Background: This paper is an attempt to estimate the percentage of erroneously identified taxa in ethnographic studies concerning the use of plants and to propose a code for recording credibility of identification in historical ethnobotany publications.
Methods: A sample of Polish-language ethnobotanical literature (45 published sources from 1874-2005) and four collections of voucher specimens (from 1894-1975) were analyzed. Errors were detected in the publications by comparing the data with existing knowledge on the distribution of plant names and species ranges. The voucher specimens were re-examined.A one-letter code was invented for quick identification of the credibility of data published in lists of species compiled from historical or ethnographic sources, according to the source of identification: voucher specimen, Latin binominal, botanical expert, obvious widespread name, folk name, mode of use, range, physical description or photograph. To test the use of the code an up-to-date list of wild food plants used in Poland was made.
Results: A significant difference between the ratio of mistakes in the voucher specimen collections and the ratio of detectable mistakes in the studies without herbarium documentation was found. At least 2.3% of taxa in the publications were identified erroneously (mean rate was 6.2% per publication), and in half of these mistakes even the genus was not correct. As many as 10.0% of voucher specimens (on average 9.2% per collection) were originally erroneously identified, but three quarters of the identification mistakes remained within-genus.The species of the genera Thymus, Rumex and Rubus were most often confused within the genus.Not all of the invented credibility codes were used in the list of wild food plants, but they may be useful for other researchers. The most often used codes were the ones signifying identification by: voucher specimen, botanical expert and by a common name used throughout the country.
Conclusions: The results of this study support the rigorous use of voucher specimens in ethnobotany, although they also reveal a relatively high percentage of misidentified taxa in the specimens studied.The invented credibility coding system may become a useful tool for communication between historical ethnobotanists, particularly in creating larger databases.