7322894067

48

to cnsure that the error in estimation of cellular abundance remained within the limits of + 10% (Uehlinger, 1964). Abundance (cells ml'¹) and specific biovolume (mm³ ml*¹), were estimated based on number of cells, as well as celi measures of length, width and thickncss (Lewis, 1976). The dominant species of the phytoplankton community were chosen as the species representing morę than 5% of the total biomass. The species determination was then used to calculate the Shannon diversity index (//'), using the formula:

H’=-Xpi log(pi)    (3.4)

where p₍ is the proportion of the total count arising from the /th species.

3.5.4 Statistical analysis

Mean comparisons were conducted using the paired T-test, comparing the diffcrent trcatments (1, 5, 10, 50, 100, 500 and 1000 pg I'¹) to the control trcatmcnt, for growth indicators, pigment and shikimate content, as well as enzymatic activity. Data were tested for normal distribution using the Shapiro-Wilk test and for homogeneity of variances. Except for the dry weight, the other variables did not fulfill these conditions and so, were logarithmically transformed. AU statistical analyzes were performed by using JMP 1(X software from SAS lnstitute.

3.6 Rcsults

3.6.1 Glyphosate pre-exposure and shikimate content

The prcsence of glyphosate in the stream water where the phytoplankton sample was initially collected depicted a concentration of Ipg l'¹ (± 0.02 pg l'¹) (data not shown). The shikimate content, meanwhile, was measured after glyphosate exposure, showing a significant inerease in phytoplankton cells exposed to 500 and 1000 pg I*¹ of glyphosate (figurę 3.1).