The role of the seed bank in recovery of temperate heath and blanket bog following wildfires

Questions - Are the germinable seed banks of upland heath and blanket bog reduced following wildfires? Are some species at particular risk? Do the impacts of wildfires on seed banks differ between heathlands and blanket bog?

Location - Northern Ireland, United Kingdom.

Methods - Vegetation surveys and seed bank sampling were conducted in 2012 at burned and unburned areas within six upland sites where large wildfires had occurred during spring 2011. Differences in seedling abundance, species richness and Jaccard similarity indices between burned and unburned areas were compared using GLMMs. Differences in the community composition were examined using pRDA.

Results - In total, 24 of the 51 species in the vegetation were detected in the germinable seed bank. Species richness and the abundance of seedlings other than Calluna vulgaris were lower in areas where wildfires had occurred. Species composition of both germinable seed banks and vegetation differed between burned and unburned areas within sites; with negative associations between burned areas and some key indicator species including Drosera rotundifolia, Eriophorum vaginatum, Empetrum nigrum, Narthecium ossifragum and Trichophorum germanicum. We did not find any evidence of significant interactions between burning and habitat, suggesting that wildfires had similar impacts on each species regardless of the habitat in which they occurred.

Conclusions - This study differs from other UK studies in that it examines impacts of wildfires at sites that have not been previously intensively managed by burning. In particular, we highlight potential impacts on N. ossifragum and D. rotundifolia, which are key components of the upland flora and, to our knowledge, were not present in previous UK studies.

Forests of the sea: predictive habitat modelling to assess the abundance of canopy forming kelp forests on temperate reefs

 Large brown seaweeds (kelps) form forests in temperate and boreal marine systems that serve as foundations to the structure and dynamics of communities. Mapping the distributions of these species is important to understanding the ecology of coastal environments, managing marine ecosystems (e.g., spatial planning), predicting consequences of climate change and the potential for carbon production. We demonstrate how combining seafloor mapping technologies (LiDAR and multibeam bathymetry) and models of wave energy to map the distribution and relative abundance of seaweed forests of Ecklonia radiata can provide complete coverage over hundreds of square kilometers. Using generalized linear mixed models (GLMMs), we associated observations of E. radiata abundance from video transects with environmental variables. These relationships were then used to predict the distribution of E. radiata across our 756.1km2 study area off the coast of Victoria, Australia. A reserved dataset was used to test the accuracy of these predictions. We found that the abundance distribution of E. radiata is strongly associated with depth, presence of rocky reef, curvature of the reef topography, and wave exposure. In addition, the GLMM methodology allowed us to adequately account for spatial autocorrelation in our sampling methods. The predictive distribution map created from the best GLMM predicted the abundance of E. radiata with an accuracy of 72%. The combination of LiDAR and multibeam bathymetry allowed us to model and predict E. radiata abundance distribution across its entire depth range for this study area. Using methods like those presented in this study, we can map the distribution of macroalgae species, which will give insight into ecological communities, biodiversity distribution, carbon uptake, and potential sequestration.

Modelos lineares generalizados mistos e equações de estimação generalizadas para dados binário aplicados em anestesiologia veterinária

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES)

Differences in habitat use of two sympatric species of Ameiva in East Costa Rica

Identifying the differences in habitat use for sympatric species is important for understanding the species preferences and the limits of population distribution. We studied the differences in the habitat use of two understudied sympatric species of Ameiva (A. festiva and A. quadrilineata) in a natural reserve of the Caribbean coast of Coast Rica. Ameiva quadrilineata showed a more restrictive habitat use pattern than A. festiva. A. quadrilineata's smaller body size may be one of the factors limiting its habitat range. Both species showed higher density in regenerated forests, while A. quadrilineata was never found in swamp forests. The air temperature and the meteorological condition at the moment of the survey also influenced the occurrence of the A. quadrilineata, while the juveniles of A. festiva were only affected by the meteorological condition. None of the studied variables seemed to affect the occurrence of A. festiva adults. The results of this study can be useful to evaluate possible changes in the species distribution patterns as a consequence of direct (i.e., deforestation) or indirect (i.e., climate change) human activities in the distribution area of these species.

Whither PQL?

Generalized linear mixed models (GLMM) are generalized linear models with normally distributed random effects in the linear predictor. Penalized quasi-likelihood (PQL), an approximate method of inference in GLMMs, involves repeated fitting of linear mixed models with “working” dependent variables and iterative weights that depend on parameter estimates from the previous cycle of iteration. The generality of PQL, and its implementation in commercially available software, has encouraged the application of GLMMs in many scientific fields. Caution is needed, however, since PQL may sometimes yield badly biased estimates of variance components, especially with binary outcomes. Recent developments in numerical integration, including adaptive Gaussian quadrature, higher order Laplace expansions, stochastic integration and Markov chain Monte Carlo (MCMC) algorithms, provide attractive alternatives to PQL for approximate likelihood inference in GLMMs. Analyses of some well known datasets, and simulations based on these analyses, suggest that PQL still performs remarkably well in comparison with more elaborate procedures in many practical situations. Adaptive Gaussian quadrature is a viable alternative for nested designs where the numerical integration is limited to a small number of dimensions. Higher order Laplace approximations hold the promise of accurate inference more generally. MCMC is likely the method of choice for the most complex problems that involve high dimensional integrals.

Gauss-Seidel Estimation of Generalized Linear Mixed Models with Application to Poisson Modeling of Spatially Varying Disease Rates

Generalized linear mixed models (GLMMs) provide an elegant framework for the analysis of correlated data. Due to the non-closed form of the likelihood, GLMMs are often fit by computational procedures like penalized quasi-likelihood (PQL). Special cases of these models are generalized linear models (GLMs), which are often fit using algorithms like iterative weighted least squares (IWLS). High computational costs and memory space constraints often make it difficult to apply these iterative procedures to data sets with very large number of cases. This paper proposes a computationally efficient strategy based on the Gauss-Seidel algorithm that iteratively fits sub-models of the GLMM to subsetted versions of the data. Additional gains in efficiency are achieved for Poisson models, commonly used in disease mapping problems, because of their special collapsibility property which allows data reduction through summaries. Convergence of the proposed iterative procedure is guaranteed for canonical link functions. The strategy is applied to investigate the relationship between ischemic heart disease, socioeconomic status and age/gender category in New South Wales, Australia, based on outcome data consisting of approximately 33 million records. A simulation study demonstrates the algorithm's reliability in analyzing a data set with 12 million records for a (non-collapsible) logistic regression model.

Tree and grazing influence on herbaceous pasture diversity in the Dehesa agroforestry system

Mediterranean Dehesas are one of the European natural habitat types of Community interest (43/92/EEC Directive), associated to high diversity levels and producer of important goods and services. In this work, tree contribution and grazing influence over pasture alpha diversity in a Dehesa in Central Spain was studied. We analyzed Richness and Shannon-Wiener (SW) indexes on herbaceous layer under 16 holms oak trees (64 sampling units distributed in two directions and in two distances to the trunk) distributed in four different grazing management zones (depending on species and stocking rate). Floristic composition by species or morphospecies and species abundance were analyzed for each sample unit. Linear mixed models (LMM) and generalized linear mixed models (GLMMs) were used to study relationships between alpha diversity measures and independent factors. Edge crown influence showed the highest values of Richness and SW index. No significant differences were found between orientations under tree crown influence. Grazing management had a significant effect over Richness and SW measures, specially the grazing species (cattle or sheep). We preliminary quantify and analyze the interaction of tree stratum and grazing management over herbaceous diversity in a year of extreme climatic conditions.

Methods for Categorical Longitudinal Survey Data: Understanding Employment Status of Australian Women

Many variables that are of interest in social science research are nominal variables with two or more categories, such as employment status, occupation, political preference, or self-reported health status. With longitudinal survey data it is possible to analyse the transitions of individuals between different employment states or occupations (for example). In the statistical literature, models for analysing categorical dependent variables with repeated observations belong to the family of models known as generalized linear mixed models (GLMMs). The specific GLMM for a dependent variable with three or more categories is the multinomial logit random effects model. For these models, the marginal distribution of the response does not have a closed form solution and hence numerical integration must be used to obtain maximum likelihood estimates for the model parameters. Techniques for implementing the numerical integration are available but are computationally intensive requiring a large amount of computer processing time that increases with the number of clusters (or individuals) in the data and are not always readily accessible to the practitioner in standard software. For the purposes of analysing categorical response data from a longitudinal social survey, there is clearly a need to evaluate the existing procedures for estimating multinomial logit random effects model in terms of accuracy, efficiency and computing time. The computational time will have significant implications as to the preferred approach by researchers. In this paper we evaluate statistical software procedures that utilise adaptive Gaussian quadrature and MCMC methods, with specific application to modeling employment status of women using a GLMM, over three waves of the HILDA survey.

Endemic zoonoses: a one health approach

Endemic zoonotic diseases remain a serious but poorly recognised problem in affected communities in developing countries. Despite the overall burden of zoonoses on human and animal health, information about their impacts in endemic settings is lacking and most of these diseases are continuously being neglected. The non-specific clinical presentation of these diseases has been identified as a major challenge in their identification (even with good laboratory diagnosis), and control. The signs and symptoms in animals and humans respectively, are easily confused with other non-zoonotic diseases, leading to widespread misdiagnosis in areas where diagnostic capacity is limited. The communities that are mostly affected by these diseases live in close proximity with their animals which they depend on for livelihood, which further complicates the understanding of the epidemiology of zoonoses. This thesis reviewed the pattern of reporting of zoonotic pathogens that cause febrile illness in malaria endemic countries, and evaluates the recognition of animal associations among other risk factors in the transmission and management of zoonoses. The findings of the review chapter were further investigated through a laboratory study of risk factors for bovine leptospirosis, and exposure patterns of livestock coxiellosis in the subsequent chapters. A review was undertaken on 840 articles that were part of a bigger review of zoonotic pathogens that cause human fever. The review process involves three main steps: filtering and reference classification, identification of abstracts that describe risk factors, and data extraction and summary analysis of data. Abstracts of the 840 references were transferred into a Microsoft excel spread sheet, where several subsets of abstracts were generated using excel filters and text searches to classify the content of each abstract. Data was then extracted and summarised to describe geographical patterns of the pathogens reported, and determine the frequency animal related risk factors were considered among studies that investigated risk factors for zoonotic pathogen transmission. Subsequently, a seroprevalence study of bovine leptospirosis in northern Tanzania was undertaken in the second chapter of this thesis. The study involved screening of serum samples, which were obtained from an abattoir survey and cross-sectional study (Bacterial Zoonoses Project), for antibodies against Leptospira serovar Hardjo. The data were analysed using generalised linear mixed models (GLMMs), to identify risk factors for cattle infection. The final chapter was the analysis of Q fever data, which were also obtained from the Bacterial Zoonoses Project, to determine exposure patterns across livestock species using generalized linear mixed models (GLMMs). Leptospira spp. (10.8%, 90/840) and Rickettsia spp. (10.7%, 86/840) were identified as the most frequently reported zoonotic pathogens that cause febrile illness, while Rabies virus (0.4%, 3/840) and Francisella spp. (0.1%, 1/840) were least reported, across malaria endemic countries. The majority of the pathogens were reported in Asia, and the frequency of reporting seems to be higher in areas where outbreaks are mostly reported. It was also observed that animal related risk factors are not often considered among other risk factors for zoonotic pathogens that cause human fever in malaria endemic countries. The seroprevalence study indicated that Leptospira serovar Hardjo is widespread in cattle population in northern Tanzania, and animal husbandry systems and age are the two most important risk factors that influence seroprevalence. Cattle in the pastoral systems and adult cattle were significantly more likely to be seropositive compared to non-pastoral and young animals respectively, while there was no significant effect of cattle breed or sex. Exposure patterns of Coxiella burnetii appear different for each livestock species. While most risk factors were identified for goats (such as animal husbandry systems, age and sex) and sheep (animal husbandry systems and sex), there were none for cattle. In addition, there was no evidence of a significant influence of mixed livestock-keeping on animal coxiellosis. Zoonotic agents that cause human fever are common in developing countries. The role of animals in the transmission of zoonotic pathogens that cause febrile illness is not fully recognised and appreciated. Since Leptospira spp. and C. burnetii are among the most frequently reported pathogens that cause human fever across malaria endemic countries, and are also prevalent in livestock population, control and preventive measures that recognise animals as source of infection would be very important especially in livestock-keeping communities where people live in close proximity with their animals.