Radiative feedbacks associated with changes in water vapor, temperature, surface albedo and clouds remain a major source of uncertainty in our understanding of climate's response to anthropogenic forcing. In this dissertation climate model data is used to investigate variations in feedbacks that result from changing CO��� forcing and the time scales on which feedbacks operate, focusing on the applicability of one method in particular, the radiative kernel technique, to these problems. This computationally efficient technique uses a uniform, incremental change in feedback variables to infer top-of-atmosphere (TOA) radiative flux changes. The first chapters explore the suitability of the linear radiative kernel technique for large forcing scenarios. We show that kernels based on the present-day climate misestimate TOA flux changes for large perturbations, translating into biased feedback estimates. We address this issue by calculating additional kernels based on a large forcing climate state with eight times present day CO��� concentrations. Differences between these and the present-day kernels result from added absorption of radiation by CO��� and water vapor, and increased longwave emission due to higher temperatures. Combining present-day and 8xCO��� kernels leads to significant improvement in the approximation of TOA flux changes and accuracy of feedback estimates. While climate sensitivity remains constant with increasing CO��� forcing when the inaccurate present-day kernels are used, sensitivity increases significantly when new kernels are used. Comparison of feedbacks in climate models with observations is one way towards understanding the disagreement among models. However, climate change feedbacks operate on time scales that are too long to be evaluated from the observational record. Rather, short-term proxies for greenhouse-gas-driven warming are often used to compute feedbacks from observations. The third chapter of this dissertation examines links between the seasonal cycle and global warming using pattern correlations of spatial distribution of feedback variables and radiative flux changes. We find strong correlations between time scales for changes in surface temperature and climate variables, but not for TOA flux anomalies, reaffirming conclusions drawn in previous work. Finally, we investigate the fitness of the radiative kernel technique for evaluation of short-term feedbacks in a comparison with the more accurate, but more computationally expensive, partial radiative perturbations. / Graduation date: 2013

Coral reefs are suffering unprecedented declines in coral cover and species diversity. These changes are often associated with  substantial shifts in community structure to new dominant organisms. Ultimately, these “phase shifts” can be persistent and very difficult to return from. Building insurance against degradation and decreasing the likelihood of reefs undergoing shifts to undesirable states will require sustainable management practices that uphold coral reef resilience. This thesis consists of five papers that contribute new knowledge useful for managing the resilience of coral reefs, and other marine ecosystems. Paper I shows how the morphology of natural substrate (dead coral colonies) can significantly influence coral recruitment patterns. Paper II focuses on larval lipid levels, a key determinant of coral dispersal potential, in a common Caribbean coral (Favia fragum). It shows that i) lipid levels exhibit a significant, non-linear reduction throughout the larval release period of F. fragum and ii) exposure to a common pollutant (copper) could potentially lead to a more rapid lipid consumption in the larvae. Paper III presents a broader analysis of the different undesirable states a coral reef can shift to as a consequence of reef degradation. It concludes that different states are caused by different driving factors and that management must explicitly acknowledge this. Paper IV proposes a suite of resilience indicators that can help managers assess when a coral-dominated reef might be moving towards a shift to an undesirable state. These indicators capture key-processes occuring on different temporal and spatial scales and signal resilience loss early enough for managers to take appropriate measures. Finally, Paper V reviews the feedback loops that reinforce the undesirable states of five important marine ecosystems and suggests certain strategies that can ease the restoration back to healthier conditions. Managing these critical feedbacks will recquire monitoring the processes underpinning these feedbacks, breaking already established feedbacks loops through large-scale management trials and acknowledging transdisciplinary solutions that move management beyond the discipline of ecology / At the time of the doctoral defense, the following papers were unpublished and had a status as follows: Paper 2: Submitted. Paper 5: In progress.

Land use/cover change (LUCC) is a major force affecting ecosystems and the services theyprovide at local, regional and global levels. Traditionally, LUCC has been approached as aseries of linear and unidirectional single cause–effect processes, but it is now increasinglyincorporating notions from complex systems to enrich this view. This study explores thepotential benefits of using the concept of regime shifts to understand LUCC. A globalcropland cover dataset was analyzed to empirically identify patterns that suggest theexistence of alternate regimes. Results indicate that in some countries cropland-dominatedand (semi)wild landscapes likely represent alternate social-ecological regimes.Furthermore, results suggest that a theoretical feedback relating market access andagglomeration of economic activates plays a strong role in maintaining these alternate regimes in some contexts. These results highlight the need for LUCC studies to go beyondthe focus on external drivers and environmental template conditions, to incorporate feedback mechanisms and the potential for regime shifts to better understand the nature of LUCC.

Smallholder livestock keeping and agriculture systems in the sub-Saharan Africa (SSA) seem to be caught into poverty traps, in as much as they cannot any longer provide ecosystem services on which local communities depend for their survival. I used a regime shifts framework to carry out a thorough assessment of these two case studies in arid and semi-arid lands and smallholder by identifying relevant traps and alternate desirable regimes. Using systems analysis and modelling, I drew casual loop diagrams of the two case studies, which helped me to identify the feedback loops that maintain the systems in undesirable traps and the external driving forces of change. A set of interventions points or leverage points were identified to change the dynamics of the systems and shift them towards more desirable regimes. Essentially, a structural change of both systems is called for if sustainable livelihoods in the rural areas of the SSA are to be seriously envisaged. Human capital investments present the main opportunity for facilitating escape from poverty by transforming farmers to non-farmers and livestock keepers to non-livestock keepers.

Land-use changes impact the energy balance of the Earth system, and feedbacks in the Earth system can dampen or amplify this perturbation. We analyze here from satellite data the response of clouds and subsequently radiation to a change of land use for the example of deforestation in the Amazon Basin. In this region, the characteristics of different cloud types over two vegetation types (forest and crop-/grasslands) were calculated for a time period of five years by using satellite data from the instruments MODIS and CERES. The cloud types are defined according to height, optical thickness, and fraction of cloud cover. For calculating the radiative forcing caused by deforestation, the dependency of spatial and temporal averages for the reflected shortwave and outgoing longwave radiation of the top of the atmosphere on vegetation types were determined as well. The results show distinct differences in cloud cover and radiative forcing over crop-/grasslands and forests for the two vegetation regimes, implying a potentially significant positive cloud feedback to deforestation.

As control systems become more integrated with high-end engineering systems as well as consumer products, they are expected to achieve specifications that may include logic rules, safety constraints, startup procedures, and so forth. Control design for such complex specifications is a relatively unexplored research area. One possible design approach is based on partitioning the state space into polytopic regions, and then formulating a certain control problem on each polytope, with the intention that the set of all controllers so obtained would collectively achieve the specification. The control problem which must be solved for each polytope is called the reach control problem, and it has been identified as turnkey to the further development of this approach. The reach control problem (RCP) is to find a state feedback to make the closed-loop trajectories of an affine (or linear) control system defined on a polytope reach and exit a prescribed facet of the polytope in finite time. This dissertation studies a number of aspects of the reach control problem, and it uses tools from convex analysis, nonsmooth analysis, and computational geometry for this study. The dissertation has three main themes. First, we formulate and solve a variant of RCP in which trajectories exit the polytope in a monotonic sense; this provides a triangulation-independent solution of RCP. Second, we develop a Lyapunov-like theory for verifying if RCP is solved using a given candidate controller. This involves the introduction of the notion of generalized flow functions, a LaSalle Principle for RCP, and several converse theorems on existence of generalized flow functions. Third, we study the relationship between affine feedbacks and continuous state feedbacks for RCP on simplices. Although the two feedback classes have been shown to be equivalent under an assumption on the triangulation of the state space, we show by a counterexample that the equivalence is no longer true under arbitrary triangulations. Then we provide for single-input systems a constructive method for the synthesis of multi-affine feedbacks for RCP on simplices.

As control systems become more integrated with high-end engineering systems as well as consumer products, they are expected to achieve specifications that may include logic rules, safety constraints, startup procedures, and so forth. Control design for such complex specifications is a relatively unexplored research area. One possible design approach is based on partitioning the state space into polytopic regions, and then formulating a certain control problem on each polytope, with the intention that the set of all controllers so obtained would collectively achieve the specification. The control problem which must be solved for each polytope is called the reach control problem, and it has been identified as turnkey to the further development of this approach. The reach control problem (RCP) is to find a state feedback to make the closed-loop trajectories of an affine (or linear) control system defined on a polytope reach and exit a prescribed facet of the polytope in finite time. This dissertation studies a number of aspects of the reach control problem, and it uses tools from convex analysis, nonsmooth analysis, and computational geometry for this study. The dissertation has three main themes. First, we formulate and solve a variant of RCP in which trajectories exit the polytope in a monotonic sense; this provides a triangulation-independent solution of RCP. Second, we develop a Lyapunov-like theory for verifying if RCP is solved using a given candidate controller. This involves the introduction of the notion of generalized flow functions, a LaSalle Principle for RCP, and several converse theorems on existence of generalized flow functions. Third, we study the relationship between affine feedbacks and continuous state feedbacks for RCP on simplices. Although the two feedback classes have been shown to be equivalent under an assumption on the triangulation of the state space, we show by a counterexample that the equivalence is no longer true under arbitrary triangulations. Then we provide for single-input systems a constructive method for the synthesis of multi-affine feedbacks for RCP on simplices.

Esta pesquisa nasce de um questionamento proveniente do cotidiano do professor de língua estrangeira: devem ser corrigidos os erros cometidos pelos alunos, enquanto eles falam em atividades de conversação? Com o intuito de responder a essa pergunta, o presente estudo aponta para os efeitos em aprendizes brasileiros de italiano de dois tipos distintos de intervenções corretivas, durante a realização de atividades de produção oral a partir de temas pré-definidos, tipologia de exercício usada para a prática de conversação em sala de aula. Na fase experimental do estudo, os seis alunos participantes foram divididos em três duplas: os alunos de uma das duplas tiveram os seus erros tratados com correções implícitas do tipo retomada (recasts), os de outra dupla receberam correções explícitas metalinguísticas e os alunos da terceira dupla não tiveram os seus erros tratados com nenhum tipo de intervenção corretiva (grupo de controle). Foram escolhidos dois elementos gramaticais como alvo das correções: o uso do verbo esserci e a flexão nominal de número. As correções implícitas previam a interrupção da fala do aluno e o fornecimento imediato da forma correta sem explicações, assim que era identificado um erro; já as correções explícitas consistiam na interrupção da fala e na explicação metalinguística sobre o erro cometido pelo aluno, sem o fornecimento da forma correta, à qual o aluno devia chegar autonomamente. Foram usados três tipos distintos de testes gramaticais em três diferentes momentos (antes, depois e um mês depois das produções orais com as intervenções corretivas), na tentativa de identificar alterações no nível de conhecimento implícito e explícito dos alunos em relação aos elementos linguísticos focalizados na pesquisa. Foram também realizados uma entrevista e um questionário, para a obtenção de dados qualitativos complementares. Não houve resultados homogêneos entre os alunos de cada dupla. Portanto, os efeitos dos dois tipos de tratamento dos erros são descritos em função do perfil de cada aluno e do tipo de erro corrigido. A esses fatores acrescentam-se o tipo e o propósito da atividade durante a qual ocorrem a conversação e o movimento corretivo. Em suma, mais do que apontar para a supremacia de um tipo de correção em detrimento do outro, o presente estudo traz uma discussão que pode auxiliar o professor de língua estrangeira, tanto a se preparar para propor atividades de produção oral em sala de aula, quanto a decidir como reagir ao identificar a ocorrência de um erro na fala de seus alunos. / This research stems from a daily question faced by foreign language teachers: should the errors made by students be corrected during their speeches in conversation activities? In order to answer this question, this study points to the effects on two distinct types of corrective interventions on Brazilian learners of Italian, while performing oral production activity with predefined themes, a type of exercise used to practice conversation in the classroom. In the empirical phase of the study, the six participants were split into three pairs: students from one of the pairs received implicit feedback (recasts), students from the second pair received explicit feedback (metalinguistic explanation) and students of the third pair did not receive any corrective intervention (control group). Two grammatical elements were chosen as correction targets: the use of the verb esserci and the inflectional rules (plural and singular) in noun phrases. Implicit feedbacks were performed through interruption of students speech and immediate supply of the right form, as soon as an error was identified, without further explanation; explicit feedbacks were made by the interruption of the speech and metalinguistic explanation about the students error, without providing the correct form to which the student should arrive autonomously. We used three different types of grammar tests at different times of the study (before, after and one month after the oral productions with corrective interventions), in an attempt to identify changes in the level of implicit and explicit knowledge of the students regarding the target structures. We also conducted interviews and questionnaires to obtain additional qualitative data. We could not find homogeneous results among students of each pair. Therefore, the effects of each type of error treatment are described according to the profile of each student and the type of error. To these factors we should add the type and purpose of the activity during which have occurred both the conversation and the corrective interventions. In short, rather than point to the supremacy of one type of feedback over the other one, this study provides a discussion that may help foreign language teachers to prepare themselves not only to propose oral production activities in the classroom, but also to decide how to react when they identify an error in the speech of their students.

Esta pesquisa nasce de um questionamento proveniente do cotidiano do professor de língua estrangeira: devem ser corrigidos os erros cometidos pelos alunos, enquanto eles falam em atividades de conversação? Com o intuito de responder a essa pergunta, o presente estudo aponta para os efeitos em aprendizes brasileiros de italiano de dois tipos distintos de intervenções corretivas, durante a realização de atividades de produção oral a partir de temas pré-definidos, tipologia de exercício usada para a prática de conversação em sala de aula. Na fase experimental do estudo, os seis alunos participantes foram divididos em três duplas: os alunos de uma das duplas tiveram os seus erros tratados com correções implícitas do tipo retomada (recasts), os de outra dupla receberam correções explícitas metalinguísticas e os alunos da terceira dupla não tiveram os seus erros tratados com nenhum tipo de intervenção corretiva (grupo de controle). Foram escolhidos dois elementos gramaticais como alvo das correções: o uso do verbo esserci e a flexão nominal de número. As correções implícitas previam a interrupção da fala do aluno e o fornecimento imediato da forma correta sem explicações, assim que era identificado um erro; já as correções explícitas consistiam na interrupção da fala e na explicação metalinguística sobre o erro cometido pelo aluno, sem o fornecimento da forma correta, à qual o aluno devia chegar autonomamente. Foram usados três tipos distintos de testes gramaticais em três diferentes momentos (antes, depois e um mês depois das produções orais com as intervenções corretivas), na tentativa de identificar alterações no nível de conhecimento implícito e explícito dos alunos em relação aos elementos linguísticos focalizados na pesquisa. Foram também realizados uma entrevista e um questionário, para a obtenção de dados qualitativos complementares. Não houve resultados homogêneos entre os alunos de cada dupla. Portanto, os efeitos dos dois tipos de tratamento dos erros são descritos em função do perfil de cada aluno e do tipo de erro corrigido. A esses fatores acrescentam-se o tipo e o propósito da atividade durante a qual ocorrem a conversação e o movimento corretivo. Em suma, mais do que apontar para a supremacia de um tipo de correção em detrimento do outro, o presente estudo traz uma discussão que pode auxiliar o professor de língua estrangeira, tanto a se preparar para propor atividades de produção oral em sala de aula, quanto a decidir como reagir ao identificar a ocorrência de um erro na fala de seus alunos. / This research stems from a daily question faced by foreign language teachers: should the errors made by students be corrected during their speeches in conversation activities? In order to answer this question, this study points to the effects on two distinct types of corrective interventions on Brazilian learners of Italian, while performing oral production activity with predefined themes, a type of exercise used to practice conversation in the classroom. In the empirical phase of the study, the six participants were split into three pairs: students from one of the pairs received implicit feedback (recasts), students from the second pair received explicit feedback (metalinguistic explanation) and students of the third pair did not receive any corrective intervention (control group). Two grammatical elements were chosen as correction targets: the use of the verb esserci and the inflectional rules (plural and singular) in noun phrases. Implicit feedbacks were performed through interruption of students speech and immediate supply of the right form, as soon as an error was identified, without further explanation; explicit feedbacks were made by the interruption of the speech and metalinguistic explanation about the students error, without providing the correct form to which the student should arrive autonomously. We used three different types of grammar tests at different times of the study (before, after and one month after the oral productions with corrective interventions), in an attempt to identify changes in the level of implicit and explicit knowledge of the students regarding the target structures. We also conducted interviews and questionnaires to obtain additional qualitative data. We could not find homogeneous results among students of each pair. Therefore, the effects of each type of error treatment are described according to the profile of each student and the type of error. To these factors we should add the type and purpose of the activity during which have occurred both the conversation and the corrective interventions. In short, rather than point to the supremacy of one type of feedback over the other one, this study provides a discussion that may help foreign language teachers to prepare themselves not only to propose oral production activities in the classroom, but also to decide how to react when they identify an error in the speech of their students.

It is increasingly recognized that individuals of the same species differ from each other and influence and respond to their environment in unique ways. This thesis deals with size variation among individuals that not only are of the same species but also of similar age. Such variation may develop even when individuals are born in the same environment, i.e. within a cohort. I have studied the sources and consequences of variation within and among cohorts from egg through early ontogeny using young-of-the-year (YOY) perch (Perca fluviatilis) as study organism. In agreement with predictions based on model results only taking exploitative interactions among individuals into account, I found that the broader the initial size distributions were, the more did the degree of size variation among individuals decrease over time. Still, with initially small size variation among individuals, in several experiments also size divergence was observed. Furthermore, size variation among individuals increased more under high compared to at low densities. Increased size variation over time may be explained by size-dependent diet shifts allowing for initially larger individuals to make an early diet shift when the first resource becomes limiting. However, as size divergence also was observed in situations with only shared resources available, it can be concluded that diet shifts are not a prerequisite for size divergence in young animal cohorts. Hence, I also suggest that mechanisms not related to competition for limiting resources, such as genetic variation, stochasticity and behavioural traits must be taken into account, especially when initial size differences are small. The importance of considering size variation among individuals within cohorts was demonstrated in a study of winter mortality in YOY perch cohorts. A large individual size in autumn was shown to increase overwinter survival within cohorts. However, late summer growth rather than average body size reached in autumn explained variation in overwinter survival between cohorts. Higher accumulation to lipid reserves and accordingly lower mortality over winter was observed in years with high growth rates late in the season. In another study I showed that apparent patterns of density-dependent growth can emerge among larval fish, but rather than a result of density-dependent resource limitation this was due to variation in size-selective predation pressure. Individuals in the right end of the size distributions grew in to a high predation pressure from cannibalistic perch when cannibal density was high, coinciding with high larval perch densities. Finally, as substantial size variation among individuals can develop within cohorts, also intra-cohort cannibalism can occur. Using a physiologically structured population model it was shown that the development of size bimodality within cohorts as a result of intra-cohort cannibalism is critically dependent on long hatching periods, high victim densities and density-dependent feedbacks on shared resources. / Det faktum att individer som tillhör samma art skiljer sig från varandra och påverkar och påverkas av sin omgivande miljö på ett unikt sätt tillskrivs allt större betydelse inom ekologin. Den här avhandlingen handlar framför allt om storleksvariation mellan individer som förutom att tillhöra samma art dessutom tillhör samma årsklass. Sådan storleksvariation kan till och med utvecklas mellan individer som föds och växer upp i samma miljö (inom en kohort). Jag har studerat orsaker bakom och konsekvenser av variation inom och mellan kohorter. Som studieorganism har jag använt mig av årsyngel av abborre (Perca fluviatilis). I överensstämmelse med förutsägelser baserade på en modell som enbart tar hänsyn till konkurrens om en gemensam resurs visade det sig såväl i ett dammexperiment som i en naturlig sjö att ju bredare den initiala storleksfördelningen var desto mer minskade graden av variation i kroppsstorlek mellan individer över tid. Å andra sidan, när den initiala variationen var relativt liten observerades i flera oberoende experiment även storleksdivergens över tid mellan individer. Variationen i storlek ökade särskilt i miljöer med höga tätheter av konsumenter (abborrar). För att förstå de bakomliggande mekanismerna av sådana täthetseffekter måste man ta hänsyn till den återkoppling som sker mellan antalet konsumenter och mängden resurser. Ökad storleksvariation över tid skulle kunna förklaras med storleksberoende dietskiften som tillåter individer med en initial storleksfördel att genomgå ett tidigt dietskifte samtidigt som tillgången av den första resursen begränsar övriga individers tillväxt. Eftersom storleksdivergens även observerades i situationer där enbart en delad resurs var tillgänglig kan man dock dra slutsatsen att dietskiften inte är en förutsättning för storleksdivergens inom kohorter. Jag föreslår därför också att mekanismer som inte är relaterade till konkurrens om en begränsad resurs, såsom inneboende variation mellan individer och variation i beteendemönster bör beaktas för att förklara uppkomsten av storleksvariation, speciellt i de fall då den initiala variationen är liten. De resultat som presenteras visar på betydelsen av att ta hänsyn till storleksvariation mellan individer. Storlek efter den första tillväxtsäsongen var viktig för att förklara vilka individer inom en kohort av årsyngel som överlever sin första vinter. För att förklara variation i vinteröverlevnad mellan kohorter (mellan år och sjöar) var däremot hög tillväxt sent på säsongen (oberoende av medelstorlek på hösten) avgörande. Högre ackumulering av fettreserver och lägre mortalitet inom kohorter av årsyngel under vintern observerades under år med hög tillväxt under den senare delen av tillväxtsäsongen. I en annan studie visade jag att mönster som tyder på täthetsberoende tillväxt kan uppkomma hos fiskyngel men att detta fenomen snarare än täthetsberoende resursbegränsning förklaras av variation i storleksberoende predationstryck. De större individerna inom kohorterna växte in i ett högt predationstryck från kannibalistiska abborrar när tätheten av kannibaler var hög, något som samkorrelerade med höga tätheter av yngel. Slutligen, då en hög grad av storleksvariation mellan individer kan utvecklas inom kohorter är även kannibalism mellan individer inom en kohort möjligt. Genom att använda mig av en så kallad fysiologiskt strukturerad populationsmodell kunde jag visa att divergerande tillväxtkurvor mellan kannibaler och deras byten (vilket resulterar i storleksbimodalitet) som ett resultat av kannibalism inom kohorter är beroende av den tid det tar för ynglen att kläcka ut, antalet bytesfiskar per kannibal samt den återkoppling som finns mellan konsumenterna och deras gemensamma resurs (djurplankton).