The behaviour of radiocaesium in woodland ecosystems. Measurement and modelling

In order to better quantify risk to non-human biota from environmental radioactivity, our understanding of the behaviour of radionuclides in the biosphere needs to be increased. Hence the aims of the present study were threefold: 1) Review Ecological Risk Assessment (ERA) methodologies, 2) Quantify radiocaesium distribution in woodland biota in the vicinity of BNFL Sellafield, 3) Use this information to develop radiocaesium food-chain and dosimetry models for terrestrial invertebrates and small-mammals. A review of ERA literature concluded that of the two main methodologies available (individual receptor and holistic ERA), individual receptor based analysis is the most credible method given in today's level of scientific understanding. lt was also concluded that the use of modelling techniques in ERA will increase in future years. Two study sites were sampled, a Picea sitchensis monoculture (Lady Wood) and a mixed deciduous site (Longrigg Wood). Mean levels of Cs-137 (all activities quoted in Bq/kg dry weight) in soils:leaf litter were 473:408 (Lady Wood) and 142:32 (Longrigg Wood). The activity of understorey vegetation varied with ranges of 17-508 and 4-48 Bq/kg in Lady Wood and Longrigg, respectively. No vegetation species had concentration ratios (CRs) > 1. The greatest range in Cs-137 activity (2-5242 Bq/kg DW, Lady Wood) was found in fungi, with Mycena galariculata and Hypholoma fasciculare attaining the highest biomagnifications (CRs = 2.6, 2.0 respectively). Due to radioanalytical constraints, only 'mixed invertebrate' samples were measured for Longrigg Wood (yearly average = 8.5 Bq/kg). No significant invertebrate body-burden differences were found between taxa or between seasons for each invertebrate group in Lady Wood. Mean yearly Cs-137 body-burdens (Bq/kg DW) were 94 (Diplopoda), 104 (Isopoda), 54 (Chilopoda), 120 (Araneae), 91 (Opilionidae) and 41 (Carabidae). No invertebrates had CRs > 1. Seasonal Cs-137 body-burdens were also measured for the woodmouse (Apodemus sylvaticus). Significant differences were found due to site (yearly means = 20 and 58 Bq/kg DW in Longrigg and Lady Wood respectively) and autumn body-burdens were significantly different to spring and summer cohorts (Lady Wood only). It was hypothesised that this was due to the ingestion of highly contaminated fungal fruiting bodies by Apodemus in autumn. Dynamic compartmental Cs-137 food-chain models were constructed for the invertebrates and Apodemus. The Apodemus model predictions (mean seasonal Cs-137 body-burdens) fell between the confidence intervals of the observed data in 5 out of 6 comparisons. The model underestimated body-burdens in autumn Lady Wood animals probably due to an underestimation of fungal ingestion. The invertebrate model was less successful due to a lack of calibration literature on the fundamental ecology of soil arthropods, although the model accurately predicted a mean Cs-137 body-burden for invertebrates in general. A dosimetric ERA model was constructed along with Imperial College, London and Westlakes Scientific Consulting to estimate dose to the invertebrates and Apodemus over the history of operations at BNFL Sellafield. The highest dose estimates from Cs-137 were 3.3E-04 mGy/d (small mammals) and 8.8E-04 mGy/d (invertebrates) after an unplanned release in 1979, but at no time during the period 1958-1997 did dose exceed the 1 mGy/d limit recommended by the International Atomic Energy Agency. However, further work may be needed to better quantify dose from the gut contents of receptor species and from radionuclides other than Cs-137. Based on this study, it is concluded that risk to terrestrial woodland biota from environmental radioactivity in the vicinity of BNFL Sellafield is likely to be very low. (author)