Brunel University, Exeter University, Bob Burn. Funded by DEFRA
Project
Research has shown that some fish populations may be affected by exposure to oestrogens in sewage effluent; specifically male roach show signs of feminisation and have reduced reproductive success. This project aims to estimate the impact of the effects of this exposure on the long-term survival of roach populations.
Challenge
My role is to build a stochastic population model to examine how long-term survival varies with exposure. This will require integrating information from different data sources and using expert knowledge when data is lacking.
The model will use a suite of statistical models describing reproductive success (built by team members using data from laboratory studies) to estimate the number of viable fry (offspring). I will obtain annual survival rates of adults from modelling the survey data. To link these two components together estimates of the number of viable fry that survive to maturity is needed. There is a lack of data to estimate this information. Working with experts I will use elicitation methods to inform this part of the population model.
Filed under: Environment, Ecotoxicology, Statistical modelling, Elicitation, Bayesian hierarchical models
