Generalized Linear Models (generalized + linear_models)

Distribution by Scientific Domains
Distribution within Mathematics and Statistics

Kinds of Generalized Linear Models

  • hierarchical generalized linear models

  • Selected Abstracts

    Modelling the species richness distribution for French Aphodiidae (Coleoptera, Scarabaeoidea)

    ECOGRAPHY, Issue 2 2004
    Jorge M. Lobo
    The species richness distribution of the French Aphodiidae was predicted using Generalized Linear Models to relate the number of species to spatial, topographic and climate variables. The entire French territory was studied, divided into 301 0.72×0.36 degree grid squares; the model was developed using 66 grid squares previously identified as well sampled. After eliminating nine outliers, the final model accounted for 74.8% of total deviance with a mean Jackknife predictive error of 10.5%. Three richest areas could be distinguished: the western head (Brittany), southwestern France, and, to a lesser extent, the northeastern region. Sampling effort should now be focused on the western head, where no square was correctly sampled, and on southwestern France, which was recognised as a diversity hotspot, both for Aphodiidae and for Scarabaeidae. The largest fraction of variability (37%) in the number of species was accounted for by the combined effect of the three groups of explanatory variables. After controlling for the effect of significant climate and topographic variables, spatial variables still explain 27% of variation in species richness, suggesting the existence of a spatial pattern in the distribution of species richness (greater diversity in western France) that can not be explained by the environmental variables considered here. We hypothesize that this longitudinal spatial pattern is due to the relevance of a western colonization pathway along the glacial-interglacial cycles, as well as by the barrier effect played by the Alps. [source]

    Predictive models of habitat preferences for the Eurasian eagle owl Bubo bubo: a multiscale approach

    ECOGRAPHY, Issue 1 2003
    Jose Antonio Martínez
    Habitat preference of eagle owls Bubo bubo were examined through comparing habitat composition around 51 occupied cliffs and 36 non-occupied cliffs in Alicante (E Spain). We employed Generalized Linear Models to examine patterns of habitat preference at three different spatial scales: nest site (7 km2), home range (25 km2), and landscape (100 km2). At the nest site scale, occupied cliffs were more rugged, had a greater proportion of forest surface in the surroundings, and were further from the nearest paved road than unoccupied cliffs. Additionally, probability of having an occupied cliff increased when there was another occupied territory in the surroundings. At both the home range scale and the landscape scale, high probabilities of presence of eagle owls were related to high percentages of Mediterranean scrubland around the cliffs, which are the preferred habitat of European rabbits Oryctolagus cuniculus, the main prey of the owls. We suggest a hierarchical process of habitat selection in the eagle owl concerning suitable trophic resources at the broadest scales and adequate sites for breeding and roosting at the smallest scale. However, it should be noted that some structural features such as the proximity of roads were not necessarily avoided by the owls, but their presence were possibly constrained by systematic killing of individuals. Our paper provides new evidence for the requirement of multi-scale approaches to gain insight into both the different limiting factors for the persistence of populations and the role of individual perception of the environment in the evolution of habitat selection. [source]

    BIOMOD , optimizing predictions of species distributions and projecting potential future shifts under global change

    GLOBAL CHANGE BIOLOGY, Issue 10 2003
    Wilfried ThuillerArticle first published online: 9 OCT 200
    Abstract A new computation framework (BIOMOD: BIOdiversity MODelling) is presented, which aims to maximize the predictive accuracy of current species distributions and the reliability of future potential distributions using different types of statistical modelling methods. BIOMOD capitalizes on the different techniques used in static modelling to provide spatial predictions. It computes, for each species and in the same package, the four most widely used modelling techniques in species predictions, namely Generalized Linear Models (GLM), Generalized Additive Models (GAM), Classification and Regression Tree analysis (CART) and Artificial Neural Networks (ANN). BIOMOD was applied to 61 species of trees in Europe using climatic quantities as explanatory variables of current distributions. On average, all the different modelling methods yielded very good agreement between observed and predicted distributions. However, the relative performance of different techniques was idiosyncratic across species, suggesting that the most accurate model varies between species. The results of this evaluation also highlight that slight differences between current predictions from different modelling techniques are exacerbated in future projections. Therefore, it is difficult to assess the reliability of alternative projections without validation techniques or expert opinion. It is concluded that rather than using a single modelling technique to predict the distribution of several species, it would be more reliable to use a framework assessing different models for each species and selecting the most accurate one using both evaluation methods and expert knowledge. [source]

    Generalized Linear Models for Insurance Data by Piet de Jong, Gillian Z. Heller

    Norman R. Draper
    No abstract is available for this article. [source]

    Generalized Linear Models in Family Studies

    Zheng WU
    Generalized linear models (GLMs), as defined by J. A. Nelder and R. W. M. Wedderburn (1972), unify a class of regression models for categorical, discrete, and continuous response variables. As an extension of classical linear models, GLMs provide a common body of theory and methodology for some seemingly unrelated models and procedures, such as the logistic, Poisson, and probit models, that are increasingly used in family studies. This article provides an overview of the principle and the key components of GLMs, such as the exponential family of distributions, the linear predictor, and the link function. To illustrate the application of GLMs, this article uses Canadian national survey data to build an example focusing on the number of close friends among older adults. The article concludes with a discussion of the strengths and weaknesses of GLMs. [source]

    Functional Generalized Linear Models with Images as Predictors

    BIOMETRICS, Issue 1 2010
    Philip T. Reiss
    Summary Functional principal component regression (FPCR) is a promising new method for regressing scalar outcomes on functional predictors. In this article, we present a theoretical justification for the use of principal components in functional regression. FPCR is then extended in two directions: from linear to the generalized linear modeling, and from univariate signal predictors to high-resolution image predictors. We show how to implement the method efficiently by adapting generalized additive model technology to the functional regression context. A technique is proposed for estimating simultaneous confidence bands for the coefficient function; in the neuroimaging setting, this yields a novel means to identify brain regions that are associated with a clinical outcome. A new application of likelihood ratio testing is described for assessing the null hypothesis of a constant coefficient function. The performance of the methodology is illustrated via simulations and real data analyses with positron emission tomography images as predictors. [source]

    Mixture Generalized Linear Models for Multiple Interval Mapping of Quantitative Trait Loci in Experimental Crosses

    BIOMETRICS, Issue 2 2009
    Zehua Chen
    Summary Quantitative trait loci mapping in experimental organisms is of great scientific and economic importance. There has been a rapid advancement in statistical methods for quantitative trait loci mapping. Various methods for normally distributed traits have been well established. Some of them have also been adapted for other types of traits such as binary, count, and categorical traits. In this article, we consider a unified mixture generalized linear model (GLIM) for multiple interval mapping in experimental crosses. The multiple interval mapping approach was proposed by Kao, Zeng, and Teasdale (1999, Genetics152, 1203,1216) for normally distributed traits. However, its application to nonnormally distributed traits has been hindered largely by the lack of an efficient computation algorithm and an appropriate mapping procedure. In this article, an effective expectation,maximization algorithm for the computation of the mixture GLIM and an epistasis-effect-adjusted multiple interval mapping procedure is developed. A real data set, Radiata Pine data, is analyzed and the data structure is used in simulation studies to demonstrate the desirable features of the developed method. [source]

    An Introduction to Generalized Linear Models, 3rd edition by DOBSON, A. J. and BARNETT, A. G.

    BIOMETRICS, Issue 2 2009
    Article first published online: 28 MAY 200
    No abstract is available for this article. [source]

    Bayesian Analysis for Generalized Linear Models with Nonignorably Missing Covariates

    BIOMETRICS, Issue 3 2005
    Lan Huang
    Summary We propose Bayesian methods for estimating parameters in generalized linear models (GLMs) with nonignorably missing covariate data. We show that when improper uniform priors are used for the regression coefficients, ,, of the multinomial selection model for the missing data mechanism, the resulting joint posterior will always be improper if (i) all missing covariates are discrete and an intercept is included in the selection model for the missing data mechanism, or (ii) at least one of the covariates is continuous and unbounded. This impropriety will result regardless of whether proper or improper priors are specified for the regression parameters, ,, of the GLM or the parameters, ,, of the covariate distribution. To overcome this problem, we propose a novel class of proper priors for the regression coefficients, ,, in the selection model for the missing data mechanism. These priors are robust and computationally attractive in the sense that inferences about , are not sensitive to the choice of the hyperparameters of the prior for , and they facilitate a Gibbs sampling scheme that leads to accelerated convergence. In addition, we extend the model assessment criterion of Chen, Dey, and Ibrahim (2004a, Biometrika91, 45,63), called the weighted L measure, to GLMs and missing data problems as well as extend the deviance information criterion (DIC) of Spiegelhalter et al. (2002, Journal of the Royal Statistical Society B64, 583,639) for assessing whether the missing data mechanism is ignorable or nonignorable. A novel Markov chain Monte Carlo sampling algorithm is also developed for carrying out posterior computation. Several simulations are given to investigate the performance of the proposed Bayesian criteria as well as the sensitivity of the prior specification. Real datasets from a melanoma cancer clinical trial and a liver cancer study are presented to further illustrate the proposed methods. [source]

    Variable Selection for Marginal Longitudinal Generalized Linear Models

    BIOMETRICS, Issue 2 2005
    Eva Cantoni
    Summary Variable selection is an essential part of any statistical analysis and yet has been somewhat neglected in the context of longitudinal data analysis. In this article, we propose a generalized version of Mallows's Cp (GCp) suitable for use with both parametric and nonparametric models. GCp provides an estimate of a measure of model's adequacy for prediction. We examine its performance with popular marginal longitudinal models (fitted using GEE) and contrast results with what is typically done in practice: variable selection based on Wald-type or score-type tests. An application to real data further demonstrates the merits of our approach while at the same time emphasizing some important robust features inherent to GCp. [source]

    Standard Errors for EM Estimates in Generalized Linear Models with Random Effects

    BIOMETRICS, Issue 3 2000
    Herwig Friedl
    Summary. A procedure is derived for computing standard errors of EM estimates in generalized linear models with random effects. Quadrature formulas are used to approximate the integrals in the EM algorithm, where two different approaches are pursued, i.e., Gauss-Hermite quadrature in the case of Gaussian random effects and nonparametric maximum likelihood estimation for an unspecified random effect distribution. An approximation of the expected Fisher information matrix is derived from an expansion of the EM estimating equations. This allows for inferential arguments based on EM estimates, as demonstrated by an example and simulations. [source]

    Neophyte species richness at the landscape scale under urban sprawl and climate warming

    Michael P. Nobis
    Abstract Aim, Land use and climate are two major components of global environmental change but our understanding of their simultaneous and interactive effects upon biodiversity is still limited. Here, we investigated the relationship between the species richness of neophytes, i.e. non-native vascular plants introduced after 1500 AD, and environmental covariates to draw implications for future dynamics under land-use and climate change. Location, Switzerland, Central Europe. Methods, The distribution of vascular plants was derived from a systematic national grid of 1 km2 quadrates (n = 456; Swiss Biodiversity Monitoring programme) including 1761 species, 122 of which were neophytes. Generalized linear models (GLMs) were used to correlate neophyte species richness with environmental covariates. The impact of land-use and climate change was thereafter evaluated by projections for the years 2020 and 2050 using scenarios of moderate and strong changes for climate warming (IPCC) and urban sprawl (NRP 54). Results, Mean annual temperature and the amount of urban areas explained neophyte species richness best, with a high predictive power of the corresponding model (cross-validated D2 = 0.816). Climate warming had a stronger impact on the potential increase in the mean neophyte species richness (up to 191% increase by 2050) than ongoing urban sprawl (up to 10% increase) independently from variable interactions and model extrapolations to non-analogue environments. Main conclusions, In contrast to other vascular plants, the prediction of neophyte species richness at the landscape scale in Switzerland requires few variables only, and regions of highest species richness of the two groups do not coincide. The neophyte species richness is basically driven by climatic (temperature) conditions, and urban areas additionally modulate small-scale differences upon this coarse-scale pattern. According to the projections climate warming will contribute to the future increase in neophyte species richness much more than ongoing urbanization, but the gain in new neophyte species will be highest in urban regions. [source]

    Evaluation of relative distance as new descriptor of yellow European eel spatial distribution

    H. Imbert
    Abstract,,, The spatial distribution of yellow European eel (Anguilla anguilla) smaller than 300 mm was analysed during the upstream colonisation process. A 9-year electric-fishing programme in the Gironde catchment (France) provided eel occurrence data in 256 sites and eel abundance data in 23 sites. Generalized linear models showed that small eel spatial distribution decreased with river slope, dam number and with downstream-upstream distance, estimated using either the distance from the tidal limit, called ,tidal distance', or the ,relative distance', calculated as the fish's position relative to the total distance between tidal limit and river source. This new descriptor should be considered in future eel distribution studies as it reveals fractal dimension in eel spatial distribution and may provide a standardised method to compare directly freshwater eel assessment between streams and catchments of different lengths. If the relevancy of this descriptor is subsequently confirmed, it may have important implications for the management of eel population conservation. [source]

    Partial regression method to fit a generalized additive model

    ENVIRONMETRICS, Issue 6 2007
    Shui He
    Abstract Generalized additive models (GAMs) have been used as a standard analytic tool in studies of air pollution and health during the last decade. The air pollution measure is usually assumed to be linearly related to the health indicator and the effects of other covariates are modeled through smooth functions. A major statistical concern is the appropriateness of fitting GAMs in the presence of concurvity. Generalized linear models (GLM) with natural cubic splines as smoothers (GLM,+,NS) have been shown to perform better than GAM with smoothing splines (GAM,+,S), in regard to the bias and variance estimates using standard model fitting methods. As nonparametric smoothers are attractive for their flexibility and easy implementation, search for alternative methods to fit GAM,+,S is warranted. In this article, we propose a method using partial residuals to fit GAM,+,S and call it the "partial regression" method. Simulation results indicate better performance of the proposed method compared to gam.exact function in S-plus, the standard tool in air pollution studies, in regard to bias and variance estimates. In addition, the proposed method is less sensitive to the degree of smoothing and accommodates asymmetric smoothers. Copyright © 2007 John Wiley & Sons, Ltd. [source]

    Advance directives and emergency department patients: ownership rates and perceptions of use

    D. McD.
    Abstract Background: Advance directives (ADs) are rarely avail­able in Australia to guide management but may become more important as our population ages. Aims: The present study aimed to determine patient knowledge, perception and ownership rates of ADs and the factors that impact upon these variables. Methods: A cross-sectional survey of emergency department patients was undertaken. The main outcome measures were: (i) prior discussion about the extent of medical treatment and ADs, (ii) knowledge and perceptions of ADs, (iii) present AD ownership rates and (iv) likelihood of future AD ownership. Generalized linear models were used for analysis. Results: Four hundred and three patients were enrolled. The mean age of patients was 73 years and 239 (59.3%) were male. Two hundred and forty patients (59.6%) had discussed the extent of treatment. Only 81 patients (20.1%) had discussed the use of an AD. One hundred and thirty-seven patients (34.0%) knew of one type of AD and 333 patients (82.6%) thought some ADs were a good idea. Only 32 patients (7.9%) owned an AD, although 276 (68.5%) would consider owning one. The main reason for never obtaining an AD was ,always wanting full treatment' (93 patients, 23.1%). Level of education was the only characteristic that impacted significantly upon an outcome measure. Patients with a higher level of education were more likely to have known and spoken about ADs, to own an AD and to consider owning one. Conclusions: AD knowledge and ownership rates were low. However, most patients perceive them favourably and many would consider owning one. Intervention strategies to improve AD awareness are indicated. This may empower patients to more effectively participate in their own advance care planning. (Intern Med J 2003; 33: 586,592) [source]

    Effect of origin, sex and sea age of Atlantic salmon on their recapture rate after river ascent

    E. Jokikokko
    Summary The recapture rate of Atlantic salmon (Salmo salar L.) after river ascent was examined by the trapping and tagging of ascending spawners in the lower reaches of the Simojoki River, which flows into the northern Baltic Sea. In 1997 and 1998, altogether 825 Carlin-tagged salmon were released to continue their upstream migration. Of these, 800 could be sexed and categorized as reared (91%) or wild (9%) salmon. In 1997, most of the ascending salmon were multi-sea-winter (MSW) fish, whereas in 1998 almost all were one-sea-winter (1SW) male grilse due to the late trapping season. About 10% of all tagged fish were recaptured, two-thirds of which were caught in the river before their descent to the sea. There was no difference in the recapture rate between salmon of wild (8.5%) or reared (9.5%) origin, or between females (11.6%) and males (9.3%). Generalized linear models for data from 1997 showed that the recapture rate increased with length and age of females, but that the opposite was true for males. River fishing did not seem to remove proportionally more early ascending salmon than fish that ascended later. [source]

    Predicting the distribution of four species of raptors (Aves: Accipitridae) in southern Spain: statistical models work better than existing maps

    Javier Bustamante
    Abstract Aim, To test the effectiveness of statistical models based on explanatory environmental variables vs. existing distribution information (maps and breeding atlas), for predicting the distribution of four species of raptors (family Accipitridae): common buzzard Buteo buteo (Linnaeus, 1758), short-toed eagle Circaetus gallicus (Gmelin, 1788), booted eagle Hieraaetus pennatus (Gmelin, 1788) and black kite Milvus migrans (Boddaert, 1783). Location, Andalusia, southern Spain. Methods, Generalized linear models of 10 × 10 km squares surveyed for the presence/absence of the species by road census. Statistical models use as predictors variables derived from topography, vegetation and land-use, and the geographical coordinates (to take account of possible spatial trends). Predictions from the models are compared with current distribution maps from the national breeding atlas and leading reference works. Results, The maps derived from statistical models for all four species were more predictive than the previously published range maps and the recent national breeding atlas. The best models incorporated both topographic and vegetation and land-use variables. Further, in three of the four species the inclusion of spatial coordinates to account for neighbourhood effects improved these models. Models for the common buzzard and black kite were highly predictive and easy to interpret from an ecological point of view, while models for short-toed eagle and, particularly, booted eagle were not so easy to interpret, but still predicted better than previous distribution information. Main conclusions, It is possible to build accurate predictive models for raptor distribution with a limited field survey using as predictors environmental variables derived from digital maps. These models integrated in a geographical information system produced distribution maps that were more accurate than previously published ones for the study species in the study area. Our study is an example of a methodology that could be used for many taxa and areas to improve unreliable distribution information. [source]

    Long-term stability of periodontal conditions achieved following guided tissue regeneration with bioresorbable membranes: case series results after 6,7 years

    Andreas Stavropoulos
    Abstract Objectives: To evaluate the results of guided tissue regeneration (GTR) treatment of intrabony defects with bioresorbable membranes after 6,7 years, and to disclose factors that may influence the long-term outcome of the treatment. Methods: Twenty-five defects in 19 patients were treated by means of polylactic acid/citric acid ester copolymer bioresorbable membranes. At baseline and after 1 and 6,7 years, the following parameters were recorded: (1) probing pocket depth (PPD), (2) gingival recession (REC), (3) probing attachment level (PAL)=PPD+REC, (4) presence/absence of plaque (PI), (5) presence/absence of bleeding on probing (BOP). Smoking habits and frequency of dental-control visits were also recorded. Significance of differences between categorical variables was evaluated with McNemar's test, and between numerical variables with the t -test for paired observations. Generalized linear models were constructed to evaluate the influence of various factors on PAL gain and PPD changes from 1 to 6,7 years. Association of smoking, frequency of dental controls, oral hygiene, and BOP with sites losing 2 mm in PAL was evaluated with Fisher's exact test. Results: At baseline, a mean PPD of 8.7±1.1 mm and a mean PAL of 9.8±1.5 mm was recorded. Statistically significant clinical improvements were observed at 1 and 6,7 years after GTR treatment. An average residual PPD of 3.8±1.1 mm and a mean PAL gain of 3.8±1.4 mm were observed after 1 year. After 6,7 years the corresponding values were 4.7±1.3 and 3.6±1.4 mm, respectively. There were no statistically significant differences between the 1- and the 6,7-year values. At the 6,7-year control, only 16% of the sites had lost 2 mm (maximum 3 mm), of the PAL gain obtained 1 year after GTR treatment. None of the sites had lost all of the attachment gained 1 year after treatment. Smoking, frequency of dental controls, oral hygiene, and BOP did not seem to influence the change of PPD and PAL gain, or the stability of PAL gain (i.e. losing PAL or not) from 1 to 6,7 years from treatment. Conclusion: Clinical improvements achieved by GTR treatment of intrabony defects by means of bioresorbable membranes can be maintained on a long-term basis. [source]

    Generalized Linear Models in Family Studies

    Zheng WU
    Generalized linear models (GLMs), as defined by J. A. Nelder and R. W. M. Wedderburn (1972), unify a class of regression models for categorical, discrete, and continuous response variables. As an extension of classical linear models, GLMs provide a common body of theory and methodology for some seemingly unrelated models and procedures, such as the logistic, Poisson, and probit models, that are increasingly used in family studies. This article provides an overview of the principle and the key components of GLMs, such as the exponential family of distributions, the linear predictor, and the link function. To illustrate the application of GLMs, this article uses Canadian national survey data to build an example focusing on the number of close friends among older adults. The article concludes with a discussion of the strengths and weaknesses of GLMs. [source]

    Seasonal Variation in Fluoride Intake: the Iowa Fluoride Study

    Barbara Broffitt MS;
    ABSTRACT Objectives: Although patterns of fluid intake change seasonally, little is known about how fluoride intake varies by season. Since even short-term increases in fluoride intake could potentially lead to more dental fluorosis, it is valuable to assess the degree of seasonal variation to determine if it increases fluoride intake to levels that could be considered a concern in young children. Methods: Questionnaires were mailed periodically to participants in the Iowa Fluoride Study beginning at 6 weeks of age and continuing for a number of years. Parents recorded the date; child's weight; estimates of the amounts of water and other beverages that their child consumed per week; the type and amount of any fluoride supplements used; and the type, amount, and frequency of dentifrice used, with an estimate of the proportion of dentifrice that was swallowed. Documented water fluoride levels from municipal sources and assay of individual sources were linked to water intake amounts. Total fluoride intake per kg body weight was estimated from water, other beverages, fluoride supplements, and ingested dentifrice. Generalized linear models compared temperature-related and seasonal effects after adjusting for the child's age. Results: Separate analyses for ages 0,12 months and 12,72 months showed different results. Children younger than 12 months of age did not exhibit significant seasonal or temperature-related variation in any of the components of fluoride intake. Children aged 12,72 months had higher fluoride intake (mg F/kg bw) from beverages in summer (P <.05), and fluoride intake from beverages increased with monthly temperature (P <.001). Conclusions: Fluoride intake from beverages for children aged 12,72 months is slightly higher in the summer and increases with mean monthly temperature. Fluoride intake from supplements and dentifrice did not change significantly with either season or temperature. [source]

    Generalized linear modelling in periglacial studies: terrain parameters and patterned ground

    Miska Luoto
    Abstract Generalized linear models (GLM) are mathematical extensions of linear models. GLM models are more flexible and better suited for analysing relationships of spatial data, which can often be poorly represented by classical Gaussian distributions such as least-square-regression techniques. This paper demonstrates GLM model-building procedures step-by-step for the distribution and abundance of active patterned ground in northern Finland. The exercise is based on data from an area of 200,km2 (800 modelling squares of 0.25,km2). Both the distribution and abundance models clearly indicate an increasing activity of patterned ground with (1) increasing soil moisture and (2) proportion of concave topography. Activity decreases with increasing altitude. We conclude that GLM techniques combined with a geographic information system can play an important role in analysing and modelling periglacial data sets. Copyright © 2004 John Wiley & Sons, Ltd. [source]

    Conidial dispersal by Alternaria brassicicola on Chinese cabbage (Brassica pekinensis) in the field and under simulated conditions

    PLANT PATHOLOGY, Issue 5 2003
    L. Y. Chen
    This study investigated conidial dispersal in the field, and effects of simulated wind and rain on the dispersal of A. brassicicola on Chinese cabbage (Brassica pekinensis). Spores were sampled using a Burkard volumetric spore sampler and rotorod samplers in a Chinese cabbage crop. Disease incidence in the field was well fitted by a Gompertz curve with an adjusted r2 of >0·99. Conidia of A. brassicicola were trapped in the field throughout the growing season. Peaks of high spore concentrations were usually associated with dry days, shortly after rain, high temperature or high wind speed. Diurnal periodicity of spore dispersal showed a peak of conidia trapped around 10·00 h. The number of conidia trapped at a height of 25 cm above ground level was greater than that at 50, 75 and 100 cm. Conidial dispersal was also studied under simulated conditions in a wind tunnel and a rain simulator. Generalized linear models were used to model these data. The number of conidia caught increased significantly at higher wind speeds and at higher rain intensities. Under simulated wind conditions, the number of conidia dispersed from source plants with wet leaves was only 22% of that for plants with dry leaves. Linear relationships were found between the number of conidia caught and the degree of infection of trap plants. [source]

    Secondary old-field succession in an ecosystem with restrictive soils: does time from abandonment matter?

    E. Martínez-Duro
    Abstract Question: Our knowledge of secondary old-field succession in Mediterranean environments is extremely poor and is non-existent for restrictive soil conditions. How these ecosystems, such as those on semi-arid gypsum outcrops, recover seems a priority for managing change and for ensuring conservation of specialized and endangered biota. We tested whether reinstallation of gypsum vegetation after cropland abandonment requires: (1) soil physical restructuring and (2) chemical readjustment to enable growth and survival of specialized gypsophilous vegetation, and more specifically how time from abandonment drives such environmental change. Location: We sampled a complete set of old fields on gypsum soils (1,60 yr since abandonment) in Villarrubia de Santiago (Toledo, Spain). Methods: Generalized linear models and model comparisons were used to analyse the effect of several environmental parameters on species abundance and richness. Ordination methods (canonical correspondence analyses and partial canonical correspondence analyses) were undertaken to evaluate compositional variation among the sampled fields. Results: Secondary old-field succession on semi-arid Mediterranean gypsum soils was controlled by a complex set of factors acting relatively independently. Surprisingly, time since abandonment explains only a small proportion of compositional variation (3%). Conversely, soil chemical features independently from time since abandonment are important for explaining differences found in old-field composition. Conclusions: Secondary succession on specialized Mediterranean soils does not follow the widely described "amelioration" process in which soil features and composition are closely related over time. Restrictive soil conditions control both structure and functioning of mature communities and also secondary succession. [source]

    Species richness,environment relationships within coastal sclerophyll and mesophyll vegetation in Ku-ring-gai Chase National Park, New South Wales, Australia

    AUSTRAL ECOLOGY, Issue 4 2003
    Andrew F. Le Brocque
    Abstract Patterns in species richness from a wide range of plant communities in Ku-ring-gai Chase National Park, New South Wales, Australia, were examined in relation to a number of environmental variables, including soil physical and chemical characteristics. Total species richness and richness of three growth-form types (trees, shrubs and ground cover) were determined in duplicate 500-m2 quadrats from 50 sites on two geological substrata: Hawkesbury Sandstone and Narrabeen shales and sandstones. Generalized linear models (GLM) were used to determine the amount of variation in species richness that could be significantly explained by the measured environmental variables. Seventy-three per cent of the variation in total species richness was explained by a combination of soil physical and chemical variables and site attributes. The environmental variables explained 24% of the variation in tree species richness, 67% of the variation in shrub species richness and 62% of the variation in ground cover species richness. These results generally support the hypothesis of an environmental influence on patterns in total species richness and richness of shrubs and ground cover species. However, tree species richness was not adequately predicted by any of the measured environmental variables; the present environment exerts little influence on the richness of this growth-form type. Historical factors, such as fire or climatic/environmental conditions at time of germination or seedling establishment, may be important in determining patterns in tree species richness at the local scale. [source]

    Patch Occupancy and Potential Metapopulation Dynamics of Three Forest Mammals in Fragmented Afromontane Forest in South Africa

    Michael J. Lawes
    We recorded patch occupancy of blue duiker ( Philantomba monticola), tree hyrax ( Dendrohyrax arboreus), and samango monkey (Cercopithecus mitis labiatus) in 199 forest patches. Their rarity is ascribed to the fragmentation and destruction of their forest habitat. Incidence functions, derived from presence and absence data, were formulated as generalized linear models, and environmental effects were included in the fitted logistic models. The small and mostly solitary hyrax and duiker persisted in smaller patches than the large and social monkey. Although this result follows expectations based on relative home-range sizes of each species, the incidence probability of the samango monkey was invariant with increasing isolation, whereas a gradual decrease with increasing isolation was observed for the hyrax and duiker. Group dynamics may inhibit dispersal and increase the isolation effect in social species such as samango monkeys. A mainland-island metapopulation model adequately describes patterns of patch occupancy by the hyrax and duiker, but the monkeys' poor dispersal ability and obvious area-dependent extirpation suggest that they exist in transient, nonequilibrium (declining) metapopulations. Through identification of large forest patches for careful protection and management, the survival of all three species,especially the monkey,could be prolonged. Because no functional metapopulation may exist for the monkey, however, this is an emergency measure. For the duiker and hyrax, larger patches should form part of a network of smaller and closer patches in a natural matrix. Resumen: Investigamos la persistencia de tres mamíferos forestales raros de tamaño mediano (2,9 kg) en los bosques fragmentados de cinturón de niebla Podocarpus en la región central de la provincia KwaZulu-Natal, Sudáfrica. Registramos la ocupación del duiker azul ( Philantomba monticola), el hyrax arborícola ( Dendrohyrax arboreus) y el mono samango (Cercopithecus mitis labiatus) en 199 parches forestales. Su rareza se atribuye a la fragmentación y destrucción de su hábitat forestal. Las funciones de incidencia, derivadas de datos de presencia y ausencia, fueron formuladas como modelos lineales generalizados, y los efectos ambientales fueron incluidos en los modelos logísticos ajustados. Los pequeños y mayormente solitarios hyrax y duiker persistieron en parches más pequeños que los monos, que son más grandes y más sociables. A pesar de que este resultado obedece a expectativas basadas en tamaños de rango de hogar relativos de cada especie, la probabilidad de incidencia del mono samango no cambió con un incremento en el aislamiento, mientras que una disminución gradual al crecer el aislamiento se observó en hyrax y duiker. Las dinámicas de grupos podrían inhibir la dispersión e incrementar el efecto de aislamiento en especies sociables como lo es el mono samango. Un modelo de metapoblación continente-isla describe adecuadamente los patrones de la ocupación de parches por hyrax y duiker; sin embargo, la pobre capacidad de dispersión de los monos y la obvia extirpación área-dependente sugiere que estos existen en metapoblaciones transitorias, desequilibradas (en disminución). Mediante la identificación de parches forestales grandes para la protección y manejo cuidadosos, la supervivencia de las tres especies ( pero especialmente la de los monos) podría ser prolongada. Sin embargo, debido a que no existen metapoblaciones funcionales de monos, esta es una medida de emergencia. Para el duiker y el hyrax, los parches grandes deberán formar parte de una red de parches más pequeños y más cercanos en una matriz natural. [source]

    Differences in spatial predictions among species distribution modeling methods vary with species traits and environmental predictors

    ECOGRAPHY, Issue 6 2009
    Alexandra D. Syphard
    Prediction maps produced by species distribution models (SDMs) influence decision-making in resource management or designation of land in conservation planning. Many studies have compared the prediction accuracy of different SDM modeling methods, but few have quantified the similarity among prediction maps. There has also been little systematic exploration of how the relative importance of different predictor variables varies among model types and affects map similarity. Our objective was to expand the evaluation of SDM performance for 45 plant species in southern California to better understand how map predictions vary among model types, and to explain what factors may affect spatial correspondence, including the selection and relative importance of different environmental variables. Four types of models were tested. Correlation among maps was highest between generalized linear models (GLMs) and generalized additive models (GAMs) and lowest between classification trees and GAMs or GLMs. Correlation between Random Forests (RFs) and GAMs was the same as between RFs and classification trees. Spatial correspondence among maps was influenced the most by model prediction accuracy (AUC) and species prevalence; map correspondence was highest when accuracy was high and prevalence was intermediate (average prevalence for all species was 0.124). Species functional type and the selection of climate variables also influenced map correspondence. For most (but not all) species, climate variables were more important than terrain or soil in predicting their distributions. Environmental variable selection varied according to modeling method, but the largest differences were between RFs and GLMs or GAMs. Although prediction accuracy was equal for GLMs, GAMs, and RFs, the differences in spatial predictions suggest that it may be important to evaluate the results of more than one model to estimate the range of spatial uncertainty before making planning decisions based on map outputs. This may be particularly important if models have low accuracy or if species prevalence is not intermediate. [source]

    Species richness patterns and metapopulation processes , evidence from epiphyte communities in boreo-nemoral forests

    ECOGRAPHY, Issue 2 2006
    Swantje Löbel
    For several epiphyte species, dispersal limitation and metapopulation dynamics have been suggested. We studied the relative importance of local environmental conditions and spatial aggregation of species richness of facultative and obligate epiphytic bryophytes and lichens within two old-growth forests in eastern Sweden. The effect of the local environment was analyzed using generalized linear models (GLM). We tested whether species richness was spatially structured by fitting variogram models to the residuals of the GLM. In addition, we analyzed the species-area relationship (area=tree diameter). Different environmental variables explained the richness of different species groups (bryophytes vs lichens, specialists vs generalists, sexual vs asexual dispersal). In most groups, the total variation explained by environmental variables was higher than the variation explained by the spatial model. Spatial aggregation was more pronounced in asexually than in sexually dispersed species. Bryophyte species richness was only poorly predicted by area, and lichen species richness was not explained by area at all. Spatial aggregation may indicate effects of dispersal limitation and metapopulation dynamics on community species richness. Our results suggest that species groups differ in habitat requirements and dispersal abilities; there were indications that presence of species with different dispersal strategies is linked to the age of the host tree. Separate analyses of the species richness of species groups that differ in the degree of habitat specialization and dispersal ability give insights into the processes determining community species richness. The poor species-area relationship, especially in lichens, may indicate species turnover rather than accumulation during the lifetime of the host tree. Epiphyte species extinctions may be mainly caused by deterministic processes, e.g. changes in habitat conditions as the host tree grows, ages and dies, rather than by stochastic population processes. [source]

    Comparative effects of pH and Vision® herbicide on two life stages of four anuran amphibian species,

    Andrea N. Edginton
    Abstract Vision®, a glyphosate-based herbicide containing a 15% (weight:weight) polyethoxylated tallow amine surfactant blend, and the concurrent factor of pH were tested to determine their interactive effects on early life-stage anurans. Ninety-six-hour laboratory static renewal studies, using the embryonic and larval life stages (Gosner 25) of Rana clamitans, R. pipiens, Bufo americanus, and Xenopus laevis, were performed under a central composite rotatable design. Mortality and the prevalence of malformations were modeled using generalized linear models with a profile deviance approach for obtaining confidence intervals. There was a significant (p < 0.05) interaction of pH with Vision concentration in all eight models, such that the toxicity of Vision was amplified by elevated pH. The surfactant is the major toxic component of Vision and is hypothesized, in this study, to be the source of the pH interaction. Larvae of B. americanus and R. clamitans were 1.5 to 3.8 times more sensitive than their corresponding embryos, whereas X. laevis and R. pipiens larvae were 6.8 to 8.9 times more sensitive. At pH values above 7.5, the Vision concentrations expected to kill 50% of the test larvae in 96-h (96-h lethal concentration [LC50]) were predicted to be below the expected environmental concentration (EEC) as calculated by Canadian regulatory authorities. The EEC value represents a worst-case scenario for aerial Vision application and is calculated assuming an application of the maximum label rate (2.1 kg acid equivalents [a.e.]/ha) into a pond 15 cm in depth. The EEC of 1.4 mg a.e./L (4.5 mg/L Vision) was not exceeded by 96-h LC50 values for the embryo test. The larvae of the four species were comparable in sensitivity. Field studies should be completed using the more sensitive larval life stage to test for Vision toxicity at actual environmental concentrations. [source]

    Response of macroinvertebrates to copper and zinc in a stream mesocosm

    Christopher W Hickey
    Abstract Metal pollution of streams and rivers is recognized as one of the major concerns for management of freshwaters. Macroinvertebrate communities were established within 12 artificial streams and exposed to three replicated concentrations of a metals mixture (copper and zinc) for 34 d. The cumulative criterion units (CCU = ,[metals]/hardness-adjusted U.S. Environmental Protection Agency [U.S. EPA] 1996 chronic criterion value) of total metals in the low, medium, and high treatments were 2.4, 5.9, and 18 CCUs. Zinc comprised approximately 75% of the CCUs in each of the treatments. Effects on taxa richness and the number of taxa in the orders Ephemeroptera, Plecoptera, and Trichoptera (EPT) were moderate at the high exposure concentration (,23% and ,26% respectively, p < 0.05). All of the five major mayfly species showed near extinction, whereas four of the seven caddisflies showed stimulation (up to +121%) and three were reduced (up to ,76%). Redundancy analysis for this metal gradient indicated that 94% of the variance in community structure was explained by three quantitative variables: total mayfly abundance, a mollusk (Potamopyrgus antipodarum) abundance, and the number of EPT individuals, indicating that multiple indices do provide improved predictors of metal stress. Most species showed a threshold response relationship, whereas some community indicators showed apparent hormetic responses (e.g., number of mayfly taxa, total taxa, and number of EPT taxa). Model concentration-response relationships with generalized linear models were used to provide threshold of 20% effective concentration values for species and community metrics. Threshold effect values ranged upwards of 1.4 CCUs, indicating that U.S.EPA chronic criteria would be protective of species and community responses. [source]

    On the use of generalized linear models for interpreting climate variability

    ENVIRONMETRICS, Issue 7 2005
    Richard E. Chandler
    Abstract Many topical questions in climate research can be reduced to either of two related problems: understanding how various different components of the climate system affect each other, and quantifying changes in the system. This article aims to justify the addition of generalized linear models to the climatologist's toolkit, by demonstrating that they offer an intuitive and flexible approach to such problems. In particular, we provide some suggestions as to how ,typical' climatological data structures may be represented within the GLM framework. Recurring themes include methods for space,time data and the need to cope with large datasets. The ideas are illustrated using a dataset of monthly U.S. temperatures. Copyright © 2005 John Wiley & Sons, Ltd. [source]