Species distribution models can be highly sensitive to algorithm configuration

Hallgren, Willow; Santana, Fabiana Soares; Low‐Choy, Samantha; Zhao, Yanguang; Mackey, Brendan

doi:10.1016/j.ecolmodel.2019.108719

Cited by 73 publications

(50 citation statements)

References 34 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

Section: Discussionmentioning

confidence: 99%

“…To simplify the interpretation of results and for a better comparison between taxa, we decided to keep the default settings of the algorithms, which may have resulted in uncertainty regarding model performance and resulting species distribution due to the differential sensitivity of each algorithm to species modelled, sampling effort and evaluation metric (Hallgren et al., 2019). We recommend investigating the sensitivity of the algorithms to their configuration settings on the resulting projected species distribution (Hallgren et al., 2019). In addition, we relied on pseudo‐absences instead of true absences and considered a greater number of randomly selected pseudo‐absence points for those taxa with a larger study region, which may have introduced uncertainty given the differential sensitivity of each algorithm to the number of pseudo‐absences (Barbet‐Massin et al., 2012; Guillera‐Arroita et al., 2015; Phillips et al., 2009).…”

Section: Discussionmentioning

confidence: 99%

See 1 more Smart Citation

Predicting range shifts of African apes under global change scenarios

Carvalho

Graham

Bocksberger

et al. 2021

Diversity and Distributions

View full text Add to dashboard Cite

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Discussionmentioning

confidence: 99%

Predicting range shifts of African apes under global change scenarios

Carvalho

Graham

Bocksberger

et al. 2021

Diversity and Distributions

View full text Add to dashboard Cite

show abstract

“…Maximum entropy (MaxEnt) [13] and support vector machines (SVMs) [14,15] are flexible and very powerful techniques. MaxEnt is a machine learning algorithm with a high capability in artificial fitting rules or functional connections (e.g., nonlinear relation) according to appearance information, usage of species' presence, and background data for the prediction of species distribution and habitat suitability [8,[16][17][18]. The MaxEnt algorithm is applied to detect the maximum entropy distribution likelihood and is used to forecast the possible distribution of a target species according to its maximum entropy under different conditions.…”

Section: Introductionmentioning

confidence: 99%

Predicting Habitat Suitability and Conserving Juniperus spp. Habitat Using SVM and Maximum Entropy Machine Learning Techniques

Boogar

Sher

Pourghasemi

et al. 2019

Water

View full text Add to dashboard Cite

Support vector machine (SVM) and maximum entropy (MaxEnt) machine learning techniques are well suited to model the habitat suitability of species. In this study, SVM and MaxEnt models were developed to predict the habitat suitability of Juniperus spp. in the Southern Zagros Mountains of Iran. In recent decades, drought extension and climate alteration have led to extensive changes in the geographical occurrence of this species and its growth and regeneration are extremely limited in this area. This study evaluated the habitat suitability of Juniperus through spatial modeling and predicts appropriate regions for future cultivation and resource conservation. We modeled the natural habitat of Juniperus for an area of 700 ha in Sepidan Area in the Fars province using (1) data regarding the presence of the species (295 samples) collected through field surveys and GPS, (2) habitat soil information and indices derived from 60 soil samples collected in the study area, and (3) climatic and topographic datasets collected from various sources. In total, 15 conditioning factors were used for this spatial modeling approach. Receiver operator characteristic (ROC) curves were applied to estimate the accuracy of the habitat suitability models produced by the SVM and MaxEnt techniques. Results indicated logical and similar area under the curve (AUC)-ROC values for the SVM (0.735) and MaxEnt (0.728) models. Both the SVM and MaxEnt methods revealed a significant relationship between the Juniperus spp. distribution and conditioning factors. Environmental factors played a vital role in evaluating the presence of Juniperus sp. as Max and Min temperatures and annual mean rainfall were the three most important factors for habitat suitability in the study area. Finally, an area with high and very high suitability for the future cultivation of Juniperus sp. and for landscape conservation was suggested based on the SVM model.

show abstract

“…Since the performance of the Maxent models is known to be notably influenced by the background point selection, environmental variable selection, feature types, and regularization parameters ( 75 , 98 – 100 ), we tested different alternatives regarding them. For the selection of background points, we tested two options: we either (i) used the 10,000 points randomly selected in the Swiss territory or (ii) used only the random points situated below 1,500 m in altitude, where tick occurrence is more likely.…”

Section: Methodsmentioning

confidence: 99%

Nested Species Distribution Models of Chlamydiales in Ixodes ricinus (Tick) Hosts in Switzerland

Rochat

Vuilleumier

Aeby

et al. 2020

Appl Environ Microbiol

View full text Add to dashboard Cite

The tick Ixodes ricinus is the vector of various pathogens, including Chlamydiales bacteria, potentially causing respiratory infections. In this study, we modelled the spatial distribution of I. ricinus and associated Chlamydiales over Switzerland from 2009 to 2019. We used a total of 2293 ticks and 186 Chlamydiales occurrences provided by a Swiss Army field campaign, a collaborative smartphone application and a prospective campaign. For each tick location, we retrieved from Swiss federal datasets the environmental factors reflecting the topography, climate and land cover. We then used the Maxent modelling technique to estimate the suitability for I. ricinus and to subsequently build the nested niche of Chlamydiales bacteria. Results indicate that I. ricinus high habitat suitability is determined by higher temperature and vegetation index (NDVI) values, lower temperature during driest months and a higher percentage of artificial and forests areas. The performance of the model was increased when extracting the environmental variables for a 100 m-radius buffer around the sampling points and when considering the climatic conditions of the two years previous to sampling date. For Chlamydiales bacteria, the suitability was favoured by lower percentage of artificial surfaces, driest conditions, high precipitation during coldest months and short distances to wetlands. From 2009 to 2018, we observed an extension of tick and Chlamydiales suitable areas, associated with a shift towards higher altitude. The importance to consider spatio-temporal variations of the environmental conditions for obtaining better prediction was also demonstrated. Importance Ixodes ricinus is the vector of pathogens, including the agent of Lyme disease, the tick borne encephalitis virus and the less known Chlamydiales bacteria at the origin of some respiratory infections. In this study, we identified the environmental factors influencing the presence of I. ricinus and Chlamydiales in Switzerland and generated maps of their distribution from 2009 to 2018. We found an important expansion of suitable areas for both the tick and the bacteria during the last decade. Results provided also the environmental factors that determine the presence of Chlamydiales within ticks. Distribution maps as generated here are expected to bring valuable informations for decision-makers to control tick-borne diseases in Switzerland and establish prevention campaigns. The methodological framework presented could be used to predict the distribution and spread of other host-pathogen couples, to identify environmental factors driving their distribution and to develop control or prevention strategies accordingly.

show abstract

Species distribution models can be highly sensitive to algorithm configuration

Cited by 73 publications

References 34 publications

Predicting range shifts of African apes under global change scenarios

Predicting range shifts of African apes under global change scenarios

Predicting Habitat Suitability and Conserving Juniperus spp. Habitat Using SVM and Maximum Entropy Machine Learning Techniques

Nested Species Distribution Models of Chlamydiales in Ixodes ricinus (Tick) Hosts in Switzerland

Contact Info

Product

Resources

About