorithms and big data analytics to address an important challenge in maritime disruption management, and illustrates how it can be engaged with information technologies and Internet of Things. Accordingly, in this thesis, we design, develop and evaluate methods to improve decision support systems (DSSs) in maritime supply chain management. We pursue three research goals in this thesis. First, the Vessel Schedule recovery Problem (VSRP) is reformulated and a bi-objective optimization approach is proposed. We employ bi-objective evolutionary algorithms (MOEAs) to solve optimization problems. An optimal Pareto front provides a valuable trade-off between two objectives (minimizing delay and minimizing financial loss) for a stakeholder in the freight ship company. We evaluate the problem in three domains, namely scalability analysis, vessel steaming policies, and voyage distance analysis, and statistically validate their performance significance. According to the experiments, the problem complexity varies in different scenarios, while NSGAII performs better than other MOEAs in all scenarios. In the second work, a new data-driven VSRP is proposed, which benefits from the available Automatic Identification System (AIS) data. In the new formulation, the trajectory between the port calls is divided and encoded into adjacent geohashed regions. In each geohash, the historical speed profiles are extracted from AIS data. This results in a large-scale optimization problem called G-S-VSRP with three objectives (i.e., minimizing loss, delay, and maximizing compliance) where the compliance objective maximizes the compliance of optimized speeds with the historical data. Assuming that the historical speed profiles are reliable to trust for actual operational speeds based on other ships' experience, maximizing the compliance of optimized speeds with these historical data offers some degree of avoiding risks. Three MOEAs tackled the problem and provided the stakeholder with a Pareto front which reflects the trade-off among the three objectives. Geohash granularity and dimensionality reduction techniques were evaluated and discussed for the model. G-S-VSRPis a large-scale optimization problem and suffers from the curse of dimensionality (i.e. problems are difficult to solve due to exponential growth in the size of the multi-dimensional solution space), however, due to a special characteristic of the problem instance, a large number of function evaluations in MOEAs can still find a good set of solutions. Finally, when the compliance objective in G-S-VSRP is changed to minimization, the regular MOEAs perform poorly due to the curse of dimensionality. We focus on improving the performance of the large-scale G-S-VSRP through a novel distributed multiobjective cooperative coevolution algorithm (DMOCCA). The proposed DMOCCA improves the quality of performance metrics compared to the regular MOEAs (i.e. NSGAII, NSGAIII, and GDE3). Additionally, the DMOCCA results in speedup when running on a cluster.

Ecological interactions such as predation, competition, and mutualism are important forces that influence species evolution. Coevolution is defined as reciprocal evolutionary change in interacting species. The Geographic Mosaic Theory of Coevolution (GMTC) provides a theoretical framework to explain how collections of populations should coevolve across space. Two fundamental aspects of the GMTC are gene flow among populations and the presence of selection mosaics, which are collections of localities with particular selection regimes. Several studies have explored how phenotypic trait matching between species evolves in pairs or small groups of species. However, ecological interactions frequently form large networks that connect dozens of species present in a given community. In networks of mutualisms, for instance, the organization of interactions may affect ecological and evolutionary processes. A next step in understanding the coevolutionary process is to investigate how aspects of the GMTC affect the evolution of species embedded in interaction networks. In this dissertation, we tried to fill this gap using a mathematical model of coevolution, complex networks tools, and information on empirical mutualistic networks. Our numerical simulations of the coevolutionary model allow us to draw three main conclusions. First, gene flow affects trait patterns generated by coevolution and may favor the emergence of trait matching depending on the selection mosaic. Second, the organization of mutualistic networks influences coevolution, but this effect may vanish when gene flow favors trait matching. Intimate mutualisms, such as protection of host plants by ants, form small and compartmentalized networks that generate higher trait matching than large and nested networks typical of mutualisms among free-living species, such as pollination. Third, habitat fragmentation resulting in the disruption of gene flow should reduce the reciprocal adaptations between interacting species and at the same time promote adaptations to the local abiotic environment. In conclusion, we show that a complex interplay between gene flow, the geographic structure of selection, and the organization of ecological networks shapes the evolution of large groups of species. Our results therefore allow predictions of how environmental impacts such as habitat fragmentation will modify the evolution of species interactions / Interações ecológicas como predação, competição e mutualismo são importantes forças que influenciam a evolução de espécies. Chamamos de coevolução a mudança evolutiva recíproca em espécies que interagem. A Teoria do Mosaico Geográfico da Coevolução (TMGC) fornece um arcabouço teórico para entender como conjuntos de populações coevoluem ao longo do espaço. Dois aspectos fundamentais da TMGC são o fluxo gênico entre populações e a presença de mosaicos de seleção, isto é, conjuntos de locais com regimes de seleção particulares. Diversos estudos exploraram como o acoplamento entre fenótipos de diferentes espécies evolui em pares ou pequenos grupos de espécies. Entretanto, interações ecológicas frequentemente formam grandes redes que conectam dezenas de espécies presentes em uma comunidade. Em redes de mutualismos, por exemplo, a organização das interações pode influenciar processos ecológicos e evolutivos. Um próximo passo para a compreensão do processo coevolutivo consiste em investigar como aspectos da TMGC influenciam a evolução de espécies em redes de interações. Nesta dissertação, tentamos preencher esta lacuna usando um modelo matemático de coevolução, ferramentas de redes complexas e informação sobre redes mutualistas empíricas. Nossas simulações numéricas do modelo coevolutivo apontam para três principais conclusões. Primeiro, o fluxo gênico influencia os padrões fenotípicos gerados por coevolução e pode favorecer a emergência de acoplamento fenotípico entre espécies dependendo do mosaico de seleção. Segundo, a organização de redes mutualistas influencia a coevolução, mas este efeito pode desaparecer quando o fluxo gênico favorece acoplamento fenotípico. Mutualismos íntimos, como proteção de plantas hospedeiras por formigas, formam redes pequenas e compartimentalizadas que geram um maior acoplamento fenotípico do que as redes grandes e aninhadas típicas de mutualismos entre espécies de vida livre, como polinização. Por fim, a fragmentação de habitat, ao extinguir o fluxo gênico, pode reduzir as adaptações recíprocas entre espécies e ao mesmo tempo tornar cada espécie mais adaptada ao seu ambiente abiótico local. Em suma, mostramos que interações complexas entre fluxo gênico, estrutura geográfica da seleção e organização de redes ecológicas moldam a evolução de grandes grupos de espécies. Dessa forma, podemos traçar previsões sobre como impactos ambientais como a fragmentação de habitat irão alterar a evolução de interações ecológicas

De nombreux insectes vivent en association avec des endosymbiotes se transmettant de mère à descendants. Du point de vue de leurs hôtes, certains de ces symbiotes sont obligatoires tandis que d'autres sont facultatifs. Alors qu'elle appartient le plus souvent à cette deuxième catégorie, la bactérie intracellulaire Wolbachia est devenue obligatoire chez certains de ses hôtes. C'est notamment le cas chez deux hyménoptères parasitoïdes de drosophiles du genre Asobara. Chez A. tabida, il a été montré que Wolbachia était nécessaire à la production d'œufs viables et que le défaut d'ovogenèse chez les femelles privées de leur symbiote est associé à une perturbation de leur homéostasie oxydative. Nous avons par conséquent étudié l'impact de composés antioxydant et pro-oxydant sur les traits d'histoire de vie de l'hôte et la densité symbiotique, afin de déterminer dans quelle mesure l'homéostasie oxydative pourrait être impliquée dans la coévolution entre Wolbachia et son hôte, au point de conduire à la dépendance de ce dernier. Chez certaines populations d'A. japonica, la coévolution avec Wolbachia induisant de la parthénogenèse thélytoque a entraîné la perte de la capacité des femelles à se reproduire sexuellement. Nous avons recherché les bases mécanistiques de cette perte de traits en comparant les transcriptomes et les profils d'hydrocarbures cuticulaires – des molécules qui peuvent notamment jouer le rôle de phéromones sexuelles – de femelles sexuées et asexuées / Many insects live in association with maternally transmitted endosymbiont. From their host point of view, some of those symbionts are obligatory, whereas others are facultative. Although it usually belongs to this second category, the intracellular bacterium Wolbachia has become obligatory in some of its hosts. This is notably the case in two drosophila parasitoid of the genus Asobara (Hymenoptera). It has been shown that Wolbachia was necessary for egg production in A. tabida, and that the failure of oogenesis in females cured of their symbiont is associated with a perturbation of their oxidative homeostasis. We therefore studied the impact of antioxidant and pro-oxidant compounds on the host life-history traits and the symbiotic density to determine the extent to which the oxidative homeostasis could be involved in host–Wolbachia coevolution, and in the evolution of the dependence of the host to its symbiont. In some populations of A. japonica, the coevolution with a thelytoky-inducing Wolbachia led to the loss of the females’ ability to sexually reproduce. We researched the mechanistic bases of this loss of traits by comparing the transcriptomes and the cuticular hydrocarbons – molecules that can play the role of sex pheromones – of sexual and asexual females

Biological rhythms are thought to have evolved to enable organisms to organise their activities according to the Earth’s predictable cycles, but quantifying the fitness advantages of rhythms is challenging and data revealing their costs and benefits are scarce. More difficult still is explaining why parasites that exclusively live within the bodies of other organisms have biological rhythms. Rhythms exist in the development and traits of parasites, in host immune responses, and in disease susceptibility. This raises the possibility that timing matters for how hosts and parasites interact and, consequently, for the severity and transmission of diseases. Despite their obvious importance in other fields, circadian rhythms are a neglected aspect of ecology and evolutionary biology. The ambitions of this thesis are to integrate chronobiology, parasitology and evolutionary theory with mathematical models to obtain a greater understanding about how and suggest why malaria parasites have rhythms as well as the effect of infection on host rhythms. First, I identify how malaria parasites lose their developmental rhythms in culture, when they lack any potential time cues from the host. Next, I characterise parasite rhythms inside the mammalian host in terms of synchrony and timing and demonstrate there is genotype by environment interactions for characteristics of parasite rhythms. Then, I investigate the effect that parasite infection has on host rhythms and show there is variation between parasite genotypes in their effect on host locomotor activity and body temperature rhythms during infections. Finally, I explore which host rhythms may be driving parasite synchrony and timing and demonstrate the importance of peripheral host rhythms for the timing of malaria parasite developmental rhythms. The data presented here provides novel and important information on the role of rhythms during disease and opens up a new arena for studying host-parasite coevolution.

Müllerian mimicry theory postulate that individuals of different species benefit from decreased per-capita attack risks by sharing similar warning signals. In species-rich mimetic assemblages, there is the formation of several distinct sympatric groups of species sharing the same warning signals, often color pattern, called mimetic rings. The coexistence of multiple rings seems paradoxical considering that selection among unpalatable species should favor convergence and thus reinforce a single color pattern. Different rings evolving in distinct habitats could explain the coexistence of multiple mimicry rings. However, the way species use the habitats might influence the emergence of multiple mimicry rings. We combined mathematical modeling and numerical simulations to explore how habitat heterogeneity, abiotic selection and habitat generalist species influence the formation of mimicry rings in a community. We showed that distinct selection pressures, derived from habitat heterogeneity, favored the formation of distinctive mimicry rings. Nevertheless, just the co-existence of species was enough to drive the emergence of the rings. Simulations in which there was just biotic or abiotic selection, time for convergence was faster than when both sources of selection acted together, suggesting conflicting selective pressures exerted by environment and co-existing species. In the presence of species that was habitat generalist, species converged to similar trait values, decreasing the distinctiveness of mimicry rings. A unique mimicry ring was formed if the different habitats optima in the community were very similar or when most species were habitat generalists. Our results suggest that multiple sympatric mimicry rings are formed by a complex interplay between abiotic and biotic selection and is only possible in groups of animals in which local species composition is strongly affected by habitat heterogeneity such as butterflies / No mimetismo Mülleriano, indivíduos de diferentes espécies, ao compartilharem um mesmo sinal de advertência, beneficiam-se mutuamente devido ao menor risco de predação. Em comunidades ricas em mímicos há a formação de grupos simpátricos de espécies denominados anéis miméticos, que compartilham os mesmos sinais de advertência, como por exemplo padrões de coloração. A coexistência de anéis miméticos parece paradoxal, uma vez que, em teoria, a seleção favoreceria a convergência das espécies impalatáveis e, portanto, geraria um único padrão de cor. A evolução de diferentes anéis miméticos em habitats distintos poderia explicar a coexistência desses anéis em uma comunidade. No entanto, a maneira como as espécies utilizam esses habitats talvez influencie também a emergência dos múltiplos anéis miméticos. Utilizamos modelos matemáticos e simulações computacionais para melhor compreender como a heterogeneidade de habitats, a seleção ambiental e espécies generalistas de habitat influenciam a formação de anéis miméticos em uma comunidade. Demonstramos que diferentes pressões seletivas, derivadas de uma heterogeneidade de habitats, podem favorecer a formação de anéis miméticos. Porém, a simples coexistência das espécies é suficiente para a emergência de anéis. Em simulações nas quais só havia seleção imposta por espécies impalatáveis ou pelo ambiente, o tempo de convergência foi mais rápido do que quando as duas forças seletivas atuavam juntas. Isto demonstra que provavelmente há conflito entre a seleção biótica e abiótica, não favorecendo o mesmo ótimo fenotípico. A presença de uma espécie generalista de habitat favoreceu a convergência de espécies para um valor fenotípico similar, diminuindo a distinção entre os anéis. Um único anel foi formado quando os diferentes fenótipos favorecidos pela seleção imposta pelo ambiente eram muito similares entre habitats, ou quando muitas espécies eram generalistas de habitat. Nosso trabalho sugere que múltiplos anéis miméticos simpátricos são formados por uma complexa interação entre seleção biótica e abiótica e que só são possíveis em grupos no qual a diversidade é estruturada em pequenas escalas espaciais, como borboletas

The work in this thesis concerns the host galaxies of a class of luminous, yet heavily-obscured, quasars at z$\sim$2 - a peak epoch of both star formation and black hole accretion. Here, we seek to characterise the star-forming properties of these obscured quasars to improve our understanding of galaxy-quasar coevolution. A key issue facing host galaxy studies among populations of the most luminous quasars is being able to disentangle the galaxy emission from that of the quasar. With combined observations from the Dark Energy Survey (DES), the VISTA Hemisphere Survey (VHS) and the UKIDSS Large Area Survey (ULAS), we exploit the quasar dust extinction in our sample to demonstrate that the quasar and galaxy emission can be separated via SED-fitting in these systems. By isolating the galaxy emission in this way, we estimate instantaneous SFRs for the galaxies in our sample, based on the restframe UV emission. In general, we find obscured quasars to reside in prodigiously star forming hosts with 25 $\lesssim$ SFR$_{\rm{UV}}$ $\lesssim$ 365 M$_{\rm{\odot}}$yr$^{-1}$. Furthermore, we show that the most luminous quasars reside in the most actively star-forming galaxies, potentially indicating the same gas supply is fuelling both star formation and accretion on to the black hole. Having isolated the galaxy emission via SED-fitting, we test our ability to model the restframe-UV emission of obscured z$\sim$2 quasar hosts in 2D. Until now, morphological studies of luminous quasar hosts have typically been limited to low redshifts or relied on space-based imaging. By making use of a multi-band modelling code however, we demonstrate that it is possible to accurately infer several galaxy properties (i.e. the position of the galaxy in the image, (X,Y), its radius, R$_{\rm{eff}}$, axis ratio, q$_{\rm{GAL}}$, angle of orientation, $\theta$, and S\'rsic index, $n$), based on current ground-based imaging, accross the full range of galaxy and quasar luminosities considered in our sample. This potentially opens the door to future ground-based morphological studies of obscured quasars at high redshift. At sub-mm wavelengths, thermal emission from cold dust peaks, meaning these wavelengths can be used to probe the dust heating by star formation, effectively giving a measure of the obscured star formation in the galaxy. Using targeted observations from SCUBA-2, we trace the 850$\mu$m emission in a sample of obscured quasars, finding evidence for prodigious star formation $ > $ 2400 M$_$yr$^$ in three of the 19 quasars in our sample. The detection rate of our obscured quasar sample is found to be consistent with that of both more heavily-obscured Hot-DOGs and UV-luminous quasars, once the samples have been matched in luminosity and redshift. Furthermore, we find evidence that several of the obscured quasars lie in overdense regions of the sky ($\sim$ 3 times denser than sub-mm blank fields).

<p><p>Plants and their pathogens coevolve, with pathogen infection and host resistance acting in evolutionary antagonism of each other. Plant-pathogen coevolution has been shown to effect genetic divergence between populations and species, resulting in localized or specialized interactions between hosts and pathogens. Because most of the studies to date investigating plant-pathogen coevolution have been carried out in managed systems and have focused on pairwise interactions, we know little about three aspects of plant pathosystems in natural settings: 1) the role in nature of the gene-for-gene paradigm for genetic determination of resistance; 2) the relationship of host community diversity and structure, and host-pathogen interaction structure, to the coevolutionary process; and 3) the factors which underlie and drive local adaptation and specialization of interactions.</p><p><p>This dissertation constitutes the results of research in which I have begun addressing these aspects in a natural plant-fungus pathosystem comprising three <i>Ipomoea</i> host species and a single rust pathogen, <i>Coleosporium ipomoeae</i>. I have expanded previous characterization of the genetics of plant resistance in one constituent host species in the system by genetic crosses to characterize the basis of resistance in two additional species, finding support for the expectation that the gene-for-gene paradigm of interaction is important in natural systems. I conducted a cross-inoculation experiment designed to assess host and pathogen variation in infectivity and resistance, to investigate patterns of community interaction structure, and the role that antagonistic coevolution may play in structuring the communities which compose pathosystems. In these experiments I found that the coevolutionary interaction in this system leads to genetic divergence and the substantial amount of host and pathogen variation I discovered, but that it tends to preserve one pattern of community interaction structure across communities. I expanded my cross-inoculation experimental design to facilitate analysis of quantitative aspects of pathogenesis by measuring the intensity of infections, to test existing hypotheses concerning local adaptation and specialization in pathosystems. In this analysis I found strong host local adaptation and pathogen local maladaptation for the qualitative interaction trait of infectivity, and I found weak host local maladaptation and pathogen local adaptation for the quantitative interaction trait of aggressiveness. I also found host specialization among pathogens, and specialized resistance among hosts, to be common in this system. In light of these results, I hypothesize that the geographic scale of host-pathogen coevolution in this system is that of the local community, and that differences between host species result in persistent but incomplete host specialization in pathogen races.</p> / Dissertation

A hierarchical symbiotic framework for policy search with genetic programming (GP) is evaluated in two control-style temporal sequence learning domains. The symbiotic formulation assumes each policy takes the form of a cooperative team between multiple symbiont programs. An initial cycle of evolution establishes a diverse range of host behaviours with limited capability. The second cycle uses these initial policies as meta actions for reuse by symbiont programs. The relationship between development and ecology is explored by explicitly altering the interaction between learning agent and environment at fixed points throughout evolution. In both task domains, this developmental diversity significantly improves performance. Specifically, ecologies designed to promote good specialists in the first developmental phase and then good generalists result in much stronger organisms from the perspective of generalization ability and efficiency. Conversely, when there is no diversity in the interaction between task environment and policy learner, the resulting hierarchy is not as robust or general. The relative contribution from each cycle of evolution in the resulting hierarchical policies is measured from the perspective of multi-level selection. These multi-level policies are shown to be significantly better than the sum of contributing meta actions.

Obligate brood parasites rely solely on other species, the hosts, to incubate their eggs and raise their offspring, which often reduces the host’s reproductive output. This reproductive cost has led to the evolution of anti-parasite adaptations among hosts, which in turn, has led to better trickery by parasites, a process termed an evolutionary arms race. The objective of this thesis was to investigate host-parasite coevolutionary arms races to address questions of host-use diversity. Host diversity varies dramatically among brood-parasitic species, but reasons for variations in host-use among brood parasites are not well understood. In Chapter 2, I address questions on host diversity specifically, whereas I address questions about coevolutionary interaction between hosts and parasites in Chapters 3, 4 and 5 using two host-parasite systems, one in New Zealand and one in North America. Chapter 2 investigates if host diversity is constrained by aggressive nest defence behaviour. I compared the nest defence behaviour of the exclusive host of the shining cuckoo Chrysococcyx lucidus lucidus on the main islands of New Zealand, the grey warbler Gerygone igata, to two other potentially suitable hosts that are not currently parasitised, the fantail Rhipidura fuliginosa and the silvereye Zosterops lateralis. The results suggest that grey warblers are as aggressive as fantails and silvereyes towards shining cuckoos at the nest and thus, host specialisation in shining cuckoos in New Zealand, at least, does not appear to be the result of nest-defence constraints imposed by potential but unused host species. Chapter 3 investigates if red-winged blackbirds Agelaius phoeniceus, a species that typically accepts the eggs of parasites, recognises, as indicated by changes in incubation behaviour, when they have been parasitised by brown-headed cowbirds Molothrus ater. Recognition without rejection suggests that rejection may be context-dependent but the results suggest that red-winged blackbirds do not recognise when their nests have been parasitised by brown-headed cowbirds, at least at the egg stage. This study was the first to investigate if hosts that almost invariably accept the eggs of parasites recognise when they have been parasitised. Chapter 4 investigated the possibility of coevolutionary arms races occurring through olfactory channels in contrast to earlier work that focussed only on visual and auditory cues. Recent research has revealed that olfactory abilities in birds are more common than previously thought. Uropygial gland secretions are posited to be a key source of avian body odour and its composition has been found to vary among species and individuals as well as between the sexes. I compared gas-chromatography (GC-FID) traces of shining cuckoo preen wax to the GC-FID traces of the grey warbler, the only host of the shining cuckoo in mainland New Zealand, as well as the preen wax of seven other species for evidence of mimicry. Preliminary results suggest there is evidence for mimicry and the potential for odour-based nestling discrimination in grey warblers. Further tests recording the response of grey warblers to odour-manipulated nestlings are necessary. Finally, in Chapter 5, I investigated the response of the song thrush Turdus philomelos, a species that rejects the eggs of the common cuckoo Cuculus canorus and conspecifics at intermediate and low frequencies, respectively, to nest-odour manipulations using the preen wax of conspecifics and heterospecifics. The results suggest song thrush do not use odour to assess the risk of parasitism at least as indicated in terms of changes in incubation behaviour. Investigations of the role of olfaction in avian brood parasite systems can provide a better understanding of brood-parasite coevolution. Only by considering all channels of communication can we be sure to completely understand the coevolutionary dynamics between brood parasites and their hosts.