Submitted by Carlos Augusto Rolim da Silva Junior (carlos_jrolim@hotmail.com) on 2016-02-17T17:38:40Z No. of bitstreams: 1 arquivototal.pdf: 1255636 bytes, checksum: db1ad020aa90913b44b90e0e13250b01 (MD5) / Made available in DSpace on 2016-02-17T17:38:40Z (GMT). No. of bitstreams: 1 arquivototal.pdf: 1255636 bytes, checksum: db1ad020aa90913b44b90e0e13250b01 (MD5) Previous issue date: 2013-12-04 / The dimensions in which species share resources can be classified into three general axis: food, habitat and time. The degree of sharing can be set indirectly throughout morphological differences, foraging mode and the use of space by sympatric species. This study presented two chapters, the first of which is entitled “Niche partitioning between the sympatric lizards Phyllopezus periosus and Phyllopezus pollicaris (Sauria: Phyllodactylidae) in a Caatinga area in northeastern Brazil”. We presented data on habitat use, perch height, period of activity, diet and morphology for both species. The second chapter was to focus on the “Demographic aspects of sympatric lizards Phyllopezus periosus and Phyllopezus pollicaris (Sauria: Phyllodactylidae) in a Caatinga area in northeastern Brazil”. In this chapter, we estimated the probability of survival (apparent survival), recapture probability and population density of P. periosus and P. pollicaris. Data collection was conducted in a Caatinga area (divided into sub-areas I, II and III) in the Salgadinho municipality, Paraíba State, from April 2012 to March 2013. The search for lizards occurred between 18:00 and 5:00 am of the following day. For each individual observed and/or captured, we recorded the following information: date and time of observation, microhabitat use, time, morphometric measurements, and gender. The marking of lizards was performed using Visible Implant Fluorescent Elastomer (polymer thick liquid that once applied subcutaneously, solidifies, however, remained flexible and visible). The analysis of apparent survival and recapture probability were performed using the model Cormack-Jolly Sebber (CJS), in the software MARK 6.2. Estimates of population density were calculated by the sums the individuals sighted, divided by the area (sub-area I) used in the study. Both species occupied predominantly saxicolous microhabitats, being observed a high overlap in the use of different microhabitats. The perch height preferentially used by P. periosus and P. pollicaris was 0 and 60 cm. However, P. periosus occupied higher perches (421 to 540 cm). The first individuals of both species were sighted from 18:02 until 04:29 hours of the the following morning. The two species showed no significant differences in the activity period. The more consumed prey by P. periosus were Coleoptera, Formicidae and Hemiptera, while the most consumed items for P. pollicaris were Coleoptera, Isoptera and Araneae. Phyllopezus pollicaris presented both the trophic niche breadth based on prey number and volume higher compared to P. periosus. However, the two species showed high overlap in their diets. We not observed differences in the volume of consumed prey. Both differed in body size and body mass, but adults of P. periosus and P. pollicaris showed no intraspecific differences in body size and body mass in relation to sex. Phyllopezus periosus presented a recapture rate of 30%, while P. pollicaris showed an average rate of 19%. Only Phyllopezus pollicaris showed variation in recapture rate over time, that it was negatively correlated with average rainfall of each month. The estimated survival in both species was constant over time. Phyllopezus periosus and P. pollicaris survival estimates showed 66% and 94%, respectively. P. pollicaris presented population density greater than P. periosus throughout the study. The highest densities observed for P. periosus were in July and November 2012, as for P. pollicaris were in the months of May and November of the same year. Both species showed variations in their population abundances depending on the sub-sampled area. While P. pollicaris was more abundant in sub-area I, P. periosus was more abundant in the sub-area III. Variations in population abundances between these two species may contribute to the coexistence of the two species in the area. / As dimensões em que as espécies animais partilham recursos podem ser classificadas em três grupos gerais: tipo de alimento, habitat e tempo. A forma mais comum de segregação é por habitat, seguido por tipo de alimento e, por fim, divisão temporal. O grau de partilha de recursos pode ser definido indiretamente através das diferenças morfológicas, do tipo de forrageamento e da utilização dos recursos espaciais por espécies simpátricas. O presente estudo foi desenvolvido em dois capítulos, o primeiro deles é intitulado de “Partição de nicho entre os lagartos Phyllopezus periosus e Phyllopezus pollicaris (Sauria: Phyllodactylidae) em simpatria em uma área de Caatinga no Nordeste do Brasil”. Foram abordados dados sobre uso do habitat, altura de empoleiramento, período de atividade, dieta e morfometria em ambas as espécies de lagartos. No segundo capítulo, intitulado de “Aspectos demográficos dos lagartos Phyllopezus periosus e Phyllopezus pollicaris (Sauria: Phyllodactylidae) em simpatria em área de Caatinga no Nordeste do Brasil”, foi estimada a probabilidade de sobrevivência (sobrevivência aparente), probabilidade de recaptura e densidade populacional de P. periosus e P. pollicaris. A coleta de dados foi realizada em uma área (dividida nas sub-áreas I, II e III) de caatinga no município de Salgadinho – PB, durante os meses de abril de 2012 a março de 2013. A busca pelos lagartos ocorreu entre 18:00 e 05:00 horas do dia seguinte. Para cada indivíduo observado e/ou capturado, eram registradas as seguintes informações: data e horário da observação, categoria de microhábitat utilizado, altura do empoleiramento, medidas morfométricas, sexo e número de marcação. A marcação dos lagartos foi realizada através do Implante Visível de Elastômero Fluorescente (polímero líquido pastoso - que depois de aplicado subcutaneamente solidifica-se, porém, se mantendo flexível e visível). As análises de sobrevivência aparente e probabilidade de recaptura foram realizadas através do modelo Cormack-Jolly Sebber (CJS), no software MARK 6.2. As estimativas de densidade populacional foram calculadas através da soma dos indivíduos avistados, divididos pela área (sub-área I) utilizada no estudo. Ambas as espécies ocuparam predominantemente microhábitats saxícolas, sendo constatada uma alta sobreposição quanto ao uso dos diferentes microhábitats. A altura de empoleiramento utilizada preferencialmente por P. periosus e P. pollicaris foi de 0 e 60 cm, entretanto, P. periosus ocupou poleiros mais altos (421 até 540 cm de altura), o que não foi observado em P. pollicaris. Os primeiros indivíduos de ambas as espécies foram avistados a partir das 18:02 até as 04:29 horas da manhã do dia seguinte e as duas espécies de lagartos não apresentaram diferenças significativas quanto ao período de atividade. As presas mais consumidas por P. periosus foram Coleoptera, Formicidae e Hemiptera, já para P. pollicaris os itens mais consumidos foram Coleoptera, Isoptera e Aranae. P. pollicaris apresentou ambas as larguras de nicho trófico (número e volume) bem superiores quando comparado com P. periosus. No entanto, as duas espécies apresentaram alta sobreposição em suas dietas. Não foi constatado diferenças quanto ao volume das presas consumidas. As duas espécies diferiram em relação ao tamanho do corpo e massa corpórea. Mas, indivíduos adultos de P. periosus e P. pollicaris não apresentaram diferenças intraespecíficas do tamanho do corpo e massa corpórea em relação ao sexo. P. periosus apresentou uma taxa de recaptura de 30%, enquanto que seu congênere apresentou uma taxa média de 19%. Apenas a espécie P. pollicaris apresentou variação na taxa de recaptura ao longo do tempo, que se mostrou negativamente correlacionada com a precipitação média de cada mês. A estimativa de sobrevivência em ambas as espécies foi constante ao longo do tempo. P. periosus e P. pollicaris apresentaram estimativas de sobrevivência de 66% e 94%, respectivamente. P. pollicaris apresentou densidade populacional maior que P. periosus durante todo o estudo. A maior densidade observada para P. periosus foi no mês de julho de 2012, já para P. pollicaris foi no mês de novembro do mesmo ano. Ambas as espécies apresentaram variações em suas abundâncias populacionais dependendo da sub-área amostrada. Enquanto P. pollicaris foi mais abundante na sub-área I, P. periosus mostrou-se mais abundante que seu congênere na sub-área III. Variações nas abundâncias populacionais entre essas duas espécies, provavelmente seja um fator que possa vir a contribuir para coexistência das duas espécies na área de estudo.

Historical biogeography of cichlid fishes within the Usumacinta province (Mexico, Guatemala, and Belize) is reviewed based on newly collected material specifically for this study. Five areas of endemism within the province were recognized using Parsimony analysis of endemicity. Dispersal-Vicariance analysis implemented in S-DIVA suggests only one of them (the Grijalva-Usumacinta) as the ancestral area for the whole fauna. Other areas of endemism within the Usumacinta province were colonized from this area by dispersals. This biogeographic interpretation suggests a long evolution of the richest cichlid fauna in Middle America in the Grijalva-Usumacinta area of endemism in a sympatric context.

Apresenta-se nesta tese os resultados de aplicações do formalismo da Mecânica Estatística em dois problemas independentes. O primeiro diz respeito a um modelo para Evolução do Acasalamento Preferencial no processo de Especiação Simpátrica; enquanto que o segundo refere-se ao desenvolvimento de um algoritmo de aprendizado por meio de Inferência Aproximativa. No problema biológico estudado, cada indivíduo em um modelo de agentes é composto por dois traços. Enquanto um é responsável pela ecologia do indivíduo, o outro dita uma aparência física descorrelacionada com a adaptabilidade. Esses traços são expressos por diferentes loci que estão ligados entre si por uma taxa de recombinação. O modelo inclui também a possibilidade de evolução da preferência sexual dos indivíduos. Foi construído para esse modelo um diagrama de fases no espaço dos parâmetros que descrevem o ambiente como, por exemplo, quantidades de recursos e deficiência do indivíduo híbrido. Foram encontradas três fases de equilíbrio: (i) emergência de Acasalamento Preferencial; (ii) extinção de um dos alelos do locus responsável pela ecologia e (iii) equilíbrio Hardy-Weinberg. Foi verificado que o acasalamento preferencial pode emergir ou mesmo ser perdido (e vice-versa) em resposta a mudanças no ambiente. Além disso, o sistema apresenta memória característica típica de transições de primeira ordem, o que permitiu a descrição desse sistema biológico por meio do arcabouço da Mecânica Estatística. Em relação à Inferência Aproximativa, está-se interessado na construção de um algoritmo de aprendizado supervisionado por meio da técnica de Propagação de Expectativas. Mais especificamente, pretende-se inferir os parâmetros que compõem um Perceptron Professor a partir do conjunto de pares - entradas e saídas - que formam o conjunto de dados disponíveis. A estimativa desses parâmetros será feita pela substituição de uma distribuição Posterior original, geralmente intratável, por uma distribuição aproximativa tratável. o algoritmo Propagação de Expectativas foi adotado para a atualização, passo a passo, dos termos que compõem essa distribuição aproximativa. Essa atualização deve ser repetida até que a convergência seja atingida. Utilizando o Teorema do Limite Central e o método de Cavidade, foi possível obter um algoritmo genérico e que apresentou desempenho bastante evidente em dois modelos estudados: o modelo do Perceptron Binário e o modelo do Perceptron Gaussiano, com desempenho ótimo em ambos os casos. / This thesis presents applications of the framework of Statistical Mechanics to two independent problems. The first corresponds to a computational model for the evolution of Assortative Mating in the Sympatric Speciation process; and the second a learning algorithm built by means of a Bayesian Inference approach. In the biological problem each individual in an agent-based model is composed of two traits. One trait, called the ecological trait, is directly related with the fitness; the other, called the marker trait, has no bearing on the fitness. The traits are determined by different loci which are linked by a recombination rate. There is also the possibility of evolution of mating preferences, which are inherited from the mother and subject to random variations. The study of the phase diagram in the spa e of parameters describing the environment (like carrying capacity and disruptive selection) reveals the existence of three phases: (i) assortative mating; (ii) extinction of one allele from ecological loci; and (iii) Hardy-Weinberg equilibrium. It was verifed that the assortative mating an emerge or even be lost (and vice-versa) acording with the environmental hanges. Moreover, the system shows memory of the initial condition, characterising a hysteresis. Hysteresis is the signature of first order phase transition, which allows the description of the system by means of the Statistical Mechanics framework. In relation to the Bayesian Inference, a supervised learning algorithm was constructed by means of the Expectation Propagation approach. The idea is to estimate the parameters which compose a Teacher Perceptron by the substitution of the original posterior distribution, intra table, by a tractable approximative distribution. The step-by-step update of the terms composing the approximative distribution was performed by using the Expectation Propagation algorithm. The update must be repeated until the convergence ocurrs. Using the Central Limit Theorem and the Cavity Approah, it was possible to get a generic algorithm that has shown a very good performance in two application scenarios: The Binary Perceptron Model and the Gaussian Perceptron Model.

At Parque Estadual do Turvo, an area of atlantic forest in the northwest of Rio Grande do Sul, Brazil, was recorded in sympatry by different sampling methods three viperids species. Bothropoides diporus the snake was more common, and also the best adapted to dry and degraded areas (64.95% of total), Bothropoides jararaca (21.65%), was sampled in degraded areas and inside forest, was the only one be sampled perched and Bothrops jararacussu (13.40%) was restricted to forested areas, particularly near water bodies. The offspring of B. jararaca and B. jararacussu inhabit the streams. Compared by a multiple linear regression the number of monthly meetings of snakes with the following variables: average monthly minimum temperature, monthly rainfall and monthly insolation during the study period. Significant difference between the number of snakes caught between the stations of the largest and lowest temperatures. There was no significant difference in the daily activity of snakes. The diet of B. diporus and B. jararaca is composed of mammals, amphibians and reptiles, and mammals the item most used by both species was not observed differences in diet between males and females of B. diporus, showed had specialized diet in mammalian, was not observed shift ontogenetic diet B. jararaca. The method provided the best results was local collectors (65.99%) followed by occasional encounters (17.53%), time constrained search (12.37%), search in roads (4.12%). Pitfall traps with drift fences were not successful in viperids in this study / No Parque Estadual do Turvo, uma área de mata atlântica no Noroeste do Estado do Rio Grande do Sul, Brasil, foram registrados em simpatria por diferentes métodos de amostragem três espécies de viperídeos. Bothropoides diporus foi a serpente mais comum, sendo também a melhor adaptada em áreas degradadas e secas (64,95% do total), Bothropoides jararaca (21,65%), foi amostrada em áreas degradadas e interior de floresta, foi a única a ser amostrada empoleirada e Bothrops jararacussu (13,40%) ficou restrita às áreas florestadas, sobretudo próxima a corpos d água. Os filhotes de B. jararaca e B. jararacussu habitam os riachos. Comparamos por uma regressão múltipla linear o número mensal de encontros de serpentes com as seguintes variáveis ambientais: temperatura média mínima mensal, pluviosidade mensal e insolação mensal durante período do estudo. Houve diferença significativa entre o número de serpentes capturadas entre as estações de maiores e menores temperaturas. Não houve diferença significativa na atividade diária das serpentes. A dieta de B. diporus e B. jararaca é composta de mamíferos, anfíbios anuros e répteis, sendo mamíferos o item mais utilizado por ambas as espécies, não foi evidenciado diferenças na dieta de machos e fêmeas de B. diporus, que mostrou-se especialista em mamíferos nem variação ontogenética na dieta de B. jararaca. O método que proporcionou os maiores resultados foi auxílio de terceiros (65,99%) seguido de encontros ocasionais (17,53%), procura visual limitada por tempo (12,37%), procura em estradas (4,12%). Armadilhas de interceptação e queda não foram eficientes para viperídeos neste estudo.

The basking shark, Cetorhinus maximus, has historically been a target of international fisheries, leading to well-documented declines in parts of its global distribution. Currently, the basking shark is listed as globally ‘Vulnerable’ and regionally ‘Endangered’ (North Pacific and Northeast Atlantic) on the IUCN Red List of Threatened Species, rendering the species an international conservation priority. Here, we assessed the global matrilineal genetic population structure and evolutionary history of the basking shark by completing the first whole mitochondrial genome sequence level survey of animals sampled from three globally widespread geographic regions: the western North Atlantic (n = 11), the eastern North Atlantic (n = 11), and within New Zealand territorial waters (n = 12). Despite the relatively large amount of sequence data assessed (~16,669 bp per individual), whole mitogenome analyses showed no evidence of population differentiation (ΦST = -0.047, P > 0.05) and very low nucleotide diversity (π = 0.0014 ± 0.000) across a global seascape. The absence of population structure across large distances and even between ocean basins is indicative of long-dispersal by this species, including an ability to cross known biogeographic barriers known to differentiate populations of other highly vagile pelagic fishes. Notably, evolutionary analyses of the mitogenome sequences revealed two globally sympatric but evolutionary divergent lineages, with a Bayesian framework estimated coalescence time of ~2.46 million years ago. Coalescent-based Bayesian skyline analysis uncovered subtle evidence of Pleistocene demographic flux for this species, including a potential decline in female effective population size. Thus, historical population changes may be responsible for the occurrence of the two highly divergent, yet sympatric lineages, as population declines may have resulted in the loss of intermediate haplotypes and resulted in an overall loss of genetic diversity. This work supports the recognition of basking sharks as a single matrilineal global population, and as such requires the application of a cooperative multiagency and international approach to fisheries management to conserve this highly vulnerable and ecologically unique species.

付記する学位プログラム名: 霊長類学・ワイルドライフサイエンス・リーディング大学院 / 京都大学 / 0048 / 新制・課程博士 / 博士(理学) / 甲第21616号 / 理博第4523号 / 新制||理||1649(附属図書館) / 京都大学大学院理学研究科生物科学専攻 / (主査)教授 岡本 宗裕, 准教授 Andrew MacIntosh, 教授 髙井 正成 / 学位規則第4条第1項該当 / Doctor of Science / Kyoto University / DGAM

During 2000-2004 I studied the population status of the Formosan clouded leopard (Neofelis nebulosa brachyurus) and the ecology of its prey and other sympatric carnivores in the largest remaining lowland primary forest in southern Taiwan. My research team and I set up 232 hair snare stations and 377 camera trap sites at altitudes of 150-3,092m in the study area. No clouded leopards were photographed in total 13,354 camera trap days. Hair snares did not trap clouded leopard hairs, either. Assessment of the prey base and available habitat indicated that prey depletion and habitat loss, plus historical pelt trade, were likely the major causes of extinction of clouded leopards in Taiwan. Using zero-inflated count models to analyze distribution and occurrence patterns of Formosan macaques (Macaca cyclopis) and 4 ungulates, we found habitat segregation among these 5 herbivore species. Formosan macaques, Reeve's muntjacs (Muntiacus reevesi micrurus), and Formosan serows (Nemorhaedus swinhoei) likely were the most important prey species of Formosan clouded leopards given their body size and high occurrence rates in lower altitudes. In contrast, sambar deer (Cervus unicolor swinhoii) tended to occur more frequently as altitude increased. Formosan macaques exhibited seasonal differences in occurrence rates and were absent at altitudes > 2,500m in winter. Only Formosan serows showed preference for cliffs and rugged terrain, while the other 4 species, except wild boars (Sus scrofa taivanus), avoided these areas. Habitat segregation in forest understory and structure were more pronounced among the 4 ungulates. Forest structure rarely affected occurrence rates of Formosan macaques on the ground. Niche relationships of the other sympatric carnivores were studied through habitat, diet, and temporal dimensions. Resource partitioning by carnivores was observed. Altitude was the strongest factor explaining the composition of the carnivore community in the local study-area scale and in the landscape scale across Taiwan. Carnivores could be divided into 2 groups: low-mid altitude consisting of Formosan ferret badgers (Melogale moschata subaurantiaca), gem-faced palm civets (Paguma larvata taivana), lesser oriental civets (Viverricula indica taivana), crab-eating mongooses (Herpestes urva formosanus), leopard cats (Prionailurus bengalensis chinensis), and feral cats (Felis catus), and the mid-high altitude group consisting of yellow-throated martens (Martes flavigula chrysospila), Siberian weasels (Mustela sibirica taivana), and Asiatic black bears (Ursus thibetanus formosanus). Carnivore richness was higher at mid altitudes where these 2 groups overlapped (i.e. mid-domain effect). The low-mid altitude carnivores were more nocturnal and tolerant of human activity and forest alteration except crab-eating mongooses, which were diurnal and avoided human encroachment. Similar to crab-eating mongooses, the mid-high altitude carnivores also avoided human encroachment and were diurnal except for Siberian weasels, which were more nocturnal. Diet summary based on their major food items for all sympatric carnivores revealed 3 groups of foragers which foraged on: invertebrates, small mammals, and plant fruits. Felidae, yellow-throated martens, and Siberian weasels preyed on small mammals. Asiatic black bears and gem-faced palm civets ate mostly plant fruits. The other 3 carnivores were mainly invertebrate foragers. These 9 carnivores partitioned resource uses in the 3 niche dimensions except for some overlap in resource use by leopard cats and feral cats. Prey base for Formosan clouded leopards and the carnivore richness in Taiwan were found to be lower in areas with higher levels of human activity. On the other hand, Formosan macaques and ungulates could become over-abundant without human hunting and top carnivore predation. Mesopredator release may occur because of vanishing top carnivores, causing reduction of the lower trophic level prey species. It is important to assess the cascading impacts of the loss of the Formosan clouded leopards and Eurasian otters (Lutra lutra chinensis) and the declining Asiatic black bears and to consider reintroduction of Formosan clouded leopards, as well as active management of the other larger mammals. These results provided baseline information for reintroduction of clouded leopards and management of their prey and generated new hypotheses regarding the ecology of these large mammals for future investigation. / Ph. D.

Natural populations exhibit a non-random mating behavior and it is assumed that mate preferences causing the non-random mating behavior play a role in sympatric speciation. In my thesis, I have proposed a new model of a mate preference that is based on ecological performance and I have named it fitness-based mating. Individuals that express this mate preference choose primarily fit partners. Fitness-based mating is modelled for haploid, diallelic populations. Individuals are distributed across two niches, and genomes are simplified to two loci. The first locus is subject to natural selection, and the second-locus genotype gives the strength of the mate preference. The population is separated into females and males, among which only females exhibit the mate preference. Ecological selection acts on both sexes alike. With the model I have investigated how female choosiness based on direct advantages offered by their partners can cause and maintain a polymorphic population. Fitness-based mating is an evolutionary successful mating strategy. It spreads in a population due to its amplifying effect on the reproductive success and on the attractiveness of its carriers. A polymorphism arises naturally in the model. The emergence of a stable polymorphism of traits underlying ecological selection is of special interest, because a polymorphism can be a precursor of speciation.

Breeding compexes of poeciliid fishes with a bisexual and two unisexual species were studied for mechanisms permitting Sympatric coexistence. The unisexuals are gynogenetic and thereby sexually dependent on the males of the bisexual species for sperm to initiate development, but inheritance is entirely maternal. Bisexual females are more abundant in headwater localities; unisexuals increase in downstream localities. Males were 10 - 18% of the total poeciliid population, regardless of the relative proportions of bisexual to unisexual females. Downstream localities were typified by greater habitat diversity, including a variety of backwater pools. The unisexuals showed a marked preference for such pools. Both field and laboratory studies showed that all three types of females as well as males preferred shaded areas with a gravel substrate. Although the four types of fish were found together, nearest neighbour data indicated that each type of female preferred its own kind. Males courted throughout the year and were indiscriminate in their choice of mates. Despite the skewed sex ratio, males were not in short supply because only a few females were sexually receptive at a given time. No significant differences existed between bisexuals and unisexuals in their relative reproductive outputs, but they were asynchronous.

Die Bedeutung ökologischer Faktoren bei der Entstehung phänotypischer sowie genetischer Vielfalt durch natürliche Selektion, besonders bei sympatrischer Artbildung, ist derzeit ein Fokus der Evolutionsforschung. Gemeinsam vorkommende und nah verwandte Arten werden daher als Modellorganismen verwendet, um die Ursachen und Mechanismen ökologischer und evolutionärer Diversifizierung zu untersuchen. Ein sympatrisches Fisch-Artenpaar, für das eine sympatrische Artbildung auf Basis genetischer Analysen vermutet wird, existiert im norddeutschen Stechlinsee. Die zwei Maränenarten sind morphologisch kaum zu unterscheiden, zeigen eine ähnliche Nahrungszusammensetzung und kommen gemeinsam im Freiwasser, allerdings in leicht unterschiedlichen Wassertiefen, vor. Die Hypothese meiner Arbeit war, dass sich die Physiologie bzw. das Verhalten der Arten in Bezug auf die wichtigsten Umweltfaktoren ihres Lebensraumes, Futterdichte, Lichtintensität und Wassertemperatur unterscheiden. Daher haben wir Fraßeffizienz, Stoffwechselraten und Temperaturpräferenzen in Abhängigkeit dieser Faktoren bei beiden Arten untersucht, nachdem diese zuvor unter identischen Laborbedingungen herangezogen wurden. Wir fanden keinen Unterschied in der Fraßeffizienz, allerdings zeigten beide Arten je nach Temperatur unterschiedliche Stoffwechselraten sowie entsprechende Unterschiede in der Temperaturpräferenz. Die Ergebnisse legen nahe, dass temperaturbedingte physiologische Anpassungen sowie die Nutzung unterschiedlicher thermischer Mikrohabitate die zwischenartliche Konkurrenz verringern und eine gemeinsame Existenz ermöglichen. Um die mögliche Bedeutung einer solchen öko-physiologischen Spezialisierung für die Artbildung innerhalb des Sees einschätzen zu können, haben wir basierend auf den Freiland- und Labordaten ein mathematisches Evolutionsmodell entwickelt. Demnach ist ein Aufspalten einer Ausgangspopulation in zwei Populationen mit unterschiedlichen Temperaturoptima wahrscheinlich. Eine ökologische und evolutionäre Diversifizierung entlang des Temperaturgradienten ist somit ein empirisch und theoretisch plausibles Szenario für die sympatrische Artbildung der Stechlinsee-Maränen. / Abstract The role of ecological factors in generating phenotypic and genetic diversity through natural selection has received increasing attention in evolutionary biology during the last decade, especially with respect to diversification in sympatry. Sympatrically occurring and closely related species are used as model systems to study the causes and mechanisms of ecological and evolutionary diversification. A sympatric species pair of coregonid fish, for which a speciation in sympatry has been suggested based on genetic analyses, coexists in the German Lake Stechlin. The two species are morphologically similar planktivores with weak divergence in diet composition that co-occur within the pelagic area of the lake at slightly different water depths. Accordingly, it was hypothesized that the species would differ in physiology and behaviour with respect to the most important environmental factors of their natural habitat, food density, light intensity and water temperature. We studied the feeding efficiencies, metabolic rates, and temperature preferences of both species previously hatched and raised under identical laboratory conditions. We found no divergence in feeding efficiency, but significant differences in temperature-related metabolic costs as well as a corresponding difference in thermal preference. These results suggest that temperature-related physiological adaptations and the associated use of slightly different thermal microhabitats reduces exploitative competition between the species and facilitates their coexistence. To evaluate a potential role of this eco-physiological specialization for a speciation in sympatry, we developed a mathematical evolutionary model, based on our field observations and laboratory experiments. The model showed that an evolutionary splitting of an ancestral into two coexisting populations with different temperature optima is likely in this system. In conclusion, an eco-physiological and evolutionary diversification along the temperature-depth gradient of the lake is an empirically and theoretically plausible scenario for the sympatric speciation of the coregonids.