Long, narrow sea surface slicks associated with internal waves are common features of exposed coastal environments. Such slicks, which may be accompanied by foam or debris, also have been shown to be associated with high concentrations of many types of planktonic organisms. Shanks (e.g., 1983, Mar. Ecol. Prog. Ser., Vol. 13) proposed that organisms in these sea-surface slicks are concentrated by currents associated with the internal waves. Although most observations have been made near the sea surface, recently published data suggest that the effects of internal waves on the distribution of the plankton and small fishes may extend throughout the water column. Still, the magnitude of internal wave-induced changes in the concentration of planktonic organisms and its relationship to the physical forcing remain largely unexplored. Using a low order analytical model we predict the effects of high-frequency internal waves on the patchiness of planktonic organisms within the water column, and test those predictions using field data.

Simple model predictions for high-frequency interfacial waves suggest that organisms that swim vertically, in opposition to the physical flow, will experience ephemeral changes in concentration throughout the water column as a wave passes. Previous modeling studies of surface-trapped particles have shown that the maximum concentration should occur over the trough of the wave. Using two interfacial wave profiles, a small-amplitude wave and a weakly nonlinear wave, we also find that increases in the concentration of planktonic organisms within the upper layer are predicted to occur over the wave trough. That is, changes in concentration of planktonic organisms within the upper layer are predicted to be out of phase with displacements of the interface. Within the lower layer, the perturbations are predicted to be in phase. The maximum wave-induced increase in concentration is predicted to increase linearly with either the wave amplitude or the wave period, depending on the swimming behavior. Maximum increases in concentration are predicted to be less than twice the background concentration.

To compare field data to these model predictions, we have computed estimates of along-isotherm fluorescence, along-isotherm beam attenuation coefficient and isotherm depth. If organisms behave completely passively, then on short time scales we would expect the organisms to be advected with the water, similar to temperature. That is, we would expect no coherent relationship between along-isotherm concentration anomalies and isotherm displacements. Data were collected using a profiling CTD\fluorometer\transmissometer instrument package during the spring and summer of 1997 in depths of 30 m or less off Mission Beach California. Spring sampling occurred during a bloom of dinoflagellates, including Lingulodinium polyedrum, Ceratium spp. and Prorocentrum spp. Profiling data were collected at a fixed station for up to 6 hours at a time. During the summer, profiling data were also collected by sampling at a near-surface drifter, drogued at 0.5-1.5 m. High-frequency internal waves with periods less than 30 minutes were often observed, sometimes accompanied by high-frequency changes in the level of fluorescence (Fig. 1).
 

Our analyses so far show that temporal variability in the relationships between along-isotherm fluorescence, along-isotherm beam attenuation coefficient and isotherm depth are often consistent with model predictions of high-frequency internal wave-induced changes in the concentration of planktonic organisms. Estimates of the phase relationship between along-isotherm fluorescence and isotherm depth at periods between 5-30 minutes suggest that these quantities are often out of phase in the upper part of the water column, with a possible phase shift deeper down (Fig. 2). This general pattern holds for both spring and summer; however, there is a considerable amount of day-to-day variability, particularly in spring. Although the phase relationship between along-isotherm beam attenuation coefficient and isotherm depth was more variable, at periods between 5-30 min, along-isotherm fluorescence and along-isotherm beam attenuation coefficient were approximately in phase throughout the water column. For data collected in a quasi-Lagrangian sampling mode, depressions of isotherms often corresponded to localized maxima in along-isotherm fluorescence indicating that increases in fluorescence were occurring over the trough of the waves (Fig. 3). Also associated with some isotherm depressions were localized increases in the along-isotherm beam attenuation coefficient. Cell counts computed for water samples collected during the spring suggest that high-frequency variability in fluorescence at a fixed depth can be due to changes in the concentration of swimming dinoflagellates. More extensive testing of the model predictions is currently being done using data from individual internal wave events.

Figure 2. Waterfall plots showing the phase relationship between along-isotherm fluorescence and isotherm depth versus the median isotherm depth on 17 days during spring and summer of 1997.  Open circles correspond to the absolute value of the median of phase estimates at periods between 5-30 min. The horizontal bars indicate plus/minus the median absolute deviation. Panels represent data collected in 30 m water in spring (top panel), 15 m of water in summer (middle panel) and 30 m of water in summer (bottom panel). Only phases at frequencies for which the squared coherency was significantly greater than zero (5% level) were used to estimate the phase in this frequency band. The number of frequencies associated with each phase estimate is shown to the right of the point estimate. The horizontal axis shows the phase in radians from -p to p . Note that p radians for a previous profile corresponds to -p radians for the next profile. The solid gray vertical lines correspond to 0 radians and the dashed gray lines correspond to +/-p radians. Dates of sampling are shown at the bottom-right of each zero line. Our simple model would predict along-isotherm fluorescence and isotherm depth to be out of phase (i.e., phase = p ) above the pycnocline and in phase (i.e., phase = 0) below the pycnocline.