The study addresses the issue of overestimating microbial diversity in environmental DNA surveys due to sequencing errors and inadequate clustering algorithms. The authors propose a new clustering method that combines single-linkage preclustering followed by average-linkage clustering based on pairwise alignments. This method significantly reduces the number of predicted operational taxonomic units (OTUs) compared to traditional methods, while maintaining the accuracy of rare biosphere detection. The new approach reduces OTU richness by 30-60% in environmental samples without altering the fraction of rare OTUs in long-tailed rank abundance curves. The study demonstrates the effectiveness of this method through simulations and real-world data analysis, showing that it can accurately predict expected OTUs in both single and pooled template preparations of known taxonomic composition.The study addresses the issue of overestimating microbial diversity in environmental DNA surveys due to sequencing errors and inadequate clustering algorithms. The authors propose a new clustering method that combines single-linkage preclustering followed by average-linkage clustering based on pairwise alignments. This method significantly reduces the number of predicted operational taxonomic units (OTUs) compared to traditional methods, while maintaining the accuracy of rare biosphere detection. The new approach reduces OTU richness by 30-60% in environmental samples without altering the fraction of rare OTUs in long-tailed rank abundance curves. The study demonstrates the effectiveness of this method through simulations and real-world data analysis, showing that it can accurately predict expected OTUs in both single and pooled template preparations of known taxonomic composition.