Seawater was collected on 13 April 2006 at the surface, in the Morbihan Bay of the Kerguelen Archipelago (49° 22’ S, 70° 12’ E), 2 km from the Port-au-Français scientific station. Samples were immediately carried to the biological laboratory of this station, where the experiments were conducted. Seawater samples (600 mL) were added to sterile 1L-Erlenmeyer flasks, together with 0.5 mL of either Arabian light crude oil or diesel fuel. Oil was added alone to seawater or with Inipol EAP 22 organic fertilizer (10% V/V; Elf Atochem, Colombes, France). The Arabian light crude used in this experiment was liquid petroleum with a high American Petroleum Institute gravity (32°), a low pour point of −28°C and a low dynamic viscosity of 31 cP at 0°C. The Inipol EAP22 is a micro-emulsion containing butoxyethanol, oleic acid and surfactants with a dynamic viscosity of 250 cP at 20°C and a pour point at 11°C.

One flask for each treatment (i.e., non-amended seawater; seawater amended with crude oil only; seawater amended with crude oil and Inipol EAP 22; seawater amended with diesel fuel only; seawater amended with diesel fuel and Inipol EAP 22) was incubated at three different temperatures (20, 10 and 4°C) for 7 weeks under orbital stirring.

A minimum of 50 mL of seawater was pre-filtered through a 3 µm-pore filter (47 mm, Polycarbonate Nuclepore^®, Whatman, Kent, UK) in order to remove large organisms and then filtered through a 0.22-µm pore filter (47 mm, Polycarbonate Nuclepore) before being stored at −80°C until analysis. Cellular lysis was performed in two steps on a lysis buffer, containing 50 mM Trizma base (Sigma-Aldrich, St. Louis, MO, USA), 40 mM ethylenediaminetetraacetic acid (Sigma-Aldrich) and 0.75 M Sucrose (Sigma-Aldrich). First, a freshly prepared lysozyme solution (final concentration 1 mg mL⁻¹) was added to the 0.22-µm pore filter and samples were incubated at 37°C for 45 min. Second, sodium dodecyl sulphate (final concentration 1%) and proteinase K (final concentration 0.2 mg mL⁻¹) were added and tubes were incubated at 50°C for 1 h.

A volume of 600 µL of the resulting cellular lysate was treated with 10 µL of a 100 mg mL⁻¹ RNase A solution (Qiagen, Hilden, Germany) before DNA extraction which was carried out using the DNeasy Tissue kit (Qiagen), as described by the manufacturer instructions.

Samples for RNA analysis were processed in parallel to DNA extraction as previously described (Ghiglione et al. 2009). Briefly, the remaining 400 µL of cellular lysate was treated with DNase I for RNA extraction using an SV Total RNA Isolation kit (Promega, Madison, WI, USA). Effective DNA digestion was verified before cDNA synthesis by negative 16S rDNA polymerase chain reaction (PCR) amplification (PCR conditions as reported below). RNA was reverse transcribed (RT) into single-strand cDNA using Moloney Murine Leukemia Virus (M-MLV) reverse transcriptase (Promega), according to the manufacturer's instructions. PCR amplification of cDNA was carried out as for DNA (see below).

The highly variable V3 region (E. coli 16S rRNA gene positions 329-533; Brosius et al. 1981) of the 16S ribosomal rRNA gene (16S rDNA) was amplified for both DNA and cDNA extracts. The bacterial universal primer set w49 forward (5'ACG GTC CAG ACT CCT ACG GG-3’; Delbès et al. 2001) and w34 reverse (5’-TTA CCG CGG CTG CTG GCA C-3’; Lee et al. 1996) were used. The reverse primer was fluorescently labelled with 5’-tetrachloro-fluorescein-CE phosphoramidite (TET, Applied Biosystems–Life Technologies, Norwalk, CT, USA) at the 5’-end position. PCR amplification of 50-µL reactions containing template, primers (final concentration 0.25 µM), dNTPs (Eurogentec, Seraing, Belgium; final concentration 0.6 mM), Pfu DNA polymerase (1 U, Promega) and Pfu reaction buffer was run in a Robocycler thermocycler (Stratagene, Agilent Technologies, Santa Clara, CA, USA). PCR steps were initial denaturation at 94°C for 1 min, followed by 25 cycles of denaturation at 94°C for 1 min, annealing at 61°C for 15 s and extending at 72°C for 30 s with a final 10 min extension at 72°C. Size (ca. 200 bp length) and concentration of PCR products were determined by agarose gel electrophoresis (2%) with a DNA size standard (Low DNA Mass Ladder, Gibco–Life Technologies, Carlsbad, CA, USA).

Successive dilutions in nuclease-free sterile water were made to obtain 10 ng µL⁻¹ of PCR product. One µL of the dilution was mixed with 18.9 µL of formamide and 0.1 µL of the internal size standard Gene-Scan-400 Rox (Applied Biosystems), denatured at 94°C for 5 min, and immediately cooled on ice for 5 min before electro-kinetical injection (5 s, 12 kV) into a capillary tube (47 cm×50 µm) filled with 5.6% of Genescan polymer in a ABI Prism 310 Genetic analyser (Applied Biosystems). Electrophoresis was carried out at 15 kV for 30 min sample⁻¹ at 30°C, and data were collected with ABI Prism 310 collection software (Applied Biosystems). TET-labelled fragments were detected by a laser with a virtual C filter (detection wavelengths: 532, 537 and 584 nm).

Total fingerprinting area was normalized among all the aligned CE-SSCP profiles obtained using the internal size standard Gene-Scan-400 Rox throughout the Statistical Analysis of Fingerprints Using Molecular Biology (SAFUM) software (Zemb et al. 2007). Similarity matrices based on Bray-Curtis distances and unweighted pair group method with arithmetic mean (UPGMA) dendrograms were generated using the Primer-E5 software package (Clarke & Warwich 2001). One-way Anosim tests for significant differences in bacterial community structure were also performed using Primer-E5. Peak detection and Simpson index values were automatically obtained from each fingerprint using the SAFUM software.

16S rDNA-based fingerprints of samples incubated with the oleophilic fertilizer clustered separately from those incubated only with crude or diesel oil (Fig. 1a, clusters 3 and 7, and Fig. 1b, clusters 6 and 7), suggesting nutrients addition to induce changes in total bacterial community structure. However, these changes occurred mostly at high temperatures (i.e., 10°C and 20°C), since fertilizer-enriched incubations made at low temperature (i.e., 4°C) induced mostly no variations in structure (Fig. 1a, clusters 3 and 4, and Fig. 1b, clusters 4 and 5). In the presence of crude oil, fertilizer-induced bacterial community structure changed between the first and the seventh week of incubation, with similar changes observed at 10 and 20°C (Fig. 1a, clusters 3 and 7). Conversely, fertilizer-induced structure remained stable over time in the presence of diesel (from the third week at 10°C), but with separated clusters obtained at 10 and 20°C (Fig. 1b, clusters 6 and 7). Compiling all data, Anosim tests showed weak structural differences induced by Inipol EAP 22 in diesel incubations (Global R=0.145; significance level of sample statistic=3.7%), and no differences in crude oil incubations. (Global R=0.067; significance level of sample statistic=16.4%).

Clusters containing mainly samples incubated at the same temperature were more frequently observed in the diesel-related dendrogram (Fig. 1b, clusters 2–6) than in the crude oil-related dendrogram (Fig. 1a, cluster 6), indicating the influence of temperature on total community structure to be more important in the presence of diesel. Anosim tests confirmed these temperature-related structural differences in all diesel incubations (Global R=0.1; significance level of sample statistic=12.3%), and for crude oil incubations the absence of such differences (Global R=0.024; significance level of sample statistic=33.3%).

In fact, a slight different impact of oil mixture alone was observed for the total bacterial community structure. Crude oil-addition induced structural changes at all temperatures, as shown by separate clusters containing controls and crude oil-amended samples (Fig. 1a, clusters 1 and 2). However, a simple diesel addition caused no systematic changes in the structure of total bacterial communities (Fig. 1b, clusters 2 and 3).

The structure of initial bacterial communities was similar to that of non-amended controls incubated at 4°C, from the first to the seventh week of incubation (Fig. 1a, b, cluster 1). This indicates a low impact of confinement in bacterial community structure along the study.

The 16S rRNA-based dendrogram provided different results in the ongoing dynamics of the active bacterial fraction. Samples collected after 1 and 3 weeks, including non-amended controls, strongly resembled one another, constituting well separated clusters in both crude oil and diesel-related dendrograms (Fig. 2a, cluster 1; Fig. 2b, cluster 2). The sample corresponding to the initial communities clustered separately from these clusters (Fig. 2a, b). Differences in active community structure related to time of incubation were significant in both crude oil (Global R=0.228; significance level of sample statistic=0.2%) and diesel incubations (Global R=0.498; significance level of sample statistic=0.1%).

Samples incubated in the presence of fertilizer primarily clustered separately from those incubated in the presence of crude oil only (Fig. 2a, clusters 1, 2, 5 and 6), whereas they mostly clustered together with those incubated with diesel only (Fig. 2b, clusters 2, 3 and 5). Thus, fertilization induced changes in active community structure depending on the oil mixture tested, but temperature had a low influence on structure: no temperature-related clusters were found either in crude oil or in diesel fuel-related 16S rRNA-based dendrograms (Fig. 2).

Simpson's index values of 16S rDNA and 16S rRNA CE-SSCP fingerprints obtained from crude oil and diesel amended cultures were relativized to values obtained from non-amended controls (Fig. 3).

Relative Simpson's index values of total bacterial communities (16S rDNA) initially decreased in the presence of both crude oil and diesel at 10°C and 20°C (Fig. 3a, c), whereas this reduction occurred later on at 4°C (Fig. 3a, c), indicating some influence of temperature in diversity dynamics along the incubation period. Diversity remained low after seven incubation weeks except at 10°C in the presence of both crude oil and diesel. In the presence of fertilizer, the diversity of the whole community approached control values after seven incubation weeks, showing minor differences among temperatures (Fig. 3b, d).

Regarding values for active bacterial populations, lower diversity reduction was generally observed after crude and diesel oil addition (Fig. 3e, g). Fertilizer addition tended to reduce diversity, and this reduction occurred earlier at 4°C (Fig. 3f, h).