Dataset: Initial prey abundance and biomass: JEMBE 2017

From Jungbluth et al. 2017 – JEMBE:
Naupliar grazing rates were measured on field-collected prey assemblages in bottle incubation experiments in the laboratory. Nauplii used in these experiments were derived from laboratory culture populations of Parvocalanus crassirostris, originally established from animals collected in Kaneʻohe Bay. At 18-h prior to the start of each experiment, adults were isolated and fed Tisochrysis lutea (formerly Isochrysis galbana Tahitian strain [Bendif et al., 2013]) at a concentration of 105–106 cells mL− 1 After 6-h, adults were removed, and eggs and nauplii were allowed to develop for 12-h in order to produce a cohort of mid-stage nauplii (N3-N4) with a narrow age-range at the beginning of each experiment. Sets of approximately 50 nauplii were isolated into small volumes (< 10 mL) of 0.2 μm filtered seawater 1–2 h prior to the start of each grazing experiment.

Seawater for the prey assemblage was collected from the central basin of the southern semi-enclosed region of Kane'ohe Bay, Oahu, Hawai'i (21°25'56"N, 157°46'47"W) on two dates: 10 March 2015 (Experiment: E1) and 22 April 2015 (Experiment: E2). Seawater was collected from ~ 2 m depth using a 5 L General Oceanics Niskin bottle deployed by hand line, and gently transferred using acid-washed silicon tubing directly from the Niskin bottle into 20 L covered (dark) polycarbonate carboys. The seawater was transported to the laboratory within 2-h of collection. The collected water was gently pre-screened (35 µm Nitex mesh), which was intended to remove all in situ nauplii and other large grazers, so that the only metazoan grazers in the bottles were the added nauplii. The < 35 µm incubation water was added to pre-washed (10% HCL rinse, followed by 3 rinses with experimental seawater) 1 L polycarbonate bottles (total volume: 1120 mL).

Nutrients were not amended in control or treatment bottles due to the expected low rates of excretion by these small biomass nauplii over the incubation duration as compared with baseline levels in Kane'ohe Bay, and also in order to minimize development of artificially high nutrients given prevailing oligotrophic conditions in the study area. Excretion rates of copepods are a function of biomass (Vidal and Whitledge, 1982, Mauchline, 1998), with excretion by nauplii roughly an order of magnitude lower than conspecific adults. At a nauplius grazer concentration of 50 nauplii in a 1 L volume, excretion rates result in values 2 to 3 orders of magnitude below the average nitrogen concentrations of 0.2–1.0 µM in Kane?ohe Bay (Drupp et al., 2011). Therefore, excretion rates in bottle incubations were expected to have negligible impacts on prey growth rates in experimental bottles, and nutrient amendment would have only altered the prey community further away from in situ conditions.

The isolated N3-N4 nauplii were transferred into triplicate < 35 µm incubation water bottles (grazing treatments) and placed on a bottle roller (4–6 rpm) to maintain prey in suspension for the duration of the incubation period. Parallel triplicate control treatments (incubation water without added nauplii) were also placed on the bottle roller. Grazing rates were measured using two densities of naupliar grazers: high (N = 92–97 nauplii L- 1) and moderate (N = 45–50 nauplii L- 1) densities. All incubations were run for a total of 24-h in the dark, with subsamples taken every six hours to examine changes in ingestion rates over time. Experiments were run at 21 °C, which is at the low end of the range of annual temperature fluctuations for this region of Kane'ohe Bay (20–29 °C in prior 5 years [Franklin et al., 2015]).

During the course of the incubation, triplicate 2-mL volumes of each subsample were measured with a Coulter Counter (Beckman-Coulter Multisizer III) with a 100 µm orifice tube, yielding a spectrum of particle sizes from 2 to 35 µm ESD, as well as quantitative abundance data. In a diverse environment with a variety of autotrophic and heterotrophic pico- to microplankton, standard cell quantification methods (e.g. epifluorescence microscopy, inverted microscopy) do not reliably preserve some components of the community (Omori and Ikeda, 1984, Sherr and Sherr, 1993), requiring a patchwork of methods to quantify the full potential suite of prey items. In the absence of large cells or of abiotic particles that may result in unreliable quantification (e.g. Harbison and McAlister, 1980), the Coulter Counter is an appropriate and more reliable means of describing how the abundance of different sized cells change over the duration of grazing incubations (Paffenhöfer, 1984), with results comparable to methods based on gut fluorescence and egg production (Kiørboe et al., 1985). Water subsamples for Coulter Counter measurements were taken directly from experimental bottles upon addition of nauplii at the start of each experiment (time 0) and at each six-hour time point, being careful to retain nauplii as experimental grazers by recovery of animals on a 35 µm cap filter and washing them back into bottles during sub-sampling with a small volume of filtered seawater.

Data on prey size (ESD) and abundance from the Coulter Counter were further processed using R (Core Team, 2013). Prey ESD was converted to biovolume (BV, µm3), then to carbon (C, pg C cell- 1) using the relationship C = 0.216 × BV0.939 (Menden-Deuer and Lessard, 2000). Averages (triplicate Coulter Counter measurements) were binned into 5 functionally relevant prey size groupings (2–5, 5–10, 10–15, 15–20, and 20–35 µm), chosen to ensure comparable data to a prior study of adult copepod grazing in Kane'ohe Bay (Calbet et al., 2000). The binned data for initial and final time points for each control and treatment bottle were used to calculate carbon ingestion (I, ng C grazer- 1 h-1) and clearance (F, mL grazer- 1 h-1) rates on each prey size group using the equations of Frost (1972), and are reported here only where F or I > 0.

Linear regressions were used to evaluate whether there was a relationship between control bottle prey biomass and incubation duration, and between measured ingestion rates (I) and incubation time. An analysis of covariance (ANCOVA) was used to test for significant (p < 0.05) effects of incubation time, predator treatment (Plow, Phigh), and experiment (E1, E2) on carbon ingestion rates (I), and for interactions between variables, accounting for random error due to differences between replicate bottles. The ANCOVAs were performed for each prey size group or total prey using the aov function in the package stat with time, predator treatment, and experiment as potentially interacting factors, and incorporating bottle replicate error as a random effect. The coefficient of variation (CV, %) was calculated for cell abundance estimates and followed by a two-way ANOVA and post-hoc Tukey test to evaluate for the effects of prey size group and incubation time on variation in cell abundance. In many studies of zooplankton grazing (e.g., Atienza et al., 2006, Calbet et al., 2009, Almeda et al., 2011), significant differences in prey growth rates between control and treatment bottles were tested, and then only significantly different conditions were considered in further interpretations of I. Here, significant differences (t-test, p < 0.05) between treatment and control prey growth rates were used to evaluate whether the significance of this test was affected by the duration of incubation. All statistical analyses were conducted using the program R (stats package) (R core team, 2016).