To clean or not to clean: Cleaning open‐source data improves extinction risk assessments for threatened plant species

Panter, Connor T.; Clegg, Rosemary; Moat, Justin; Bachman, Steven P.; Klitgård, Bente B.; White, Rachel L.

doi:10.1111/csp2.311

Cited by 22 publications

(31 citation statements)

References 51 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Our assessment’s accuracy and sensitivity (80.7% and 98.0%, respectively) is in line with global and regional estimates [ 37 , 53 , 54 , 65 ] and can be thus considered reliable and robust, as it agrees with the findings of [ 39 , 40 ], regarding the taxa previously assessed (n = 238). We found that the vast majority of the Greek endemics are considered as threatened ( Figure 3 ), which is in harmony with the previous and now outdated assessments conducted in Greece (Greek endemic taxa assessed: n = 238; ca.…”

Section: Discussionsupporting

confidence: 88%

“…After Greek endemic taxa were assigned to an IUCN category, we estimated the total number of taxa recorded and the proportion of taxa assessed under each IUCN category, under Criteria A and B separately, and by combining both criteria, i.e., a Greek endemic taxon would, for example, be categorized as CR (CR END ) if it is assessed as CR by at least one of the two criteria. We should note that this approach does not substitute full Red List assessments, but in the absence of such official, time-consuming and resource-intensive assessments [ 27 ], it provides a fast, robust and reliable alternative using the two most commonly applied IUCN criteria [ 22 , 56 , 57 , 64 ] (besides, Criteria C and D can be considered as special cases of Criteria A and B: the sub-criteria for Criterion C rely on population reduction as in Criterion A and Criterion D2 relies on the estimated AOO as in Criterion B [ 57 ]) following the IUCN guidelines, that may serve as a baseline for more in-depth conservation assessments in the future [ 28 , 53 , 54 ], while also contributing towards Aichi Biodiversity Target 2 and assisting in effective, evidence-based conservation decision making [ 65 ].…”

Section: Methodsmentioning

confidence: 99%

See 1 more Smart Citation

Extinction Risk Assessment of the Greek Endemic Flora

Kougioumoutzis

Kokkoris

Panitsa

et al. 2021

Biology

View full text Add to dashboard Cite

Human-induced biodiversity decline has been on the rise for the past 250 years, due to various causes. What is equally troubling, is that we are unaware which plants are threatened and where they occur. Thus, we are far from reaching Aichi Biodiversity Target 2, i.e., assessing the extinction risk of most species. To that end, based on an extensive occurrence dataset, we performed an extinction risk assessment according to the IUCN Criteria A and B for all the endemic plant taxa occurring in Greece, one of the most biodiverse countries in Europe, in a phylogenetically-informed framework and identified the areas needing conservation prioritization. Several of the Greek endemics are threatened with extinction and fourteen endemics need to be prioritized, as they are evolutionary distinct and globally endangered. Mt. Gramos is identified as the most important conservation hotspot in Greece. However, a significant portion of the identified conservation hotspots is not included in any designated Greek protected area, meaning that the Greek protected areas network might need to be at least partially redesigned. In the Anthropocene era, where climate and land-use change are projected to alter biodiversity patterns and may force many species to extinction, our assessment provides the baseline for future conservation research, ecosystem services maintenance, and might prove crucial for the timely, systematic and effective aversion of plant extinctions in Greece.

show abstract

Section: Discussionsupporting

confidence: 88%

Section: Methodsmentioning

confidence: 99%

Extinction Risk Assessment of the Greek Endemic Flora

Kougioumoutzis

Kokkoris

Panitsa

et al. 2021

Biology

View full text Add to dashboard Cite

show abstract

“…Given the well-documented issues with digitally available occurrence records (Maldonado et al, 2015;Meyer et al, 2016;Zizka et al, 2019), it may seem reasonable to assume that AA methods will always perform better with carefully cleaned and georeferenced data. For example, Panter et al (2020) obtained more reliable preliminary assessments for species in Bolivia after manually cleaning GBIF data.…”

Section: How Clean Must Occurrence Data Be?mentioning

confidence: 99%

“…How clean must occurrence data be? Issue with the quality of occurrence record data from online databases are well known (Meyer et al, 2016;Panter et al, 2020;Paton et al, 2020). Species occurrence records are therefore thoroughly checked and geo-referenced during Red List assessments, requiring significant time investment.…”

Section: Introductionmentioning

confidence: 99%

Evidence-based guidelines for automated conservation assessments of plant species

Walker¹,

Leão²,

Bachman³

et al. 2021

Preprint

View full text Add to dashboard Cite

Assessing species’ risk of extinction is a vital first step in setting conservation priorities. However, assessment endeavours like the IUCN Red List of Threatened Species still have significant gaps in their coverage of some taxonomic groups. Automated assessment (AA) methods are gaining popularity to rapidly fill these gaps, taking advantage of improvements in computing and digitally available information. However, implicit choices made when developing and reporting automated assessment methods could prevent their successful adoption or, even worse, their predictions could lead to poor allocation of conservation resources.We systematically explored how the choice of data cleaning, taxonomic group, training sample, and automation method affected predicted threat status. We used occurrence records from GBIF to generate assessments for three distinct taxonomic groups using four different automated assessment methods. We measured each method’s performance and coverage after applying increasingly stringent cleaning to the input occurrence data. We used these results to build evidence-based guidelines for developing and reporting automated assessments.Automatically cleaned data from GBIF resulted in comparable performance to occurrence records cleaned manually by an expert. However, all types of data cleaning removed species and limited the coverage of automated assessments. This limitation was more severe for some groups of species than others. Overall, machine learning-based methods performed well on all taxonomic groups, even with minimal data cleaning.We recommend using a machine learning-based method on minimally cleaned data to get the best compromise between performance and species coverage. However, our results demonstrate that the optimal data cleaning, training sample, and automation method depends on the focal group of species. Therefore, we recommend evaluating new AA methods across multiple groups and providing additional context with extinction risk predictions for users to make informed decisions.

show abstract

“…These data inaccuracies result in wrong geographical information in species records, and are common problems in GBIF and other biological databases (Maldonado et al., 2015; Zizka et al., 2019). If not addressed, such inaccuracies can distort results of analyses such as species distribution modelling (SDM; Sporbert et al., 2019) or richness assessments (Walther & Moore, 2005) and conservation assessments (Panter et al., 2020; Zizka et al., 2021).…”

Section: Introductionmentioning

confidence: 99%