Iyer, V.; Prasad, A.

What constitutes a single species is ultimately arbitrary, resulting in idiosyncrasies in taxonomic praxis determining species counts. One consequence of this arbitrariness is taxonomic synonymization, wherein variation determined not unique enough to constitute distinct species gets merged into one. Synonymization is non-random and subject to taxon and trait-specific biases. We explore these trends using a comprehensive dataset of the world\'s marine species for all large phyla and Animalia. We match all accepted species-names from the present day with their synonyms and calculate metrics of synonymization including the time lag between the years of descriptions of a species and its newest synonym. We find that although species description trends across phyla show no consistent, discernible trends, the fraction of names that gets synonymised remains initially constant across phyla, with a subsequent gradual decrease. Most accepted species described around the early 20th century possess synonyms created in the same year, indicating synonyms arising due to new taxonomic combinations, suggesting a distinct period of taxonomic praxis. This period was followed by a period of most species descriptions lacking synonyms on moving towards the present. Across taxa, we find the mean lag to be strongly negatively correlated to time, with time not being a limiting factor determining this trend. We also find the number of synonyms per accepted name to also be strongly negatively correlated to time in a taxon-independent manner. Our study considers when the diversity characterising species in the form of once-unique species-names becomes known to science. Our results suggest that across taxa, species described more recently encompass less biological variability, indicating a gradual change in how morphologically delineated species which comprise the vast majority of species described across time are described.