Methods

Field collections

In late December 2003, we collected seawater and resident plankton from several sites in the Ross Sea polynya (Fig. 1) aboard the research vessel Nathaniel B. Palmer. At station A (ca. 73°24' S, 173°14 E), in ice-free waters at the northern edge of the polynya, we used a Zodiac inflatable boat to collect surface seawater and resident plankton in an acid-cleaned 50-l polyethylene carboy. The seawater and plankton were collected by submerging an acid-cleaned 10-l polyethylene carboy from the bow of the work boat while slowly underway, then transferring the seawater from the 10-l carboy into the 50-l carboy on the work boat, using a protocol designed to avoid contamination of the seawater during the transfer process. The 50-l carboy of seawater and resident plankton were then stored for ~2 h in a shipboard cold room maintained at 2°C, prior to subsam-pling and processing for the shipboard iron-light manipulation experiment described below.

Station 25 (ca. 76°05' S, 170°08' E), in the central polynya, was one of several stations where we collected near-surface seawater and resident plankton in 10-l Niskin bottles lowered on a rosette. Whole seawater samples from these stations were maintained in polystyrene culture flasks inside a shipboard incubator at ~0°C, prior to return to the Hollings Marine Laboratory (South Carolina, USA) for isolation of Phaeocystis cultures. In the nearby, ice-free waters at station B (ca. 76°02' S, 169°53' E), we collected -4,000 l of seawater from the upper water column through acid-rinsed polyethylene tubing using an electrically operated Jabsco polypropylene-Santo-prene double-diaphragm pump. The tubing inlet was lowered to -10 m water depth from the side of the ship using a polyaramid line and epoxy-coated end weight, then seawater was pumped aboard while slowly underway at 1-2 knots. The pumped seawater was filtered through in-line, acid-rinsed 1-im/0.2-im high-capacity pleated-polypropylene cartridge filters (Cole-Parmer, Inc.), and then collected in acid-cleaned 125-l blue polyethylene barrels with airtight polyethylene screw closures. Subsequent measurements of iron in subsamples of this filtered seawater showed dissolved Fe concentrations of 0.07 ± 0.02 nM (n = 4). The barrels of low-iron filtered seawater (hereafter 'Station B seawater') were returned to the Hollings Marine Laboratory for use in the laboratory dose-response iron-addition experiment described below.

Shipboard iron-light manipulation experiment

Shipboard microscopy suggested that phytoplank-ton in seawater collected from Station A was dominated by healthy colonies of P. antarctica; this was later verified by high-performance liquid chromatography (HPLC) pigment analysis (see section ''Laboratory dose-response iron-addition experiment''). Using the 50-l of cleanly collected seawater and resident plankton, we carried out a bioassay experiment that was designed to evaluate the effects of iron and light availability on the growth rate of colonial Phaeocystis antarctica. The

Fig. 1 Location of sampling sites in the Ross Sea, Antarctica, where seawater and plankton were collected for the experiments described in this paper. Seawater and resident plankton used in the shipboard iron-light experiment were collected at station A; uncontaminated seawater used in the laboratory iron experiment was collected at station B; the P. antarctica culture used in the laboratory iron experiment was isolated from water collected at station 25

Fig. 1 Location of sampling sites in the Ross Sea, Antarctica, where seawater and plankton were collected for the experiments described in this paper. Seawater and resident plankton used in the shipboard iron-light experiment were collected at station A; uncontaminated seawater used in the laboratory iron experiment was collected at station B; the P. antarctica culture used in the laboratory iron experiment was isolated from water collected at station 25

seawater was gently mixed in the 50-l carboy, then transferred into acid-cleaned 1.2-1 polycarbonate bottles, leaving -200 ml headspace in each bottle, under class-100 filtered air using the rigorous trace-metal clean techniques described by Sedwick et al. (2000, 2002). Three iron treatments were used for these experimental incubation bottles: (1) 'high-Fe', to which 3 nM Fe was added as a 17.9 iM solution of ferric nitrate (the added nitrate was negligible relative to the ambient seawater nitrate concentration of -20 iM); (2) 'control', to which there was no Fe added (initial ambient dissolved Fe concentration = 0.38 ± 0.03 nM); and (3) 'low-Fe', to which 3 nM of the iron-chelating ligand deferoxamine (DFOB or desferal) was added as a 38.4 iM solution of deferoxamine mesylate salt (Sigma Chemical).

After filling, the bottles were tightly sealed with polyvinyl chloride (PVC) tape, then set in circulating surface seawater at -0°C in deckboard incubator baths shaded to two light levels using neutral-density screening: (1) 'low-light', shaded to -15% of incident deckboard irradiance, and (2) 'high-light', shaded to -50% of incident deckboard irradiance. The low-light treatment was chosen so as to fall within the range of mean in situ irradiance experienced by phytoplankton in the surface mixed layer, whereas the high-light treatment was intended to approximate the expected maximum value for mean in-situ irradiance (see discussion in section ''Conclusions and directions for future research''). Our shipboard experiment thus incubated bottles containing seawater and resident plankton with six different treatments: (1) low-light, low-iron; (2) low-light, control-iron (this treatment was intended to approximate in situ irradiance and iron concentration); (3) low-light, high-iron; (4) high-light, low-iron; (5) high-light, control-iron; and (6) high-light, high-iron. Duplicate bottles for each treatment were harvested over an 80-h period, and subsampled for dissolved macronutrients, chlorophyll a, phytoplankton pigments, and other parameters.

Laboratory dose-response iron-addition experiment

At the Hollings Marine Laboratory, a pure culture of colonial P. antarctica was isolated from the native phytoplankton assemblage collected at Station 25, and subsequently maintained in nutrient-replete semicontinuous batch cultures, using Station B seawater amended with L1 medium (Guillard and Hargraves 1993). For the dose-response iron-addition experiment described here, we used a culture of this P. antarctica strain (predominantly colonial) that had been maintained for two months in L1 medium diluted 457-fold with station B seawater.

Eleven days before commencing the experiment, 50 ml of this culture was diluted to 1 l with 0.2 im-filtered station B seawater in an acid-cleaned polycarbonate bottle, using a stringent trace-metal clean protocol. A peristaltic pump was used to transfer the Station B seawater from the 125-l polyethylene barrel (in which it had been stored since collection) through acid-cleaned silicone tubing and an acid-rinsed 0.2-im CritiCap Supor capsule filter (Pall Corporation) into the 1-l polycarbonate bottle under class-100 filtered air. Subsamples of this 0.2-^m filtered station B seawater were collected for subsequent iron measurements, which confirmed that dissolved Fe concentrations were relatively low (0.17 ± 0.1 nM, n = 4). Based on the concentrations of FeCl3 (11.65 iM) and EDTA (11.71 iM) in the L1 medium, the resultant 1-l inoculum of colonial P. antarctica had initial dissolved Fe and EDTA concentrations of approximately 1.42 and 1.25 nM, respectively.

After acclimating this P. antarctica inoculum to an irradiance of -20 iE m-2 s1 for 11 days at -0°C, approximately 230 ml was removed for measurements of 'initial' pigment concentrations and qualitative observations, which showed the inoculum to be dominated by colonies rather than solitary cells. The remaining 770 ml of the P. antarctica inoculum was added to 37.8 l of station B seawater (-0.17 nM dissolved Fe) that had been filtered (0.2 iM Supor capsule filter) then chilled to -0°C in an acid-cleaned 50-l polyethylene carboy. The calculated dissolved Fe concentration in the resulting large-volume P. antarctica inoculum was -0.20 nM, which was subsequently confirmed by measurement of 0.22 ± 0.00 nM (n = 2) dissolved Fe in filtered subsamples, whereas the final concentration of EDTA in this solution was negligibly low

(-0.02 nM). This large-volume inoculum was gently mixed, and then used to fill 32 acid-cleaned 1.2-l polycarbonate incubation bottles, leaving 200 ml of headspace in each bottle. In addition, subsamples were taken from the 50-l carboy for measurements of initial chlorophyll a and dissolved macronutrient concentrations. In order to avoid inadvertent contamination with iron, the preparation of the large-volume P. antarctica inoculum and filling of the 1.2-l bottles were carried out using stringent trace-metal clean techniques, inside a plastic enclosure under positive pressure of class-100 filtered air; in addition, all materials contacting the station B filtered seawater and P. antarctica inoculums were rigorously cleaned using methods similar to those described by Sedwick et al. (2000).

Of the 32 incubation bottles, eight received no iron additions ('control treatments'), eight received an addition of 0.2 nM Fe ('+0.2 nM treatments'), eight received an addition of 0.6 nM Fe ('+0.6 nM treatments'), and eight received an addition of 1.8 nM Fe ('+1.8 nM treatments'). Iron was added as an aqueous 17.9 pM solution of ferric nitrate in 0.1% hydrochloric acid; the added nitrate and hydrochloric acid had negligible effect on the nitrate concentration and pH of the inoculum. The four iron treatments correspond to initial dissolved Fe concentrations of 0.22 nM ('control treatments'), 0.42 nM ('+0.2 nM treatments'), 0.82 nM ('+0.6 nM treatments'), and 2.02 nM ('+1.8 nM treatments'). The 32 incubation bottles were then placed in an incubator maintained at an irradi-ance of -20 iE m-2 s1 and a temperature of -0°C. For each different iron treatment, duplicate 1-l bottles were terminally sampled after incubation periods of 16, 25 and 31 days, with subsam-ples taken for measurements of dissolved macronutrients, chlorophyll a, and other chemical and biological parameters. These sampling times were chosen by following the concentration of particulate chlorophyll a in dedicated monitoring bottles (one of the eight bottles prepared for each iron treatment), which were repeatedly subsam-pled for chlorophyll measurements during the course of the experiment. Axenic conditions were not maintained in this experiment; however, in an effort to minimize bacterial contamination, the low-iron seawater used to prepare the experimental inoculums was filtered through the 0.2-pm CritiCap Supor capsule filter immediately prior to use, and the 50-l inoculum carboy and 1-l incubation bottles were thoroughly rinsed with this filtered seawater under class-100 air before filling.

Analytical methods

Dissolved inorganic nitrate + nitrite, phosphate and silicic acid were determined in samples filtered through 0.45-pm Acrodisc Supor syringe filters (Pall Corporation), using standard flow analysis methods by the Marine Science Institute Analytical Laboratory, University of California, Santa Barbara. Chlorophyll a was measured by fluorometry in a 90% acetone extract of particles collected under 1-2 psi vacuum on a Whatman GF/F filter, using a Turner Designs 10-AU fluorometer and standard Joint Global Ocean Flux Study (JGOFS) protocols (Knap et al. 1996). Water samples for the analysis of iron were collected and processed using the rigorous trace-metal clean protocols described by Sedwick et al. (1997, 2000, 2005), which are required to avoid inadvertent sample contamination. Dissolved iron was determined in 0.4 pm-filtered, acidified water samples by flow injection analysis, modified after the method of Measures et al. (1995), as described by Sedwick et al. (2005). Phytoplank-ton pigments were measured in particles collected on Whatman GF/F filters by high-performance liquid chromatography using the methods described by DiTullio and Geesey (2002).

Interpretation of experimental results

As in previous studies that have used bioassay growout experiments to examine resource limitation of phytoplankton growth (e.g. Martin et al. 1990; De Baar et al. 1990; DiTullio et al. 1993; Fitzwater et al. 1996; Hutchins et al. 2001; Sedwick et al. 2002), we infer relative net growth rates in our incubation bottles from the accumulation of algal biomass. Because changes in light intensity and iron availability are known to effect variations in the ratio of chlorophyll a to cell carbon for Phaeocystis antarctica (Van Leeuwe and Stefels 1998; Stefels and van Leeuwe 1998;

Schoemann et al. 2005), measurements of chlorophyll a may not provide a reliable estimate of relative changes in cell numbers (i.e., growth rate) in our incubation bottles. In addition, there are potential problems in using particulate organic carbon (POC) to estimate changes in the cellular biomass of colonial Phaeocystis, as a result of the POC that resides in the colonial mucilage rather than in the Phaeocystis cells (Schoemann et al. 2005). Thus, for the experiments reported here, we estimate relative net growth rates from net decreases (drawdown) in dissolved nitrate + nitrite, which are assumed to be proportional to net increases in cell carbon and cell number of colonial Phaeocystis antarctica, assuming balanced, exponential growth.

The experiments described here are principally diagnostic in nature: where cellular biomass was significantly enhanced in bottles after resource (iron or light) amendment, relative to control (or other) treatments, we infer that algal growth rates in the control (or other) treatments were limited by a deficiency in that resource. The statistical significance of differences between mean values of parameters measured in different treatments were assessed using a two-tailed t-test for comparisons between two treatments, or a one-way analysis of variance (ANOVA) for comparisons between three or more treatments, at a confidence level of 95% (P = 0.05).

0 0

Post a comment