Microplastic (MP; < 5mm) is ubiquitous in marine environments and is likely transported by biotic benthic-pelagic coupling. Mussels are key benthic-pelagic couplers, concentrating particles from the water column into dense and nutrient rich biodeposits. This study examined how MP affects benthic-pelagic coupling processes of mussels exposed to feeding regimes with and without MP by measuring four attributes of biodeposits: 1) morphology, 2) quantity of algal and MP particles, 3) sinking rate, and 4) resuspension velocity. We found interacting effects of particle treatment and biodeposit type on biodeposit morphology. Biodeposits from the algae treatment contained more algal cells on average than biodeposits from the MP treatment. Biodeposits from the MP treatment sank 34-37% slower and resuspended in 7-22% slower shear velocities than biodeposits from the algae treatment. Decreases in sinking and resuspension velocities of biodeposits containing MP may increase dispersal distances, thus decreasing in-bed nutrient input and increasing nutrient subsidies for other communities.

Pacific blue mussels (Mytilus trossulus; 35 ± 2 mm) were collected from Argyle Lagoon (48.519401, -123.013180) on San Juan Island in Washington State, U.S.A. in August 2019. Byssal threads and epibionts were removed upon collection and mussels were acclimatized at 11-13°C in flow-through seawater tables at Friday Harbor Laboratories (FHL), University of Washington. Two feeding treatments were tested, algae and MP + algae. The algae treatment used Dunaliella spp., grown in culture at FHL, in concentrations ranging 10,000 – 20,000 cells mL^-1 between trials (concentration was consistent within trials; concentration previously shown to not affect CR; Harris and Carrington 2019). The microplastic + algae (or MP) treatment was the same as the algae treatments, but with the addition of fluorescent violet polyethylene spheres 32-38 µm (Item # UVPMS-BV-1.00; Cosphereic; Harris and Carrington 2019). The spheres were soaked in Tween-20, a surfactant that reduces hydrophobicity and clumping, for 24 h prior to experimentation. Microplastic concentrations ranged from 0 – 675 particles mL^-1 (concentration previously shown to not affect CR; Harris and Carrington 2019). Mussels were placed in treatment containers (1 mussel per container with 1 L of aerated FSW) to feed for 1 h. Water samples (1.5 mL) were taken from each container at 0, 30, and 60 minutes to calculate mussel clearance rate. Particle concentrations were quantified with a flow cytometer (Guava C6, EMP Millipore, Hayward, CA) using a RedR vs side scatter plot where the two types of particles fluoresced at different intensities and granularities. 

Biodeposit classification and measurements: Biodeposits from and associated mussel were collected and transferred to a 200 mL beaker of FSW after experimental feeding treatments where the mussel continued to excrete biodeposits for an additional 24 h. Biodeposits were then selected from each mussel for one of three experimental measurements: particle quantification, sinking rate, or resuspension velocity. Selected biodeposits were photographed and measured for length and width using ImageJ and volume was calculated. All biodeposit classifications were based on morphology. 

Particle quantification: Each biodeposit selected for particle quantification was homogenized with a pipette in a 1.5 mL microcentrifuge tube with 0.5 mL FSW. Algal cells (live and whole) and MP particles were counted in each homogenate using a hemocytometer under a compound microscope. 

Sinking rate: Sinking experiments were conducted in a 1 L graduated cylinder filled with FSW at 15°C. Biodeposits were placed a few centimeters below the water surface to avoid complications with surface tension. Biodeposits were initially allowed to sink 10 cm to reach terminal velocity, which was measured as the time to sink an additional 10 cm. 

Resuspension velocity: Mussel biodeposit resuspension velocity was measured in a flume (Rolling Hills Water Tunnel 2436; El Segundo, CA) filled with seawater held at 11-13°C and flow was manipulated by an external computer. Twenty-four biodeposits were placed 6 cm apart from each other in a 4 x 6 grid pattern at the bottom of the flume working section (40 cm x 40 cm x 2 m, width x height x length). Free stream velocity was ramped up to 3 cm s^-1 (shear velocity of 0.3 cm s^-1) for 10 minutes and the biodeposits remaining were recorded. This procedure was repeated at progressively higher velocities, up to 64 cm s^-1 (shear velocity of 6.4 cm s^-1) or until all biodeposits left the grid and were resuspended.