Influences of temporal independence of data on modelling species distributions

Ko, Chia‐Ying; Ko, Jerome Chie‐Jen; Lin, Ruey‐Shing; Lee, Pei-Fen

doi:10.1016/j.baae.2013.04.003

Cited by 5 publications

(4 citation statements)

References 33 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In particular, the layers can be used in Species Distribution Modelling (SDM) (Peterson et al., 2011) to predict the distribution of species at the global scale (Chefaoui et al., 2015; Fragkopoulou et al., 2022), including non‐native species (Assis et al., 2015), address niche‐based questions (Hu et al., 2021; Lee‐Yaw et al., 2016; Song et al., 2021) and phylogeographic hypotheses (Neiva et al., 2014), identify biodiversity hotspots (Fragkopoulou et al., 2022) and support the conservation and management of marine biodiversity (Boavida et al., 2016; Hobday et al., 2010). In the scope of SDM, by letting users filter historical data into two time periods (decades 2000–2010 and 2010–2020), the current version allows generating independent data for temporal cross‐validation, which can assist in evaluating model performance and prediction error (Ko et al., 2013). Moreover, the development of biologically meaningful variables for future climate change scenarios (e.g.…”

Section: Usage Notesmentioning

confidence: 99%

Bio‐ORACLE v3.0. Pushing marine data layers to the CMIP6 Earth System Models of climate change research

Assis,

Fernández Bejarano,

Salazar

et al. 2024

Global Ecol Biogeogr

View full text Add to dashboard Cite

MotivationImpacts of climate change on marine biodiversity are often projected with species distribution modelling using standardized data layers representing physical, chemical and biological conditions of the global ocean. Yet, the available data layers (1) have not been updated to incorporate data of the Sixth Phase of the Coupled Model Intercomparison Project (CMIP6), which comprise the Shared Socioeconomic Pathway (SSP) scenarios; (2) consider a limited number of Earth System Models (ESMs), and (3) miss important variables expected to influence future biodiversity distributions. These limitations might undermine biodiversity impact assessments, by failing to integrate them within the context of the most up‐to‐date climate change projections, raising the uncertainty in estimates and misinterpreting the exposure of biodiversity to extreme conditions. Here, we provide a significant update of Bio‐ORACLE, extending biologically relevant data layers from present‐day conditions to the end of the 21st century Shared Socioeconomic Pathway scenarios based on a multi‐model ensemble with data from CMIP6. Alongside, we provide R and Python packages for seamless integration in modelling workflows. The data layers aim to enhance the understanding of the potential impacts of climate change on biodiversity and to support well‐informed research, conservation and management.Main Types of Variable ContainedSurface and benthic layers for, chlorophyll‐a, diffuse attenuation coefficient, dissolved iron, dissolved oxygen, nitrate, ocean temperature, pH, phosphate, photosynthetic active radiation, total phytoplankton, total cloud fraction, salinity, silicate, sea‐water direction, sea‐water velocity, topographic slope, topographic aspect, terrain ruggedness index, topographic position index and bathymetry, and surface layers for air temperature, mixed layer depth, sea‐ice cover and sea‐ice thickness.Spatial Location and GrainGlobal at 0.05° resolution.Time Period and GrainDecadal from present‐day to the end of the 21st century (2000–2100).Major Taxa and Level of MeasurementMarine biodiversity associated with surface and epibenthic habitats.Software FormatA package of functions developed for Python and R software.

show abstract

Section: Usage Notesmentioning

confidence: 99%

Bio‐ORACLE v3.0. Pushing marine data layers to the CMIP6 Earth System Models of climate change research

Assis,

Fernández Bejarano,

Salazar

et al. 2024

Global Ecol Biogeogr

View full text Add to dashboard Cite

show abstract

“…Species distribution modeling [9] has become a useful method for predicting amphibian ranges, based on the relationships between species records and environmental variables. However, the state-of-the-art solutions reported by many authors cover very large geographical areas, which are not precise enough to allow us to accurately predict the absence or presence of given amphibian species in a single water reservoir.…”

Section: Current State Of Knowledgementioning

confidence: 99%

Predicting Presence of Amphibian Species Using Features Obtained from GIS and Satellite Images

Blachnik

Sołtysiak

Dąbrowska

2019

IJGI

View full text Add to dashboard Cite

The construction of transport infrastructure is often preceded by an environmentalimpact assessment procedure, which should identify amphibian breeding sites and migrationroutes. However, the assessment is very difficult to conduct because of the large number ofhabitats spread out over a vast expanse, and the limited amount of time available for fieldwork.We propose utilizing local environmental variables that can be gathered remotely using only GISsystems and satellite images together with machine learning methods. In this article, we introducesix new and easily extractable types of environmental features. Most of the features we proposecan be easily obtained from satellite imagery and spatial development plans. The proposed featurespace was evaluated using four machine learning algorithms, namely: a C4.5 decision tree,AdaBoost, random forest and gradient-boosted trees. The obtained results indicated that theproposed feature space facilitated prediction and was comparable to other solutions. Moreover,three of the new proposed features are ranked most important; these are the three dominantproperties of the surroundings of water reservoirs. One of the new features is the percentageaccess from the edges of the reservoir to open areas, but it affects only a few species. Furthermore,our research confirmed that the gradient-boosted trees were the best method for the analyzeddataset.

show abstract

“…Four categories of environmental variables (a total of 14 variables), including topography, climate, vegetation, and human disturbance, with high correlations between species distributions and environments were chosen by Ko et al (2009) and Ko, Ko, Lin, and Lee (2013) and used in the following five different individualspecies distribution models: logistic regression (LR), discriminant analysis (DA), genetic algorithm for rule-based prediction (GARP), artificial neural network (ANN), and maximum entropy (MAXENT). These models are widely used and have been shown to provide effective conservation management strategies based on estimates of the spatial distributions of those species requiring attention (Elith et al, 2006;Meynard & Quinn, 2007;Phillips & Dudik, 2008).…”

Section: Species Distribution Modelsmentioning

confidence: 99%

An assessment of the efficiency of protection status through determinations of biodiversity hotspots based on endemic bird species, Taiwan

Murphy

Root

et al. 2014

Journal for Nature Conservation

Self Cite

View full text Add to dashboard Cite

Influences of temporal independence of data on modelling species distributions

Cited by 5 publications

References 33 publications

Bio‐ORACLE v3.0. Pushing marine data layers to the CMIP6 Earth System Models of climate change research

Bio‐ORACLE v3.0. Pushing marine data layers to the CMIP6 Earth System Models of climate change research

Predicting Presence of Amphibian Species Using Features Obtained from GIS and Satellite Images

An assessment of the efficiency of protection status through determinations of biodiversity hotspots based on endemic bird species, Taiwan

Contact Info

Product

Resources

About