Comparison of sampling and analytical methods for monitoring of cyanobacteria-dominated surface waters.

Different sampling methods can often produce different results when cyanobacterial blooms are intense and there are surface scums. Accordingly, five commonly used sampling methods and four sampling times were compared for monitoring of Oscillatoria and Microcystis populations. The different methods and times led to significantly different results for cyanobacterial biomass and its variance. Sampling methods that tended to be more specific to surface scums resulted in much higher and more variable measurements. Although sampling later in the day greatly reduced variances, the measured biomass was much lower than at earlier times of day, which was attributed to cyanobacterial downward migration in the afternoon. These trends were common for Oscillatoria and Microcystis. In order to overcome such problems, the median and median absolute deviation (MAD) were proposed, instead of the arithmetic mean and standard deviation (SD), for presenting a central tendency of measured values. Although the mean ± SD has been widely used to express a central tendency, it is too sensitive to outlying values that are very common during cyanobacterial blooms. The large differences in the mean values between the sampling methods and times were reduced by using the median values. Furthermore, the median ± MAD revealed the real data distribution more effectively. [ABSTRACT FROM AUTHOR]