Recorder effort and GBIF-calculated habitat suitability affected the 2016-2018 occurrences of JTM in all data sources (Figure 2, Table S4). Unsurprisingly, occurrence records from GBIF were present in areas of high predicted habitat suitability, but GBIF also mostly recorded JTM where recorder effort was high (Figure 2a and 2e). Likewise, occurrence records from iNaturalist and Flickr were present in areas of relatively high recorder effort and GBIF-calculated habitat suitability (Figures 3b and 3f, 3c and 3g). However, occurrence records from Instagram were more likely to be in areas with low GBIF-calculated habitat suitability. While Instagram records were more likely to be in well-recorded areas, the effect of recorder effort was less than for the other data sources (figures 3d and 3h).