Within the reported software, we identified 93 software tools used in metabarcoding bioinformatic pipelines (Table S3), of which 27% (25) were software ‘packages’. When taking into account distinct functions within packages, a total of 169 unique tools were recorded, however, this is likely an inaccurate picture given low reporting rates of functions used within software packages across all steps. There is a clear increase in the number of different software and software functions employed across all papers over time (Fig. 5a). Examining the diversity of software used within tasks over time, controlling for the number of papers published, there is limited improvement in homogeneity and a decrease in dominance of software (Figs 5c and 5d). Given that the number of metabarcoding publications is increasing year-on-year, there is thus a concomitant increase in the diversity of software used for a given task, and previously commonly used software are being used less (Figs 5c and 5d). These trends reflect that while new software tools are constantly being made available for metabarcoding, uptake is not consistent across the field and while some researchers use more recent tools, many researchers continue to use older methods, diversifying the field.