Threshold of gross primary production for planktonic metabolic balance in the Southern Ocean: An experimental test

The proposed threshold planktonic gross primary production (GPP) value for O2 of 2.05 mmol m23 d21 separating net heterotrophic from net autotrophic communities in the Southern Ocean was tested experimentally using large mesocosms (20 m3). A set of eight mesocosms was moored in Johnson’s Dock (62839.5769S, 60822.4089W, Livingston Island, Antarctica) and a gradient of GPP was experimentally generated by imposing four light levels (100%, 50%, 25%, and 10%) in the presence or absence of nutrient additions (0.1 mol NH4Cl, 0.1 mol of F6Na2Si, and 0.01 mol KH2PO4 per mesocosm per day). The experimental treatments resulted in a broad range of chlorophyll a (Chl a) (0.31–93.5 mg m23) and GPP (O2, 0.17–16.7 mmol m23 d21). Community respiration (R) increased with increasing GPP, but not proportionately, resulting in a range of P : R ratios ranging from 0.12 in intensely shaded communities to 1.3 in those receiving high irradiance and nutrient additions, with the compensation irradiance for community metabolism (i.e., percentage irradiance at P : R 5 1) being reached at 83% of the ambient irradiance. The experimental estimate of the threshold GPP for metabolic balance of the community investigated (i.e., P : R 5 1) was 2.2 (SE 6 0.016) mmol O2 m23 d21, thereby validating the estimate of 2.05 mmol O2 m23 d21 derived in the past from comparative analyses of planktonic metabolism across the Southern Ocean. The validation of this threshold suggests that net heterotrophic planktonic communities may be more prevalent in the Southern Ocean than hitherto believed. The Southern Ocean holds a significant potential for atmospheric CO2 uptake through thermodynamic processes (Watson and Orr 2003), yet its actual contribution to anthropogenic CO2 uptake is minor (Sabine et al. 2004), as models suggest that the biota in the Southern Ocean act as a CO2 source (Watson and Orr 2003) despite high reported spring blooms (e.g., Ducklow 2003). However, these blooms also support an important heterotrophic activity, carbon demands of which may exceed primary production at places (Odate et al. 2002; Agustı́ et al. 2004). Indeed, Agustı́ et al. (2004) report, on the basis of a comparative analysis of planktonic metabolism in the Southern Ocean, that planktonic respiration (R) exceeds gross primary production (GPP) in communities where GPP , 2 mmol O2 m23 d21. This finding is consistent with a general tendency for unproductive planktonic communities to be net heterotrophic (i.e., GPP , R, Duarte and Agustı́ 1998). The existence of such thresholds is important to help delineate the area of the Southern Ocean supporting net heterotrophic communities at any one time, the metabolism of which must be supported by organic carbon excedents produced in the past or elsewhere, and to ascertain the critical inputs of limiting resources such as iron required to elevate GPP above this threshold, thereby yielding net autotrophic (GPP . R) communities. However, Agustı́ et al. (2004) derived the GPP threshold of O2 of 2 mmol m23 d21 separating net autotrophic from net heterotrophic communities on the basis of a comparative analysis of reports of planktonic metabolism in the Southern Ocean, and the validity and predictive power of this threshold remains to be validated. Recently, Duarte et al. (2004) reported a GPP threshold of 3.95 mmol O2 m23 d21 for GPP 5 R for the Mediterranean littoral, and validated this threshold through an independent mesocosm experiment where GPP was increased through a series of nutrient inputs. The experimental manipulation of planktonic communities in large mesocosms, albeit cumbersome, has been proven feasible in Antarctic waters (Agustı́ and Duarte 2000). We therefore used a series of large (20 m3) mesocosms to experimentally generate a gradient of GPP, and examined the metabolism of the planktonic community to test the proposed threshold GPP value of 2.05 mmol O2 m23 d21 separating net heterotrophic from net autotrophic communities (Agustı́ et al. 2004). The ESEPAC (Estı́mulo Experimental de la Producción Antarctica Costera) experiment was conducted in Johnson’s Dock (62839.5769S; 60822.4089W, Livingston Island, Antarctica), a sheltered bay receiving glacial melt water (Agustı́ and Duarte 2000). The experimental manipulations aimed at generating a gradient of GPP, encompassing the predicted threshold value of 2 mmol O2 m23 d21. As the communities present in these coastal waters are already productive ones (3.6 6 1.0 mmol O2 m23 d21; Agustı́ and Duarte, 2000), the manipulations required to achieve the required gradient of GPP included changes in the growth conditions both to stimulate (nutrient additions) and depress (shading) primary production. A floating platform holding eight mesocosms, consisting of large (14 m tall, 2.3 m diameter) ultraviolet (UV)-stabilized polyethylene bags, was moored in the deepest sector of the bay (about 25 m) on 21 January 2000, with the top meter of the mesocosms extending above the water level as to avoid accidental intrusions of water by breaking waves. The mesocosms were filled with ambient, unscreened water (from a depth range of 0.5 to 3 m) on 23 January, resulting in an average volume of 20 m3 in the mesocosms. The natural communities enclosed included grazers (mainly copepods, salps, and amphipods) at concentrations comparable to those in the ambient waters. The outer surface and top of duplicate mesocosms were covered with neutral screens