Non-target Analysis and Chemometric Evaluation of a Passive Sampler Monitoring of Small Streams

Complex multivariate datasets are
generated in environmental non-target
screening (NTS) studies covering different sampling locations and
times. This study presents a comprehensive chemometrics-based data
processing workflow to reveal hidden data patterns and to find a subset
of discriminating features between samples. We used ANOVA-simultaneous
component analysis (ASCA) to disentangle the influence of spatial
and seasonal effects as well as their interaction on a multiclass
dataset. The dataset was obtained by a Chemcatcher passive sampler
(PS) monitoring campaign of three small streams and one major river
over four sampling periods from spring to summer. Monitoring of small
streams is important as they are impacted by non-point source introduction
of organic micropollutants (OMPs). The use of a PS provides a higher
representativeness of sampling, and NTS broadens the range of detectable
OMPs. A comparison of ASCA results of target analysis and NTS showed
for both datasets a dominant influence of different sampling locations
and individual temporal pollution patterns for each river. With the
limited set of target analytes, general seasonal pollution patterns
were apparent, but NTS data provide a more holistic view on site-specific
pollutant loads. The similarity of temporal pollution patterns of
two geographically close small streams was revealed, which was not
observed in undecomposed data analysis like principal component analysis
(PCA). With a complementary partial least squares-discriminant analysis
(PLS-DA) and Volcano-based prioritization strategy, 223 site- and
45 season-specific features were selected and tentatively identified.