Aim: Freshwater mussels are considered among the most at-risk taxa in the world. As such, comprehensive monitoring assessments of what abiotic and biotic factors influence mussel occupancy will be vital for guiding effective conservation. Here, we analyzed vertebrate and mussel eDNA metabarcoding data to explore the influence of biotic (i.e., host fish diversity, predator presence, and community composition) and abiotic (i.e., drainage size, forest cover, stream order) factors on freshwater mussel populations.

Location: This study utilized water samples and tactile survey data collected from streams throughout Fort Johnson, Louisiana.

Methods: We first evaluated the effectiveness of environmental DNA (eDNA) metabarcoding for characterizing freshwater communities, based on previous conventional tactile surveys. Next, we used eDNA metabarcoding analysis for freshwater mussels and vertebrate species alongside remote sensing data to within an occupancy modeling framework to assess how various biotic and abiotic variables impact freshwater mussel eDNA occupancy.

Results: Our eDNA metabarcoding survey largely agreed with both historical and contemporary surveys on Fort Johnson, while uniquely detecting Pleurobema riddellii, a proposed threatened species under the US Endangered Species Act. We also found that eDNA detections and occupancy had strong seasonal variation, with increased read abundance and diversity in the spring. Vertebrate, fish, and predator diversity, as a function of habitat quality, were strongly predictive of mussel occupancy, supporting the concept of land managers focusing on the entire ecosystem for mussel conservation. Lastly, we found that percent forest cover and drainage basin size interact to influence mussel eDNA occupancy, informing habitat preferences for mussel species of interest (i.e., the mussels preferred larger drainage sizes and perennial streams).

Conclusions: Our results demonstrate that combining eDNA metabarcoding of target and non-target species with occupancy modeling can provide insights into the ecology of freshwater mussels and is a useful tool to improve their conservation and management.

Two separate eDNA metabarcoding libraries, one for freshwater mussels and one for vertebrates, were prepared using the same eDNA samples. For each library, a two-step PCR protocol was used following the Illumina 16S metagenomic sequencing library preparation guidelines, modified to target either a 180–200 bp fragment of mussel 16S ribosomal ribonucleic acid (rRNA) (Coghlan et al., 2021) or an 85–­117 bp fragment of vertebrate 12S rRNA (Riaz et al., 2011; Table 1). For the first PCR (PCR1), reactions were performed in triplicate for all samples, negative controls (sterile, molecular grade water), and positive controls consisting of Inflated heelsplitter (Potamilus inflatus (Lea, 1831); 0.05 ng/μL) and European glass lizard (Pseudopus apodus (Pallas, 1775); 0.1 ng/μL) genomic DNA for the mussel and vertebrate libraries respectfully. Each 25 μL PCR1 reaction for the mussel metabarcoding library consisted of 3 μL of DNA template, 12.5 μL of Q5® High-Fidelity 2X Master Mix (NEB), 0.5 μL of 50 mg/ml bovine serum albumin, 6 μL of sterile, molecular grade water, and 1.5 μL of each 10 μM primer. The vertebrate library followed the same formula except each reaction used 5 μL of DNA template and 4 μL of sterile, molecular grade water. Thermocycling conditions of the mussel metabarcoding library for PCR1 included a 98°C incubation step for 5 min, followed by 35 cycles of 98°C for 10 s, 55°C for 1 min, and 72°C for 20 s, with a final extension at 72°C for 2 min. Thermocycling conditions of the vertebrate metabarcoding library for PCR1 included a 98°C incubation step for 5 min, followed by 35 cycles of 98°C for 10 s, 58°C for 30 s, and 72°C for 30 s, with a final extension at 72°C for 7 min. Triplicate reactions for each sample were pooled and amplification was verified via gel electrophoresis. Pooled PCR1 products were size-selected and cleaned with AMPure XP beads (Beckman Coulter) using a 0.7× and 0.75× bead:sample ratio (right side) for the mussel and vertebrate samples, respectively, followed by a 1.2× bead:sample ratio (left side) for both libraries, and confirmed via gel electrophoresis. Unique dual indexes (10 bp) and Illumina sequencing adapters were then annealed to the amplified products using IDT for Illumina UD Indexes (Table 1) with a subsequent PCR step (PCR2).

PCR2 was performed in duplicate for each sample, including negative and positive controls. PCR2 reaction and cycling protocols were identical for both libraries with each 25 μL PCR2 reaction consisting of 2 μL of PCR1 template, 12.5 μL of Q5® High-Fidelity 2X Master Mix (New England Biolabs), 5.5 μL of sterile, molecular grade water, and 2 μL of each 10 μM IDT-Illumina UD index primer. Thermocycling conditions for PCR2 started with a 98°C incubation step for 3 min, followed by 10 cycles of 98°C for 30 s, 55°C for 30 s, and 72°C for 30 s, with a final extension at 72°C for 5 min. Duplicate reactions for each sample were pooled and amplification was verified via gel electrophoresis. Pooled PCR2 products were size-selected and cleaned with AMPure XP beads (Beckman Coulter) using a 1.0× and 1.1× bead:sample ratio (left side only) for mussel and vertebrate libraries, respectively, and confirmed via gel electrophoresis. PCR2 samples were normalized based on DNA concentrations  quantified via Qubit dsDNA HS kit (Invitrogen) and pooled into sub-libraries by PCR plate. AMPure XP bead cleanups using 0.9× and 1.0× bead:sample ratio (left side only) for mussel and vertebrate libraries, respectively, were performed again on each sub-library to concentrate DNA and ensure primer removal prior to sequencing. Sub-libraries were then normalized, pooled, and submitted for sequencing at the W. M. Keck Center (University of Illinois, Urbana, IL) on the Illumina NovaSeq platform using the SP flow-cell with 2 × 250 bp paired-end reads, with an output of approximately 750 million reads.