The few anxious days reading through my PhD thesis in preparation of my viva turned out to be time well spent, especially as after my viva examination I had to leave the room, and when I was invited back I was greeted by “Congratulations Dr Beriro”! Many people may dread the PhD viva and perhaps I’m slightly bonkers to think otherwise, but I found the experience rather cathartic. Having spent the past 4 year preparing work toward my thesis, it was a great relief to finally get the whole thing off my chest with two people who had read and enjoyed my work.
My thesis title was: “Gene expression programming models of human ingestion bioaccessibility of benzo[a]pyrene in coking-works soils”My EPSRC and University of Nottingham funded research incidentally started out life in Chile for one year, reading and reviewing literature on modelling land contamination problems e.g. ground gas, hydrocarbons in soil, groundwater flow. My expectation at the time was that there would be a perfect dataset for me to use and explore using a novel stochastic symbolic regression method called Gene Expression Programming. The method looks for patterns in data and expresses them as mathematical equations. These equations can then be used to predict unknown samples and tell us something about the process being modelled. The method uses the principles of biological evolution to ‘evolve’ multiple ‘generations’ of ‘populations’ comprising hundreds of models. The ‘fittest’ of these models ‘survive’ and are selected for further evaluation.
Back to the search for data…there was of course no perfect dataset BUT after returning to the Nottingham and attending a regional contaminated land group meeting, I listened to a talk by Dr Mark Cave where he explained how he was modelling bioaccessibility. I asked whether we could do some work together and within a few months I was working with Dr Michael Watts , Dr Christopher Vane , Dr Joanna Wragg and Dr Mark Cave at BGS. The opportunity to create my own dataset from soils sampled from the former Avenue Coking and Chemical Works near Chesterfield became a reality. My goal was to determine, analytically, what proportion of Benzo[a]pyrene was released from the soil when exposed to simulated gastro-intestinal fluids. The laboratory work was completed in the Centre for Environmental Geochemistry.
The importance of contaminant bioaccessibility is that when children use garden for play they eat, by hand-to-mouth contact, a small amount of soil. This soil may contain contaminants that are harmful to their health (carcinogenic, mutagenic etc.). The assumption in generic risk assessment is that all of the contaminants in soil will be absorbed and have the potential to do harm. In reality this is not true and some will be excreted with remnant of the soil – the amount dissolved prior to absorption is the bioaccessible fraction. BGS have worked tirelessly over the past 10 years and more on developing in vitro models for measuring bioaccessibility (their inorganic method is now an industry standard and their organic method won a Brownfield Briefing industry award). These tests are now used across the globe to reduce the conservatism in risk assessment and lower the costs (environmental, social and economic) of risk-based land management. Modelling these data with physico-chemical soil properties can reduce the reliance on laboratory testing and increase the potential use of bioaccessibility in risk assessment.
Part of my job as an Environmental Geochemist is to continue to measuring and modelling bioaccessibility in the Medical Geology sector at BGS. Thanks to the generous support of my supervisors (Prof Paul Nathanail and Dr Bob Abrahart) at the University of Nottingham (School of Geography) and BGS colleagues, I now practice what I preached (to many less than willing friends).
The PhD process, the collaboration with BGS and the viva all make for an interesting and enjoyable journey. Viva the Viva!
By Darren Beriro
BGS Environmental Geochemist