Modeling Forest Species Distributions in a Human-Dominated Landscape in Northeastern, USA (original) (raw)
Related papers
We describe our collective efforts to develop and apply methods for using FIA data to model forest resources and wildlife habitat. Our work demonstrates how flexible regression techniques, such as generalized additive models, can be linked with spatially explicit environmental information for the mapping of forest type and structure. We illustrate how these maps of forest structure can be used to model wildlife habitat, focusing on the prediction of suitable habitat for cavity-nesting birds in forest systems in the Intermountain West.
Landscape Ecology, 2005
Current methods of vegetation analysis often assume species response to environmental gradients is homogeneously monotonic and unimodal. Such an approach can lead to unsatisfactory results, particularly when vegetation pattern is governed by compensatory relationships that yield similar outcomes for various environmental settings. In this paper we investigate the advantages of using classification tree models (CART) to test specific hypotheses of environmental variables effecting dominant vegetation pattern in the Piedmont. This method is free of distributional assumptions and is useful for data structures that contain non-linear relationships and higher-order interactions. We also compare the predictive accuracy of CART models with a parametric generalized linear model (GLM) to determine the relative strength of each predictive approach. For each method, hardwood and pine vegetation is modeled using explanatory topographic and edaphic variables selected based on historic reconstructions of patterns of land use. These include soil quality, potential soil moisture, topographic position, and slope angle. Predictive accuracy was assessed on independent validation data sets. The CART models produced the more accurate predictions, while also emphasizing alternative environmental settings for each vegetation type. For example, relic hardwood stands were found on steep slopes, highly plastic soils, or hydric bottomlands -alternatives not well captured by the homogeneous GLM. Our results illustrate the potential utility of this flexible modeling approach in capturing the heterogeneous patterns typical of many ecological datasets.
Additions of landscape metrics improve predictions of occurrence of species distribution models
Species distribution models are used to aid our understanding of the processes driving the spatial patterns of species' habitats. This approach has received criticism, however, largely because it neglects landscape metrics. We examined the relative impacts of landscape predictors on the accuracy of habitat models by constructing distribution models at regional scales incorporating environmental variables (climate, topography, vegetation, and soil types) and secondary species occurrence data, and using them to predict the occurrence of 36 species in 15 forest fragments where we conducted rapid surveys. We then selected six landscape predictors at the landscape scale and ran general linear models of species presence/absence with either a single scale predictor (the probabilities of occurrence of the distribution models or landscape variables) or multiple scale predictors (distribution models ? one landscape variable). Our results indicated that distribution models alone had poor predictive abilities but were improved when landscape pre-dictors were added; the species responses were not, however, similar to the multiple scale predictors. Our study thus highlights the importance of considering landscape metrics to generate more accurate habitat suitability models.
Environmental Management, 2009
To achieve the overall objective of restoring natural environment and sustainable resource usability, each forest management practice effect needs to be predicted using a simulation model. Previous simulation efforts were typically confined to public land. Comprehensive forest management practices entail incorporating interactions between public and private land. To make inclusion of private land into management planning feasible at the regional scale, this study uses a new method of combining Forest Inventory and Analysis (FIA) data with remotely sensed forest group data to retrieve detailed species composition and age information for the Missouri Ozark Highlands. Remote sensed forest group and land form data inferred from topography were integrated to produce distinct combinations (ecotypes). Forest types and size classes were assigned to ecotypes based on their proportions in the FIA data. Then tree species and tree age determined from FIA subplots stratified by forest type and size class were assigned to pixels for the entire study area. The resulting species composition map can improve simulation model performance in that it has spatially explicit and continuous information of dominant and associated species, and tree ages that are unavailable from either satellite imagery or forest inventory data. In addition, the resulting species map revealed that public land and private land in Ozark Highlands differ in species composition and stand size. Shortleaf pine is a co-dominant species in public land, whereas it becomes a minor species in private land. Public forest is older than private forest. Both public and private forests have deviated from historical forest condition in terms of species composition. Based on possible reasons causing the deviation discussed in this study, corresponding management avenues that can assist in restoring natural environment were recommended.
2009
Remote sensing provides critical information for broad scale assessments of wildlife habitat distribution and conservation. However, such efforts have been typically unable to incorporate information about vegetation structure, a variable important for explaining the distribution of many wildlife species. We evaluated the consequences of incorporating remotely sensed information about horizontal vegetation structure into current assessments of wildlife habitat distribution and conservation. For this, we integrated the new NLCD tree canopy cover product into the US GAP Analysis database, using avian species and the finished Idaho GAP Analysis as a case study. We found: (1) a 15-68% decrease in the extent of the predicted habitat for avian species associated with specific tree canopy conditions, (2) a marked decrease in the species richness values predicted at the Landsat pixel scale, but not at coarser scales, (3) a modified distribution of biodiversity hotspots, and (4) surprising results in conservation assessment: despite the strong changes in the species predicted habitats, their distribution in relation to the reserves network remained the same. This study highlights the value of area wide vegetation structure data for refined biodiversity and conservation analyses. We discuss further opportunities and limitations for the use of the NLCD data in wildlife habitat studies.
Studies in Computational Intelligence, 2008
Sustainable management efforts are currently hindered by a lack of basic information about the spatial distribution of species on large landscapes. Based on complex ecological databases, computationally advanced species distribution models can provide great progress for solving this ecological problem. However, current lack of knowledge about the ecological relationships that drive species distributions reduces the capacity for classical statistical approaches to produce accurate predictive maps. Advancements in machine learning, like classification and bagging algorithms, provide a powerful tool for quickly building accurate predictive models of species distributions even when little ecological knowledge is readily available. Such approaches are also well known for their robustness when dealing with large data sets that have low quality. Here, we use Random Forests (Salford System's Ltd. and R language), a highly accurate bagging classification algorithm originally developed by L. Breiman and A. Cutler, to build multi-species avian distribution models using data collected as part of the Kenai National Wildlife Refuge Long-term Ecological Monitoring Program (LTEMP). Distribution maps are a useful monitoring metric because they can be used to document range expansions or contractions and can also be linked to population estimates. We utilize variable radius point count data collected in 2004 and 2006 at 255 points arranged in a 4.8 km resolution, systematic grid spanning the 7722 km 2 spatial extent of Alaska's Kenai National Wildlife Refuge. We build distribution models for 41 bird species that are present within 200m of 2-56% of the sampling points resulting in models that represent species which are both rare and common on the landscape. All models were built using 2 Dawn R. Magness, Falk Huettmann, and John M. Morton a common set of 157 environmental predictor variables representing topographical features, climatic space, vegetation, anthropogenic variables, spatial structure, and 5 randomly generated neutral landscape variables for quality assessment. Models with that many predictors have not been used before in avian modeling, but are commonly used in similar types of applications in commercial disciplines. Random Forests produced strong models (ROC >0.8) for 16 bird species, marginal models (0.7 >ROC <0.8) for 13 species, and weak models (ROC <0.7) for 11 species. The ability of Random Forests to provide accurate predictive models was independent of how common or rare a bird was on the landscape. Random Forests did not rank any of the 5 neutral landscape variables as important for any of the 41 bird species. We argue that for inventory and monitoring programs the interpretive focus and confidence in reliability should be placed in the predictive ability of the map, and not in the assumed ecological meaning of the predictors or their linear relationships to the response variable. Given this focus, computer learning algorithms would provide a very powerful, cost-saving approach for building reliable predictions of species occurrence on the landscape given the current lack of knowledge on the ecological drivers for many species. Land management agencies need reliable predictions of current species distributions in order to detect and understand how climate change and other landscape drivers will affect future biodiversity.
Study on selecting sensitive environmental variables in modelling species spatial distribution
Annals of GIS, 2016
This study explores the effects of different environmental variables on the accuracy of species distribution models. Forest inventory and analysis data sets were used to generate absence and pseudo-absence points of chestnut oak (Quercus prinus) in the central and southern Appalachian mountain region of the US. We simulate chestnut oak distribution using different criteria for selecting environmental variables: (1) the selection of sensitive variables using factor analysis and the calculation of a sensitivity index, (2) principal components analysis. Factor analysis to environmental variables at both occurrence and pseudo-absence points was conducted to calculate the sensitivity index for each environmental variable. The identification of sensitive variables may use the factor loadings of first one or two factors of environmental variables. Modelling with sensitive variables (mean Kappa > 0.60; mean true skill statistic (TSS) > 0.60) can enhance model accuracy more than using PCA variables or all available environmental variables (mean Kappa ranges from 0.45 to 0.65; mean TSS ranges from 0.40 to 0.70). Modelling with leading principal components (larger than 90% variations) can achieve similar or higher accuracy than modelling with all variables. The influence of redundant information on species modelling varies with the model used. Our results suggest that selecting environmental variables using a sensitivity index defined by factor analysis may improve model accuracy and reduce redundant information in species modelling. The proposed method for selecting sensitive variables is easy to implement and has strong ecological interpretability.
Modelling the spatial distribution of tree species with fragmented populations from abundance data
Community Ecology, 2009
Spatial distribution modelling can be a useful tool for elaborating conservation strategies for tree species characterized by fragmented and sparse populations. We tested five statistical models-Support Vector Regression (SVR), Multivariate Adaptive Regression Splines (MARS), Gaussian processes with radial basis kernel functions (GP), Regression Tree Analysis (RTA) and Random Forests (RF)-for their predictive performances. To perform the evaluation, we applied these techniques to three tree species for which conservation measures should be elaborated and implemented: one Mediterranean species (Quercus suber) and two temperate species (Ilex aquifolium and Taxus baccata). Model evaluation was measured by MSE, Goodman-Kruskal and sensitivity statistics and map outputs based on the minimal predicted area criterion. All the models performed well, confirming the validity of this approach when dealing with species characterized by narrow and specialized niches and when adequate data (more than 40-50 samples) and environmental and climatic variables, recognized as important determinants of plant distribution patterns, are available. Based on the evaluation processes, RF resulted the most accurate algorithm thanks to bootstrap-resampling, trees averaging, randomization of predictors and smoother response surface.
Forest landscape models (FLMs) are important tools for testing basic ecological theory and for exploring forest changes at landscape and regional scales. However, the ability of these models to accurately predict changes in tree species' distributions and their spatial pattern may be significantly affected by the formulation of site-scale processes that simulate gap-level succession including seedling establishment, tree growth, competition, and mortality. Thus, the objective of this study is to evaluate the effects of site-scale processes on landscape-scale predictions of tree species' distributions and spatial patterns.