Rafael de Andrade Moral is Professor of Statistics at Maynooth University. His research interests include the development and application of statistical modelling techniques to Ecology, Wildlife Management, and Agriculture. He is also interested in the computational implementation of statistical models, especially as R code. 

Wildlife Monitoring Is About How We Count Needles in a Haystack

Monitoring wildlife is both essential and fundamentally messy. Ecosystems (both natural and human-altered) can support endangered, damaging, and economically valuable species, yet estimating animal abundance remains difficult, because detection is imperfect, data are either very expensive to obtain or very sparse, and a number of assumptions often dominate inference. Wildlife counting methods are like looking for a needle in a haystack with the needle usually hiding from you, or occasionally attacking you… This brings about conceptual, technological, and societal challenges: what should indicators indicate? How many indicators are too many? How relevant is spatial and/or temporal heterogeneity? In this talk, I will frame wildlife monitoring as a statistical problem with escalating levels of difficulty and will show how to estimate animal abundance when we can identify individuals, see them but not identify who’s who, or only infer their presence from indirect evidence. Across these settings, I will show how statistical models can be adapted to accommodate real-world constraints. Real case studies include monitoring deer, collared peccaries, foxes, giant anteaters, and wild boars. I will also illustrate how movement, behaviour, and landscape structure can affect data collection, and showcase a counterintuitive result of a common practice aimed at avoiding bias that ends up introducing more bias in camera trap surveys. Ultimately, I will attempt to substantiate an argument for a pragmatic approach to wildlife monitoring that acknowledges conceptual, technological, and societal constraints, while equipping statisticians to ask better questions of imperfect data in the real world.

Ruth King is the Thomas Bayes' Chair of Statistics at the School of Mathematics at the University of Edinburgh. Prior roles held were at the Universities of St Andrews (EPSRC postdoctoral fellow; lecturer; reader); Cambridge (research fellow); and Bristol (BSc; PhD).  Her research interests primarily lie within the areas of statistical ecology and epidemiology, developing new statistical methods and tools to answer scientific questions of interest motivated by real data problems. Particular research areas include Bayesian inference; state-space models; capture-recapture models; multiple systems estimation; abundance estimation; missing data and individual heterogeneity. She is an elected Fellow of the Learned Society of Wales (2017); Royal Society of Edinburgh (2018); Institute of Mathematical Statistics (2022) and Academy of the Mathematical Sciences (2026). She was awarded the Barnett Award by the RSS in 2022 for her contributions to the area of environmental statistics. In 2023 she was appointed as the Director of the Bayes Centre, the University of Edinburgh's Innovation Hub for Data Science and AI supporting the realisation of relevant innovation for societal and economic benefit. 

Statistical models: don’t jump the gun!

The rise and prominence of “Data Science” has highlighted what statisticians have known for many years – data are useful and can be used to answer scientific hypotheses of interest and support informed decision-making. However, to apply/develop relevant statistical analyses, extract meaningful information from data and interpret this appropriately it is important to have an understanding of core statistical/modelling principles, data collection process and properties of the data/model. 

Computational and technological advances have permitted increasingly complex models to be developed and fitted to data. Research often then focuses on “how” such models can be efficiently fitted to the data. However, simply because we can fit such complex models to data, the question of whether we should is often neglected. I will present real case studies of capture-recapture data in ecology and epidemiology with the focus of “going back to basics” and consider whether we should fit the given specified model to the data. This will involve interrogating both the data and model to understand their associated properties or assumptions given domain knowledge. In these situations, I will demonstrate the bias that would have been introduced had we not stopped to question the appropriateness of the model and its associated assumptions.