Tritium and 14C background levels in pristine aquatic systems and their potential sources of variability.

Tritium and (14)C are currently the two main radionuclides discharged by nuclear industry. Tritium integrates into and closely follows the water cycle and, as shown recently the carbon cycle, as does (14)C (Eyrolle-Boyer et al., 2014a, b). As a result, these two elements persist in both terrestrial and aquatic environments according to the recycling rates of organic matter. Although on average the organically bound tritium (OBT) activity of sediments in pristine rivers does not significantly differ today (2007-2012) from the mean tritiated water (HTO) content on record for rainwater (2.4 ± 0.6 Bq/L and 1.6 ± 0.4 Bq/L, respectively), regional differences are expected depending on the biomass inventories affected by atmospheric global fallout from nuclear testing and the recycling rate of organic matter within watersheds. The results obtained between 2007 and 2012 for (14)C show that the levels varied between 94.5 ± 1.5 and 234 ± 2.7 Bq/kg of C for the sediments in French rivers and across a slightly higher range of 199 ± 1.3 to 238 ± 3.1 Bq/kg of C for fish. This variation is most probably due to preferential uptake of some organic carbon compounds by fish restraining (14)C dilution with refractory organic carbon and/or with old carbonates both depleted in (14)C. Overall, most of these ranges of values are below the mean baseline value for the terrestrial environment (232.0 ± 1.8 Bq/kg of C in 2012, Roussel-Debet, 2014a) in relation to dilution by the carbonates and/or fossil organic carbon present in aquatic systems. This emphasises yet again the value of establishing regional baseline value ranges for these two radionuclides in order to account for palaeoclimatic and lithological variations. Besides, our results obtained from sedimentary archive investigation have confirmed the delayed contamination of aquatic sediments by tritium from the past nuclear tests atmospheric fallout, as recently demonstrated from data chronicles (Eyrolle-Boyer et al., 2014a,b). Thus Sedimentary archives can be successfully used to reconstruct past (14)C and OBT levels. Additionally, sediment repositories potentially represent significant storages of OBT that may account for in case of further remobilisation. We finally show that floods can significantly affect the OBT and (14)C levels within suspended particles or sediments depending on the origin of particles reinforcing the need to acquire baseline value range at a regional scale.