Dataset: Acartia tonsa mortality with CTD data: RV Sharp

This is mortality data was obtained from two week-long cruises (1301 in August and 1302 in September) during which Niskin samples were taken from the mid-bay of the Chesapeake from 9 stations in a box formation; 3 stations in a northern transect across the bay (N1-N3), 3 in a midline transect (M1-M3), and 3 in a southern transect (S1-S3).

The Niskin bottles used were General Oceanics 1010x External Spring Water Sampler with a 10L capacity (part number: 101010X). 12 of these were deployed rosette style on the SBE 32 Carousel Water Sampler from the starboard winch of the RV Sharp, along with the SBE 9plus unit which was attached to the rosette. On each downcast, instrument readings were sent from the SBE 9plus unit on the rosette to the SBE 11plus V2 Deck Unit. Based on these data, sampling depths were selected which fell into three areas of interest: above, within, and below the pycnocline.  If a pycnocline was not evident and density seemed consistent, only two samples were taken. On the upcast, Niskins were triggered to close electronically from the dry lab.

For mortality assessment, three Niskin bottles were triggered at each chosen depth to give a total sampled volume of 30L. Once on board, Niskin bottles were carefully drained through sieves with 64µm mesh, stained, and preserved following the procedure below, as developed by Elliott and Tang. Sieves with 64µm mesh were selected to catch all life stages of the copepod Acartia tonsa since Acartia tonsa eggs are about 75µm in diameter and all subsequent life stages are larger.

Sieves were placed in a water bath with the salinity and temperature to which the animals were accustomed. To reduce stress on the copepods, tubes were connected to the Niskin stopcock and squeezed briefly to force air bubbles to flow out; once the water bath began to fill, the flow rate was reduced. Overflow from the water bath was collected in a squirt bottle for later transferring the animals.

Once one Niskin drained, the sieve tube was transferred to the next Niskin to begin draining while the bottom of the drained Niskin was emptied into a wide-mouth plastic beaker, poured into the corresponding sieve, and rinsed with the water collected earlier. After 30L were drained, the contents of each sieve were gently transferred to numbered jars for incubation; date, time, station, depth sampled, and cast numbers were entered into a log.

Neutral red stain was added based on the volume in the jar- the desired concentration was 150 μl per 100ml (for a stock solution of 0.1g Neutral Red powder per 10 ml DI water). The jars were returned to the water baths and incubated in the shade for 15-20 minutes.

After incubation, each sample was vacuum filtered onto a 64μm mesh filter which was transferred to a small petri dish. The dish was capped and labeled with the date, CTD number, and depth of the sample, then wrapped in parafilm to seal. The samples were placed in a Ziploc baggie labeled with the date and given a burst of Flash Freeze, then stored in a -20°C freezer for later analysis.

After returning from the cruises, samples were stored in -20°C freezers; one set (1301) with David Elliott at the Virginia Institute of Marine Science, College of William and Mary, Gloucester Point, VA, USA and the other (1302) with Jamie Pierson at University of Maryland Center for Environmental Science, Horn Point Laboratory, Cambridge, MD, USA. Samples were individually processed using the procedure developed by Elliott and Tang:

The Petri dish containing the sample was thawed at room temperature, then the mesh was submerged in filtered seawater and gently shaken to resuspend the sample. The sample was then transferred to a counting wheel where it was checked for density and subsampled if necessary.

The sample was then slightly acidified by adding 10% HCl drop-wise until the stained animals distinctly changed from faded pink to bright pink. The sample was then tallied for live or dead status and general stage under dissecting microscope with darkfield illumination. Copepods which were bright red were considered alive at the time of staining; copepods which had no stain, were cloudy white, or were only light pink were considered dead.

To account for individual discrepancies between counters, a calibration count was done during which David Elliott and Catherine Fitzgerald processed the same sample and tallied the live and dead one after the other, with no comparisons until all five samples were completed (a full profile of 3 samples, plus two random samples).

Data were entered into an Excel spreadsheet and checked for transcription errors, then imported into MatLab for data analysis.