Krill Ocean Acidification Physiology Data

Brief description

Antarctic krill (Euphausia superba) have a keystone role in the Southern Ocean, as the primary prey of Antarctic predators. Any decreases in krill abundance could result in a major ecological regime shift, but there is currently limited information on how climate change may affect krill. Increasing anthropogenic carbon dioxide (CO2) emissions are causing ocean acidification, as absorption of atmospheric CO2 in seawater alters ocean chemistry. Ocean acidification increases mortality and negatively affects physiological functioning in some marine invertebrates, and is predicted to occur most rapidly at high latitudes. Here we show that, in the laboratory, adult krill are able to survive, grow, store fat, mature, and maintain respiration rates when exposed to near-future ocean acidification (1000 – 2000 μatm pCO2) for one year. Despite differences in seawater pCO2 incubation conditions, adult krill are able to actively maintain the acid-base balance of their body fluids in near-future pCO2, which enhances their resilience to ocean acidification.

Lineage

Maintenance and Update Frequency: notPlanned

Statement: 4.3. Materials and Methods
4.3.1. Experimental conditions
Live krill were collected on the RSV Aurora Australis via rectangular mid-water trawl on 22nd – 23rd February 2015 (66–03°S, 59–25°E and 66–33°S, 59–35°E). Krill were held in shipboard aquaria using standard maintenance methods (King et al. 2003) before being transferred to the Australian Antarctic Division’s (AAD) Krill Aquarium in Tasmania (seawater temperature 0.5°C and pH 8.1). Seawater was supplied to aquarium tanks via a seawater recirculating system (Kawaguchi et al. 2010).
For ocean acidification experiments, 0.5°C seawater was supplied from a 70 L header tank and equilibrated with air (control) or CO2-enriched air (elevated pCO2 treatments) before delivery to experimental tanks. The CO2-enriched air was monitored using mass flow controllers (Horiba STEC SEC-E-40) and air valves, to regulate flow rates of atmospheric air and pure CO2. Five experimental 300 L tanks were maintained at five pCO2 levels; control 400 μatm pCO2 (pH 8.1), 1000 μatm pCO2 (pH 7.8), 1500 μatm pCO2 (pH 7.6), 2000 μatm pCO2 (pH 7.4) and 4000 μatm pCO2 (pH 7.1).
Appropriate tank size and the best possible animal husbandry were high priorities in such a long-term study. As krill are a pelagic species, large sized (300 L) experimental tanks were needed to emulate wild conditions as closely as possible in a laboratory. Our experimental design was limited by the space and resources needed for these large tanks, and our observational units (CO2 treatment tanks) could not be replicated. We did not however, observe any visual evidence to suggest that ‘tank’ effects were confounding our results.
Two hundred krill were randomly assigned to each experimental tank on 25th January 2016, corresponding to a density of 0.6 individuals L-1. This density is in the range of 0.5 – 2 individuals L-1 which has been successfully used in previous experiments at the AAD krill aquarium (Brown et al. 2013; Höring et al. 2018). The experiment ran for 46 weeks from the 25th Jan 2016 – 12th Dec 2016 covering all four seasons. Mortality rates in all pCO2 treatments (ranging from 0.03 – 0.2 % day-1) were much lower than previously reported for Antarctic krill in shipboard aquaria (2% day-1; King et al. 2003) and in other pCO2 studies on Pacific krill (0.5 % day-1; Cooper et al. 2017) and northern Atlantic krill (5% day-1; Sperfeld et al. 2014).
The pCO2 levels of the CO2-enriched air and seawater were monitored daily using a LI820 CO2 gas analyzer and associated computer software (version 2.0.0), and daily pH levels of experimental tanks were measured manually using a pH meter (Mettler Toledo SevenGo Duo Pro). A three-point calibration of the pH meter was undertaken daily using Radiometer Analytical IUPAC Standard buffers of pH 7.000, 7.413 and 9.180. Total alkalinity (AT) and dissolved inorganic carbon (DIC) were measured weekly using a Kimoto ATT-05 Total Alkalinity Titrator. Salinity was measured weekly using a Profiline™ Cond 197i Conductivity Meter, WTW. The average total pH (pHT), pCO2, calcite and aragonite saturation (ΩC and ΩA) values over the 46 week experiment were calculated in CO2SYS (Pierrot et al. 2006) using our measured salinity, temperature, alkalinity and DIC data, and using equilibrium constants of Merhbach, as modified by Dickson and Millero (Dickson et al. 2007). Average levels of pCO2 were 8 – 169 μatm below target levels for the 400 – 2000 μatm treatments, and 123 μatm above the target level for the 4000 μatm treatment. Seawater temperature and AT were stable in all treatments, while DIC increased with increasing pCO2. Seawater chemistry in the experimental aquarium is shown in detail in Appendix V.
Krill were fed six days per week with a mixed microalgal diet of the Antarctic flagellate Pyramimonas gelidicola at a final concentration of 2 x 104 cells ml-1, and Reed Mariculture Inc. (USA) cultures of the diatom Thalassiosira weissflogii (8.8 x 103 cells ml-1), flagellate Pavlova lutheri (4.5 x 104 cells ml-1) and flagellate Isochrysis galbana (5.5 x 104 cells ml-1) (Brown et al. 2010; Höring et al. 2018).
Light was controlled in the laboratory to ensure that the photoperiod mimicked the seasonally changing light regime of the Southern Ocean (66°S, 30 m depth). Photoperiod was altered monthly, with a maximum of 100 lux light intensity in February and minimum intensity (24 hr darkness) in August (Appendix VI). Light was provided by twin fluorescent tubes and was controlled via standard aquarium procedures (Kawaguchi et al. 2010).

4.3.2. Survival
Each pCO2 treatment was checked daily for mortalities, which were recorded and placed in vials of 10% formalin. Daily mortality data were used to calculate the percentage of krill still surviving at the end of each experimental week in each treatment using the equation:

Percentage of krill remaining in the previous week- (Number of mortalities during the current week)/(Number of krill remaining in tank) × 100

Krill that were sampled for experimental purposes were not counted as mortalities, but were subtracted from the number of krill remaining in the tank each week. This ensured that the remaining number of krill used to calculate survival percentages reflected actual experimental mortality.

4.3.3. Total length
Krill lengths (mm) were obtained from krill in each pCO2 treatment in weeks 1, 2, 4, 5, 26, 39, 41, 43 and 46. Sample sizes (n) for length measurements for each week and treatment are shown in Appendix VIIA. Individuals were sexed using microscopy and the length of each specimen was measured from the tip of the rostrum to the tip of the uropod (measurement Standard Length 1; Kirkwood 1984). Length data from frozen krill and live krill were combined.

4.3.4. Lipid class analysis (triacylglycerols)
Lipid analysis focused on triacylglycerols which are the main storage fat in krill and, therefore, drive overall lipid concentrations and lipid class composition of krill (Hellessey et al. 2018). Krill were sampled for lipid analysis from all pCO2 treatments in weeks 1, 2, 4, 5, 26, 39, 41 and 43. Individual krill were placed in cryo-tubes and immediately stored in a –80°C freezer.
Lipid class analysis was carried out on 4 – 5 krill from pCO2 treatments 400, 1000, 1500 and 2000 on each sampling week (n = 3 for the 4000 μatm tank in weeks 39, 41, and 43 due to increased mortality and lower numbers of krill in that treatment). Sample sizes (n) for each week and treatment are shown in Appendix VIIB. The wet mass (g), total length (measurement Standard Length 1; Kirkwood 1984), and sex for each krill was obtained, and krill were kept frozen during this process to prevent sample degradation. A dry mass (g DM) was obtained later by multiplying the wet mass by 0.2278 to account for the 77.2% water content in the organism (Virtue et al. 1993a). Total lipid extracts of krill specimens were obtained using a modified Bligh and Dyer method (Bligh & Dyer 1959; Ericson et al. 2018a). Lipid class composition and content were determined using an Iatroscan MK-5 TLC/FID Analyser using standard methods (Hellessey et al. 2018).

4.3.5. Sexual maturation
The maturity stages of individual krill were identified during weeks 39, 41, 43 and 46 (n = 5 for 400 – 2000 μatm pCO2 treatments, n = 3 for the 4000 μatm pCO2 treatment). Adult krill undergo sexual regression in winter, so these measurements occurred at the end of the experiment to capture the onset of maturity during late spring/early summer.
The sex and maturity stage of each krill was identified via microscopy (using the staging key in Appendix VIII). Each maturity stage was given a maturity score with higher numbers denoting greater maturity (Appendix VIII). After staging, individual krill were placed in a cryopreservation tube with 10% formalin.
4.3.6. Ovarian development
On the final day of the experiment (12th Dec 2016, Week 46), krill left in each experimental tank were preserved in 10% formalin. These samples were used to determine the ovarian development of eight randomly selected females from each of the 400, 1000, 1500 and 2000 μatm pCO2 treatments. Only two females remained in the 4000 μatm pCO2 treatment, therefore only two replicates could be obtained for this tank.
The ovary was dissected out of each organism and a single lobe was placed on a microscope slide with a drop of distilled and deionized water and lightly squashed (Cuzin-Roudy & Amsler 1991). Photographs were taken of the ovary section and the lengths of the largest cells (across the longest axis of the cell) were measured using the computer software Image J (https://imagej.nih.gov/ij/). The cell size and photographs were used to determine the maturation stage of krill ovaries using the key in Appendix IX (modified from Cuzin-Roudy & Amsler 1991). When an ovary was transitioning from one stage to another, a 0.5 value was used (e.g. 4.5). Photographic examples of different ovary stages are shown in Appendix X.

4.3.7. Respiration rate
Respirometry measurements were carried out in experimental week 38. Respirometry vessels (2L) with pre-fitted O2 mini sensors were filled with seawater sourced from the inlet hose of each experimental tank and placed in a 0°C water bath. Each vessel was connected to O2 computer software (version OXY10v3_50TX) via an optic fiber probe.
Ten krill were sampled from each experimental tank (n = 50 total) and the total length (measurement Standard Length 1; Kirkwood 1984) and wet mass (g) were obtained for each individual. A dry mass (g DM) was obtained by multiplying the wet mass by 0.2278 to account for the 77.2% water content in the organism (Virtue et al. 1993a).
Each krill was then placed into a respirometry vessel completely filled with experimental seawater, with no air spaces in the vessels. Oxygen saturation (%) was logged at 5 min intervals in each respirometry vessel over 22-hrs (9AM – 7AM the following day), using the computer software. The software was calibrated at 0°C and the atmospheric pressure at the time of measurement. After 22-hrs krill were removed from the vessels and returned to their experimental tanks.
Only measurements of O2 saturation (%) taken between 12PM – 7AM were considered for analysis, to ensure that krill had three hours at the beginning of respiration trials to settle into a normal rhythm of respiration before data was collected. Oxygen saturation (100 %) for seawater at 0ºC and 35.1 salinity units (‰) was converted to O2 ml L-1 using the equation in Fox (1907) to obtain a value of 8.035 ml L-1. This was used to convert the O2 saturation (%) at each logged time point to milliliters of O2 (O2 ml) in each 2L respirometry vessel using the equation:

O_2 ml in respirometry vessel = (% O_2 saturation)/100 × (8.035 × 2)

Values for O2 ml in each respirometry vessel between 12PM and 7AM were used to create regression equations which were used to compute the O2 ml used in each respirometry vessel during this period. This value was divided by the krill dry mass (mg), converted to µl O2 mg DM-1, then divided by 19 hrs to obtain the µl O2 mg DM hr-1.

4.3.8. Haemolymph pH
The haemolymph pH of five krill from each experimental tank was measured in week 46. Haemolymph pH was measured in situ by inserting a pH microelectrode directly into the pericardial cavity. This ensured that air contact with the haemolymph was minimised, as contact with air may alter the CO2 concentration and pH of the body fluids (Riebesell et al. 2011). A Unisense pH Microelectrode (model pH-50, tip diameter 40 – 60 µm) and Unisense Reference Electrode connected to a Unisense pH/mV Meter and computer software (SensorTrace Logger) were used to complete measurements. The pH microelectrode and reference electrode were calibrated using the SensorTrace software via a three-point calibration using Radiometer Analytical IUPAC Standard pH buffers 7.000, 7.413 and 9.180. The buffers were chilled to the seawater temperature in which haemolymph measurements were conducted (0 – 0.5°C). The pH of these buffers at 0°C were used for calibrations (pH 7.12, 7.53 and 9.46 respectively).
Krill were individually removed from their 300 L tanks and placed under a compound microscope in a refrigerated microscope stage, submerged in seawater from the tank they originated from. The pH of this seawater was measured using the microprobe and reference probe, and a portable pH meter (Mettler Toledo SevenGo Duo Pro) to ensure that the measurements matched to within < 0.05 pH units before proceeding.
Live krill were restrained within the microscope stage using acrylic blocks, designed to expose the integument that links the krill carapace to the abdomen. A micromanipulator was used to position the microelectrode relative to the animal. A camera connected to the compound microscope was also used to magnify the krill carapace-abdomen joint and view the real-time image on a computer monitor to ensure the accuracy of microelectrode placement.
The microelectrode was inserted through the integument underneath the carapace and into the pericardial cavity between the thorax and first abdominal segment. The reference probe remained in the seawater surrounding the krill during this process. Some resistance was observed as the microelectrode pierced the integument, causing a slight tear in the body wall as the probe penetrated the integument, ensuring that electrical conductivity was maintained between the reference probe and microelectrode.
The SensorTrace Logger software logged the pH of the haemolymph, and the pH was recorded once the reading had stabilised after approximately 1 min. The microelectrode was then withdrawn from the abdomen and haemolymph was observed leaking into the surrounding seawater as positive pressure from within the animal pushed it outwards. The krill was removed from the microscope stage and preserved in 10% formalin.

4.3.9. Statistical analyses
Data were analysed in the RStudio statistics package (version 0.99.893) using one-way ANOVA with pCO2 treatment as a factor, or two-way ANOVA with pCO2 and Week as factors. Dunnett comparisons (carried out using the RStudio ‘multcomp’ package) were used to identify significant differences between the control treatment (400 μatm pCO2) and all other factor levels, while Tukey Post-hoc comparisons were used to compare all factor levels with one another. Polynomial contrasts were used to identify linear, quadratic and cubic trends in the data. Type 3 Sums of Squares (SS) were used when data was unbalanced and Type 1 SS were not appropriate. Data were log or square root transformed when assumptions of normality or homogeneity of variances were not met. Data in tables are expressed as mean ± standard deviation. For all analyses, α was set at 0.05 and all tests were two tailed. The RStudio packages ‘ggplot2’, ‘plyr’ and ‘dplyr’ were used to produce all figures.