Are shrubs really a sign of declining ecosystem function? Disentangling the myths and truths of woody encroachment in Australia

Since European settlement, there has been a dramatic increase in the density, cover and distribution of woody plants in former grassland and open woodland. There is a widespread belief that shrub encroachment is synonymous with declines in ecosystem functions, and often it is associated with landscape degradation or desertification. Indeed, this decline in ecosystem functioning is considered to be driven largely by the presence of the shrubs themselves. This prevailing paradigm has been the basis for an extensive program of shrub removal, based on the view that it is necessary to reinstate the original open woodland or grassland structure from which shrublands are thought to have been derived. We review existing scientific evidence, particularly focussed on eastern Australia, to question the notion that shrub encroachment leads to declines in ecosystem functions. We then summarise this scientific evidence into two conceptual models aimed at optimising landscape management to maximise the services provided by shrub-encroached areas. The first model seeks to reconcile the apparent conflicts between the patch- and landscape-level effects of shrubs. The second model identifies the ecosystem services derived from different stages of shrub encroachment. We also examined six ecosystem services provided by shrublands (biodiversity, soil C, hydrology, nutrient provision, grass growth and soil fertility) by using published and unpublished data. We demonstrated the following: (1) shrub effects on ecosystems are strongly scale-, species- and environment-dependent and, therefore, no standardised management should be applied to every case; (2) overgrazing dampens the generally positive effect of shrubs, leading to the misleading relationship between encroachment and degradation; (3) woody encroachment per se does not hinder any of the functions or services described above, rather it enhances many of them; (4) no single shrub-encroachment state (including grasslands without shrubs) will maximise all services; rather, the provision of ecosystem goods and services by shrublands requires a mixture of different states; and (5) there has been little rigorous assessment of the long-term effectiveness of removal and no evidence that this improves land condition in most cases. Our review provides the basis for an improved, scientifically based understanding and management of shrublands, so as to balance the competing goals of providing functional habitats, maintaining soil processes and sustaining pastoral livelihoods.

Exercise efficiency relates with mitochondrial content and function in older adults

Chronic aerobic exercise has been shown to increase exercise efficiency, thus allowing less energy expenditure for a similar amount of work. The extent to which skeletal muscle mitochondria play a role in this is not fully understood, particularly in an elderly population. The purpose of this study was to determine the relationship of exercise efficiency with mitochondrial content and function. We hypothesized that the greater the mitochondrial content and/or function, the greater would be the efficiencies. Thirty-eight sedentary (S, n = 23, 10F/13M) or athletic (A, n = 15, 6F/9M) older adults (66.8 ± 0.8 years) participated in this cross sectional study. V˙O2peak was measured with a cycle ergometer graded exercise protocol (GXT). Gross efficiency (GE, %) and net efficiency (NE, %) were estimated during a 1-h submaximal test (55% V˙O2peak). Delta efficiency (DE, %) was calculated from the GXT. Mitochondrial function was measured as ATPmax (mmol/L/s) during a PCr recovery protocol with (31)P-MR spectroscopy. Muscle biopsies were acquired for determination of mitochondrial volume density (MitoVd, %). Efficiencies were 17% (GE), 14% (NE), and 16% (DE) higher in A than S. MitoVD was 29% higher in A and ATPmax was 24% higher in A than in S. All efficiencies positively correlated with both ATPmax and MitoVd. Chronically trained older individuals had greater mitochondrial content and function, as well as greater exercise efficiencies. GE, NE, and DE were related to both mitochondrial content and function. This suggests a possible role of mitochondria in improving exercise efficiency in elderly athletic populations and allowing conservation of energy at moderate workloads.

Temperature Prediction Model for Bone Drilling Based on Density Distribution and In Vivo Experiments for Minimally Invasive Robotic Cochlear Implantation

Surgical robots have been proposed ex vivo to drill precise holes in the temporal bone for minimally invasive cochlear implantation. The main risk of the procedure is damage of the facial nerve due to mechanical interaction or due to temperature elevation during the drilling process. To evaluate the thermal risk of the drilling process, a simplified model is proposed which aims to enable an assessment of risk posed to the facial nerve for a given set of constant process parameters for different mastoid bone densities. The model uses the bone density distribution along the drilling trajectory in the mastoid bone to calculate a time dependent heat production function at the tip of the drill bit. Using a time dependent moving point source Green's function, the heat equation can be solved at a certain point in space so that the resulting temperatures can be calculated over time. The model was calibrated and initially verified with in vivo temperature data. The data was collected in minimally invasive robotic drilling of 12 holes in four different sheep. The sheep were anesthetized and the temperature elevations were measured with a thermocouple which was inserted in a previously drilled hole next to the planned drilling trajectory. Bone density distributions were extracted from pre-operative CT data by averaging Hounsfield values over the drill bit diameter. Post-operative [Formula: see text]CT data was used to verify the drilling accuracy of the trajectories. The comparison of measured and calculated temperatures shows a very good match for both heating and cooling phases. The average prediction error of the maximum temperature was less than 0.7 °C and the average root mean square error was approximately 0.5 °C. To analyze potential thermal damage, the model was used to calculate temperature profiles and cumulative equivalent minutes at 43 °C at a minimal distance to the facial nerve. For the selected drilling parameters, temperature elevation profiles and cumulative equivalent minutes suggest that thermal elevation of this minimally invasive cochlear implantation surgery may pose a risk to the facial nerve, especially in sclerotic or high density mastoid bones. Optimized drilling parameters need to be evaluated and the model could be used for future risk evaluation.

Density-dependent dynamics of a dominant rain forest tree change with juvenile stage and time of masting

Although negative density dependence (NDD) can facilitate tree species coexistence in forests, the underlying mechanisms can differ, and rarely are the dynamics of seedlings and saplings studied together. Herein we present and discuss a novel mechanism based on our investigation of NDD predictions for the large, grove-forming ectomycorrhizal mast fruiting tree, Microberlinia bisulcata (Caesalpiniaceae), in an 82.5-ha plot at Korup, Cameroon. We tested whether juvenile density, size, growth and survival decreases with increasing conspecific adult basal area for 3245 ‘new’ seedlings and 540 ‘old’ seedlings (< 75-cm tall) during an approximately 4-year study period (2008–2012) and for 234 ‘saplings’ (≥ 75-cm tall) during an approximately 6-year study period (2008–2014). We found that the respective densities of new seedlings, old seedlings and saplings were positively, not and negatively related to increasing BA. Maximum leaf numbers and heights of old seedlings were negatively correlated with increasing basal areas, as were sapling heights and stem diameters. Whereas survivorship of new seedlings decreased by more than one-half with increasing basal area over its range in 2010–2012, that of old seedlings decreased by almost two-thirds, but only in 2008–2010, and was generally unrelated to conspecific seedling density. In 2010–2012 relative growth rates in new seedlings’ heights decreased with increasing basal area, as well as with increasing seedling density, together with increasing leaf numbers, whereas old seedlings’ growth was unrelated to either conspecific density or basal area. Saplings of below-average height had reduced survivorship with increasing basal area (probability decreasing from approx. 0.4 to 0.05 over the basal area range tested), but only sapling growth in terms of leaf numbers decreased with increasing basal area. These static and dynamic results indicate that NDD is operating within this system, possibly stabilizing the M. bisulcata population. However, these NDD patterns are unlikely to be caused by symmetric competition or by consumers. Instead, an alternative mechanism for conspecific adult–juvenile negative feedback is proposed, one which involves the interaction between tree phenology and ectomycorrhizal linkages.

An extracytoplasmic function sigma factor operon regulates Myxococcus xanthus developmental gene expression

Myxococcus xanthus is a Gram-negative soil bacterium that undergoes multicellular development when high-density cells are starved on a solid surface. Expression of the 4445 gene, predicted to encode a periplasmic protein, commences 1.5 h after the initiation of development and requires starvation and high density conditions. Addition of crude or boiled supernatant from starving high-density cells restored 4445 expression to starving low-density cells. Addition of L-threonine or L-isoleucine to starving low-density cells also restored 4445 expression, indicating that the high-density signaling activity present in the supernatant might be composed of extracellular amino acids or small peptides. To investigate the circuitry integrating these starvation and high-density signals, the cis- and trans-acting elements controlling 4445 expression were identified. The 4445 transcription start site was determined by primer extension analysis to be 58 by upstream of the predicted translation start site. The promoter region contained a consensus sequence characteristic of e&barbelow;xtrac&barbelow;ytoplasmic f&barbelow;unction (ECF) sigma factor-dependent promoters, suggesting that 4445 expression might be regulated by an ECF sigma factor-dependent pathway, which are known to respond to envelope stresses. The small size of the minimum regulatory region, identified by 5′-end deletion analysis as being only 66 by upstream of the transcription start site, suggests that RNA polymerase could be the sole direct regulator of 4445 expression. To identify trans-acting negative regulators of 4445 expression, a strain containing a 4445-lacZ was mutagenized using the Himar1-tet transposon. The four transposon insertions characterized mapped to an operon encoding a putative ECF sigma factor, ecfA; an anti-sigma factor, reaA; and a negative regulator, reaB. The reaA and the reaB mutants expressed 4445 during growth and development at levels almost 100-fold higher than wild type, indicating that these genes encode negative regulators. The ecfA mutant expressed 4445-lacZ at basal levels, indicating that ecfA is a positive regulator. High Mg2+ concentrations over-stimulated this ecfA pathway possibly due to the depletion of exopolysaccharides and assembled type IV pili. These data indicate that the ecfA operon encodes a new regulatory stress pathway that integrates and transduces starvation and cell density cues during early development and is also responsive to cell-surface alterations.^

Dissolution rate spectra data of calcite single crystal and micrite material

The large discrepancy between field and laboratory measurements of mineral reaction rates is a long-standing problem in earth sciences, often attributed to factors extrinsic to the mineral itself. Nevertheless, differences in reaction rate are also observed within laboratory measurements, raising the possibility of intrinsic variations as well. Critical insight is available from analysis of the relationship between the reaction rate and its distribution over the mineral surface. This analysis recognizes the fundamental variance of the rate. The resulting anisotropic rate distributions are completely obscured by the common practice of surface area normalization. In a simple experiment using a single crystal and its polycrystalline counterpart, we demonstrate the sensitivity of dissolution rate to grain size, results that undermine the use of "classical" rate constants. Comparison of selected published crystal surface step retreat velocities (Jordan and Rammensee, 1998) as well as large single crystal dissolution data (Busenberg and Plummer, 1986) provide further evidence of this fundamental variability. Our key finding highlights the unsubstantiated use of a single-valued "mean" rate or rate constant as a function of environmental conditions. Reactivity predictions and long-term reservoir stability calculations based on laboratory measurements are thus not directly applicable to natural settings without a probabilistic approach. Such a probabilistic approach must incorporate both the variation of surface energy as a general range (intrinsic variation) as well as constraints to this variation owing to the heterogeneity of complex material (e.g., density of domain borders). We suggest the introduction of surface energy spectra (or the resulting rate spectra) containing information about the probability of existing rate ranges and the critical modes of surface energy.

Geochemistry, lithology, and bulk density of Cricket Flat paleosol samples

The high-resolution marine isotope climate record indicates pronounced global cooling during the Langhian (16-13.8 Ma), beginning with the warm middle Miocene climatic optimum and ending with significant Antarctic ice sheet expansion and the transition to "icehouse" conditions. Terrestrial paleoclimate data from this interval is sparse and sometimes conflicting. In particular, there are gaps in the terrestrial record in the Pacific Northwest during the late Langhian and early Serravallian between about 14.5 and 12.5 Ma. New terrestrial paleoclimate data from this time and region could reconcile these conflicting records. Paleosols are particularly useful for reconstructing paleoenvironment because the rate and style of pedogenesis is primarily a function of surface environmental conditions; however, complete and well-preserved paleosols are uncommon. Most soils form in erosive environments that are not preserved, or in environments such as floodplains that accumulate in small increments; the resulting cumulic soils are usually thin, weakly developed, and subject to diagenetic overprinting from subsequent soils. The paleosol at Cricket Flat in northeastern Oregon is an unusually complete and well-preserved paleosol from a middle Miocene volcanic sequence in the Powder River Volcanic Field. An olivine basalt flow buried the paleosol at approximately 13.8 ± 0.6 Ma, based on three 40Ar/39Ar dates on the basalt. We described the Cricket Flat paleosol and used its physical and chemical profile and micromorphology to assess pedogenesis. The Cricket Flat paleosol is an Ultisol-like paleosol, chemically consistent with a high degree of weathering. Temperature and rainfall proxies suggest that Cricket Flat received 1120 ± 180 mm precipitation y-1 and experienced a mean annual temperature of 14.5 ± 2.1 °C during the formation of the paleosol, significantly warmer and wetter than today. This suggests slower cooling after the middle Miocene climatic optimum than is seen in the existing paleosol record.

Density of 9 sediment cores from the Weddell Sea

The wet bulk density is one of the most important parameters of the physical and geological properties of marine sediments. The density is connected directly with sedimentation history and a few sedirnent properties. Knowledge of the fine scale density-depth structure is the base for many model calculations, for both sedimentological and palaeoclimatic research. A density measurement system was designed and built at the Alfred Wegener Institute in Bremerhaven for measuring the wet buk density of sediment cores with high resolution in a non-destructive way. The density is deterrnined by measuring the absorption of Gamma-rays in the sediment. This principle has been used since the 50's in materials research and in the geosciences. In the present case, Cs137 is used as the radioactive source and the intensity is measured by a detector system (scintillator and photomultiplier). Density values are obtainable in both longitudinal core sections and planar cross-sections (the latter are a function of the axial rotation angle). Special studies on inhomogenity can be applied with core rotation. Detection of ice rafted debris (IRD) is made possible with this option. The processes that run the density measurement system are computer controlled. Besides the absorption measurement the core diameter at every measurement point is determined with a potentiometric system. The data values taken are stored on a personal computer. Before starting routine measurements on the sediment cores, a few experiments conceming the statistical aspects of the gamma-ray signal and its accuracy were carried out. These experiments led to such things as the optimum operational parameters. A high spatial resolution in the mm-range is possible with the 4mm-thin gamma-ray measurements. Within five seconds the wet bulk density can be deterrnined with an absolute accuracy of 1%. A comparison between data measured with the new system and conventional measurements on core samples after core splitting shows an agreement within +I- 5% for most of the values. For this thesis, density determinations were carried out on ten sediment cores. A few sediment characteristics are obtainable from using just the standard measurement results without core rotation. In addition to differentes and steps in the absolute density range, variations in the "frequency" of the density-depth structure can be detected due to the close spatial measurement interval and high resolution. Examples from measurements with small (9°) and great (90°) angle increments show that abrupt and smooth transitional changes of sedirnent layers as well as ice rafted debris of several dimensions can be detected and distiflguished clearly. After the presentation of the wet bulk density results, a comparison with data from other investigations was made. Measurements of the electrical resistivity correlated very well with the density data because both parameters are closely related to the porosity of the sedirnent. Additionally, results from measurements of the magnetic susceptibility and from ultra-sonic wave velocity investigations were considered for a integrative interpretation. The correlation of these both parameters and wet bulk density data is strongly dependent on the local (environmental) conditions. Finally, the densities were compared with recordings from sediment-echographic soundings and an X-ray computer tomography analysis. The individual results of all investigations were then finally combined into an accurate picture of the core. Problems of ambiguity, which exist when just one Parameter is determined alone, can be reduced more or less according to the number of parameters and sedimentary characteristics measured. The important role of the density data among other parameters of such an integrated interpretation is evident. Evidence of this role include the high resolution of the measurement, the excellent accuracy and the key position within methods and parameters concerning marine sediments.

Asymptotically Optimum Estimation of a Probability in Inverse Binomial Sampling under General Loss Functions

The optimum quality that can be asymptotically achieved in the estimation of a probability p using inverse binomial sampling is addressed. A general definition of quality is used in terms of the risk associated with a loss function that satisfies certain assumptions. It is shown that the limit superior of the risk for p asymptotically small has a minimum over all (possibly randomized) estimators. This minimum is achieved by certain non-randomized estimators. The model includes commonly used quality criteria as particular cases. Applications to the non-asymptotic regime are discussed considering specific loss functions, for which minimax estimators are derived.

An integral model for natural regeneration of Pinus pinea L. in the Northern Plateau (Spain) = Modelo integral de regeneración natural para Pinus pinea L. en la Meseta Norte (España)

Natural regeneration is an ecological key-process that makes plant persistence possible and, consequently, it constitutes an essential element of sustainable forest management. In this respect, natural regeneration in even-aged stands of Pinus pinea L. located in the Spanish Northern Plateau has not always been successfully achieved despite over a century of pine nut-based management. As a result, natural regeneration has recently become a major concern for forest managers when we are living a moment of rationalization of investment in silviculture. The present dissertation is addressed to provide answers to forest managers on this topic through the development of an integral regeneration multistage model for P. pinea stands in the region. From this model, recommendations for natural regeneration-based silviculture can be derived under present and future climate scenarios. Also, the model structure makes it possible to detect the likely bottlenecks affecting the process. The integral model consists of five submodels corresponding to each of the subprocesses linking the stages involved in natural regeneration (seed production, seed dispersal, seed germination, seed predation and seedling survival). The outputs of the submodels represent the transitional probabilities between these stages as a function of climatic and stand variables, which in turn are representative of the ecological factors driving regeneration. At subprocess level, the findings of this dissertation should be interpreted as follows. The scheduling of the shelterwood system currently conducted over low density stands leads to situations of dispersal limitation since the initial stages of the regeneration period. Concerning predation, predator activity appears to be only limited by the occurrence of severe summer droughts and masting events, the summer resulting in a favourable period for seed survival. Out of this time interval, predators were found to almost totally deplete seed crops. Given that P. pinea dissemination occurs in summer (i.e. the safe period against predation), the likelihood of a seed to not be destroyed is conditional to germination occurrence prior to the intensification of predator activity. However, the optimal conditions for germination seldom take place, restraining emergence to few days during the fall. Thus, the window to reach the seedling stage is narrow. In addition, the seedling survival submodel predicts extremely high seedling mortality rates and therefore only some individuals from large cohorts will be able to persist. These facts, along with the strong climate-mediated masting habit exhibited by P. pinea, reveal that viii the overall probability of establishment is low. Given this background, current management –low final stand densities resulting from intense thinning and strict felling schedules– conditions the occurrence of enough favourable events to achieve natural regeneration during the current rotation time. Stochastic simulation and optimisation computed through the integral model confirm this circumstance, suggesting that more flexible and progressive regeneration fellings should be conducted. From an ecological standpoint, these results inform a reproductive strategy leading to uneven-aged stand structures, in full accordance with the medium shade-tolerant behaviour of the species. As a final remark, stochastic simulations performed under a climate-change scenario show that regeneration in the species will not be strongly hampered in the future. This resilient behaviour highlights the fundamental ecological role played by P. pinea in demanding areas where other tree species fail to persist.

Contributions to Bayesian network learning with applications to neuroscience

Neuronal morphology is a key feature in the study of brain circuits, as it is highly related to information processing and functional identification. Neuronal morphology affects the process of integration of inputs from other neurons and determines the neurons which receive the output of the neurons. Different parts of the neurons can operate semi-independently according to the spatial location of the synaptic connections. As a result, there is considerable interest in the analysis of the microanatomy of nervous cells since it constitutes an excellent tool for better understanding cortical function. However, the morphologies, molecular features and electrophysiological properties of neuronal cells are extremely variable. Except for some special cases, this variability makes it hard to find a set of features that unambiguously define a neuronal type. In addition, there are distinct types of neurons in particular regions of the brain. This morphological variability makes the analysis and modeling of neuronal morphology a challenge. Uncertainty is a key feature in many complex real-world problems. Probability theory provides a framework for modeling and reasoning with uncertainty. Probabilistic graphical models combine statistical theory and graph theory to provide a tool for managing domains with uncertainty. In particular, we focus on Bayesian networks, the most commonly used probabilistic graphical model. In this dissertation, we design new methods for learning Bayesian networks and apply them to the problem of modeling and analyzing morphological data from neurons. The morphology of a neuron can be quantified using a number of measurements, e.g., the length of the dendrites and the axon, the number of bifurcations, the direction of the dendrites and the axon, etc. These measurements can be modeled as discrete or continuous data. The continuous data can be linear (e.g., the length or the width of a dendrite) or directional (e.g., the direction of the axon). These data may follow complex probability distributions and may not fit any known parametric distribution. Modeling this kind of problems using hybrid Bayesian networks with discrete, linear and directional variables poses a number of challenges regarding learning from data, inference, etc. In this dissertation, we propose a method for modeling and simulating basal dendritic trees from pyramidal neurons using Bayesian networks to capture the interactions between the variables in the problem domain. A complete set of variables is measured from the dendrites, and a learning algorithm is applied to find the structure and estimate the parameters of the probability distributions included in the Bayesian networks. Then, a simulation algorithm is used to build the virtual dendrites by sampling values from the Bayesian networks, and a thorough evaluation is performed to show the model’s ability to generate realistic dendrites. In this first approach, the variables are discretized so that discrete Bayesian networks can be learned and simulated. Then, we address the problem of learning hybrid Bayesian networks with different kinds of variables. Mixtures of polynomials have been proposed as a way of representing probability densities in hybrid Bayesian networks. We present a method for learning mixtures of polynomials approximations of one-dimensional, multidimensional and conditional probability densities from data. The method is based on basis spline interpolation, where a density is approximated as a linear combination of basis splines. The proposed algorithms are evaluated using artificial datasets. We also use the proposed methods as a non-parametric density estimation technique in Bayesian network classifiers. Next, we address the problem of including directional data in Bayesian networks. These data have some special properties that rule out the use of classical statistics. Therefore, different distributions and statistics, such as the univariate von Mises and the multivariate von Mises–Fisher distributions, should be used to deal with this kind of information. In particular, we extend the naive Bayes classifier to the case where the conditional probability distributions of the predictive variables given the class follow either of these distributions. We consider the simple scenario, where only directional predictive variables are used, and the hybrid case, where discrete, Gaussian and directional distributions are mixed. The classifier decision functions and their decision surfaces are studied at length. Artificial examples are used to illustrate the behavior of the classifiers. The proposed classifiers are empirically evaluated over real datasets. We also study the problem of interneuron classification. An extensive group of experts is asked to classify a set of neurons according to their most prominent anatomical features. A web application is developed to retrieve the experts’ classifications. We compute agreement measures to analyze the consensus between the experts when classifying the neurons. Using Bayesian networks and clustering algorithms on the resulting data, we investigate the suitability of the anatomical terms and neuron types commonly used in the literature. Additionally, we apply supervised learning approaches to automatically classify interneurons using the values of their morphological measurements. Then, a methodology for building a model which captures the opinions of all the experts is presented. First, one Bayesian network is learned for each expert, and we propose an algorithm for clustering Bayesian networks corresponding to experts with similar behaviors. Then, a Bayesian network which represents the opinions of each group of experts is induced. Finally, a consensus Bayesian multinet which models the opinions of the whole group of experts is built. A thorough analysis of the consensus model identifies different behaviors between the experts when classifying the interneurons in the experiment. A set of characterizing morphological traits for the neuronal types can be defined by performing inference in the Bayesian multinet. These findings are used to validate the model and to gain some insights into neuron morphology. Finally, we study a classification problem where the true class label of the training instances is not known. Instead, a set of class labels is available for each instance. This is inspired by the neuron classification problem, where a group of experts is asked to individually provide a class label for each instance. We propose a novel approach for learning Bayesian networks using count vectors which represent the number of experts who selected each class label for each instance. These Bayesian networks are evaluated using artificial datasets from supervised learning problems. Resumen La morfología neuronal es una característica clave en el estudio de los circuitos cerebrales, ya que está altamente relacionada con el procesado de información y con los roles funcionales. La morfología neuronal afecta al proceso de integración de las señales de entrada y determina las neuronas que reciben las salidas de otras neuronas. Las diferentes partes de la neurona pueden operar de forma semi-independiente de acuerdo a la localización espacial de las conexiones sinápticas. Por tanto, existe un interés considerable en el análisis de la microanatomía de las células nerviosas, ya que constituye una excelente herramienta para comprender mejor el funcionamiento de la corteza cerebral. Sin embargo, las propiedades morfológicas, moleculares y electrofisiológicas de las células neuronales son extremadamente variables. Excepto en algunos casos especiales, esta variabilidad morfológica dificulta la definición de un conjunto de características que distingan claramente un tipo neuronal. Además, existen diferentes tipos de neuronas en regiones particulares del cerebro. La variabilidad neuronal hace que el análisis y el modelado de la morfología neuronal sean un importante reto científico. La incertidumbre es una propiedad clave en muchos problemas reales. La teoría de la probabilidad proporciona un marco para modelar y razonar bajo incertidumbre. Los modelos gráficos probabilísticos combinan la teoría estadística y la teoría de grafos con el objetivo de proporcionar una herramienta con la que trabajar bajo incertidumbre. En particular, nos centraremos en las redes bayesianas, el modelo más utilizado dentro de los modelos gráficos probabilísticos. En esta tesis hemos diseñado nuevos métodos para aprender redes bayesianas, inspirados por y aplicados al problema del modelado y análisis de datos morfológicos de neuronas. La morfología de una neurona puede ser cuantificada usando una serie de medidas, por ejemplo, la longitud de las dendritas y el axón, el número de bifurcaciones, la dirección de las dendritas y el axón, etc. Estas medidas pueden ser modeladas como datos continuos o discretos. A su vez, los datos continuos pueden ser lineales (por ejemplo, la longitud o la anchura de una dendrita) o direccionales (por ejemplo, la dirección del axón). Estos datos pueden llegar a seguir distribuciones de probabilidad muy complejas y pueden no ajustarse a ninguna distribución paramétrica conocida. El modelado de este tipo de problemas con redes bayesianas híbridas incluyendo variables discretas, lineales y direccionales presenta una serie de retos en relación al aprendizaje a partir de datos, la inferencia, etc. En esta tesis se propone un método para modelar y simular árboles dendríticos basales de neuronas piramidales usando redes bayesianas para capturar las interacciones entre las variables del problema. Para ello, se mide un amplio conjunto de variables de las dendritas y se aplica un algoritmo de aprendizaje con el que se aprende la estructura y se estiman los parámetros de las distribuciones de probabilidad que constituyen las redes bayesianas. Después, se usa un algoritmo de simulación para construir dendritas virtuales mediante el muestreo de valores de las redes bayesianas. Finalmente, se lleva a cabo una profunda evaluaci ón para verificar la capacidad del modelo a la hora de generar dendritas realistas. En esta primera aproximación, las variables fueron discretizadas para poder aprender y muestrear las redes bayesianas. A continuación, se aborda el problema del aprendizaje de redes bayesianas con diferentes tipos de variables. Las mixturas de polinomios constituyen un método para representar densidades de probabilidad en redes bayesianas híbridas. Presentamos un método para aprender aproximaciones de densidades unidimensionales, multidimensionales y condicionales a partir de datos utilizando mixturas de polinomios. El método se basa en interpolación con splines, que aproxima una densidad como una combinación lineal de splines. Los algoritmos propuestos se evalúan utilizando bases de datos artificiales. Además, las mixturas de polinomios son utilizadas como un método no paramétrico de estimación de densidades para clasificadores basados en redes bayesianas. Después, se estudia el problema de incluir información direccional en redes bayesianas. Este tipo de datos presenta una serie de características especiales que impiden el uso de las técnicas estadísticas clásicas. Por ello, para manejar este tipo de información se deben usar estadísticos y distribuciones de probabilidad específicos, como la distribución univariante von Mises y la distribución multivariante von Mises–Fisher. En concreto, en esta tesis extendemos el clasificador naive Bayes al caso en el que las distribuciones de probabilidad condicionada de las variables predictoras dada la clase siguen alguna de estas distribuciones. Se estudia el caso base, en el que sólo se utilizan variables direccionales, y el caso híbrido, en el que variables discretas, lineales y direccionales aparecen mezcladas. También se estudian los clasificadores desde un punto de vista teórico, derivando sus funciones de decisión y las superficies de decisión asociadas. El comportamiento de los clasificadores se ilustra utilizando bases de datos artificiales. Además, los clasificadores son evaluados empíricamente utilizando bases de datos reales. También se estudia el problema de la clasificación de interneuronas. Desarrollamos una aplicación web que permite a un grupo de expertos clasificar un conjunto de neuronas de acuerdo a sus características morfológicas más destacadas. Se utilizan medidas de concordancia para analizar el consenso entre los expertos a la hora de clasificar las neuronas. Se investiga la idoneidad de los términos anatómicos y de los tipos neuronales utilizados frecuentemente en la literatura a través del análisis de redes bayesianas y la aplicación de algoritmos de clustering. Además, se aplican técnicas de aprendizaje supervisado con el objetivo de clasificar de forma automática las interneuronas a partir de sus valores morfológicos. A continuación, se presenta una metodología para construir un modelo que captura las opiniones de todos los expertos. Primero, se genera una red bayesiana para cada experto y se propone un algoritmo para agrupar las redes bayesianas que se corresponden con expertos con comportamientos similares. Después, se induce una red bayesiana que modela la opinión de cada grupo de expertos. Por último, se construye una multired bayesiana que modela las opiniones del conjunto completo de expertos. El análisis del modelo consensuado permite identificar diferentes comportamientos entre los expertos a la hora de clasificar las neuronas. Además, permite extraer un conjunto de características morfológicas relevantes para cada uno de los tipos neuronales mediante inferencia con la multired bayesiana. Estos descubrimientos se utilizan para validar el modelo y constituyen información relevante acerca de la morfología neuronal. Por último, se estudia un problema de clasificación en el que la etiqueta de clase de los datos de entrenamiento es incierta. En cambio, disponemos de un conjunto de etiquetas para cada instancia. Este problema está inspirado en el problema de la clasificación de neuronas, en el que un grupo de expertos proporciona una etiqueta de clase para cada instancia de manera individual. Se propone un método para aprender redes bayesianas utilizando vectores de cuentas, que representan el número de expertos que seleccionan cada etiqueta de clase para cada instancia. Estas redes bayesianas se evalúan utilizando bases de datos artificiales de problemas de aprendizaje supervisado.

Role of surface trap states on two-dimensional electron gas density in InAlN/AlN/GaN heterostructures

In order to clarify the effect of charged dislocations and surface donor states on the transport mechanisms in polar AlInN/AlN/GaN heterostructures, we have studied the current-voltage characteristics of Schottky junctions fabricated on AlInN/AlN/GaN heterostructures. The reverse-bias leakage current behaviour has been interpreted with a Poole-Frenkel emission of electrons from trap states near the metal-semiconductor junction to dislocation induced states. The variation of the Schottky barrier height as a function of the AlN layer thickness has been measured and discussed, considering the role of the surface states in the formation of the two dimensional electron gas at AlN/GaN interface.

Electrodynamic Tethers For Planetary And De-orbiting Missions

Nuevas aplicaciones tecnológicas y científicas mediante amarras electrodinámicas son analizadas para misiones planetarias. i) Primero, se considera un conjunto de amarras cilíndricas en paralelo (veleros electrosolares) para una misión interplanetaria. Los iones provenientes del viento solar son repelidos por el alto potencial de dichas amarras generando empuje sobre el velero. Para conocer el intercambio de momento que provocan los iones sobre las amarras se ha considerado un modelo de potencial estacionario. Se ha analizado la transferencia orbital de la Tierra a Júpiter siguiendo un método de optimización de trayectoria indirecto. ii) Una vez que el velero se encuentra cerca de Júpiter, se ha considerado el despliegue de una amarra para diferentes objetivos científicos. iia) Una amarra podría ser utilizada para diagnóstico de plasmas, al ser una fuente efectiva de ondas, y también como un generador de auroras artificiales. Una amarra conductora que orbite en la magnetosfera jovial es capaz de producir ondas. Se han analizado las diferentes ondas radiadas por un conductor por el que circula una corriente constante que sigue una órbita polar de alta excentricidad y bajo apoápside, como ocurre en la misión Juno de la NASA. iib) Además, se ha estudiado una misión tentativa que sigue una órbita ecuatorial (LJO) por debajo de los intensos cinturones de radiación. Ambas misiones requiren potencia eléctrica para los sistemas de comunicación e instrumentos científicos. Las amarras pueden generar potencia de manera más eficiente que otros sistemas que utlizan paneles solares o sistemas de potencia de radioisótopos (RPS). La impedancia de radiación es necesaria para determinar la corriente que circula por todo el circuito de la amarra. En un modelo de plasma frío, la radiación ocurre principalmente en los modos de Alfven y magnetosónica rápida, mostrando un elevado índice de refracción. Se ha estudiado la impedancia de radiación en amarras con recubrimiento aislante para los dos modos de radiación y cada una de las misiones. A diferencia del caso ionosférico terrestre, la baja densidad y el intenso campo magnético que aparecen en el entorno de Júpiter consiguen que la girofrecuencia de los electrones sea mucho mayor que la frecuencia del plasma; esto hace que el espectro de potencia para cada modo se modifique substancialmente, aumentando la velocidad de Alfven. Se ha estimado también la impedancia de radiación para amarras sin aislante conductor. En la misión LJO, un vehículo espacial bajando lentamente la altitud de su órbita permitiría estudiar la estructura del campo magnético y composición atmosférica para entender la formación, evolución, y estructura de Júpiter. Adicionalmente, si el contactor (cátodo) se apaga, se dice que la amarra flota eléctricamente, permitiendo emisión de haz de electrones que generan auroras. El continuo apagado y encendido produce pulsos de corriente dando lugar a emisiones de señales, que pueden ser utilizadas para diagnóstico del plasma jovial. En Órbita Baja Jovial, los iones que impactan contra una amarra polarizada negativamente producen electrones secundarios, que, viajando helicoidalmente sobre las líneas de campo magnético de Júpiter, son capaces de alcanzar su atmósfera más alta, y, de esta manera, generar auroras. Se han identificado cuáles son las regiones donde la amarra sería más eficiente para producir auroras. iic) Otra aplicación científica sugerida para la misión LJO es la detección de granos cargados que orbitan cerca de Júpiter. Los electrones de alta energía en este ambiente pueden ser modelados por una distribucción no Maxwelliana conocida como distribución kappa. En escenarios con plasmas complejos, donde los campos eléctricos en Júpiter pueden acelerar las cargas hasta velocidades que superen la velocidad térmica, este tipo de distribuciones son muy útiles. En este caso las colas de las distribuciones de electrones siguen una ley de potencias. Se han estudiado las fluctuaciones de granos cargados para funciones de distribución kappa. iii) La tesis concluye con el análisis para deorbitar satélites con amarras electrodinámicas que siguen una Órbita Baja Terrestre (LEO). Una amarra debe presentar una baja probabilidad de corte por pequeño debris y además debe ser suficientemente ligero para que el cociente entre la masa de la amarra y el satélite sea muy pequeño. En este trabajo se estiman las medidas de la longitud, anchura y espesor que debe tener una amarra para minimizar el producto de la probabilidad de corte por el cociente entre las masas de la amarra y el satélite. Se presentan resultados preliminares del diseño de una amarra con forma de cinta para deorbitar satélites relativamente ligeros como Cryosat y pesados como Envisat. Las misiones espaciales a planetas exteriores y en el ámbito terrestre plantean importantes retos científico-tecnológicos que deben ser abordados y solucionados. Por ello, desde el inicio de la era espacial se han diseñando novedosos métodos propulsivos, sistemas de guiado, navegación y control más robustos, y nuevos materiales para mejorar el rendimiento de los vehículos espaciales (SC). En un gran número de misiones interplanetarias y en todas las misiones a planetas exteriores se han empleado sistemas de radioisótopos (RPS) para generar potencia eléctrica en los vehículos espaciales y en los rovers de exploración. Estos sistemas emplean como fuente de energía el escaso y costoso plutonio-238. La NASA, por medio de un informe de la National Academy of Science (5 de Mayo del 2009), expresó una profunda preocupación por la baja cantidad de plutonio almacenado, insuficiente para desarrollar todas las misiones de exploración planetaria planeadas en el futuro [81, 91]. Esta circustancia ha llevado a dicha Agencia tomar la decisión de limitar el uso de estos sistemas RPS en algunas misiones de especial interés científico y una recomendación de alta prioridad para que el Congreso de los EEUU apruebe el reestablecimiento de la producción de plutonio-238, -son necesarios cerca de 5 kg de este material radiactivo al año-, para salvaguardar las misiones que requieran dichos sistemas de potencia a partir del año 2018. Por otro lado, la Agencia estadounidense ha estado considerando el uso de fuentes de energía alternativa; como la fisión nuclear a través del ambicioso proyecto Prometheus, para llevar a cabo una misión de exploración en el sistema jovial (JIMO). Finalmente, dicha misión fue desestimada por su elevado coste. Recientemente se han estado desarrollando sistemas que consigan energía a través de los recursos naturales que nos aporta el Sol, mediante paneles solares -poco eficientes para misiones a planetas alejados de la luz solar-. En este contexto, la misión JUNO del programa Nuevas Fronteras de la NASA, cuyo lanzamiento fue realizado con éxito en Agosto de 2011, va a ser la primera misión equipada con paneles solares que sobrevolará Júpiter en el 2015 siguiendo una órbita polar. Anteriormente se habían empleado los antes mencionados RPS para las misiones Pioneer 10,11, Voyager 1,2, Ulysses, Cassini-Huygens y Galileo (todas sobrevuelos excepto Galileo). Dicha misión seguirá una órbita elíptica de alta excentricidad con un periápside muy cercano a Júpiter, y apoápside lejano, evitando que los intensos cinturones de radiación puedan dañar los instrumentos de navegación y científicos. Un tether o amarra electrodinámica es capaz de operar como sistema propulsivo o generador de potencia, pero también puede ser considerado como solución científicotecnológica en misiones espaciales tanto en LEO (Órbita Baja Terrestre) como en planetas exteriores. Siguiendo una perspectiva histórica, durante las misiones terrestres TSS-1 (1992) y TSS-1R (1996) se emplearon amarras estandard con recubrimiento aislante en toda su longitud, aplicando como terminal anódico pasivo un colector esférico para captar electrones. En una geometría alternativa, propuesta por J. R. Sanmartín et al. (1993) [93], se consideró dejar la amarra sin recubrimiento aislante (“bare tether”), y sin colector anódico esférico, de forma que recogiera electrones a lo largo del segmento que resulta polarizado positivo, como si se tratara de una sonda de Langmuir de gran longitud. A diferencia de la amarra estandard, el “bare tether” es capaz de recoger electrones a lo largo de una superficie grande ya que este segmento es de varios kilómetros de longitud. Como el radio de la amarra es del orden de la longitud de Debye y pequeño comparado con el radio de Larmor de los electrones, permite una recolección eficiente de electrones en el régimen OML (Orbital Motion Limited) de sondas de Langmuir. La corriente dada por la teoría OML varía en función del perímetro y la longitud. En el caso de una cinta delgada, el perímetro depende de la anchura, que debe ser suficientemente grande para evitar cortes producidos por debris y micrometeoritos, y suficientemente pequeño para que la amarra funcione en dicho régimen [95]. En el experimento espacial TSS-1R mencionado anteriormente, se identificó una recolección de corriente más elevada que la que predecía el modelo teórico de Parker- Murphy, debido posiblemente a que se utilizaba un colector esférico de radio bastante mayor que la longitud de Debye [79]. En el caso de una amarra “bare”, que recoge electrones a lo largo de gran parte de su longitud, se puede producir un fenómeno conocido como atrapamiento adiabático de electrones (adiabatic electron trapping) [25, 40, 60, 73, 74, 97]. En el caso terrestre (LEO) se da la condición mesotérmica en la que la amarra se mueve con una velocidad muy superior a la velocidad térmica de los iones del ambiente y muy inferior a la velocidad térmica de los electrones. J. Laframboise y L. Parker [57] mostraron que, para una función de distribución quasi-isotrópica, la densidad de electrones debe entonces ser necesariamente inferior a la densidad ambiente. Por otra parte, debido a su flujo hipersónico y a la alta polarización positiva de la amarra, la densidad de los iones es mayor que la densidad ambiente en una vasta región de la parte “ram” del flujo, violando la condición de cuasi-neutralidad,-en una región de dimensión mayor que la longitud de Debye-. La solución a esta paradoja podría basarse en el atrapamiento adiabático de electrones ambiente en órbitas acotadas entorno al tether. ABSTRACT New technological and scientific applications by electrodynamic tethers for planetary missions are analyzed: i) A set of cylindrical, parallel tethers (electric solar sail or e-sail) is considered for an interplanetary mission; ions from the solar wind are repelled by the high potential of the tether, providing momentum to the e-sail. An approximated model of a stationary potential for a high solar wind flow is considered. With the force provided by a negative biased tether, an indirect method for the optimization trajectory of an Earth-to-Jupiter orbit transfer is analyzed. ii) The deployment of a tether from the e-sail allows several scientific applications in Jupiter. iia) It might be used as a source of radiative waves for plasma diagnostics and artificial aurora generator. A conductive tether orbiting in the Jovian magnetosphere produces waves. Wave radiation by a conductor carrying a steady current in both a polar, highly eccentric, low perijove orbit, as in NASA’s Juno mission, and an equatorial low Jovian orbit (LJO) mission below the intense radiation belts, is considered. Both missions will need electric power generation for scientific instruments and communication systems. Tethers generate power more efficiently than solar panels or radioisotope power systems (RPS). The radiation impedance is required to determine the current in the overall tether circuit. In a cold plasma model, radiation occurs mainly in the Alfven and fast magnetosonic modes, exhibiting a large refraction index. The radiation impedance of insulated tethers is determined for both modes and either mission. Unlike the Earth ionospheric case, the low-density, highly magnetized Jovian plasma makes the electron gyrofrequency much larger than the plasma frequency; this substantially modifies the power spectrum for either mode by increasing the Alfven velocity. An estimation of the radiation impedance of bare tethers is also considered. iib) In LJO, a spacecraft orbiting in a slow downward spiral under the radiation belts would allow determining magnetic field structure and atmospheric composition for understanding the formation, evolution, and structure of Jupiter. Additionally, if the cathodic contactor is switched off, a tether floats electrically, allowing e-beam emission that generate auroras. On/off switching produces bias/current pulses and signal emission, which might be used for Jovian plasma diagnostics. In LJO, the ions impacting against the negative-biased tether do produce secondary electrons, which racing down Jupiter’s magnetic field lines, reach the upper atmosphere. The energetic electrons there generate auroral effects. Regions where the tether efficiently should produce secondary electrons are analyzed. iic) Other scientific application suggested in LJO is the in-situ detection of charged grains. Charged grains naturally orbit near Jupiter. High-energy electrons in the Jovian ambient may be modeled by the kappa distribution function. In complex plasma scenarios, where the Jovian high electric field may accelerate charges up superthermal velocities, the use of non-Maxwellian distributions should be considered. In these cases, the distribution tails fit well to a power-law dependence for electrons. Fluctuations of the charged grains for non-Mawellian distribution function are here studied. iii) The present thesis is concluded with the analysis for de-orbiting satellites at end of mission by electrodynamic tethers. A de-orbit tether system must present very small tether-to-satellite mass ratio and small probability of a tether cut by small debris too. The present work shows how to select tape dimensions so as to minimize the product of those two magnitudes. Preliminary results of tape-tether design are here discussed to minimize that function. Results for de-orbiting Cryosat and Envisat are also presented.

Electronic structure of copper nitrides as a function of nitrogen content

he nitrogen content dependence of the electronic properties for copper nitride thin films with an atomic percentage of nitrogen ranging from 26 ± 2 to 33 ± 2 have been studied by means of optical (spectroscopic ellipsometry), thermoelectric (Seebeck), and electrical resistivity measurements. The optical spectra are consistent with direct optical transitions corresponding to the stoichiometric semiconductor Cu3N plus a free-carrier contribution, essentially independent of temperature, which can be tuned in accordance with the N-excess. Deviation of the N content from stoichiometry drives to significant decreases from − 5 to − 50 μV/K in the Seebeck coefficient and to large enhancements, from 10− 3 up to 10 Ω cm, in the electrical resistivity. Band structure and density of states calculations have been carried out on the basis of the density functional theory to account for the experimental results.

Dynamic analysis of payloads and structures with intermediate modal density

La necesidad de desarrollar técnicas para predecir la respuesta vibroacústica de estructuras espaciales lia ido ganando importancia en los últimos años. Las técnicas numéricas existentes en la actualidad son capaces de predecir de forma fiable el comportamiento vibroacústico de sistemas con altas o bajas densidades modales. Sin embargo, ambos rangos no siempre solapan lo que hace que sea necesario el desarrollo de métodos específicos para este rango, conocido como densidad modal media. Es en este rango, conocido también como media frecuencia, donde se centra la presente Tesis doctoral, debido a la carencia de métodos específicos para el cálculo de la respuesta vibroacústica. Para las estructuras estudiadas en este trabajo, los mencionados rangos de baja y alta densidad modal se corresponden, en general, con los rangos de baja y alta frecuencia, respectivamente. Los métodos numéricos que permiten obtener la respuesta vibroacústica para estos rangos de frecuencia están bien especificados. Para el rango de baja frecuencia se emplean técnicas deterministas, como el método de los Elementos Finitos, mientras que, para el rango de alta frecuencia las técnicas estadísticas son más utilizadas, como el Análisis Estadístico de la Energía. En el rango de medias frecuencias ninguno de estos métodos numéricos puede ser usado con suficiente precisión y, como consecuencia -a falta de propuestas más específicas- se han desarrollado métodos híbridos que combinan el uso de métodos de baja y alta frecuencia, intentando que cada uno supla las deficiencias del otro en este rango medio. Este trabajo propone dos soluciones diferentes para resolver el problema de la media frecuencia. El primero de ellos, denominado SHFL (del inglés Subsystem based High Frequency Limit procedure), propone un procedimiento multihíbrido en el cuál cada subestructura del sistema completo se modela empleando una técnica numérica diferente, dependiendo del rango de frecuencias de estudio. Con este propósito se introduce el concepto de límite de alta frecuencia de una subestructura, que marca el límite a partir del cual dicha subestructura tiene una densidad modal lo suficientemente alta como para ser modelada utilizando Análisis Estadístico de la Energía. Si la frecuencia de análisis es menor que el límite de alta frecuencia de la subestructura, ésta se modela utilizando Elementos Finitos. Mediante este método, el rango de media frecuencia se puede definir de una forma precisa, estando comprendido entre el menor y el mayor de los límites de alta frecuencia de las subestructuras que componen el sistema completo. Los resultados obtenidos mediante la aplicación de este método evidencian una mejora en la continuidad de la respuesta vibroacústica, mostrando una transición suave entre los rangos de baja y alta frecuencia. El segundo método propuesto se denomina HS-CMS (del inglés Hybrid Substructuring method based on Component Mode Synthesis). Este método se basa en la clasificación de la base modal de las subestructuras en conjuntos de modos globales (que afectan a todo o a varias partes del sistema) o locales (que afectan a una única subestructura), utilizando un método de Síntesis Modal de Componentes. De este modo es posible situar espacialmente los modos del sistema completo y estudiar el comportamiento del mismo desde el punto de vista de las subestructuras. De nuevo se emplea el concepto de límite de alta frecuencia de una subestructura para realizar la clasificación global/local de los modos en la misma. Mediante dicha clasificación se derivan las ecuaciones globales del movimiento, gobernadas por los modos globales, y en las que la influencia del conjunto de modos locales se introduce mediante modificaciones en las mismas (en su matriz dinámica de rigidez y en el vector de fuerzas). Las ecuaciones locales se resuelven empleando Análisis Estadístico de Energías. Sin embargo, este último será un modelo híbrido, en el cual se introduce la potencia adicional aportada por la presencia de los modos globales. El método ha sido probado para el cálculo de la respuesta de estructuras sometidas tanto a cargas estructurales como acústicas. Ambos métodos han sido probados inicialmente en estructuras sencillas para establecer las bases e hipótesis de aplicación. Posteriormente, se han aplicado a estructuras espaciales, como satélites y reflectores de antenas, mostrando buenos resultados, como se concluye de la comparación de las simulaciones y los datos experimentales medidos en ensayos, tanto estructurales como acústicos. Este trabajo abre un amplio campo de investigación a partir del cual es posible obtener metodologías precisas y eficientes para reproducir el comportamiento vibroacústico de sistemas en el rango de la media frecuencia. ABSTRACT Over the last years an increasing need of novel prediction techniques for vibroacoustic analysis of space structures has arisen. Current numerical techniques arc able to predict with enough accuracy the vibro-acoustic behaviour of systems with low and high modal densities. However, space structures are, in general, very complex and they present a range of frequencies in which a mixed behaviour exist. In such cases, the full system is composed of some sub-structures which has low modal density, while others present high modal density. This frequency range is known as the mid-frequency range and to develop methods for accurately describe the vibro-acoustic response in this frequency range is the scope of this dissertation. For the structures under study, the aforementioned low and high modal densities correspond with the low and high frequency ranges, respectively. For the low frequency range, deterministic techniques as the Finite Element Method (FEM) are used while, for the high frequency range statistical techniques, as the Statistical Energy Analysis (SEA), arc considered as more appropriate. In the mid-frequency range, where a mixed vibro-acoustic behaviour is expected, any of these numerical method can not be used with enough confidence level. As a consequence, it is usual to obtain an undetermined gap between low and high frequencies in the vibro-acoustic response function. This dissertation proposes two different solutions to the mid-frequency range problem. The first one, named as The Subsystem based High Frequency Limit (SHFL) procedure, proposes a multi-hybrid procedure in which each sub-structure of the full system is modelled with the appropriate modelling technique, depending on the frequency of study. With this purpose, the concept of high frequency limit of a sub-structure is introduced, marking out the limit above which a substructure has enough modal density to be modelled by SEA. For a certain analysis frequency, if it is lower than the high frequency limit of the sub-structure, the sub-structure is modelled through FEM and, if the frequency of analysis is higher than the high frequency limit, the sub-structure is modelled by SEA. The procedure leads to a number of hybrid models required to cover the medium frequency range, which is defined as the frequency range between the lowest substructure high frequency limit and the highest one. Using this procedure, the mid-frequency range can be define specifically so that, as a consequence, an improvement in the continuity of the vibro-acoustic response function is achieved, closing the undetermined gap between the low and high frequency ranges. The second proposed mid-frequency solution is the Hybrid Sub-structuring method based on Component Mode Synthesis (HS-CMS). The method adopts a partition scheme based on classifying the system modal basis into global and local sets of modes. This classification is performed by using a Component Mode Synthesis, in particular a Craig-Bampton transformation, in order to express the system modal base into the modal bases associated with each sub-structure. Then, each sub-structure modal base is classified into global and local set, fist ones associated with the long wavelength motion and second ones with the short wavelength motion. The high frequency limit of each sub-structure is used as frequency frontier between both sets of modes. From this classification, the equations of motion associated with global modes are derived, which include the interaction of local modes by means of corrections in the dynamic stiffness matrix and the force vector of the global problem. The local equations of motion are solved through SEA, where again interactions with global modes arc included through the inclusion of an additional input power into the SEA model. The method has been tested for the calculation of the response function of structures subjected to structural and acoustic loads. Both methods have been firstly tested in simple structures to establish their basis and main characteristics. Methods are also verified in space structures, as satellites and antenna reflectors, providing good results as it is concluded from the comparison with experimental results obtained in both, acoustic and structural load tests. This dissertation opens a wide field of research through which further studies could be performed to obtain efficient and accurate methodologies to appropriately reproduce the vibro-acoustic behaviour of complex systems in the mid-frequency range.