Fleroff goes digital: georeferenced records from "Flora des Gouvernements Wladimir" (Fleroff, 1902)

Global Biodiversity Information Facility (GBIF) has uneven data coverage across taxonomic, spatial and temporal dimensions. Temporal imbalances in the data coverage are particularly dramatic. Thus, 188.3M GBIF records were made in 2020, more than the whole lot of the currently available pre-1986 electronic data. This underscores the importance of reliable and precise biodiversity spatial data collected in early times. Biological collections certainly play a key role in our knowledge of biodiversity in the past. However, digitisation of historical literature is underway, being a modern trend in biodiversity data mining. The grid dataset for the flora of Vladimir Oblast, Russia, includes many historical records borrowed from the "Flora des Gouvernements Wladimir" by Alexander F. Fleroff (also known as Flerov or Flerow). Intensive study of Fleroff's collections and field surveys exactly in the same localities where he worked, showed that the quality of his data is superb. Species lists collected across hundreds of localities form a unique source of reliable information on the floristic diversity of Vladimir Oblast and adjacent areas for the period from 1894 to 1901. Since the grid dataset holds generalised data, we made precise georeferencing of Fleroff's literature records and published them in the form of a GBIF-mediated dataset.

A dataset, based on "Flora des Gouvernements Wladimir. I. Pflanzengeographische Beschreibung des Gouvernements Wladimir" by Fleroff (1902), includes 8,889 records of 654 taxa (mainly species) from 366 localities. The majority of records originate from Vladimir Oblast (4,611 records of 534 taxa from 195 localities) and Yaroslavl Oblast (2,013 records of 409 taxa from 66 localities), but also from Nizhny Novgorod Oblast (942 records), Ivanovo Oblast (667 records) and Moscow Oblast (656 records). The leading second-level administrative units by the number of records are Pereslavsky District (2,013 records), Aleksandrovsky District (1,318 records) and Sergievo-Posadsky District (599 records). Georeferencing was carried out, based on the expert knowledge of the area, analysis of modern satellite images and old topographic maps. For 2,460 records, the georeferencing accuracy is 1,000 m or less (28%), whereas for 6,070 records it is 2,000 m or less (68%). The mean accuracy of records of the entire dataset is 2,447 m. That accuracy is unattainable for most herbarium collections of the late 19^th century. Some localities of rare plants discovered by Fleroff and included into the dataset were completely lost in the 20^th century due to either peat mining or development of urban areas.