Genetics in statistical ecology
Rachel Fewster, University of Aukland
It is often said that genetic data are easier to assemble than to analyse, as advances in technology outpace the necessary statistical developments. In ecology as well as in other fields, genetic data offer many opportunities for novel statistical work. I will give some examples showing how genetics can be useful to ecologists and interesting to statisticians. One role unique to genetics is the ability to determine where an individual animal has come from, in the sense of discriminating between various source populations. This can be useful when investigating stock interactions of marine mammals, or when protecting conservation sanctuaries from invasive pests. In islands around New Zealand, genetic methods can reliably discriminate between rodent populations sometimes as little as 200m apart. Genetic methods also have great potential in mark-recapture studies. I will show how genetic data may be combined with other recapture methods, such as photo-identification, acknowledging the non-independence of the different data sources. This is particularly relevant in expensive marine surveys, where the combination of two recapture methods may substantially improve the precision of abundance estimates at relatively little extra cost.