Assessing the effect of sample bias correction in species distribution models

Dubos, Nicolas; Préau, Clémentine; Lenormand, Maxime; Papuga, Guillaume; Montsarrat, Sophie; Denelle, Pierre; Louarn, Marine Le; Heremans, Stien; Roel, May; Roche, Philip; Luque, Sandra

doi:10.1016/j.ecolind.2022.109487

Cited by 25 publications

(22 citation statements)

References 72 publications

(109 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The number and types of environmental variables, sampling effort, sampling bias, and amount and extent of species occurrence data affect prediction probabilities [ 11 ], and MaxEnt relatively can handle these issues more adequately than the other modelling tools. Species distribution models (SDMs) can predict high occurrence probability for individual species at some geographic space/places [ 7 , 13 ], which depict an existence of more optimal environmental conditions, and can be called as a suitable habitat [ 14 ].…”

Section: Introductionmentioning

confidence: 99%

Species distribution modelling of Monotheca buxifolia (Falc.) A. DC.: Present distribution and impacts of potential climate change

Ali¹,

Khan²,

Khan³

et al. 2023

Heliyon

View full text Add to dashboard Cite

Section: Introductionmentioning

confidence: 99%

Species distribution modelling of Monotheca buxifolia (Falc.) A. DC.: Present distribution and impacts of potential climate change

Ali¹,

Khan²,

Khan³

et al. 2023

Heliyon

View full text Add to dashboard Cite

“…Moreover, species records are often constrained by the variety of their sources and spatial biases caused by unequal sampling efforts 103 and by uneven field accessibility 104 . The adequacy of sample bias correction methods remains uncertain 105 and field validation is still considered to be the best standard practice to assess models’ reliability 106 . Unfortunately, field validation is sometimes impossible to use, especially in geographically extensive scale studies 105 .…”

Section: Discussionmentioning

confidence: 99%

“…The adequacy of sample bias correction methods remains uncertain 105 and field validation is still considered to be the best standard practice to assess models’ reliability 106 . Unfortunately, field validation is sometimes impossible to use, especially in geographically extensive scale studies 105 . In this study some geographical regions also seemed to be inadequately sampled and for that reason the spatial filtering on various scales of topographic heterogeneity was conducted to reduce sampling bias.…”

Section: Discussionmentioning

confidence: 99%

Loss of fungal symbionts and changes in pollinator availability caused by climate change will affect the distribution and survival chances of myco-heterotrophic orchid species

Kolanowska

2023

Sci Rep

View full text Add to dashboard Cite

The first comprehensive species distribution models for orchid, its fungal symbionts and pollinator are presented. To evaluate impact of global warming on these organisms three different projections and four various climate change scenarios were analysed. The niche modelling was based on presence-only records of Limodorum abortivum, two species of Russula and three insects pollinating orchid (Anthophora affinis, Bombus terrestris, Rhodanthidium septemdentatum). Two sets of orchid predictions were examined—the first one included only climatic data and the second one was based on climate data and data on future distribution of orchid fungal symbionts. Overall, a poleward range shift is predicted to occur as a result of climate change and apparently global warming will be favorable for L. abortivum and its potential geographical range will expand. However, due to the negative effect of global warming on fungal symbionts of L. abortivum, the actual extension of the suitable niches of the orchid will be much limited. Considering future possibility of cross-pollination, the availability of A. affinis for L. abortivum will decrease and this bee will be available in the worst case scenarios only for 21% of orchid populations. On the other hand, the overlap of orchid and the buff-tailed bumblebee will increase and as much as 86.5% of plant populations will be located within B. terrestris potential range. Also the availability of R. septemdentatum will be higher than currently observed in almost all analysed climate change projections. This study showed the importance of inclusion of ecological factors in species distribution models as the climate data itself are not enough to estimate the future distribution of plant species. Moreover, the availability of pollen vectors which is crucial for long-term survival of orchid populations should be analysed in context of climate changes.

show abstract

“…The data for performance evaluation was partitioned spatially, using 5‐folds for block‐cross validation (generated with the blockCV R package, (Valavi et al, 2019; Figure S1). The effect of sample bias correction (non‐random pseudo‐absence generation) was quantified using the relative overlap index (Dubos et al, 2021). This index informs how sample bias corrections affect spatial predictions (mean Schoener's D between uncorrected and corrected individual models) relative to inter‐model variability (mean overlap between all pairwise combinations of model replicates).…”

Section: Methodsmentioning

confidence: 99%

Protection gaps in Amazon floodplains will increase with climate change: Insight from the world's largest scaled freshwater fish

Dubos

Lenormand

Castello

et al. 2022

Aquatic Conservation

View full text Add to dashboard Cite

1. The Amazon floodplains represent important surfaces of highly valuable ecosystems, yet they remain neglected from protected areas. Although the efficiency of the protected area network of the Amazon basin may be jeopardized by climate change, floodplains are exposed to important consequences of climate change but are omitted from species distribution models and protection gap analyses.2. The present and future (2070) distribution of the giant bony-tongue fish Arapaima spp. (Arapaimidae) was modelled accounting for climate and habitat requirements, and with a consideration of dam presence (already existing and planned constructions) and hydroperiod (high-and low-water stages). The amount of suitable environment that falls inside and outside the current network of protected areas was quantified to identify spatial conservation gaps.3. We predict that climate change will cause a decline in environmental suitability by 16.6% during the high-water stage, and by 19.4% during the low-water stage. About 70% of the suitable environments of Arapaima spp. remain currently unprotected. The gap is higher by 0.7% during the low-water stage. The lack of protection is likely to increase by 5% with future climate change effects. Both existing and projected dam constructions may hamper population flows between the central, Bolivian and Peruvian parts of the basin. 4. We highlight protection gaps mostly in the south-western part of the basin and recommend the extension of the current network of protected areas in the floodplains of the upper Ucayali, Juruà and Purus rivers and their tributaries. This study has shown the importance of integrating hydroperiod and dispersal barriers in forecasting the distribution of freshwater fish species, and stresses the urgent need to integrate floodplains within the protected area networks.

show abstract

Assessing the effect of sample bias correction in species distribution models

Cited by 25 publications

References 72 publications

Species distribution modelling of Monotheca buxifolia (Falc.) A. DC.: Present distribution and impacts of potential climate change

Species distribution modelling of Monotheca buxifolia (Falc.) A. DC.: Present distribution and impacts of potential climate change

Loss of fungal symbionts and changes in pollinator availability caused by climate change will affect the distribution and survival chances of myco-heterotrophic orchid species

Protection gaps in Amazon floodplains will increase with climate change: Insight from the world's largest scaled freshwater fish

Contact Info

Product

Resources

About