Conservation research and policies tend to be significantly restricted wherever relevant data on biodiversity is sparse, scattered or non-curated. Thus, the usefulness of occurrence data, for the study of biodiversity, depends not only on the availability but also on data quality. Notwithstanding the increase in the global availability of primary biodiversity data, they have numerous shortfalls, from incomplete or partially erroneous documentation to spatial and temporal biases (Hortal et al. 2015, Aubry et al. 2017). Also, many non-digitized specimen collections, scientific publications and grey literature are locked as printed or digital publications.
We integrated existing knowledge, from dispersed sources of biodiversity data, namely Global Biodiversity Information Facility (GBIF), natural history collections, wildlife survey reports, species checklist and other scientific literature. This procedure allowed an update of Mozambique’s checklist of terrestrial mammals (Neves et al. 2018). Despite the potential from digital data to overcome gaps of knowledge, a relevant constraint on creating or updating species checklist is the dificulty to access spatially-disperse collections and examine every specimens upon which occurrences are based. To partly overcome this impediment, we developed a species selection process for specimen data from GBIF and museums (Fig. 1). The aim was to categorise the species detected in more than one data source as species with the well-supported occurrence. In addition to the number of collectors, we also accounted for the number of records collected and presented in Smithers and Tello (1976), the last checklist produced for Mozambique’ mammals.
A species-occurrence record was considered well-supported and included into the species checklist when was:
independently recorded by different collectors or
recorded by a single collector but listed in Smithers and Tello (1976).
independently recorded by different collectors or
recorded by a single collector but listed in Smithers and Tello (1976).
An additional list was produced which contained species with questionable occurrence in the country. Species entered this 'questionable occurrence' list when they were:
not listed in Smithers and Tello (1976), and a single record supported its presence in the country;
not listed in Smithers and Tello (1976) and multiple records exist, but were all cited by a single author; or
registered with a single record in Smithers and Tello (1976).
not listed in Smithers and Tello (1976), and a single record supported its presence in the country;
not listed in Smithers and Tello (1976) and multiple records exist, but were all cited by a single author; or
registered with a single record in Smithers and Tello (1976).
We compiled more than 17000 records, resulting in a total of 217 species (14 orders, 39 families and 133 genera) with supported occurrence in Mozambique and 23 species with questionable reported occurrence (Table 1). The proposed approach for species selection can be adapted and function as a powerful tool to update species checklists of countries facing similar lack of knowledge regarding their biodiversity. The capacity to pinpointing species and specimens in need of occurrence and taxonomic re-evaluation is of great value to optimise collection’s study and to boost collaboration between curators and researchers. Lastly, considering that most records integrated are from European and North American institutions, this work would significantly improve with the integration of data from African institutions. Therefore, an effort should be made to make these essential collections accessible online.