全球变化知识资源中心

The continuing advancement of new technologies is making available novel sources of occurrence data, such as photo-sharing platforms. Using three Iberian Argiope spider species as models, this study aimed to compare the quality of three data sources - existing scientific literature, a biodiversity-focused photo-sharing citizen science project, and a photo-sharing social network - and the results derived from niche and distribution analyses. Data sources were compared in terms of sample size, spatial clustering, and coverage, degree of climatic niche completeness, and in their consistency in the characterisation of response curves. Possible biases towards big cities and protected natural areas were identified, and species distribution models were parameterised and compared. Photo records provided the largest number of occurrences, whereas data from scientific literature showed the worst spatial coverage. Records from the citizen science project yielded the highest degree of niche completeness. All data sources showed some bias towards big cities and protected areas. Data from existing scientific literature yielded the worst range representations and, accordingly, the most different distribution models, whereas the models based on two photo-vouchered sources were very similar. Range and niche analyses using records from different sources may yield disparate results. Occurrence data from existing scientific literature may not be enough to develop robust niche analyses nor to characterise recent range changes of poorly studied taxa. Photo-sharing platforms provide updated chorological information and have the potential to increase sample sizes dramatically, which is an important data property in distribution and niche modelling.