Spatial bias in the GBIF database and its effect on modeling species' geographic distributions

Beck, Jan; Böller, Marianne; Erhardt, Andreas; Schwanghart, Wolfgang

doi:10.1016/j.ecoinf.2013.11.002

Cited by 488 publications

(397 citation statements)

References 29 publications

Supporting

Mentioning

388

Contrasting

Unclassified

Order By: Relevance

“…However, we corroborate the results of previous studies (e.g., Yesson et al 2007, Beck et al 2014 in finding that such databases are spatially biased for the AF and CER anurans. We encourage the incorporation and continuous updating of new collections to these open-access resources.…”

Section: Discussionsupporting

confidence: 83%

“…In contrast to these qualitative data derived from GBIF and SpeciesLink, we found that the point-occurrence density data indicate that the data are not randomly distributed across both hotspots, thereby reducing our confidence in our knowledge of the anurans that occur in them. This is not unanticipated because these databases are prone to errors and geographical bias (Yesson et al 2007, Beck et al 2014. The highest density of occurrence records within the AF and CER is in southeastern Brazil (mainly in the state of São Paulo), and doubtless is a result of a recent initiative to advance the biodiversity knowledge of the state of São Paulo, with an increase in anuran surveys (the Biota-FAPESP Program, see Rossa-Feres et al 2011).…”

Section: Discussionmentioning

confidence: 99%

“…Examples of studies using GBIF and SpeciesLink databases can be found in Vasconcelos et al (2012), Beck et al (2014), and references therein. From June to October 2013, we downloaded the point observations species-by-species, based on anuran lists available for the AF (Haddad et al 2013) and CER (Valdujo et al 2012).…”

Section: Methodsmentioning

confidence: 99%

“…However, much of the raw data (i.e., the geographic coordinates) has not been captured electronically or is only available in electronic files of individual herpetological collections. Consequently, these data are not readily accessible and their lack of availability may limit the advance of basic and applied ecological studies (Beck et al 2014 and references therein). The Global Biodiversity Information Facility (GBIF; Yesson et al 2007) and the SpeciesLink (http://splink.…”

Section: Introductionmentioning

confidence: 99%

“…cria.org.br/) projects seek to compile and organize point-occurrence data from biological surveys and biocollections globally and provide open access to this information via an online data search portal. It is not unusual for there to be a spatial bias and/or incomplete dataset for any taxonomic group included in these databases owing to sampling error, and data storage and mobilization (Yesson et al 2007, Beck et al 2014. Nevertheless, data derived from these sources are being used increasingly for purposes such as taxonomic revisions (Pennisi 2000), environmental niche modeling (Vasconcelos et al 2012), compilation of "redlists" of threatened species (Shaffer et al 1998), and biodiversity assessment (Ponder et al 2001, Yesson et al 2007 and references therein).…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

The utility of open-access biodiversity information in representing anurans in the Brazilian Atlantic Forest and Cerrado

Vasconcelos¹,

Nascimento²

2014

Phyllomedusa

View full text Add to dashboard Cite

The utility of open-access biodiversity information in representing anurans in the Brazilian Atlantic Forest and Cerrado. Many geographic coordinates of species are only available in biological collections, and sometimes it is difficult to access these data. However, recent initiatives promise to compile and organize such biodiversity data at a global scale. We evaluated the effectiveness of open-access biodiversity information for forecasting the occurrence of anurans in two Brazilian hotspots -the Atlantic Forest (AF) and the Cerrado (CER). We compiled all available point-occurrence records for anuran species in both hotspots and identified the regions in each of the two hotspots having the highest and lowest number of anuran occurrences based on information from the Global Biodiversity Information Facility (GBIF) and the SpeciesLink project. A total of 13,130 point-occurrence records were recovered for 409 anurans from the AF (~77% of the known species) and 12,729 records for 176 species (~85% of the known species) from the CER. Density of point occurrence data is not randomly distributed within the hotspots. The greatest density of Atlantic Forest records is in southeastern Brazil, and the densest areas in Cerrado occur in the transitional zones to the Atlantic Forest. Comparison of these results with previous studies based on traditional museum information revealed that many important collections of anurans from these hotspots are not included in GBIF and SpeciesLink.

show abstract

Section: Discussionsupporting

confidence: 83%

Section: Discussionmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

The utility of open-access biodiversity information in representing anurans in the Brazilian Atlantic Forest and Cerrado

Vasconcelos¹,

Nascimento²

2014

Phyllomedusa

View full text Add to dashboard Cite

show abstract

Species data for understanding biodiversity dynamics: The what, where and when of species occurrence data collection

Petersen

Speed

Grøtan

et al. 2021

Ecol Sol and Evidence

View full text Add to dashboard Cite

1. The availability and quantity of observational species occurrence records have greatly increased due to technological advancements and the rise of online portals, such as the Global Biodiversity Information Facility (GBIF), coalescing occurrence records from multiple datasets. It is well-established that such records are biased in time, space and taxonomy, but whether these datasets differ in relation to origin have not been assessed. If biases are specific to different types of datasets, and the relative contribution from these datasets have changed over time, these shifting biases will have implications for interpretations of results and, consequentially, for management and conservation measures. 2. We examined observational GBIF records from Norway to test potential differences in taxonomic, time and land-cover biases between 10 different datasets, with a focus on red-listed and non-native species. 3. The datasets differ in their taxonomic coverage, with datasets dominated by citizen scientist recorders focusing greatly on birds. The number of records has increased over time; in particular, citizen science datasets have had a sharp increase in recent years. 4. The different datasets (including division of the datasets by conservation status) showed differences in geographical coverage. Anthropogenic land covers have more records than would be expected by chance in the majority of cases. Remote areas have fewer records than would be expected, underlining the prevalence of a roadside bias. 5. Accounting for biases in opportunistic species occurrence records need to be a dynamic rather than static process, as the taxonomic and geographical biases have changed over time and differ between datasets, depending on origin and inherent characteristics. Data-collection programmes should be designed to counteract the biases of the specific datasets, and methods to account for the biases in existing data should be developed. When utilizing compiled, open-source data, care must be taken to ensure complementarity between the datasets, both regarding time and space. Incorporating strengths and accounting for biases between datasets can strengthen the integration This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.

show abstract

Using macroecological species distribution models to estimate changes in the suitability of sites for threatened species reintroduction

Bellis

Longden

Styles

et al. 2021

Ecol Sol and Evidence

View full text Add to dashboard Cite

1. Species reintroductions often aim to establish populations of threatened taxa over the long-term. However, climate change may jeopardize reintroduction efforts by altering the conditions of a recipient site beyond the tolerances of the focal species. To aid the selection of recipient sites that will retain their suitability under climate change, species distribution models (SDMs) have been proposed as a method of locating areas with a greater likelihood of facilitating species persistence. 2. We applied SDMs to predict macroclimatic suitability changes for 13 threatened plant and invertebrate species considered for reintroduction at four lowland raised bog sites undergoing restoration. We estimated suitability based on current and projected future conditions under two greenhouse gas concentration scenarios-one low (RCP2.6) and one high (RCP8.5)-using three general circulation models, for the period 2041-2060. 3. When considering current predicted suitability, our models indicated that nine species were viable candidates for reintroduction to at least one of the restoration sites. But when accounting for potential future changes in suitability, the number of candidates was reduced to seven species, based on the RCP8.5 climate change scenario. While three of the sites received consistently similar predictions of suitability across species and scenarios, the most northerly site, Red Moss, received divergent suitability predictions for some species. This site is predicted to remain suitable for Metrioptera brachyptera and Genista anglica under at least one scenario despite substantial losses forecast across the rest of their U.K. ranges, suggesting that it could act as a macroclimatic refuge as climate change advances. 4. The findings presented here made a valuable contribution to the reintroduction planning process, by facilitating the prioritization of reintroduction efforts towards species with a greater likelihood of establishing long-term populations at the prospective recipient sites. This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.

show abstract

Spatial bias in the GBIF database and its effect on modeling species' geographic distributions

Cited by 488 publications

References 29 publications

The utility of open-access biodiversity information in representing anurans in the Brazilian Atlantic Forest and Cerrado

The utility of open-access biodiversity information in representing anurans in the Brazilian Atlantic Forest and Cerrado

Species data for understanding biodiversity dynamics: The what, where and when of species occurrence data collection

Using macroecological species distribution models to estimate changes in the suitability of sites for threatened species reintroduction

Contact Info

Product

Resources

About