Search

Summary: Generalised linear mixed regression models are fundamental in statistics. Modelling random effects that are shared by individuals allows for correlation among those individuals. There are many methods and statistical packages available for analysing data using these models. Most require some form of numerical or analytic approximation because the likelihood function generally involves intractable integrals over the latents. The Bayesian approach avoids this issue by iteratively sampling the full conditional distributions for various blocks of parameters and latent random effects. Depending on the choice of the prior, some full conditionals are recognisable while others are not. In this paper we develop a novel normal approximation for the random effects full conditional, establish its asymptotic correctness and evaluate how well it performs. We make the case for hierarchical binomial and Poisson regression models with canonical link functions, for hierarchical gamma regression models with log link and for other cases. We also develop what we term a sufficient reduction (SR) approach to the Markov Chain Monte Carlo algorithm that allows for making inferences about all model parameters by replacing the full conditional for the latent variables with a considerably reduced dimensional function of the latents. We expect that this approximation could be quite useful in situations where there are a very large number of latent effects, which may be occurring in an increasingly 'Big Data' world. In the sequel, we compare our methods with INLA, which is a particularly popular method and which has been shown to be excellent in terms of speed and accuracy across a variety of settings. Our methods appear to be comparable to theirs in terms of accuracy, while INLA was faster, for the settings we considered. In addition, we note that our methods and those of others that involve Gibbs sampling trivially handle parameters that are functions of multiple parameters, while INLA approximations do not. Our primary illustration is for a three‐level hierarchical binomial regression model for data on health outcomes for patients who are clustered within physicians who are clustered within particular hospitals or hospital systems. [ABSTRACT FROM AUTHOR]

Abstract Background Meta-analyses are used to summarise the results of several studies on a specific research question. Standard methods for meta-analyses, namely inverse variance random effects models, have unfavourable properties if only very few (2 – 4) studies are available. Therefore, alternative meta-analytic methods are needed. In the case of binary data, the “common-rho” beta-binomial model has shown good results in situations with sparse data or few studies. The major concern of this model is that it ignores the fact that each treatment arm is paired with a respective control arm from the same study. Thus, the randomisation to a study arm of a specific study is disrespected, which may lead to compromised estimates of the treatment effect. Therefore, we extended this model to a version that respects randomisation. The aim of this simulation study was to compare the “common-rho” beta-binomial model and several other beta-binomial models with standard meta-analyses models, including generalised linear mixed models and several inverse variance random effects models. Methods We conducted a simulation study comparing beta-binomial models and various standard meta-analysis methods. The design of the simulation aimed to consider meta-analytic situations occurring in practice. Results No method performed well in scenarios with only 2 studies in the random effects scenario. In this situation, a fixed effect model or a qualitative summary of the study results may be preferable. In scenarios with 3 or 4 studies, most methods satisfied the nominal coverage probability. The “common-rho” beta-binomial model showed the highest power under the alternative hypothesis. The beta-binomial model respecting randomisation did not improve performance. Conclusion The “common-rho” beta-binomial appears to be a good option for meta-analyses of very few studies. As residual concerns about the consequences of disrespecting randomisation may still exist, we recommend a sensitivity analysis with a standard meta-analysis method that respects randomisation.

A multilevel model for nominal data in the framework of generalized linear mixed models (GLMM) is developed to account for the inherent dependencies among observations within clusters. Motivated by a dataset from the Flash EuroBarometer survey (FEBS), the random region and country effects are incorporated into the linear predictor of a GLMM to accommodate the nested clusterings. The fixed effects are estimated by maximizing the penalized likelihood function, whereas the random variance component parameters are predicted via the restricted maximum likelihood (REML) estimation method. The model is employed to analyse the FEBS data. A Monte Carlo simulation study is conducted to evaluate the performance of estimators. [ABSTRACT FROM AUTHOR]

Background: Meta-analyses are used to summarise the results of several studies on a specific research question. Standard methods for meta-analyses, namely inverse variance random effects models, have unfavourable properties if only very few (2 – 4) studies are available. Therefore, alternative meta-analytic methods are needed. In the case of binary data, the "common-rho" beta-binomial model has shown good results in situations with sparse data or few studies. The major concern of this model is that it ignores the fact that each treatment arm is paired with a respective control arm from the same study. Thus, the randomisation to a study arm of a specific study is disrespected, which may lead to compromised estimates of the treatment effect. Therefore, we extended this model to a version that respects randomisation. The aim of this simulation study was to compare the "common-rho" beta-binomial model and several other beta-binomial models with standard meta-analyses models, including generalised linear mixed models and several inverse variance random effects models. Methods: We conducted a simulation study comparing beta-binomial models and various standard meta-analysis methods. The design of the simulation aimed to consider meta-analytic situations occurring in practice. Results: No method performed well in scenarios with only 2 studies in the random effects scenario. In this situation, a fixed effect model or a qualitative summary of the study results may be preferable. In scenarios with 3 or 4 studies, most methods satisfied the nominal coverage probability. The "common-rho" beta-binomial model showed the highest power under the alternative hypothesis. The beta-binomial model respecting randomisation did not improve performance. Conclusion: The "common-rho" beta-binomial appears to be a good option for meta-analyses of very few studies. As residual concerns about the consequences of disrespecting randomisation may still exist, we recommend a sensitivity analysis with a standard meta-analysis method that respects randomisation. [ABSTRACT FROM AUTHOR]

Aim: Ground lichens have declined in boreal and temperate Europe. The potential causes of the decline were explored by analysing the response of lichen cover to forest management, site and stand variables as well as co-existing plants in different boreal vegetation subzones. Location: Finland. Methods: Understorey vegetation was surveyed on a systematic network of 1721 sample plots in forests on mineral soil. The response of macrolichen cover to explanatory variables was analysed by generalised linear mixed models (GLMMs). Results: Ground lichens favoured old Pinus sylvestris forests on xeric sites with sufficient light conditions. Intensive forest management regimes, such as regeneration cutting and soil ploughing decreased lichen cover, while lighter scarification methods had less effect on lichens. Lichens benefitted from intermediate cutting on sites with low bryophyte cover. Lichens responded similarly to site and stand variables in all boreal subzones, showing that despite heavy reindeer grazing these response patterns apply also in the northern boreal subzone. Lichens showed both negative (Vaccinium myrtillus, Empetrum nigrum and Calluna vulgaris) and positive (Vaccinium vitis-idaea) responses to increasing dwarf shrub cover. Generally, lichens negatively responded to increasing bryophyte cover, indicating either that bryophytes benefit from a decline of lichens or that there is real competition between these groups. The negative relationship between bryophytes and lichens strengthened as the tree canopy becomes denser. Conclusions: Our study provides new quantitative insights into the effects of forest management and changes in forest structure as underlying factors for lichen decline both inside and outside the reindeer-herding area (i.e., the northern boreal subzone). These factors include increased canopy cover with increased shading as well as disturbance caused by regeneration cutting and soil preparation. The possible effects of the long-term legacy of nitrogen deposition, lack of forest fires and climate change are also discussed. [ABSTRACT FROM AUTHOR]

This article describes a retrospective research evaluation of data generated within the international music therapy trial TIME-A. In this retrospective evaluation, the main objective was to investigate the effects of individual music therapy sessions on resilience in young children with autism spectrum disorder (ASD). Video excerpts of individual music therapy sessions over a period of five months with 13 children between the ages of four and six with ASD were analysed. Six children received weekly music therapy sessions and seven children received music therapy sessions three times a week. An improvisational music therapy model was used. Short sections of music therapy session videos were selected for microanalysis. A five-second time sampling method was used. For every five-second interval, observed behaviour indicative of resilience was documented using video annotation software. Generalised Linear Mixed Models (GLMM) were chosen for the statistical analysis. Out of the 11 response variables, seven yielded a significant p-value in the full-null model comparison. Intrapersonal protective resilience factors, such as the ability to express emotions, awareness of others and goal-directed behaviour were strengthened by music therapy. In addition, treatment intensity had a significant effect on one variable and verbal ability a significant effect on four variables. Results suggest that music therapy enhances the likelihood of resilience in young children with ASD. The importance of increasing resilience for children with ASD rather than reducing symptoms is emphasised. GLMM models of statistical analysis are recommended for future music therapy research. [ABSTRACT FROM AUTHOR]

The paper outlines several approaches for dealing with meta-analyses of count outcome data. These counts are the accumulation of occurred events, and these events might be rare, so a special feature of the meta-analysis is dealing with low counts including zero-count studies. Emphasis is put on approaches which are state of the art for count data modelling including mixed log-linear (Poisson) and mixed logistic (binomial) regression as well as nonparametric mixture models for count data of Poisson and binomial type. A simulation study investigates the performance and capability of discrete mixture models in estimating effect heterogeneity. The approaches are exemplified on a meta-analytic case study investigating the acceptance of bibliotherapy. [ABSTRACT FROM AUTHOR]

Mixed models are a useful way of analysing longitudinal data. Random effects terms allow modelling of patient specific deviations from the overall trend over time. Correlation between repeated measurements are captured by specifying a joint distribution for all random effects in a model. Typically, this joint distribution is assumed to be a multivariate normal distribution. For Gaussian outcomes misspecification of the random effects distribution usually has little impact. However, when the outcome is discrete (e.g. counts or binary outcomes) generalised linear mixed models (GLMMs) are used to analyse longitudinal trends. Opinion is divided about how robust GLMMs are to misspecification of the random effects. Previous work explored the impact of random effects misspecification on the bias of model parameters in single outcome GLMMs. Accepting that these model parameters may be biased, we investigate whether this affects our ability to classify patients into clinical groups using a longitudinal discriminant analysis. We also consider multiple outcomes, which can significantly increase the dimensions of the random effects distribution when modelled simultaneously. We show that when there is severe departure from normality, more flexible mixture distributions can give better classification accuracy. However, in many cases, wrongly assuming a single multivariate normal distribution has little impact on classification accuracy. [ABSTRACT FROM AUTHOR]

The Generalised linear mixed model (GLMM) is widely used for modelling environmental data. However, such data are prone to influential observations, which can distort the estimated exposure-response curve particularly in regions of high exposure. Deletion diagnostics for iterative estimation schemes commonly derive the deleted estimates based on a single iteration of the full system holding certain pivotal quantities such as the information matrix to be constant. In this paper, we present an approximate formula for the deleted estimates and Cook's distance for the GLMM, which does not assume that the estimates of variance parameters are unaffected by deletion. The procedure allows the user to calculate standardised DFBETAs for mean as well as variance parameters. In certain cases such as when using the GLMM as a device for smoothing, such residuals for the variance parameters are interesting in their own right. In general, the procedure leads to deleted estimates of mean parameters, which are corrected for the effect of deletion on variance components as estimation of the two sets of parameters is interdependent. The probabilistic behaviour of these residuals is investigated and a simulation based procedure suggested for their standardisation. The method is used to identify influential individuals in an occupational cohort exposed to silica. The results show that failure to conduct post model fitting diagnostics for variance components can lead to erroneous conclusions about the fitted curve and unstable confidence intervals.

Terminal tackle regulations can be a valuable tool for fisheries management, especially in multispecies fisheries where bycatch and discards are common issues. In the Gulf of Maine, recreational anglers frequently discard critically depleted Atlantic cod Gadus morhua L. as bycatch when targeting the abundant haddock Melanogrammus aeglefinus (L.) stock. The present study investigated species catch composition, catch rates and animal welfare across various terminal tackle setups, and aspects of capture and handling with the goal of reducing bycatch and discard mortality. Overall, 2558 cod and 4266 haddock were captured with six terminal tackle setups in the western Gulf of Maine from April to October 2018. Along with angler experience and capture depth, lure type primarily influenced species catch composition and catch‐per‐unit‐effort; hook types additionally influenced hook‐removal times and physical injuries to fish. Results indicate that using baited hooks can both promote haddock catch and reduce cod bycatch, with specific hook types promoting increased survival. [ABSTRACT FROM AUTHOR]

• Methodological proposal to study the relation between acoustic and visual correlates of prominence in prominence perception. • Spontaneous speech samples are used from a television talent show to overcome limitations of previous studies, characterized by using animated agents and gestures elicited in controlled experimental settings. • Estimation of generalised linear mixed models (GLMMs) for the multimodal perception of prominence, and model selection by means of Akaike Information Criterion (AIC). Research on prominence perception has made use of animated agents and controlled speech in experimental settings, but these methodologies have disregarded some aspects of the acoustic and visual correlates of prominence. To overcome these limitations we propose a new methodological approach using spontaneous speech data. For this, we created a small database with extracts from a television talent show and neutralised the prominence-lending properties of the acoustic cues of prominence in the speech signal. In our pilot study twelve naïve listeners marked words for binary prominence (prominent vs. non-prominent) in two modalities, i.e. audio-only and audiovisual, under three conditions involving neutralisation of (a) fundamental frequency, (b) intensity, and (c) both fundamental frequency and intensity. Additionally, the marks of two trained listeners served as control condition. Different generalised linear mixed models were estimated and compared using the Akaike Information Criterion (AIC). The most parsimonious model was then examined using traditional null-hypothesis testing in order to provisionally establish the effects of our independent variables on prominence marking. We argue that spontaneous speech can be successfully applied to the study of the multimodal perception of prominence. [ABSTRACT FROM AUTHOR]

Fair educational assessment in linguistically diverse contexts poses new challenges which call for the need to evaluate the impact of context-related language factors on student performance. Based on data from the Basque Autonomous Community in Spain, this research analyses the effect of different factors on a mathematics achievement test. Using generalised linear mixed models, the results explain differences in achievement among student groups based on home language, educational linguistic model, and test language through sociolinguistic status variables and correspondence between home language and test language. The paper highlights the importance of language-related factors and points out that achievement test scores in linguistically diverse contexts should be interpreted with caution to avoid unfair comparisons. [ABSTRACT FROM AUTHOR]

Summary: In longitudinal data where the timing and frequency of the measurement of outcomes may be associated with the value of the outcome, significant bias can occur. Previous results depended on correct specification of the outcome process and a somewhat unrealistic visit process model. In practice, this will never exactly be the case, so it is important to understand to what degree the results hold when those assumptions are violated in order to guide practical use of the methods. This paper presents theory and the results of simulation studies to extend our previous work to more realistic visit process models, as well as Poisson outcomes. We also assess the effects of several types of model misspecification. The estimated bias in these new settings generally mirrors the theoretical and simulation results of our previous work and provides confidence in using maximum likelihood methods in practice. Even when the assumptions about the outcome process did not hold, mixed effects models fit by maximum likelihood produced at most small bias in estimated regression coefficients, illustrating the robustness of these methods. This contrasts with generalised estimating equations approaches where bias increased in the settings of this paper. The analysis of data from a study of change in neurological outcomes following microsurgery for a brain arteriovenous malformation further illustrate the results. Likelihood‐based methods perform well when the timing and frequency of the measurement of longitudinal outcomes depend on their values. [ABSTRACT FROM AUTHOR]

• Emotional data were consistent across sessions for both CATA and IRT Go/No-go. • Emotion CATA selection frequency was generally mirrored in IRT Go/No-go. • Some CATA responses, with no time limit to respond, differed from IRT Go/No-go. • Go/No-go response time measures provided insights on redundant emotion terms. • Product discrimination varied slightly across methods, and differed by data format. Adapting implicit approaches to capture consumer responses when tasting foods is of recent interest in sensory and consumer science. Implicit consumer responses are reported to be more spontaneous than explicitly gathered data. Traditionally, emotional response to foods is captured using explicit methods like check-all-that-apply (CATA). The present study aimed to compare an implicit response time (IRT) Go/No-go approach with traditional CATA to profile consumer emotional responses. Participants (n = 104) were consumers of, or willing to consume, cow's milk and plant-based milk alternatives (PBMA). Emotional responses for two cow milk and five PBMA products were evaluated across two sessions using IRT and CATA. The cow milk products were replicated across the sessions to allow consistency of response across sessions to be evaluated for each method. Data were collected using a bespoke single page web application (JavaScript, ECMA 2015). Data consistency across sessions (Spearman correlation (ρ)), emotion selection frequency (ρ and generalised linear models) and product discrimination (linear mixed models and correspondence analysis) were compared across the IRT and CATA approaches. Results showed high data consistency from both IRT and CATA across the two sessions (ρ > 0.89). Emotion selection frequency was also comparable across IRT and CATA. Interestingly, CATA was differentiating more between cow's milk products and IRT within the PBMA space. However, further investigations showed that fewer participants provided different responses in CATA than when under time pressure in the IRT Go/No-go. Additional investigations on the performance of explicit versus implicit methods, or their combination, are required across different product matrices to identify the optimum approach to capture consumer product experience. [ABSTRACT FROM AUTHOR]

Summary: Accelerometry is a low‐cost and noninvasive method that has been used to discriminate sleep from wake, however, its utility to detect sleep stages is unclear. We detail the development and comparison of methods which utilise raw, triaxial accelerometry data to classify varying stages of sleep, ranging from sleep/wake detection to discriminating rapid eye movement sleep, stage one sleep, stage two sleep, deep sleep and wake. First‐ and second‐order hidden Markov models (HMMs) with time‐homogeneous and time‐varying transition probability matrices, along with continuous acceleration observations in the form of a Gaussian‐observation HMM and K‐means classified acceleration in a discrete‐observation HMM were explored. In addition, generalised linear mixed models (GLMMs) with binary and multinomial responses and logit link functions were considered as was whether incorporating adjoining acceleration information into the models improved prediction. Model predictions were compared to the reference‐standard in sleep detection (polysomnography) and outcome accuracies were calculated. Consistently, HMMs yielded greater sleep stage detection than GLMMs but there was little difference between first‐ and second‐order HMMs. Varying degrees of difference were observed when comparing Gaussian‐observation HMMs to discrete‐observation HMMs, and time‐varying HMMs yielded greater discrimination than time‐homogeneous HMMs, as did models which considered adjoining acceleration information. These results suggest that wrist‐worn accelerometry data may be able to detect sleep stages but that further investigation is required to optimise classification accuracy. Development and comparison of hidden Markov models and generalized linear mixed models which use raw, triaxial wrist accelerometry data to determine sleep stages. [ABSTRACT FROM AUTHOR]

We present a novel method for the estimation of variance parameters in generalised linear mixed models. The method has its roots in Harville (J Am Stat Assoc 72(358):320–338, 1977)'s work, but it is able to deal with models that have a precision matrix for the random effect vector that is linear in the inverse of the variance parameters (i.e., the precision parameters). We call the method SOP (separation of overlapping precision matrices). SOP is based on applying the method of successive approximations to easy-to-compute estimate updates of the variance parameters. These estimate updates have an appealing form: they are the ratio of a (weighted) sum of squares to a quantity related to effective degrees of freedom. We provide the sufficient and necessary conditions for these estimates to be strictly positive. An important application field of SOP is penalised regression estimation of models where multiple quadratic penalties act on the same regression coefficients. We discuss in detail two of those models: penalised splines for locally adaptive smoothness and for hierarchical curve data. Several data examples in these settings are presented. [ABSTRACT FROM AUTHOR]

Despite the pivotal role of payroll staff within political parties' central offices, research on the staff expenses of parties remains scarce. In this article, we study the relative staff expenses of political parties, that is staff costs as a percentage of total annual expenses. We analyse which factors explain the differences between parties' relative staff expenses, based on a dataset of 590 individual observations representing 52 parties from seven European countries. The multivariate model shows that relative staff expenses are higher among left-oriented parties and that they increase with party age, party membership figures and the number of years a party has been in government, while they decrease with party income. Relative staff expenses also decrease with the effective number of parties in the party system, and they are lower in election years. [ABSTRACT FROM AUTHOR]

Summary: We demonstrate the use of our R package, gammSlice, for Bayesian fitting and inference in generalised additive mixed model analysis. This class of models includes generalised linear mixed models and generalised additive models as special cases. Accurate Bayesian inference is achievable via sufficiently large Markov chain Monte Carlo (MCMC) samples. Slice sampling is a key component of the MCMC scheme. Comparisons with existing generalised additive mixed model software shows that gammSlice offers improved inferential accuracy, albeit at the cost of longer computational time. [ABSTRACT FROM AUTHOR]

Abstract: Calcareous grassland is one of the most species rich and diverse habitats within Europe, but has faced decline due to agricultural intensification and abandonment. In recent years, conservation organisations have altered grazing practices in an attempt to maintain floristic components. However, there has been little consideration of the effects of changes in grazing practices on invertebrates. This study determines the impacts of commonly used grazing practices in upland calcareous grasslands on spiders in relation to vegetation structural complexity. Typical grazing management regimes were examined in three regions of upland calcareous grassland in Britain. Spiders were sampled from April to August 2014 and vegetation structural complexity was recorded in 2 × 2 m quadrats paired with pitfall traps sequentially throughout the sample period. There were three distinct spider assemblages among the grazing regimes; ungrazed, heavy sheep grazed and one shared between cattle and light sheep (which had a comparable grazing intensity). The distinct spider assemblages among grazing regimes can be attributed to the interaction of grazing and habitat structure. Increased vegetation structural complexity in ungrazed regimes resulted in an assemblage dominated by ‘sheet web weavers’ (dominated by Linyphiidae). In contrast, reduced vegetation structural complexity and homogeneity in heavy sheep grazing resulted in an assemblage dominated by ‘other hunters’ (including Oedothorax and Erigone genera). Grazing regime alters vegetation structural complexity and is important in supporting distinct spider assemblages. This research indicates that low intensity conservation grazing regimes, in addition to no grazing, should be promoted across upland calcareous grassland landscapes to maintain heterogeneity. [ABSTRACT FROM AUTHOR]

To gain regulatory approval, a new medicine must demonstrate that its benefits outweigh any potential risks, ie, that the benefit-risk balance is favourable towards the new medicine. For transparency and clarity of the decision, a structured and consistent approach to benefit-risk assessment that quantifies uncertainties and accounts for underlying dependencies is desirable. This paper proposes two approaches to benefit-risk evaluation, both based on the idea of joint modelling of mixed outcomes that are potentially dependent at the subject level. Using Bayesian inference, the two approaches offer interpretability and efficiency to enhance qualitative frameworks. Simulation studies show that accounting for correlation leads to a more accurate assessment of the strength of evidence to support benefit-risk profiles of interest. Several graphical approaches are proposed that can be used to communicate the benefit-risk balance to project teams. Finally, the two approaches are illustrated in a case study using real clinical trial data. [ABSTRACT FROM AUTHOR]

Objectives. The effects of a range of types of social integration and patterns of change in social integration over time were examined directly in relation to depressive symptoms in a large sample of the Korean population aged ≥45 years. Methods. Data from the Korean Longitudinal Study of Aging (KLoSA) between 2006 and 2012 were assessed using longitudinal data analysis. We included 10 242 research subjects at baseline (2006) and based the primary analysis on generalised linear mixed models to examine association between social integration and depressive symptom. Results. The odds ratio (OR) for depressive symptoms in individuals at the lowest level of social integration was 1.539-fold higher (95% confidence interval (CI) 1.360-1.742) that that for those at highest level of social integration. Results of subgroup analysis according to gender revealed a similar trend. A five-class linear solution fit the data best; Class 1 (lowest constant social integration level, 10.5% of the sample) was significantly associated with the highest risk of depressive symptoms (OR 1.933, 95% CI 1.706-2.190). Conclusions. The results of the present study provide a scientific basis for the specific association between the level of social integration and changes in social integration pattern with the risk of depressive symptoms in current practice. Therefore, interventions to provide emotional support for older adults via social integration may be important to protect against depressive symptoms. [ABSTRACT FROM AUTHOR]

Forests on tribal lands in the western United States have seen the return of low-intensity surface fires for several decades longer than forests on non-tribal lands. We examined the surface fire regime in a ponderosa pine-dominated (Pinus ponderosa) forest on the Hualapai tribal lands in the south-western United States. Using fire-scarred trees, we inferred temporal (frequency and seasonality) and spatial (synchrony) attributes and regulators of the fire regime over three land-use periods (historical, suppression, modern) between 1702 and 2007. Patterns of fire frequency and synchrony were similar, but fire seasonality was dissimilar, between the historical and modern periods. Logistic regression and generalised linear mixed models identified a suite of variables representing fuels, climate and human land uses that were associated with the probability of a site burning. Combined, these results allow for valuable insights regarding past fire spread and variability in fire frequency throughout our study area. In some respects, the current distinct fire regime in our study area, which predominately consists of prescribed fires implemented since the 1960s, resembles the past frequent surface fire regime that occurred here and in similar forest types on non-tribal lands in the south-western United States. Our results will be useful for informing adaptive management throughout the region as climate warms. We studied the surface fire regime in a forest on tribal lands where prescribed burning has occurred since the 1960s. Fire frequency and synchrony were similar, but fire seasonality was dissimilar, between the historical and modern land-use periods. Fuels, climate and human land uses regulated the fire regime over time. [ABSTRACT FROM AUTHOR]

Cities are unique ecosystems emerging and growing worldwide due to ongoing urbanising trends. The urban-rural gradient is an excellent setting to evaluate the effect of urbanisation on the distribution of species, a matter of public health concern in the case of disease vectors. Despite this, such distributions are affected by other co-occurring variables, mainly meteorological, that may be confounded by the urbanisation gradient due to the urban heat island effect and maritime climatological conditions in the case of coastal cities. To aid in the design of ecological studies within the urban-rural transition zone, a mapping protocol was designed and applied to Buenos Aires City and its surroundings. Based on road density and district-level population counts, a detailed (1 km2 pixel) urbanisation map was obtained which, combined with a temperature map, rendered a final urbanisation × temperature product with six classes. The resulting zonation was tested by modelling the distribution of the vector mosquitoes Aedes aegypti and Culex pipiens in artificial containers. The selected model explained the occurrence of mosquitoes 59% better than chance as a function of the urbanisation × temperature categories and the natural lighting condition of the container. This novel zonation approach allows partitioning of environmental heterogeneity prior to the selection of study sites to avoid confounding gradients and provides multiple advantages, such as making comparisons across cities easier, extrapolating the results of site-scale experiments and identifying priority areas for control measures. [ABSTRACT FROM AUTHOR]

Background:We hypothesised that the relative position of surrounding vegetation patches would account for most floristic patterns observed in patches of exposed gold mine spoil (EMS) scattered over a 60-year-old, abandoned waste heap. Aims:We assessed the extent to which the structure of plant species assemblages from 53 EMS patches was related to distance from woodlands, heathlands and grasslands by controlling for the effect of spatial autocorrelation, patch size, slope and aspect. Results:The most discrepant in terms of species composition were the EMS patches located farthest away from heathlands and those embedded in woodlands. Zoochorous species richness declined with distance from the nearest heathland. Anemochorous, competitive-stress tolerant and total species richness increased with proximity to the nearest woodland. Conversely, competitive species richness and total species cover were larger in EMS patches that were farther away from the nearest woodland, but closer to grasslands. Depending on the identity of colonising species, the woodland stands could act as propagule source, ecological filters or seed dispersal amplifiers. Conclusion:The relative importance of proximity to source habitats for plant community assembly can change dramatically under the confounding or contrasting effects of ‘third-party’ habitat patches, which are interposed in the landscape. [ABSTRACT FROM AUTHOR]

Abstract: In Vitro Fertilization (IVF) units need to decrease multiple pregnancies without affecting their overall success rate. In this study we propose a mathematical model to evaluate an embryo’s potential ability to implant in the uterus. Embryos are graded by the embryologist based on the number of blastomeres, evenness of growth and degree of fragmentation. Therefore, the following variables were considered: number of blastomeres produced by division of the egg after fertilisation (blastomeres), symmetry and fragmentation of the embryo (grade). This model evaluates the embryos assigning them a score which represents their quality. The main result derived from this model is the estimation of the significant improvement in the implantation rate due to the increase in blastomere values and the decrease in grade factor values. But the increase from two–three to four produces more improvement in the implantation rate than two–three to five–six blastomeres. First, statistical models were used to study embryo traceability from transfer to implantation and to evaluate the effect of the quality of the embryos (embryo score) and women’s age on implantation potential. This score was obtained by making predictions from the fitted model which was used to rank embryos in terms of implantation potential. Then we totalled the scores of embryos that had been transferred to each woman for obtaining the Embryo Quality Index (EQI). In addition, we studied the effects of EQI and women’s age on pregnancy. Finally, statistical techniques such as Receiver Operating Characteristics (ROC) and bootstrap procedures were used to assess the accuracy of this model. This embryo score is a quick, efficient and accurate tool to optimise embryo selection for transfers on the second day after fertilisation. This tool is especially useful for transfers involving non-top embryos. [Copyright &y& Elsevier]

Abstract: In the Alps, many forests are under abandonment because of direct protection or the unprofitability of timber production. The objective of this study was to analyse the effect of silviculture abandonment on the species richness and composition of epiphytic and dead wood-dwelling lichens in silver fir woodlands. An abandoned and a currently non-intensively managed forest watershed of the eastern Italian Alps that were matched for topography, stand structure, plant community and history of use were selected and randomly sampled for the lichen diversity of their silver fir mature stands. Generalised linear mixed models were used for the analyses of the species richness, whereas the species composition patterns were explored by redundancy analysis. The diversity of both of the lichen guilds were explained by the microhabitat conditions related to the contrasting management regimes, but they had opposite responses to the abandonment of the silvicultural activities. The epiphytic lichen communities were richer in species in the managed than in the abandoned silver fir woodlands. However, the effect of a 50year management cessation on the epiphytic lichens may change in the long term when natural disturbances create gaps and increase the attributes of old-growth forests. The dead wood-dwelling lichen species richness was higher in the abandoned forest because an increased availability of types of dead wood is usually rare in managed forests. The dead wood-dwelling lichen communities of the managed forests were a subset of a more diverse community that was developing in the abandoned forest. Our results indicated that, in non-intensively managed silver fir mature stands, lichen diversity may benefit from the retention of logs and snags, which provide habitat features typical of old-growth forests. [Copyright &y& Elsevier]

Abstract: A typical model for geostatistical data when the observations are counts is the spatial generalised linear mixed model. We present a criterion for optimal sampling design under this framework which aims to minimise the error in the prediction of the underlying spatial random effects. The proposed criterion is derived by performing an asymptotic expansion to the conditional prediction variance. We argue that the mean of the spatial process needs to be taken into account in the construction of the predictive design, which we demonstrate through a simulation study where we compare the proposed criterion against the widely used space-filling design. Furthermore, our results are applied to the Norway precipitation data and the rhizoctonia disease data. [Copyright &y& Elsevier]

A reversible jump algorithm for Bayesian model determination among generalised linear models, under relatively diffuse prior distributions for the model parameters, is proposed. Orthogonal projections of the current linear predictor are used so that knowledge from the current model parameters is used to make effective proposals. This idea is generalised to moves of a reversible jump algorithm for model determination among generalised linear mixed models. Therefore, this algorithm exploits the full flexibility available in the reversible jump method. The algorithm is demonstrated via two examples and compared to existing methods. [ABSTRACT FROM AUTHOR]

Abstract: The aim of this study was to develop statistical models for first order branchiness in young planted forest stands of pedunculate oak (Quercus robur L.) and European beech (Fagus sylvatica L.), and to give an ecological and silvicultural interpretation to these models. The reported models focus on the lower most-valuable stem part (i.e. until 6m height), and cover different tree development classes to capture the development of branchiness over time. For each species 30 study plots were selected spread over two nearby forests in Flanders (northern Belgium), minimising site and genetic variability. Branches were counted on a total of 399 oak and 376 beech trees. On a subsample of 30 trees per species (one tree per plot), detailed non-destructive branch measurements were performed, yielding data for 555 oak and 438 beech branches. For both species, models for tree self-pruning (i.e. total branch number and dead branch portion), branch mortality and branch architecture (i.e. branch diameter and branch insertion angle) were built. A generalised linear mixed modelling approach was adopted. The models for total branch number and dead branch portion may be interpreted in terms of four processes contributing to self-pruning: (1) stand and tree development, (2) tree competitive status, (3) stand density and (4) site humidity. The reported models reveal similar self-pruning rates in oak and beech, but with different driving factors: early branch dying and slow shedding for oak and the other way around for beech. Mortality of individual branches is further determined by branch position and branch dimension. Branch diameter and branch insertion angle of both species are mainly related to branch cord length and relative branch position. All modelled effects are consistent with known ecological and ecophysiological processes. Silvicultural implications for stand establishment and early tree selection are discussed. The reported models can be used to fine-tune operational silvicultural choices for quality timber production. This is a first step towards the integration of branchiness models for oak and beech into forest growth simulators. [Copyright &y& Elsevier]

The clumping growth form of tropical bamboos suggests a consolidator strategy in the face of intense competition rather than an ability to exploit disturbance. We investigated the annual growth and culm demographic responses to disturbance by fire and flood of vegetatively mature clumps of a riparian stand of Bambusa arnhemica. Linear-mixed models were employed to control for the non-independence of culm fates within clumps and clump growth among years. As the stand was monodominant and the species is gregariously monocarpic, responses can be interpreted independently of interspecific competition and trade-offs with sexual reproduction. Disturbance depressed clump growth but the culm demographic responses to fire and flood were quite different. Few culms were destroyed during the fire but damage depressed their subsequent survival, leading to declines in clump basal area. Fire also triggered the release of ramet buds, but the additional recruits were small and short-lived. Prolonged and early flooding in one of the study years depressed culm recruitment in clumps low on the lagoon bank but there was possible compensatory recruitment in the following year. There was no convincing evidence that sparse clumps recruited culms better than dense clumps, though culm recruitment was negatively correlated with culm survival. Fire and prolonged flooding are inevitable elements of the environment occupied by B. arnhemica, but we interpret the species’ response as persisting in the face of disturbance rather than exploiting the opportunities created by it. [ABSTRACT FROM AUTHOR]

There is growing evidence that the sex ratios of wild vertebrate populations are determined by mechanisms that are directly influenced by environmental characteristics. The Trivers–Willard (TWH) and extrinsic modification (EMH) hypotheses postulate differing determinants of mammalian offspring sex ratios. TWH states that mothers allocate resources according to their current condition and sex-specific offspring costs. EMH states that environmental forces that affect maternal condition determine offspring sex ratios, independently of maternal tactics of sex-biased allocation. We statistically assessed support for each of these hypotheses using long-term life histories of the allied rock-wallaby, Petrogale assimilis; a continuously breeding, polygynous, weakly dimorphic marsupial. We showed that birth sex ratios were equal and independent of maternal and environmental conditions. However, secondary sex ratios were male-biased under good environmental conditions and for high quality mothers or mothers in good condition. Sex differences in offspring survival contributed to these biases: (1) environmental conditions strongly influenced survival to pouch emergence (in support of EMH) and (2) maternal quality affected survival to the end of maternal care (in support of TWH). Environmental effects on survival were more important than maternal factors over the entire period of maternal care and contributed most to male-biased sex ratios at pouch emergence. In contrast, maternal mass was the best predictor of sex ratios at the end of maternal care—the life history stage where offspring body mass differed between the sexes. [ABSTRACT FROM AUTHOR]

The article discusses foraging habitat selection during reproduction by the Ortolan Bunting Emberiza hortulana. The last known breeding population of the Ortolan Bunting is in Valais, Switzerland but in 2007, there was no confirmed breeding and the remaining population consisted mostly of unpaired males. Six males were ringed and fitted with radio tags, four of which were tracked, along with one female. The Ortolan Buntings were observed to avoid Riparian vegetation and favor treated Maize fields which usually have more bare ground. Furthermore, food accessibility seemed to be more important than food abundance. It is noted that a fire in the area in 1979 may have triggered changes that made the environment unsuitable for foraging.

Highly structured generalised response models, such as generalised linear mixed models and generalised linear models for time series regression, have become an indispensable vehicle for data analysis and inference in many areas of application. However, their use in practice is hindered by high-dimensional intractable integrals. Quasi-Monte Carlo (QMC) is a dynamic research area in the general problem of high-dimensional numerical integration, although its potential for statistical applications is yet to be fully explored. We survey recent research in QMC, particularly lattice rules, and report on its application to highly structured generalised response models. New challenges for QMC are identified and new methodologies are developed. QMC methods are seen to provide significant improvements compared with ordinary Monte Carlo methods. [ABSTRACT FROM AUTHOR]

This article discusses mapping techniques for modeling extreme risk areas and outliers. According to the authors, extreme risk areas may be in close proximity in a smooth spatial surface or they may be distinct from neighboring sites. They model smooth surfaces with a conditional autoregressive method and extreme isolated risks with a discrete clustering component. They incorporate the spatially correlated risk as a baseline surface and discrete components are beyond the baseline risk surface. They suggest that both types of extreme risk are important but discrete risks may be useful for surveillance.

Summary: This paper provides a unified algorithm to explicitly calculate the maximum likelihood estimates of parameters in a general setting of generalised linear mixed models (GLMMs) in terms of Gauss-Hermite quadrature approximation. The score function and observed information matrix are expressed explicitly as analytically closed forms so that Newton-Raphson algorithm can be applied straightforwardly. Compared with some existing methods, this approach can produce more accurate estimates of the fixed effects and variance components in GLMMs, and can serve as a basis of assessing existing approximations in GLMMs. A simulation study and practical example analysis are provided to illustrate this point. [ABSTRACT FROM AUTHOR]

The variation of the age composition of the landings of lesser sandeel in the Danish industrial fishery in the North Sea over the period from 1984–1993 is analysed by continuation-ratio logits and generalised linear models. The analysis takes the multinomial characteristics of the age composition into account and allows both fixed and random effects to be quantified. It also allows an estimate of over-dispersion, i.e. the extent to which the variance exceeds the multinomial variance expected if the samples had been drawn at random from each stratum. The results show that the proportion of 2+ sandeel in the samples is significantly lower in the start and end of the fishing season. This suggests that the older sandeel are available to the fishery for a shorter time period that the 1-group. Significant differences are found in the age composition between the four laboratories involved in the age determination. Although the variation between ICES statistical rectangles is substantial there is a significant difference between the age composition in the northern and southern part of the North Sea. However, only one of the three finer geographical stratifications proposed to improve the assessment results in significant differences in age composition between sub-strata. Additional unidentified sources of variation result in a large and significant over-dispersion. [ABSTRACT FROM PUBLISHER]

From 1970 to 2000, the yearly Baltic Sea River Runoff (BSRR) together with the silicate concentrations in the Baltic (BS) and North Sea (NS) were observed. We found that the concentrations of silicates were consistently higher in the BS. Furthermore, long-term changes in both ecosystems indicated an overall decrease of Si with increased BSRR. Silicate concentrations were also modelled in two setups of Generalised Linear Mixed Models. Selected exceptional (high and low) BSRR events were used as a predictor in the models. The results suggest that the whole BS acts as a silicate source for the studied, adjacent NS areas. Modelling indicated that the periods with low BSRR show generally higher silicate levels in BS water. In the NS, however, the silicate concentrations decrease with low BSRR periods and increase at high. We point out that the increase of rainfall, and consequently the BSRR, has predominantly taken place in the northern parts of the BS basin, where the ground silicate concentrations are lower than in the southern BS. Therefore, we present, as a conceptual model and a subject for further research, that the BS potentially has acted as a concentrating basin for silicates during periods of normal and low rainfall, while episodic high runoff periods flush the BS surface water including the silicates through the Danish Sounds and into the adjacent parts of the NS. In a post hoc test, we found that also the reproductive success of the NS herring showed a correlation to silicate concentrations thus partially demonstrating the existence of a Si-related, bottom-up regulation. • The Baltic Sea DSi content was higher than in the North Sea, thus BS outflows acted as a DSi supply for the NS ecosystem. • The ground silicate concentrations are lower in the northern BS than in the south, where large sandy areas predominate. • Since 1980s, the increase of rainfall and BS river runoffs has predominantly taken place in the northern BS watershed. • Due to the dilution effect, the BS seawater contains now less DSi, thus BS outflows have less relevance for DSi of the NS. [ABSTRACT FROM AUTHOR]

Reef fish occupy numerous functional groups and play integral roles in maintaining the health of coral reef ecosystems. However, the management and conservation of these fish groups, and thus coral reefs, are hindered by a paucity of information regarding their spatial distribution patterns and the environmental drivers behind these patterns in the Philippines. Therefore, this study quantified the spatial abundance patterns of selected functional fish groups around two coastal bays in the Philippines, and determined which factors influenced these patterns on both small, localised and larger regional scales. A modified Reef Check methodology was utilised to determine the abundance of fish groups and the substrate composition for 15 coral reefs situated in five delineated regions. The abundance of herbivorous, coralivorous, and top predatory fishes were highly variable, both within and between reefs. Significant differences in fish group abundance were found at the reef, but not at the regional level. Commercially important fish groups (Scaridae, Lutjanidae, and Epinephelinae) were dominated by small size classes (less than 20 cm total length). Sites were also found to be variable in their substrate composition and subsistence fishing effort. Generalised Linear Mixed Models highlighted that localised influences of fishing and hard coral cover, rather than larger-scale regional influences, were the most influential on the abundance patterns of herbivorous, coralivorous and top predatory fish groups. This may suggest that the current management approach of localised small-scale marine protected areas (MPAs) may benefit some of these fish communities. However, top predatory fish likely display far-reaching movement patterns that are not encompassed by these small MPAs, and further management efforts are recommended. • Functional reef fish group abundance varied locally within and between coral reefs. • No patterns in reef fish abundance were identified regionally between two bays. • Variation in fish abundance was driven by localised fishing and benthos factors. • Over-fishing was identified and management options discussed. [ABSTRACT FROM AUTHOR]