Mytilus trossulus (Gould 1850) were collected during the winter months of December-February, 2015-2016, from Penn Cove Shellfish, located off the coast of Whidbey Island, Washington, USA. Mussels were kept in 50-liter aquaria with recirculating, 0.2 μm filtered seawater (pH = 8.1, T = 10°C, Sal = 31 PSU) for up to two weeks, and were fed Shellfish Diet 1800 (Reed Mariculture, Campbell,CA) up to 5% of wet tissue mass day-1 at an algal concentration of 2000 cells ml-1.

Mussels were allowed to attach to mica sheets over the course of 4 hours. Byssal threads were cut away from the animal at the proximal region’s interface with the shell. Threads from individuals that made less than three attachments were not included in a treatment group. Mica sheets with plaque attachments were stored dry at room temperature (~21°C, ~30-40% RH) for up to two weeks and then moved into treatment conditions.

Mica sheets with more than three threads attached were incubated in one of six seawater pH treatments (pHNBS target = 1.0, 3.0, 5.0, 7.0, 8.0, 12.0) and allowed to age for either 0.17, 1, 3, 5, 8, 12, or 20 days. Constant pH treatment levels of pH 3.0-8.0 were achieved by bubbling 3 liter containers of filtered seawater with a dynamically controlled mixture of air and CO2 gas (O’Donnell et al. 2013). Seawater pH was monitored in each container with a Durafet III pH electrode (Honeywell, Fort Washington, PA; accuracy ± 0.01), attached to a UDA2182 analyzer that controlled a solenoid valve in line with a CO2 gas canister.

Treatments were constantly bubbled with air to maintain a dissolved oxygen concentration above 8 mg L-1, which was monitored with a Honeywell DirectLine DL5000 equilibrium probe (accuracy ± 0.1). Salinity was measured daily with a Honeywell DL4000 conductivity cell (accuracy ± 1 PSU). pH (NBS scale), dissolved oxygen (mg L-1), and temperature (°C) were logged every ten minutes. Endpoint pH treatments of pH 1.0 and 12.0 were accomplished through the addition of either 1N phosphoric acid or a mixture of 0.5M potassium hydroxide and 0.5M potassium carbonate. Additions were accomplished using the pH stat system described above with the addition of drip irrigators. These treatments were not intended to accurately mimic the carbonate chemistry regime found in nearshore environments, but rather served to bookend the response curve generated through the manipulation of pCO2.

Each individual adhesive plaque was pulled until failure according to the protocol outlined in Bell & Gosline (1996). Plaques were pulled at a 90° angle relative to the substrate to ensure uniformity across samples. For each test, a hemostat was used to grip the distal region, 1 mm above the attachment plaque, and attached to a 10 N digital force gauge (OMEGA, Stamford, CT; accuracy ± 0.01 N) mounted on a motor-driven testing frame. Extension rate was 10 mm min-1 and force (N) was recorded at 20 Hz. All threads were rehydrated (>5 mins) in their respective seawater treatment prior to testing. Each plaque was imaged before tensile testing using an AmScope MU1000 camera (Irvine, CA) attached to a dissection microscope.

Plaque attachment area (planform area) in mm2 was measured using AmScopeX imaging software by tracing the outline of each plaque. Adhesion strength (kPa) was recorded as the maximum force required to remove a plaque from the substrate, normalized by the attachment area (Burkett et al. 2009). Work of adhesion (N m-1) was calculated as the area under the force-extension curve (Hamada et al. 2017). The mean of 3-5 plaque pulls was reported for each mussel. The failure mode of each plaque was scored visually as either an adhesive failure (plunger failure), a peeling failure (failure propagated from one side of the plaque to the other), or tearing failure (part of the plaque remained after failure) following the guidelines of Young & Crisp (1982).

Nanoscale surface characteristics of the cuticle of adhesive plaques were determined using a Bruker (Billerica, MA) Dimension ICON atomic force microscope (AFM). A ScanAsyst-Air probe with silicon-nitride tip was used to approach the surface of dry plaque samples, scanning 1 µm2 regions of the cuticle’s topography. Scans of the plaque surface were haphazardly taken 1 mm away from the distal root, avoiding large topographical features on the plaque’s surface and parts of the distal region that innervated the plaque architecture. A region free of large topographical features was chosen from each 1 µm2 scan for analysis, moving the tip to that section and recording a 10 nm2 adhesion image. DMT modulus (GPa) was calculated within each scan as the slope of the force curve during tip-sample separation (Young et al. 2011), with a resolution of 512 samples per line and calibrated against a fused silica standard (Veeco, Plainview, NY).

The over 260,000 measurements obtained from each 10 nm2 scan were averaged to get a representative stiffness of the cuticle at that position. Three samples were taken per plaque, averaging the mean of three plaques from each mussel.

Detailed methods and results are provided in George and Carrington, 2018.

dataset_1.csv