The major goal of the present project is to develop and explore suitable statistical tools for identifying relevant trade-offs between biological reality, data limitations and model complexity to assess wildlife population dynamics. On the one hand, by usi ng a recently developed unifying state-space models framework, I will provide an integrated framework for the two most widely-used methods i.e. capture- recapture and distance sampling methods to better integrate data and biology. On the other hand, by usi ng estimation criteria based on parameter-redundancy and near-identifiability notions, I will investigate robustness and reliability of a broad range of applied statistical models. I will produce a taxonomy, diagnostic procedures and computer software to b etter achieve parsimony regarding model complexity. As a result, this project will participate to the creation and improvement of methodologies for modelling and estimation of population sizes and demographic parameters such as survival, recruitment and di spersal, for a better understanding of the structure and changes of a population over time and space. By, for example, providing bio-indicators, this work appears to be crucial to give relevant insights into the increasing human impact on the environment a s it allows us to both analyse and forecast phenomena such as climatic change, habitat destruction and over exploitation of natural resources, all evidence of the so-called biodiversity crisis. This proposal is supported by two of the most advanced British Universities in Ecological Statistics, with the aim to durably create a durable and attractive research network in Europe, between France, Scotland and England which are currently highly active in research in this domain. This network has the potential to provide the foundation of a wider grouping in the future, involving other member states of the Community.
Fields of science
Call for proposal
See other projects for this call