Spatial Heterogeneity of Phytoplankton in a Near-Shore Environment

Abstract

An analysis of variance model for phytoplankton distribution is described, which allows the partition of the variance of chlorophyll observations into effects due to differences between stations, to differences between replicate samples at the same station, and to handling and analytical error. When the model was applied to observations of chlorophyll made in St. Margaret's Bay, N.S., it was found that, on the average, the variance between replicate samples on the same station was of the same magnitude as the analytical error variance. Significant differences were revealed between stations even at densities as high as 10 stations per 0.0625 sq miles. As the sampling area was increased (but the number of stations kept constant), the between-station variance rose rapidly, to a density of 10 stations per sq mile, after which it remained relatively constant. The maximum value of the between-station component of the variance was about 10 times the average for the between-replicate component. Heterogeneity of plankton distribution gave rise to coefficients of variation (logarithmic) up to 70% for single observations of chlorophyll concentration, for which the analytical technique has a precision of better than 10%.Analysis of chlorophyll samples taken at closely spaced stations (0.1 miles apart) along an 8-mile transect revealed considerable variability in distribution. Autocorrelation analysis of the data showed the presence of plankton patches 0.5–1.5 miles in diameter. Coherence analysis provided some evidence for correlation between the distribution of chlorophyll and of temperature and salinity. The dominant frequency in coherence spectra was −3 cycles/mile. Comparison of the autocorrelation function for chlorophyll with those for temperature and salinity indicated the influence of both deterministic and random processes on the chlorophyll distribution.Measurements taken at nine stations during the course of a plankton bloom showed that although there were differences between the stations, the time variation in chlorophyll concentration followed broadly the same course at all the stations. The variation between stations in total quantity of chlorophyll in the water column was small enough to indicate that measurements taken at a single station would be quite adequate to study major seasonal trends.