Sample Collection

In this experiment, plankton were treated as two groups: those exposed to microbial biofilms (2024) and those not (2022). To collect biofilm, “Biofilm Colonizers,” sections of PVC piping covered in stainless steel mesh on either side and attached with metal band clamps, were deployed on the benthos by HOV Alvin on sites with high diffuse flow (confirmed by temperature probe reading higher than ambient temperatures) and left to be colonized by vent biofilm-forming bacteria for 10 d before recovery. For the 2024 (biofilm) observations, segments of colonized mesh were cut to size using flame-sterilized tweezers and scissors, secured in the custom aluminum inserts, and placed in the bottom of the plankton observation chamber. This design was intended to test response to a substratum-bound cue assuming a tactile trigger, but the biofilm was flocculent and became suspended when the chamber was filled, likely bringing it into contact with larvae throughout the observation pool. For the 2022 (no biofilm) observations, the insert was not used.

Larvae were collected using a McLane Laboratories large-volume water pump (WTS-LV) with a filter assembly to capture larvae onto a large (19 cm diameter) filter of 63 μm mesh. The filter system was designed with an expanded head space and thick insulation to maximize retrieval of live specimens (Beaulieu et al. 2009). The pump was mounted on a lander to facilitate manual positioning via HOV Alvin. The lander was placed at about 10 m distance from the active vent site, and the inlet was positioned at 1.5 m above the seafloor, where the water temperature was 2°C and pressure was 252 bar. The pump was programmed to run at a rate of 25 L per minute for 21 h to end just before release from the seafloor to maintain larval health, capturing samples of roughly 31,000 L. The lander ascended to the surface at a rate of approximately 45 m per minute. Due to the warm surface waters of the eastern tropical Pacific, care had to be taken to recover the sample quickly, as prolonged exposure to elevated temperature can damage the larvae. At this site, our CTD data showed that temperature was below 5°C at depths greater than 1000 m, reached 10°C at about 300 m, and was above 25°C at the surface. After the pump was recovered and secured on deck, the pump filter assembly was immediately placed in chilled seawater (4°C). Gastropod larvae were sorted manually from the samples, on ice under dissecting microscopes, for 1–2 h, and identified by trained experts. All available gastropod larvae (150–400 μm) and 6–12 copepods were placed in the plankton observation chamber. Copepods, being larger and more active swimmers, were added as a visual aid to ensure the camera remained in focus and conditions in the chamber remained habitable. The chamber was topped off with filtered bottom seawater, sealed, and fastened to the mount to prepare for recording behavior. In 2022, one of six recovered pump samples yielded actively swimming gastropod larvae (pump 8), whereas in 2024, one of seven pump samples yielded swimming larvae (pump 5). In 2022, three gastropod larvae were introduced into the chamber (Lepetodrilus sp., Laeviphitus sp., and “unknown”), and all revived sufficiently to swim actively in the chamber. In 2024, after the biofilm insert was placed in the chamber, nine gastropod larvae were introduced (identified with certainty Lepetodrilus sp. and Peltospira sp., and provisionally Laeviphitus sp., Echinopelta sp., Clypeosectus sp. and “unknown”); two of these individuals revived, but it was not possible to distinguish which ones.

Recording set-up

Deep sea larvae were loaded and sealed in a high-pressure chamber within two hours of surfacing. The chamber was kept in a dark 4 degree (Celsius) cold room for duration of the experiments. Bottom pressure (252 bar) was slowly reached over the course of 15 min inside the chamber. Once at pressure, pump flowrate was kept at 0.05-0.2 milliliters per minute (mL/min). Videos were recorded as TIFF stacks using a frame rate of 30-60 fps. Occasionally, the chamber was manually flipped to attempt to stimulate larvae into swimming. Experiments continued until larvae stopped swimming, between 2-3 hrs. Recording was paused, breaking up footage into "Sessions" due to limited writing speed of our data storage equipment and followed by waiting periods of up to 3 minutes while memory was written to disk. "Sessions" were broken up into "partitions" whenever the camera frame of reference was moved, or if the chamber was flipped during recording.

Differences between cruises

AT50-06 (2022): Camera resolution was limited to 944x950 and chamber viewport was partially in view at 798 pixels per centimeter. Pressure chamber and camera were secured but detached one another. Framerate was set to 60 frames per seconds and sessions were limited to 3 minutes. Chamber was flipped a total of four times.

AT50-20 (2024): Camera resolution was set to the full 2048x2048 pixels, putting the whole chamber viewport in view at 563 pixels per centimeter. Pressure chamber was secured on a custom 3d-printed mount, attached to an aluminum railed that attached the mounted camera, keeping centered and fixed on the chamber. Framerate was lowered to 30 frames per second, and sessions were extended to last 30 minutes. Chamber was flipped a total of 2 times. Gastropods were exposed to vent-specific microbial biofilm for the duration of this experiment.