1,721,010 research outputs found
Recommended from our members
Statistical modeling with counts of bats
Get PDFCount data are often the most available indices for bat abundance. Counts of bats are useful for estimating differences in population size, between different habitats, or at different times. But such estimates are not without complications. We used pre-existing bat count data sets, and simulated bat count data to survey a variety of methods for interpreting bat counts. These methods included, evaluation of habitat preferences through AIC model selection of generalized linear models, evaluation of differences in abundance using hierarchical models, and evaluation of spatially replicated time series dynamics using additive mixed-models. All methods proved useful under the special circumstances that accommodated model assumptions. Generalized linear models required the most restrictive assumptions, while hierarchical and additive models allowed many assumptions to be relaxed. We identified several areas where current modeling practices might be improved
Recommended from our members
Synthesizing multiple data sources to understand the population and community ecology of California trees
Get PDFIn this work, I answer timely questions regarding tree growth, tree survival, and community change in California tree species, using a variety of sophisticated statistical and remote sensing tools. In Chapter 1, I address tree growth for a single tree species with a thorough explanation of hierarchical state-space models for forest inventory data. Understanding tree growth as a function of tree size is important for a multitude of ecological and management applications. Determining what limits growth is of central interest, and forest inventory permanent plots are an abundant source of long-term information but are highly complex. Observation error and multiple sources of shared variation make these data challenging to use for growth estimation. I account for these complexities and incorporate potential limiting factors into a hierarchical state-space model. I estimate the diameter growth of white fir in the Sierra Nevada of California from forest inventory data, showing that estimating such a model is feasible in a Bayesian framework using readily available modeling tools. In this forest, white fir growth depends strongly on tree size, total plot basal area, and unexplained variation between individual trees. Plot-level resource supply variables do not have a strong impact on inventory-size trees. This approach can be applied to other networks of permanent forest plots, leading to greater ecological insights on tree growth. In Chapter 2, I expand my state-space modeling to examine survival in seven tree species, as well as investigating the results of modeling them in aggregate and comparing with the individual species models. Declining tree survival is a complex, well-recognized problem, but studies have been largely limited to relatively rare old-growth forests or low-diversity systems, and to models which are species-aggregated or cannot easily accommodate yearly climate variables. I estimate survival models for a relatively diverse second-growth forest in the Sierra Nevada of California using a hierarchical state-space framework. I account for a mosaic of measurement intervals and random plot variation, and I directly include yearly stand development variables alongside climate variables and topographic proxies for nutrient limitation. My model captures the expected dependence of survival on tree size. At the community level, stand development variables account for decreasing survival trends, but species-specific models reveal a diversity of factors influencing survival. Species time trends in survival do not always conform to existing theories of Sierran forest dynamics, and size relationships with survival differ for each species. Within species, low survival is concentrated in susceptible subsets of the population and single estimates of annual survival rates do not reflect this heterogeneity in survival. Ultimately only full population dynamics integrating these results with models of recruitment can address the potential for community shifts over time. In Chapter 3, I combine statistical modeling with remote sensing techniques to investigate whether topographic variables influence changes in woody cover. In the North Coast of California, changes in fire management have resulted in conversion of oak woodland into coniferous forest, but the controls on this slow transition are unknown. Historical aerial imagery, in combination with Object-Based Image Analysis (OBIA), allows us to classify land cover types from the 1940s and compare these maps with recent cover. Few studies have used these maps to model drivers of cover change, partly due to two statistical challenges: 1) appropriately accounting for spatial autocorrelation and 2) appropriately modeling percent cover which is bounded between 0 and 100 and not normally distributed. I study the change in woody cover in California's North Coast using historical and recent high-spatial-resolution imagery. I classify the imagery using eCognition Developer and aggregate the resulting maps to the scale of a Digital Elevation Model (DEM) in order to understand topographic drivers of woody cover change. I use Generalized Additive Models (GAMs) with a quasi-binomial probability distribution to account for spatial autocorrelation and the boundedness of the percent woody cover variable. I find that historical woody cover has a consistent positive effect on current woody cover, and that the spatial term in the model is significant even after controlling for historical cover. Specific topographic variables emerge as important for different sites at different scales, but no overall pattern emerges across sites or scales for any of the topographic variables I tested. This GAM framework for modeling historical data is flexible and could be used with more variables, more flexible relationships with predictor variables, and larger scales. Modeling drivers of woody cover change from historical ecology data sources can be a valuable way to plan restoration and enhance ecological insight into landscape change. I conclude that these techniques are promising but a framework is needed for sensitivity analysis, as modeling results can depend strongly on variable selection and model structure. (Abstract shortened by UMI.
Recommended from our members
Individual heterogeneity in life history processes: Estimation and applications of demographic models to stage-structured arthropod populations
Get PDFLife history variation is a general feature of natural populations. Most studies assume that local processes occur identically across individuals, ignoring any genetic or phenotypic variation in life history traits. In part, this is because a realistic treatment of individual heterogeneity results in very complex population models. Fitting models with individual heterogeneity to real data is further complicated by random effects in groups of the data, observations set at specific intervals, and the non-independence of data following a cohort of individuals through time. In this dissertation, I assume that individuals differ in the duration they spend in each developmental stage and also in the amount of time they live. Stage durations and survival times follow probability distributions with parameters specific to populations and stages. Parameters of these distributions may also include random effects when considering a subset of sampled populations and covariates such as temperature. In the first chapter I formulate a model and likelihood for variable development, using the time-to-event model framework. In the second chapter I use this model to ask whether field populations of herbivorous arthropods (Tetranychus pacificus) form host-associations on different cultivars of the same host species. In the third chapter I incorporate variable development with variable survival and ongoing reproduction in a stage-structured population model. I explore the ability of the approximate Bayesian computation framework to fit such a complex model to data, evaluating posterior distributions and model performance
Recommended from our members
Inferring species distributions from semi-structured biodiversity observations
Get PDFEstimating the spatiotemporal distributions of species and understanding how variation in those distributions is explained by the environment are central goals in ecology. Observations of animals generated by participatory science (or "citizen science") are an increasingly important resource for ecologists interested in estimating species distributions because they are high-volume and high-resolution. However, statistical inference with these data is more challenging than inference with data collected under standardized sampling, because participatory science observations contain substantial unmeasured variation in sampling effort and observer behavior. Ecologists need tools and methodological guidance that support the estimation of computationally efficient, flexible statistical models useful for robust inference with participatory science data. In this dissertation, I advance the field of species distribution modeling with participatory science data via contributions across three chapters. First, I present a new software tool, nimbleEcology, that supports the efficient and flexible estimation of hierarchical ecological models, alongside a brief review of the use of such models in ecology and three worked examples of model estimation. Second, I undertake a comparison of two modeling approaches useful for estimating relative abundance from participatory science data, making practical recommendations for model selection. Finally, I apply these methodological developments to data obtained from an important participatory science dataset, eBird, to investigate how common birds respond to drought in California's Central Valley ecoregion. This project demonstrates the application of modeling principles to an important ecological case study and produces new evidence to characterize critical dimensions of birds' drought responses
Recommended from our members
Latent Variable Models: Maximum Likelihood Estimation and Microbiome Data Analysis
Get PDFData analysis often involves modeling complex relationships among many variables, some of which are unobserved. This type of analysis is usually tackled by latent variable models, which are graphical models consisting of both observed variables and latent variables. In this work, we delve into the computational aspect and the application aspect of latent variable models. On the computational side, we unify and extend stochastic gradient based maximum likelihood estimation methods for latent variable models under a framework called Hierarchical Model Stochastic Gradient Descent (HMSGD). Numerical studies have shown that certain extensions are more computationally efficient compared to the Monte Carlo Expectation Maximization (MCEM) algorithm. On the application side, we develop a non-parametric graphical model for microbiome data, and apply the framework to analyze the statistical properties of rarefaction, a popular normalization technique in microbiome data analysis. We show that rarefaction helps guarantee validity of permutation inference. We introduce the sample rarefaction efficiency index as a preliminary data-driven indicator of statistical efficiency of rarefied data compared to original data. Using the nonparametric graphical model, we propose a rarefaction-based nonparametric statistical testing procedure, the combined correlation permutation test, to assess whether library sizes are associated with microbial compositions conditioning on the grouping variable of interest. Case studies have shown that such associations are not uncommon in practice
Recommended from our members
A Statistical Investigation of Species Distribution Models and Communication of Statistics Across Disciplines
Get PDFEcologists commonly make strong parametric assumptions when formulating statistical models. Such assumptions have sparked repeated debates in the literature about statistical identifiability of species distribution and abundance models, among others. Part I of this dissertation draws upon the econometrics literature to introduce a broader view of the identifiability problem than has been taken in ecological debates. In particular we use a simulation approach to illustrate the concepts of non-parametric and parametric identifiability and their implications for ecologists. The fact that all models are approximations has very different implications for these two cases of identifiability. When non-parametric identifiability holds, even a mis-specified parametric model provides a useful approximation to the truth, and the fit of alternative models can be compared. When non-parametric identifiability does not hold, parametric assumptions create artificial identifiability, and alternative models cannot be distinguished empirically.Joint species distribution models (JSDMs) have become a popular tool for helping ecologists understand properties of a community while accounting for relationships between species. Part II of this dissertation stress tests a foundational JSDM to understand how well properties of the community are estimated in the presence of model mis-specification. Community diversity metrics summarize community characteristics that ecologists have historically been interested in, so it is of interest to ask whether estimation of these more complicated metrics is robust to inevitable model mis-specification.Being a statistician is a "hands-on" job that requires communicating with stakeholders and researchers in a variety of fields. Part III of this dissertation leverages the communication skills I have built while working at the intersection of ecology and statistics to teach statistics students how to write about statistical analyses in an accessible way that is still faithful to the data. A pedagogical approach is described that builds upon that of traditional writing and science communication. This approach adds to the solid foundation with concrete examples in the context of statistics, particular focus on the nuances of statistical language, and a focus on narrative that carries throughout the data analysis process itself
compareMCMCs: An R package for studying MCMC efficiency
Get PDFInitial release associated with paper in Journal of Open Source Software
Going Beyond Counting First Authors in Author Co-citation Analysis
Get PDFThe present study examines one of the fundamental aspects of author co-citation analysis (ACA) - the way co-citation
counts are defined. Co-citation counting provides the data on which all subsequent statistical analyses and mappings
are based, and we compare ACA results based on two different types of co-citation counting - the traditional type that
only counts the first one among a cited work's authors on the one hand and a non-traditional type that takes into
account the first 5 authors of a cited work on the other hand. Results indicate that the picture produced through this non-traditional author co-citation counting contains more coherent author groups and is therefore considerably clearer. However, this picture represents fewer specialties in the research field being studied than that produced through the traditional first-author co-citation counting when the same number of top-ranked authors is selected and analyzed. Reasons for these effects are discussed
Variations on the Author
Get PDF“Variations on the Author” discusses two of Eduardo Coutinho’s recent films (Um Dia na Vida, from 2010, and Últimas Conversas, posthumously released in 2015) and their contribution to the general question of documentary authorship. The director’s filmography is characterized by a consistent yet self-effacing form of authorial self-inscription: Coutinho often features as an interviewer that rather than express opinions propels discourses; an interviewer that is good at listening. This mode of self-inscription characterizes him as an author who is not expressive but who is nonetheless markedly present on the screen. In Um Dia na Vida, however, Coutinho is completely absent form the image, while Últimas Conversas, on the contrary, includes a confessional prologue that moves the director from the margins to the center of his films. This article examines the ways in which these works stand out in the filmography of a director who offers new insights into the notion of cinematic authorship
- …
