Habitat suitability estimated by niche models is largely unrelated to species abundance

Dallas, Tad; Hastings, Alan

doi:10.1111/geb.12820

Cited by 95 publications

(128 citation statements)

References 41 publications

Supporting

Mentioning

118

Contrasting

Order By: Relevance

“…In this vein, on the basis of ideas suggested by Maguire () and Brown (), it has been proposed that abundance in a site should be related to the position of the environment of the site in niche space, with the highest population densities in those sites with environments closest to the center of the fundamental niche (Yáñez‐Arenas et al , Martínez‐Meyer et al 2013, Ureña‐Aranda et al 2015), where fitness is expected to be highest. This would be the niche distance–abundance (NDA) relationship (Dallas and Hastings ). Niche distance relationships also exist for the genetic structure of populations (Lira‐Noriega and Manthey ).…”

Section: Introductionmentioning

confidence: 99%

On population abundance and niche structure

2019

View full text Add to dashboard Cite

Recent published evidence indicates a negative correlation between density of populations and the distance of their environments to a suitably defined ‘niche centroid’. This empirical observation lacks theoretical grounds. We provide a theoretical underpinning for the empirical relationship between population density and position in niche space, and use this framework to understand the circumstances under which the relationship will fail. We propose a metapopulation model for the area of distribution, as a system of ordinary differential equations coupled with a dispersal kernel. We present an analytical approximation to the solution of the system as well as R code to solve the full model numerically. We use this tool to analyze various scenarios and assumptions. General and realistic demographic assumptions imply a good correlation between position in niche space and population abundance. Factors that modify this correlation are: transitory states, a heterogeneous spatial structure of suitability, and Allee effects. We also explain why the raw output of the niche modeling algorithm MaxEnt is not a good predictor of environmental suitability. Our results elucidate the empirical results for spatial patterns of population size in niche terms, and provide a theoretical basis for a structured theory of the niche.

show abstract

Section: Introductionmentioning

confidence: 99%

On population abundance and niche structure

2019

View full text Add to dashboard Cite

show abstract

“…Nevertheless, abundance-suitability relationships may be complex (Dallas & Hastings, 2018 (Acevedo et al, 2017). The relationships illustrated from this study indicate that wild turkey populations in Mississippi may be limited by low amounts of hardwood forest at local scales.…”

Section: Discussionmentioning

confidence: 83%

Machine learning of large‐scale spatial distributions of wild turkeys with high‐dimensional environmental data

Farrell

Wang

Rush

et al. 2019

Ecology and Evolution

View full text Add to dashboard Cite

Species distribution modeling often involves high‐dimensional environmental data. Large amounts of data and multicollinearity among covariates impose challenges to statistical models in variable selection for reliable inferences of the effects of environmental factors on the spatial distribution of species. Few studies have evaluated and compared the performance of multiple machine learning (ML) models in handling multicollinearity. Here, we assessed the effectiveness of removal of correlated covariates and regularization to cope with multicollinearity in ML models for habitat suitability. Three machine learning algorithms maximum entropy (MaxEnt), random forests (RFs), and support vector machines (SVMs) were applied to the original data (OD) of 27 landscape variables, reduced data (RD) with 14 highly correlated covariates being removed, and 15 principal components (PC) of the OD accounting for 90% of the original variability. The performance of the three ML models was measured with the area under the curve and continuous Boyce index. We collected 663 nonduplicated presence locations of Eastern wild turkeys (Meleagris gallopavo silvestris) across the state of Mississippi, United States. Of the total locations, 453 locations separated by a distance of ≥2 km were used to train the three ML algorithms on the OD, RD, and PC data, respectively. The remaining 210 locations were used to validate the trained ML models to measure ML performance. Three ML models had excellent performance on the RD and PC data. MaxEnt and SVMs had good performance on the OD data, indicating the adequacy of regularization of the default setting for multicollinearity. Weak learning of RFs through bagging appeared to alleviate multicollinearity and resulted in excellent performance on the OD data. Regularization of ML algorithms may help exploratory studies of the effects of environmental factors on the spatial distribution and habitat suitability of wildlife.

show abstract

“…The mapping between ENM-derived habitat suitability and abundance also requires better characterization, as available evidence suggests conflicting relationships (VanDerWal et al 2009, Weber et al 2017, Dallas and Hastings 2018. It is possible that this relationship varies by context and species, and thus should be incorporated into an integrative framework (cf.…”

Section: (B) Pictorial Representations Of Data/modelsmentioning

confidence: 99%

Inference of biogeographic history by formally integrating distinct lines of evidence: genetic, environmental niche and fossil

Hoban

Dawson

Robinson³

et al. 2019

Ecography

View full text Add to dashboard Cite

A primary focus of historical biogeography is to understand changes in species ranges, abundance and genetic connectivity, and changes in community composition. Traditionally, biogeographic inference has relied on distinct lines of evidence, including DNA sequences, fossils and hindcasted ecological niche models. In this review we propose that the development of integrative modeling approaches that leverage multiple distinct data types from diverse disciplines has the potential to revolutionize the field of biogeography. Although each data type contains information on a distinct aspect of species' biogeographic histories, few studies formally integrate multiple types in analysis. For example, post hoc congruence among analyses based on different data types (e.g. fossils and genetics) is commonly assumed to indicate likely biogeographic histories. Unfortunately, analyses of different data often reach discordant conclusions. Thus, fundamental and unresolved debates continue regarding speed and timing of postglacial migration, location and size of glacial refugia, and degree of long distance dispersal. Formal statistical integration can help address these issues. More specifically, formal integration can leverage all available evidence, account for inherent biases associated with different data types, and quantify data and process uncertainty. Novel, quantitative integration of data and models across fields is now possible due to recent advances in cyberinfrastructure, spatial modeling, online and aggregated ecological databases, data processing and quantitative methods. Our purpose is to make the case for and give examples of rigorous integration of genetic, fossil and environmental/ occurrence data for inferring biogeographic history. In particular, we 1) review the need for such a framework; 2) explain common data types and approaches used to infer biogeographic history (and the challenges with each); 3) review state-of-the-art examples of data integration in biogeography; 4) lay out a series of novel, suggested improvements on current methods; and 5) provide an outlook on technical feasibility and future opportunities. Ecography E4 awardGlossary ABC (approximate Bayesian computation; also see Box 1) -a simulation-based statistical method that approximates the likelihood of a model through comparisons of summary statistics calculated from simulated and observed data. Biogeographic history -the study of processes and patterns related to: geographic range shifts, contractions and expansions; demographic fluctuations; locations of refugia; and migration patterns. Bottleneck -a drastic reduction in population size that often also results in a loss of genetic diversity; colonization at a continental scale often involves serial bottlenecks. Coalescent theory -a body of theory that uses probabilistic models to reconstruct the genealogy of a sample backward in time from the sampled lineages (present) to the most recent common ancestor (past). DRM (dynamic range model) -a hierarchical Bayesian approach to estimate paramet...

show abstract

Habitat suitability estimated by niche models is largely unrelated to species abundance

Cited by 95 publications

References 41 publications

On population abundance and niche structure

On population abundance and niche structure

Machine learning of large‐scale spatial distributions of wild turkeys with high‐dimensional environmental data

Inference of biogeographic history by formally integrating distinct lines of evidence: genetic, environmental niche and fossil

Contact Info

Product

Resources

About