DNA barcoding, which uses short DNA sequences to identify species, has been tested with mixed results. A comprehensive study of cypraeid marine gastropods (cowries) found that the overall error rate for species identification was 4%, but this dropped significantly when only traditionally recognized species were analyzed, leading to error rates of up to 17%. The study highlights that DNA barcoding performs poorly in incompletely sampled groups, where species delineation relies on thresholds to differentiate intraspecific variation from interspecific divergence. However, substantial overlap between these variations was found, making thresholds unreliable for identifying closely related species in understudied groups. The study used a large dataset of cowrie sequences, including over 2,000 individuals across 263 taxa, to evaluate barcoding performance. It found that when using evolutionary significant units (ESUs) as the basis for phylogenetic analysis, identification accuracy was high, with error rates as low as 4%. In contrast, traditional species-based analysis led to higher error rates due to increased overlap between intraspecific and interspecific variation. The study also examined the effectiveness of thresholds for species discovery in partially sampled groups. It found that thresholds were not effective in minimizing errors, as they often led to false positives and false negatives. The optimal threshold value for minimizing error in cowries was found to be 2.6%, resulting in a 17% error rate. In turbinids and limpets, thresholds were more effective, with error rates as low as 7% and 0%, respectively. The study concluded that DNA barcoding is most effective in well-sampled, taxonomically well-understood groups. However, its use in understudied groups is limited due to the overlap between intraspecific and interspecific variation. The study emphasized the need for comprehensive, taxonomically thorough databases to ensure accurate species identification and discovery. It also warned against the use of thresholds for species delineation, as they are not reliable in most major clades. Instead, more sophisticated methods incorporating population genetics and sister group attributes are needed for accurate species identification.DNA barcoding, which uses short DNA sequences to identify species, has been tested with mixed results. A comprehensive study of cypraeid marine gastropods (cowries) found that the overall error rate for species identification was 4%, but this dropped significantly when only traditionally recognized species were analyzed, leading to error rates of up to 17%. The study highlights that DNA barcoding performs poorly in incompletely sampled groups, where species delineation relies on thresholds to differentiate intraspecific variation from interspecific divergence. However, substantial overlap between these variations was found, making thresholds unreliable for identifying closely related species in understudied groups. The study used a large dataset of cowrie sequences, including over 2,000 individuals across 263 taxa, to evaluate barcoding performance. It found that when using evolutionary significant units (ESUs) as the basis for phylogenetic analysis, identification accuracy was high, with error rates as low as 4%. In contrast, traditional species-based analysis led to higher error rates due to increased overlap between intraspecific and interspecific variation. The study also examined the effectiveness of thresholds for species discovery in partially sampled groups. It found that thresholds were not effective in minimizing errors, as they often led to false positives and false negatives. The optimal threshold value for minimizing error in cowries was found to be 2.6%, resulting in a 17% error rate. In turbinids and limpets, thresholds were more effective, with error rates as low as 7% and 0%, respectively. The study concluded that DNA barcoding is most effective in well-sampled, taxonomically well-understood groups. However, its use in understudied groups is limited due to the overlap between intraspecific and interspecific variation. The study emphasized the need for comprehensive, taxonomically thorough databases to ensure accurate species identification and discovery. It also warned against the use of thresholds for species delineation, as they are not reliable in most major clades. Instead, more sophisticated methods incorporating population genetics and sister group attributes are needed for accurate species identification.