“…In addition, continuous expert supervision ensures that all species are taken into account and limits the tendency of volunteers to focus only on the rarest species (Lepczyk 2005, Lewandowski and Specht 2015, MacKenzie et al 2017, thus avoiding the risk of increasing data imbalance in the Wikiplantbase #Liguria dataset. Similarly to the use of photographs as a tool for recognition and confirmation of identifications, widely used to limit identification errors in several projects (e.g., Puky 2006, Swanson et al 2015, Smith and Davis 2019, apps (e.g., Goëau et al 2013, Kress et al 2018, Nugent 2018) and communities (e.g., Forum Acta Plantarum), during the floristic surveys hundreds of herbarium specimens have been collected (i.e., near the 30% of observed taxa). Despite the use of voucher specimens for the verification of data would require expert time to process (Crall et al 2011), the herbarium samples offer advantages that make them more suitable for our type of data collection than the photographs; for example, some key morphological characters are difficult to be observed in the photos, unless the operator has professional equipment and is adequately instructed on the details necessary for identification (which vary from species to species).…”