Home > Research > Projects > Re-shaping models to forecast faecal pathogen r...
View graph of relations

Re-shaping models to forecast faecal pathogen risk to humans

Project: Research


Since the early 20th century simple approaches (termed 1st-order kinetics) have been used to describe the population decline of bacteria in research fields as diverse as medicine, food biotechnology and environmental microbiology. When used to describe populations of faecal bacteria and pathogens in livestock faeces, these kinetics are commonly referred to as 'die-off', reflecting the generally held view that populations decline after faeces has been deposited. Consequently any potential risk of transfers to the wider environment and humans is thought to lessen with the passing of time after faeces deposition and thus models and policies reflect this. However, this project will combine new data with modelling to show that describing population change dynamics in terms of a 1st-order decline is flawed because of it failing to account for population increases that occur under fluctuating environmental conditions. This will reveal a major underestimation of diffuse source bacterial risks from cattle to soil and water quality, with increased threats to public health that may worsen if combined with expected climate change outcomes. Knowledge of microbial dynamics is partial in the sense that much work has focused on controlled laboratory experiments. Field relevant die-off profiles need to incorporate the interaction of a suite of climatic drivers thought to impact on microbial persistence (e.g. temperature, UV, episodic rewetting). New research under field conditions has speculated that bacterial growth in faecal material may be a significant factor for protracting bacterial persistence once outside the animal host. However, a key limitation of such studies is the analysis of only a few faecal samples during the immediate period post defecation. Higher resolution sampling is imperative to understand growth potential better. Our aim is to 'reshape' the credibility of modelled faecal bacteria population dynamics to ensure they reflect a more appropriate representation of growth within the mathematical profiling of bacterial persistence. The project focuses on cattle faeces as an example and quantifies the extent of error between the two approaches of accounting for, or ignoring, bacteria growth when calculating budgets of faecal bacteria deposited on pasture. The project draws on data provided by replicated field experiments. This empirical data underpins the development of the model to account for bacterial growth in dairy faeces. Our approach derives seasonal population change profiles for E. coli (a key faecal indicator organism) by collecting 16 time-series samples per seasonal experiment. Eight of these samples are collected at high resolution within the first 10 days of the experiment, including two samples collected at half day intervals on day 0. This represents a marked change from previous low frequency sampling regimes. We include a critical examination of modelled approximations of bacterial die-off using different hypothetical farm scenarios to evaluate the cumulative implications of the revised 'die-off' pattern on bacterial budgets on pasture. To facilitate this we use an existing empirical model (assuming 1st-order 'die-off' and livestock excretion rates) and amend the existing model code to reflect growth as observed through the field experiments. The output represents one of the first attempts to quantify the degree of error associated with assumptions of 1st-order decline using field relevant data and will set a precedent for acknowledging the potential under- or over- estimation of terrestrial faecal bacteria reservoirs contributed to by grazing cattle. The project assesses two main issues: (1) the magnitude of population increase of faecally derived bacteria within faeces on pasture through contrasting seasons; and (ii) the degree of under- or over- estimation of faecal bacteria burden to land if the population increase phase is ignored in modelled approximations of microbial decline.
Effective start/end date1/03/1230/09/14

Research outputs