Search

In the shallow micro-tidal cove Kertinge Nor, Denmark, a series of field campaigns were conducted from April 1995 to September 1996. During these campaigns, the effect of benthic grazing on phytoplankton concentrations was studied using a dual approach. In the first approach, the density, size distribution and in situ growth of 3 dominant benthic grazers were determined to assess grazing potential and its realisation. In the second approach, the realised grazing potential was estimated from the decline in area-specific chlorophyll a (chl a) concentrations after breakdown of stratification. Forcing functions were monitored continuously during the field campaigns using meteorological data, current meters, temperature loggers and CTD point measurements. In addition, measurements from a local monitoring program in Kertinge Nor were included in the data analysis. Stratification of the water column in the cove was mainly governed by wind speed and solar radiation and occurred 50 to 75% of the time. The potential grazing pressure of the benthic suspension feeders varied but was always greater than that required to graze the entire volume of the cove per day. Using both approaches, it was estimated that realised grazing was ~50% of the potential. The lack of realisation of the grazing potential could be attributed to a lack of mixing of the water column, which resulted in strong vertical gradients in concentrations of chl a. The primary mode of decoupling between benthic suspension feeders and phytoplankton was the stratification of the water column, which created refuges for the phytoplankton.

A filter feeding bivalve, represented by the mussel Mytilus edulis, is modelled as a pump, consisting of the bands of lateral cilia along the gdl filaments, and a system of afferent and efferent canals. The relation between pump pressure a H p and pumping rate V was approached through studies of the relations between system resistances and V. The most important resistances distinguished were the frictional resistance m the canal system n H l , the exit loss 'H,,, and the back pressure a H l 2 , imposed on the pump as a means to study the performance characteristics of the gill system. Removal of the laterofrontal cini from the flow path by treatment with serotonin did not measurably affect the frictional resistance, but the cirri presumably constituted a flow-rate-independent resistance nH1, , arising from their beating against the flow. Thus, in steady state, a H p = A H , + AH,, + L H , ~ -t &H1,. A H , and AH,, were linear functions of V, whereas AH,, was a quadratic function of V. At values of V up to the maximum flow rates, n H , and AH,, dominated the system resistance, and the pump characteristic was therefore approximately linear. The interfilament canals and the exhalant siphon presumably constitute the major resistances, in an optimally pumping 'standard' 35 mm mussel, thus constituting head losses of about 0.4 mm and 0.5 mm H,O, respectively. aHi1 amounted to S0.4 mm H 2 0 . The operating point at zero back pressure was estimated at about 1.4 mm H,O. Removal of the laterofrontal cirri from the flow path drastically decreased the efficiency with which S 6 pm algal particles were retained by the gills, but it had no or little effect on the retention of 14 urn particles, supporting previous lndlcations that the laterofrontal cirri do not act as sieves of the through current, but rather as modulators in the process of particle retention. Power output of the 35 mm mussel at zero back pressure was 14 pW, only slightly lower than the maximum of 15 pW. The work done by the pump constituted about 1.5 % of the total aerobic metabolic rate of the mussel.

Correlations between rates of ventilation and oxygen consumption in filter feeders are usually interpreted in terms of causal relations, but authors disagree on whether ventilation is regulated to meet respiratory needs or whether oxygen consumption reflects energetic costs of water transport. It seems, however, that neither causality applies, the correlations being determined by physical characteristics of the ventilatory systems. The ventilatory currents both in a tube dwelling worm, Urechis caupo, and in bivalves, such as Mytilus edulis, are laminar, and oxygen in the water is available to the animals solely by diffusion. Various estimates of the diffusive boundary layer 6, developing at the interface between the ventilatory current and the body surface of U. caupo and mantle cavity in M. edulis, were in good agreement. The estimates of 6 amounted to about 0.1 mm, or a small fraction of the thickness of the ventilatory current, indicating that the larger part of the dissolved oxygen in the current is unavailable to the animals. At low rates of ventilation oxygen uptake is governed by the condition of diffusion in the ambient water. Oxygen consumption increases steeply with increasing ventilation until threshold levels are reached, and processes within the body become limiting for the rate of oxygen consumption. Oxygen consumption approaches independency of ventilation at ventilatory rates amounting to only one-fifth or less of the rates of the undisturbedly filtering animal. Regulation of respiration by means of ventilatory regulation thus seems incompatible with filter feeding in habitats that require high rates of water processing. The energetic costs of ventilation are insignificant. Presumably, in most habitats, water processing at low costs is a prerequisite for a life based on filter feeding.