Environmental Science & Technology 
Originally, the OWF area was characterized by low organic 
matter and medium to coarse sediment.“” The macrobenthic 
community was, similar to the low ref area, dominated by 
Nephtys cirrosa, characterized by an overall low abundance and 
diversity.” However, after the construction of the OWF and the 
introduction of hard substrates, new fouling organisms, like 
filter-feeding organisms, could colonize the turbines, resulting in 
an increase of fecal pellets, and thus a higher organic matter and 
a lower grain size fraction near the turbines. 797 This effect, 
better known as the reef effect, coincides with a shift to a more 
Gne-sediment-associated community like Abra alba°®°° and 
could explain why a higher and different abundance of 
metabolites was found in the OWFs. Additionally, an increased 
fraction of fine sand (<250 um) was also observed at the Belgian 
OWFSs compared to the low and ship ref areas (Figure 7). 
However, no increase in the TOC was observed between the 
impact and reference areas. 
c 
: 
un 
| 
“ 
= 
X 
N 
N x: A 
< & S ES Ka 
SS SS 
N KO 
2 ke 
X 
SF 
Figure 7. % of fine sand in each location at the BPNS with a particle 
grain size ranging between 0 and 250 um. 
Natural occurring chemicals could also be present in the 
HiRef, LowRef, and ShipRef areas. Some compounds, such as 
PAHs, have natural sources, but shifts in the benthic community, 
whether induced by anthropogenic or shipping activities, may 
also result in the detection of specific metabolites. 
3.10. Toward the Understanding of OWF Chemical 
Fingerprint 
A wide variety of chemicals were found, with a clear difference 
jetween OWFs. However, only a small portion has been 
;dentified. Compound identification in NTS is a challenging task 
with existing libraries and tools. For the LC-HRMS, because the 
data are collected in data-dependent mode, MS/MS data are not 
available for all features. Reinjecting the samples might help to 
increase the collection of MS/MS data, further enhancing the 
identification, but the collection will still be limited by the 
availability of libraries. Additionally, compounds for which no 
MS/MS data could be collected indicate that they did not reach 
a certain threshold, potentially related to low occurrence or low 
‘onizability under the used ESI conditions. 
Because the majority of the compounds are unidentified, it 
:emains difficult to assess the risk of these (unidentified) 
ieachates in the marine environment. Therefore, further 
screening is necessary to understand whether the detected 
compounds are persistently present in the OWFSs and to what 
extent. In addition, target analysis can be used as a 
pubs.acs.org/est 
AT TC 2 
complementary approach for a quantitative determination of 
suspected chemicals. Biological effect monitoring or dedicated 
exposure experiments could help to understand the ecotoxico- 
logical effects of the mixture of leachates without the immediate 
need to identify the separate compounds (e.g., Alter et al.°°), but 
to ensure a safe environment, hazardous substances should be 
identified and monitored on a regular basis, which would require 
further research. 
The low overlap in chemical compounds between the OWFs 
would also indicate that each OWF has a unique chemical 
fingerprint. Therefore, to identify all potential hazardous 
{eachates that can be released from OWFSs in the environment, 
it is paramount to include different OWFs, built at different 
moments in time and using different types of technology, to 
avoid overlooking important chemicals so that a good and 
robust assessment can be obtained. 
Once the compounds posing a risk to the ecosystem are 
identified and quantified, it is advisable to develop a monitoring 
program based on targeted analysis to understand their temporal 
and spatial distribution, so that actions to reduce emissions can 
be taken, if necessary.“ If the obtained list of compounds to 
monitor is extended, then it will also increase the monitoring 
cost. A risk assessment should then be considered to prioritize 
the compounds, and reduce the list to an acceptable level, taking 
into account the compounds with the highest persistence, 
bioaccumulation potential and toxicity (e.g., Yulikayani et al.”°). 
@ ASSOCIATED CONTENT 
© Supporting Information 
The Supporting Information is available free of charge at 
https://pubs.acs.org/doi/10.1021/acs.est.5c17939. 
Number of compounds detected by GC-MS and LC-— 
RMS that were identified as OWF impact, OWF 
hotspot, ship reference impact, or high reference impact, 
supporting the results illustrated in Figure 3; for GC—-MS, 
: he Kovats index is also reported (S1) (XLSX) 
‘dentification, R-score, predicted chemical family, and 
potential applications (S2) (XLSX) 
Retention time and detected mass of several PAH and 
phenol standards used to identify specific compounds 
within the NTS (S3) (XLSX) 
Additional details on the chemicals used, analytical 
methods, and QA/QC (S4) (PDF) 
ME AUTHOR INFORMATION 
Corresponding Author 
David Vanavermaete — Flanders Research Institute for 
Agriculture, Fisheries and Food, Animal Sciences Unit— 
Aquatic Environment and Quality, Ostend 8400, Belgium; 
3 orcid.org/0000-0002-2087-9062; 
Email: david.vanavermaete@ilvo.vlaanderen.be 
Authors 
Pablo Zapata-Corella — IFREMER, Chemical Contamination 
»f Marine Ecosystems (CCEM), Nantes 44311, France; 
_) orcid.org/0000-0003-0041-007X 
Karien De Cauwer — Royal Belgian Institute of Natural 
Sciences, Operational Directorate Natural Environment, 
Brussels 1000, Belgium 
Javier Castro-Jimenez — IFREMER, Chemical Contamination 
9f Marine Ecosystems (CCEM), Nantes 44311, France; 
5 orcid.org/0000-0001-8456-3932 
https://doi.org/10.1021/acs.est.5c17939 
Environ. Sci. Technol. XXXX, XXX, XXX—XXX