Negative priming effect on organic matter mineralisation in NE Atlantic slope sediments.

The priming effect (PE) is a complex phenomenon which describes a modification (acceleration or retardation) in the mineralisation rate of refractory organic matter (OM) following inputs of labile material. PEs are well-studied in terrestrial ecosystems owing to their potential importance in the evolution of soil carbon stocks but have been largely ignored in aquatic systems despite the fact that the prerequisite for their occurrence, i.e. the co-existence of labile and refractory OM, is also true for sediments. We conducted stable isotope tracer experiments in continental margin sediments from the NE Atlantic (550-950 m) to study PE occurrence and intensity in relation to labile OM input. Sediment slurries were treated with increasing quantities of the (13)C-labelled diatom Thalassiosira rotula and PE was quantified after 7, 14 and 21 days. There was a stepwise effect of diatom quantity on its mineralisation although mineralisation efficiency dropped with increasing substrate amounts. The addition of diatomaceous OM yielded a negative PE (i.e. retardation of existing sediment OM mineralisation) at the end of the experiment regardless of diatom quantity. Negative PE is often the result of preferential utilisation of the newly deposited labile material by the microbial community ("preferential substrate utilization", PSU) which is usually observed at excessive substrate additions. The fact that PSU and the associated negative PE occurred even at low substrate levels in this study could be attributed to limited amounts of OM subject to priming in our study area (~0.2% organic carbon [OC]) which seems to be an exception among continental slopes (typically >0.5%OC). We postulate that PEs will normally be positive in continental slope sediments and that their intensity will be a direct function of sediment OC content. More experiments with varying supply of substrate targeting C-poor vs. C-rich sediments are needed to confirm these hypotheses.