Studies of the microbial P-cycle during a Lagrangian phosphate-addition experiment in the Eastern Mediterranean

Abstract

Microbial uptake of orthophosphate was studied before and during a Lagrangian experiment where orthophosphate was added to the surface mixed layer in the Cyprus Gyre, Eastern Mediterranean, a region previously hypothesized to be characterized by P-limited growth of both phytoplankton and heterotrophic bacteria. The addition of ca. 110 nM orthophosphate to a ca. 16 km2 patch in situ led, within 1 day, to an increase in particulate-P from 8 to ca. 15 nM, a result in good agreement with a previous microcosm bioassay indicating this system to have a maximum capacity for orthophosphate consumption of between 10 and 25 nM phosphate. In samples of unperturbed water taken before the addition, outside, or below the experimental patch, orthophosphate turnover time (Tt) was <4 h, argued to be consistent with the assumption of diffusion-limited phytoplankton growth. Upon addition, Tt increased to 94 h. Estimates of maximum potential uptake rate (Vmax) for orthophosphate in unperturbed water exceeded by more than one order of magnitude the biological P-requirement (ν) as obtained from stoichiometric conversion of C-based primary and bacterial production values to estimated P-requirement. Upon addition of orthophosphate, Vmax decreased to a level comparable to ν. The observations are consistent with the assumption of P-starved cells before and P-replete cells with excess external orthophosphate after the addition. Orthophosphate uptake in unperturbed water was dominated by<1 μm organisms (mean ±SD between samples 0.56±0.03 μm). In samples with higher turnover time, orthophosphate uptake was shifted towards larger organisms, culminating after 5 days with a near doubling in mean size (1.08 μm). The size distribution of particulate-P standing stock had a mean size of 10 μm, indicating the presence of a substantial biomass of micro-organisms larger than those involved in P-uptake. Comparison of the measured particulate-P with microscope-based biomass estimates indicated a microbial food web dominated by heterotrophic organisms (70% of particulate-P), distributed with ca. 25% of total particulate-P in heterotrophic bacteria, ca. 40% in heterotrophic flagellates, and ca. 5% in ciliates. Concentration of bioavailable phosphate (Sn) estimated from the relationship Sn=νTi indicated Sn values<1 nM PO4 before the addition, increasing afterwards. Estimates of the sum Kt+Sn for the 0.6–0.2 μm size fraction were in the range 1–7 nM PO4 before and outside patch, suggesting this sum to be dominated by the half-saturation constant Kt. Kt+Sn increased to 69 nM after addition, then dropped over the following week back to background levels. As reported elsewhere in this volume, there was a decline in the observed chlorophyll concentrations, but a positive response in copepods. Less clear than the effects at the level of osmotroph physiology were the subsequent responses expected in the food web. Two possible mechanisms are discussed: (1) a positive response in bacterial production and the subsequent food chain of bacterial predators, and (2) a positive response in phytoplankton predators due to a shift in food quality rather than in food quantity.