Testing the performance of environmental DNA metabarcoding for surveying highly diverse tropical fish communities: A case study from Lake Tanganyika

Abstract

Abstract Background and Aims Environmental DNA (eDNA) metabarcoding provides a highly sensitive method of surveying freshwater fish communities, although studies to date have largely been restricted to temperate ecosystems. Due to limited reference sequence availability and challenges identifying closely related and rare species in diverse tropical ecosystems, the effectiveness of metabarcoding methods for surveying tropical fish communities from eDNA samples remains uncertain. To address this, we applied an eDNA metabarcoding approach to survey Lake Tanganyika's (LT) species‐rich littoral fish communities. Materials and Methods As this system contains many closely related species, particularly cichlid fishes, we used four primer sets including a cichlid‐specific primer set (Cichlid_CR). A reference database was built for the 12s, 16s, and control region for 358 fish species including over 93% of known cichlids. Results and Discussion In silico and in situ results demonstrated wide variability in the taxonomic resolution of assignments by each primer with the cichlid‐specific marker (Cichlid_CR) enabling greater species‐level assignments for this highly diverse family. A greater number of non‐cichlid teleost species were detected at sites compared to the visual survey data. For cichlid species however, sequencing depth substantially influenced species richness estimates obtained from eDNA samples, with increased depths producing estimates comparable to that obtained from the visual survey data. Conclusions Our study highlights the importance of sequencing depth and local reference databases when undertaking metabarcoding studies within diverse ecosystems, as well as demonstrating the potential of eDNA metabarcoding for surveying diverse tropical fish communities, even those containing closely related species within evolutionary radiations.

Keywords

• cichlids
• environmental DNA
• Lake Tanganyika
• metabarcoding
• reference database
• tropical freshwaters