RESUMEN
Marine oil spills are typically followed by complex clean-up and monitoring operations of the shorelines affected. In this regard, determination of the changes in the chemical composition of the spilled product is crucial for evaluation purposes. The "GC-MS fingerprint" approach requires the identification of several key parameters in order to differentiate between weathering processes. In this context, we performed a 900-day study (eight sampling campaigns) of natural attenuation on a rocky shore affected by a heavy fuel oil spill. The residues coating the rocks were studied by extraction, separation and evaluation of SARA fractions, followed by GC-MS analysis for saturates and aromatics. The results identified a group of diagnostic ratios with irregular capabilities to differentiate between volatilization, photodegradation, and biodegradation (using n-alkanes, isoprenoids and PAHs), while a second group of primarily stable ratios (using dibenzothiphenes, tricyclics and hopanes) was also obtained. Notably, this is the first time that some of these ratios have been used for marine spill monitoring purposes. The ratios were applied not only to evaluate weathering at the study site, but also to obtain a comparison with the degradation of floating oil slicks-a demonstration that weathering of the latter was quicker than that of oil on the shore rocks. Additionally, Pyrolysis-GC-MS analysis of the asphaltene fraction of residues coating the rocks revealed moderate changes in the composition of this initially recalcitrant fraction. These changes were consistent with those found in the proportion of resins and asphaltene fractions over time.