Species distribution modeling with Gaussian processes: A case study with the youngest stages of sea spawning whitefish (Coregonus lavaretus L. s.l.) larvae

Vanhatalo, Jarno; Veneranta, Lari; Hudd, Richard

doi:10.1016/j.ecolmodel.2011.12.025

Cited by 28 publications

(44 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Even though this approach allows flexible modeling by describing the randomness of response variables with different probabilistic models (Nelder and Wedderburn, 1972), it is still limited by its parametric assumptions. Hence, it may fail to accurately describe a species' response to environmental conditions (Vanhatalo et al, 2012;Kotta et al, 2019). Here, we propose a semi-parametric JSDM model represented with multivariate Gaussian processes (GPs).…”

Section: Introductionmentioning

confidence: 99%

“…GPs are flexible semi-parametric regression models where the regression function is estimated without restrictive parametric assumptions about its form (O'Hagan, 1978;. Our model integrates the main strengths of semi-parametric single species models (Vanhatalo et al, 2012;Golding and Purse, 2016) and generalized linear model based JSDMs. First, we model the species responses to environmental covariates with additive multivariate GPs.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Additive Multivariate Gaussian Processes for Joint Species Distribution Modeling with Heterogeneous Data

Vanhatalo¹,

Hartmann²,

Veneranta³

2020

Bayesian Anal.

Self Cite

View full text Add to dashboard Cite

Species distribution models (SDM) are a key tool in ecology, conservation and management of natural resources. Two key components of the state-ofthe-art SDMs are the description for species distribution response along environmental covariates and the spatial random effect that captures deviations from the distribution patterns explained by environmental covariates. Joint species distribution models (JSDMs) additionally include interspecific correlations which have been shown to improve their descriptive and predictive performance compared to single species models. However, current JSDMs are restricted to hierarchical generalized linear modeling framework. Their limitation is that parametric models have trouble in explaining changes in abundance due, for example, highly nonlinear physical tolerance limits which is particularly important when predicting species distribution in new areas or under scenarios of environmental change. On the other hand, semi-parametric response functions have been shown to improve the predictive performance of SDMs in these tasks in single species models.Here, we propose JSDMs where the responses to environmental covariates are modeled with additive multivariate Gaussian processes coded as linear models of coregionalization. These allow inference for wide range of functional forms and interspecific correlations between the responses. We propose also an efficient approach for inference with Laplace approximation and parameterization of the interspecific covariance matrices on the euclidean space. We demonstrate the benefits of our model with two small scale examples and one real world case study. We use cross-validation to compare the proposed model to analogous semi-parametric single species models and parametric single and joint species models in interpolation and extrapolation tasks. The proposed model outperforms the alternative models in all cases. We also show that the proposed model can be seen as an extension of the current state-of-the-art JSDMs to semi-parametric models.MSC 2010 subject classifications: Primary 60G15, 60K35; secondary 62P12.

show abstract

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Additive Multivariate Gaussian Processes for Joint Species Distribution Modeling with Heterogeneous Data

Vanhatalo¹,

Hartmann²,

Veneranta³

2020

Bayesian Anal.

Self Cite

View full text Add to dashboard Cite

show abstract

“…Gaussian processes (GPs, also referred to as Gaussian random fields) provide a flexible approach to fitting complex statistical models (Rasmussen & Williams 2006) and offer solutions to many of the issues related to SDMs. GPs have seen occasional use in ecology for modelling population dynamics (Patil 2007;Sigourney, Munch & Letcher 2012) and have recently been proposed as an alternative approach for SDM (Vanhatalo, Veneranta & Hudd 2012). Whilst GP models have so far been applied to presence/absence data for SDM, GP models can also be fitted with likelihoods applicable for count and presence-only data (Diggle et al 2013).…”

Section: Introductionmentioning

confidence: 99%

Fast and flexible Bayesian species distribution modelling using Gaussian processes

2016

View full text Add to dashboard Cite

Summary1. Species distribution modelling (SDM) is widely used in ecology, and predictions of species distributions inform both policy and ecological debates. Therefore, methods with high predictive accuracy and those that enable biological interpretation are preferable. Gaussian processes (GPs) are a highly flexible approach to statistical modelling and have recently been proposed for SDM. GP models fit smooth, but potentially complex response functions that can account for high-dimensional interactions between predictors. We propose fitting GP SDMs using deterministic numerical approximations, rather than Markov chain Monte Carlo methods in order to make GPs more computationally efficient and easy to use. 2. We introduce GP models and their application to SDM, illustrate how ecological knowledge can be incorporated into GP SDMs via Bayesian priors and formulate a simple GP SDM that can be fitted efficiently. This model can be fitted either by learning the hyperparameters or by using a fixed approximation to them. Using a subset of the North American Breeding Bird Survey data set, we compare the out-of-sample predictive accuracy of these models with several commonly used SDM approaches for both presence/absence and presence-only data. 3. Predictive accuracy of GP SDMs fitted by Laplace approximation was greater than boosted regression trees, generalized additive models (GAMs) and logistic regression when trained on presence/absence data and greater than all of these models plus MaxEnt when trained on presence-only data. GP SDMs fitted using a fixed approximation to hyperparameters were no less accurate than those with MAP estimation and on average 70 times faster, equivalent in speed to GAMs. 4. As well as having strong predictive power for this data set, GP SDMs offer a convenient method for incorporating prior knowledge of the species' ecology. By fitting these methods using efficient numerical approximations, they may easily be applied to large data sets and automatically for many species. An R package, GRaF, is provided to enable SDM users to fit GP models.

show abstract

“…GPs are a family of stochastic processes, which define probability distribution over functions. They are a flexible tool for modelling dependency between observations in space, time and covariate space (Golding, Purse, & Warton, 2016;Rasmussen & Williams, 2006;Vanhatalo, Veneranta, & Hudd, 2012;Vanhatalo et al, 2013). A GP is defined by its mean and covariance function.…”

Section: Discussionmentioning

confidence: 99%

Hierarchical Bayesian model reveals the distributional shifts of Arctic marine mammals

Mäkinen

Vanhatalo

2018

Diversity and Distributions

Self Cite

View full text Add to dashboard Cite

Aim Our aim involved developing a method to analyse spatiotemporal distributions of Arctic marine mammals (AMMs) using heterogeneous open source data, such as scientific papers and open repositories. Another aim was to quantitatively estimate the effects of environmental covariates on AMMs’ distributions and to analyse whether their distributions have shifted along with environmental changes. Location Arctic shelf area. The Kara Sea. Methods Our literature search focused on survey data regarding polar bears (Ursus maritimus), Atlantic walruses (Odobenus rosmarus rosmarus) and ringed seals (Phoca hispida). We mapped the data on a grid and built a hierarchical Poisson point process model to analyse species’ densities. The heterogeneous data lacked information on survey intensity and we could model only the relative density of each species. We explained relative densities with environmental covariates and random effects reflecting excess spatiotemporal variation and the unknown, varying sampling effort. The relative density of polar bears was explained also by the relative density of seals. Results The most important covariates explaining AMMs’ relative densities were ice concentration and distance to the coast, and regarding polar bears, also the relative density of seals. The results suggest that due to the decrease in the average ice concentration, the relative densities of polar bears and walruses slightly decreased or stayed constant during the 17‐year‐long study period, whereas seals shifted their distribution from the Eastern to the Western Kara Sea. Main conclusions Point process modelling is a robust methodology to estimate distributions from heterogeneous observations, providing spatially explicit information about ecosystems and thus serves advances for conservation efforts in the Arctic. In a simple trophic system, a distribution model of a top predator benefits from utilizing prey species’ distributions compared to a solely environmental model. The decreasing ice cover seems to have led to changes in AMMs’ distributions in the marginal Arctic region.

show abstract

Species distribution modeling with Gaussian processes: A case study with the youngest stages of sea spawning whitefish (Coregonus lavaretus L. s.l.) larvae

Cited by 28 publications

References 17 publications

Additive Multivariate Gaussian Processes for Joint Species Distribution Modeling with Heterogeneous Data

Additive Multivariate Gaussian Processes for Joint Species Distribution Modeling with Heterogeneous Data

Fast and flexible Bayesian species distribution modelling using Gaussian processes

Hierarchical Bayesian model reveals the distributional shifts of Arctic marine mammals

Contact Info

Product

Resources

About