If we had applied the same type of ad spectra correction as Babin et al. (2003b) did, we would have obtained a different average value of Sd – 0.0098 nm−1

(±0.0028 nm−1). This last value is still smaller but closer to the values reported by Babin et al. (2003b). Such a hypothetical average Sd would also be comparable to the values reported by Bricaud et al. (1998) for their oceanic samples (note that Bricaud et al. in their work also forced ad spectra to reach CX-4945 clinical trial zero at 750 nm). These authors estimated the average Sd to be 0.0110 nm−1 (± 0.0020 nm−1) when they took into account all their data and also an average value of 0.0100 nm−1 (±0.0010 nm−1) for the group of samples with Chl a restricted to the range between 1 and 10 mg m−3. Figure 4c shows the spectra of TSA HDAC order the mass-specific absorption coefficients of detritus ad*(λ) for all samples recorded as well as the average spectrum and range of its SD. The variability in this case is distinctly greater than the variability of ap* presented earlier. Examples of average

ad* values and the corresponding CVs can be found in row 9 of Table 2. Those CV values lie between 100% and 125%, which indicates that the relationships between ad(λ) and SPM are weak in the case of our data; this is also illustrated by the spread

of the data points (ad(440) vs. SPM) in Figure 5f. Additionally, for comparison, in Figure 5f we also present two curves plotted according to the results of Babin et al. (2003b). They reported an average value of ad* (443) of 0.067 m2 g−1 (±0.022 m2 g−1) for their Baltic samples and an average PAK5 ad*(443) value of 0.041 m2 g−1 (±0.023 m2 g−1) for all their samples from coastal waters around Europe. These cited average ad* values converted to wavelength 440 nm (with the help of the already-mentioned average slopes of ad spectra) would reach values of 0.070 and 0.043 m2 g−1 for Baltic and all coastal samples respectively. Our average ad*(440) value for southern Baltic samples (0.056 m2 g−1) lies between those two values, but the variability in the case of our samples is much higher (SD = 0.058 m2 g−1) (note here, too, that the differences in the infrared signal correction should only have a minor influence on the magnitude of ad at 440 nm). In the case of the absorption coefficient of detritus normalized to POC (see the values in the last row of Table 2) and also normalized to Chl a and POM (not shown), the variability is even greater than the variability of ad*(λ).