Practice makes the expert: The importance of training volunteers in the generation of phenological data from photographs of biodiversity observation platforms

grouped in globose verticils; but it can be difficult to distinguish when an apparently empty bracteole corresponds to the early stages of a flower bud and when it corresponds to an imma- ture fruit. For that reason, we suppressed the phenophases "immature fruit" and "mature fruit" and only used "dry fruit", which in both species is easily identifiable by its change in coloration from green to brown. The illustrated guide was used to train a subgroup of volunteers, hereafter referred to as trained volunteers. The volunteers were asked to classify the photographs, with sessions lasting approximately 30 minutes per species. During each session, the chosen vegetative and repro- ductive structures were described and the structures to be scored clearly explained. At the end of each session, the group of experts scored a sample of 30 records of the select species with the help of the volunteers. This random sample corresponded to 3% of the total records for each species. Each trained volunteer was provided with the illustrated guide for later consultation and a set of post-training photographs that they classified without assistance to identify possi- ble errors or omissions during the training session. The training session was complemented by an additional session dedicated to questions by volunteers. The subgroup of volunteers with- out training did not receive any information nor material. On the Naturalista platform, when a user registers an observation, it may consist of one or more photographs with the same georeference and date assigned (Fig 2). For the proposes of

this paper, the unit of observation was the record, not the individual photographs. The data with the total number of observations of each species available for Mexico (up to July 19, 2020) were downloaded from the Naturalista website (www.Naturalista.mx), Leonotis nepetifolia (1,234 records) and Nicotiana glauca (1,583 records), and randomly assigned into 29 sub- groups of similar size, hereafter databases. Fig 1. Flowchart followed to generate the phenology of species from records of iNaturalist. https://doi.org/10.1371/journal.pone.0282750.g001 PLOS ONE The training of volunteers as a promoter of phenological studies PLOS ONE | https://doi.org/10.1371/journal.pone.0282750 March 7, 2023 5 / 17