Abstract
Estimates of extant liverwort species range from 4,500 to 9,000, with estimates in the past decade converging on 5,000 to 6,000. Potential problems and pitfalls of deriving species estimates are addressed, including binomial accumulation, the impact of synonymy, taxonomic inflation, the impact of unrevised species-rich genera, species concepts and cryptic species. We present a revised mean estimate of 7,500 for the number of liverwort species based on estimating rates of synonymy in a sample of recently monographed and revised taxa. This estimate does not include infraspecific names and may underestimate global diversity as a result. We also present a databased estimate of about 8,500 species derived from the Early Land Plants Today data set. We argue higher estimates are supported by: 1) the number of published species has not reached a plateau and new species continue to be discovered; 2) not all regions have been thoroughly explored and with equal intensity, with survey effort being historically biased toward northern temperate regions; 3) synonymy rates are not uniform across taxonomic groups; 4) novel discovery of species outpaces new species derived from elevation in rank (taxonomic inflation); and 5) species numbers are not necessarily distorted by large unrevised genera. A standardized global worldwide liverwort checklist with strong community participation coupled with the critical need for ongoing monographs and revisions, will aid in arriving at a clearer estimates of liverwort diversity. We promote and encourage interest in the topic using an evidence-based approach and quantitative data.
