Estimating the spatial pattern of human-caused forest fires using a generalized linear mixed model with spatial autocorrelation in South Korea

Most forest fires in Korea are spatially concentrated in certain areas and are highly related to human activities. These site-specific characteristics of forest fires are analyzed by spatial regression analysis using the R-module generalized linear mixed model (GLMM), which can consider spatial autocorrelation. We examined the quantitative effect of topology, human accessibility, and forest cover without and with spatial autocorrelation. Under the assumption that slope, elevation, aspect, population density, distance from road, and forest cover are related to forest fire occurrence, the explanatory variables of each of these factors were prepared using a Geographic Information System-based process. First, we tried to test the influence of fixed effects on the occurrence of forest fires using a generalized linear model (GLM) with Poisson distribution. In addition, the overdispersion of the response data was also detected, and variogram analysis was performed using the standardized residuals of GLM. Second, GLMM was applied to consider the obvious residual autocorrelation structure. The fitted models were validated and compared using the multiple correlation and root mean square error (RMSE). Results showed that slope, elevation, aspect index, population density, and distance from road were significant factors capable of explaining the forest fire occurrence. Positive spatial autocorrelation was estimated up to a distance of 32 km. The kriging predictions based on GLMM were smoother than those of the GLM. Finally, a forest fire occurrence map was prepared using the results from both models. The fire risk decreases with increasing distance to areas with high population densities, and increasing elevation showed a suppressing effect on fire occurrence. Both variables are in accordance with the significance tests.     

Effect of different substrata on rhizome growth,leaf biometry and shoot density of Posidonia oceanica

The effects of different substratum typologies on Posidonia oceanica growth and morphology were estimated in four Sicilian meadows using Generalized and Linear Mixed Models combined with retrodating and biometric analyses. Substratum exerted a multiple effect, resulting in different biometric features for P. oceanica shoots settled on rock from those growing on sand and matte. On rock, values for growth rate, leaf length and shoot surface were lower than those on other substrata, with 42%, 23% and 32% the highest degree of difference respectively. The present study may have interesting methodological consequences for the comprehensive understanding of the causative variables potentially affecting meadows features and their health status. The importance of substratum in the prediction of likely biometry changes in P. oceanica meadows, means that knowledge of substratum type should receive due attention in the future to derive reliable estimates of meadow status.     

Site-specific prediction of West Nile virus mosquito abundance in Greater Toronto Area using generalized linear mixed models

Mosquito surveillance programs provide a primary means of understanding mosquito vector population dynamics for the risk assessment of human exposure to West Nile virus (WNv). The lack of spatial coverage and missing observations in mosquito surveillance data often challenge our efforts to predict this vector-borne disease and implement control measures. We developed a WNv mosquito abundance prediction model in which local meteorological and environmental data were synthesized with entomological data in a generalized linear mixed modeling framework. The discrete nature of mosquito surveillance data is accommodated by a Poisson distributional assumption, and the site-specific random effects of the generalized linear mixed model (GLMM) capture any fluctuation unexplained by a general trend. The proposed Poisson GLMMs efficiently account for the nested structure of mosquito surveillance data and incorporate the temporal correlation between observations obtained at each trap by a first-order autoregressive model. In the case study, Bayesian inference of the proposed models is illustrated using a subset of mosquito surveillance data in the Greater Toronto Area. The relevance of the proposed GLMM tailored to WNv mosquito surveillance data is highlighted by the comparison of model performance in the presence of inevitable but quantifiable uncertainties.     

Soil seed bank longevity of the exotic annual grass Bromus rubens in the Mojave Desert,USA

Although arid environments are often considered among the least invaded terrestrial biomes, the impacts of exotic plant species can be severe and long lasting. Bromus rubens (red brome) is an exotic annual grass species in the Mojave Desert known to outcompete native plant species, alter habitat, and promote accumulation of fuel that contributes to increasing fire frequency and severity. We assessed longevity of the exotic B. rubens seeds in the soil by burying seeds at four depths (0, 2, 5, and 10 cm) and recovering seeds 6, 12, 18, and 24 months after burial. Seed viability was reduced with greater burial depth and greater time since burial. A relatively small proportion of seeds retained viability for two years, suggesting that while the B. rubens seed bank can be large, it is relatively short-lived. Although B. rubens apparently relies more on the annual production, dispersal, and germination of seeds than on a long-lived seed bank for its annual recruitment, the numerous seeds produced by individual plants indicate that even a small proportion of seeds remaining viable for more than a year can aid recruitment from the seed bank and is an important factor in understanding population dynamics.     

Benefits and drawbacks of two modelling approaches for a generalist carnivore: can models predict where Wile E. Coyote will turn up next?

Wildlife ecologists frequently make use of limited information on locations of a species of interest in combination with readily available GIS data to build models to predict space use. In addition to a wide range of statistical data models that are more commonly used, machine learning approaches provide another means to develop predictive spatial models. However, comparison of output from these two families of models for the same data set is not often carried out. It is important that wildlife managers understand the pitfalls and limitations when a single set of models is used with limited GIS data to try to predict and understand species distribution. To illustrate this, we carried out two sets of models (generalized linear mixed models (GLMMs) and boosted regression trees (BRTs)) to predict geographic occupancy of the eastern coyote (Canis latrans) on the island of Newfoundland, Canada. This exercise is illustrative of common spatial questions in wildlife research and management. Our results show that models vary depending on the approach (GLMM vs. BRT) and that, overall, BRT had higher predictive ability. Although machine learning has been criticized because it is not explicitly hypothesis-driven, it has been used in other areas of spatial modelling with success. Here, we demonstrate that it may be a useful approach for predicting wildlife space use and to generate hypotheses when data are limited. The results of this comparison can help to improve other models for species distributions and also guide future sampling and modelling initiatives.