What happened in the experiment
The Institute of Molecular and Cell Biology at the University of Tartu, with Norwegian partners, ran a three-month experiment in controlled tanks that mimic the formation of sea ice. When crude oil and a dispersant were added, the community changed: cold-adapted genera such as Oleispira, Bermanella and Pseudoalteromonas became dominant. The first two carry a broad set of genes linked to alkane degradation, while Pseudoalteromonas has limited hydrocarbon-degradation capability.[1]
The expected second step did not follow. Although the proportion of oil-degrading bacteria rose, the biodegradation of crude oil did not accelerate markedly. The researchers give the reason themselves: microbial abundance in sea ice is extremely low and the conditions are harsh. The community changed; the workload did not.[1]
How much each result can carry
This design is well suited to seeing a shift in community composition. Establishing who multiplied, by sequencing, can be done reliably in a closed tank followed for three months. The question of rate is different: whether degradation sped up depends on the measured mass loss being separable from the experiment's noise, and that separation narrows when cell numbers are low. Two results from the same experiment cannot be read with the same strength.[1]
The researchers' explanation also draws the boundary of the negative result. If the absence of acceleration comes from low cell density, the finding says that the dispersant produced no measurable effect at that density rather than that it does not work. Another explanation remains standing: the dispersant may not genuinely make the oil available to bacteria inside the brine channels of the ice, in which case cell number is not the limiting factor. The write-up does not choose between the two.[1]
The design that would make the result usable
That distinction does not stay on paper. Whether an oil spill in the Arctic should be met with a dispersant is argued precisely by reference to laboratory findings of this kind. What is in hand is a single rate measured at a single microbial density. Under spill conditions, mixing, currents and light differ, so natural abundance may differ from this experiment's as well.[1]
The design that would test this is clear. If the same mesocosm setup is repeated with the inoculated cell density varied, and the degradation rate rises with density, the negative result belongs to the system's biomass rather than to the dispersant. If the rate stays independent of density, the explanation is not to be found in bacterial numbers. A published dose-response curve would separate those two cases and make today's negative result usable.[1]