The Diatom Diversity and Ecological Status of a Tufa-Depositing River through eDNA Metabarcoding vs. a Morphological Approach—A Case Study of the Una River (Bosnia and Herzegovina)

Tufa deposits in karst rivers are unique habitats created by mutual interactions between specific environmental and biotope features and inhabited by diatoms as a highly abundant and diverse algal group. This pilot study aimed to investigate the diversity of diatom communities on tufa depositing habitats and assess the Una River's ecological status using a comparative molecular and morphological approach for diatom identification. The 312 base pairs of the <i>rbcL</i> gene were barcoded and analyzed using MiSeq reads and amplicon sequence variants (ASVs) obtained by the DADA2 pipeline. The reference database Diat.barcode v7 was used for taxonomic assignment. The morphological identification of the diatoms was carried out in parallel. In total, the combined dataset revealed 46 taxa identified at genus rank, 125 on the subgenus, and 145 on combined taxonomy rank. The metabarcoding approach mostly leads to a lower number of identified taxa at species rank (58 in molecular vs. 119 in optical inventory), resulting in higher values of beta diversity and heterogeneity in diatom assemblages in samples obtained by morphological approach. Despite the high percentage of taxonomically not assigned diatom ASVs to the species rank, high Shannon diversity index values and a similar number of taxa per locations compared to the morphological approach were obtained. Taxa <i>Achnanthidium minutissimum</i> (Kützing) Czarnecki, <i>Achnanthidium pyrenaicum</i> (Hustedt) H.Kobayasi, <i>Amphora pediculus</i> (Kützing) Grunow, <i>Diatoma vulgaris</i> Bory, <i>Navicula cryptotenella</i> Lange-Bertalot, and <i>Navicula tripunctata</i> (O.F.Müller) Bory were identified at all locations in both inventories. Although limited consistency in the diatom abundances between the two inventory datasets was found, a similar grouping of samples was observed connected to the river's longitudinal gradient. The data obtained using molecular approach in most sites indicated a mostly lower ecological status (good or moderate) compared to the data obtained from the morphological approach (high, good, and moderate). The potential of environmental DNA (eDNA) diatom metabarcoding for water monitoring and diversity studies is undeniable, but to fully realize the benefits of these methods in the future, it is essential to standardize protocols and expand the reference database for species found in specific habitats, such as tufa deposits.