Optimising biodiversity assessment protocols is useful for both conservation purposes, and for pure ecological research. I have worked on optimising such methods on a range of invertebrate taxa. During one of my undergraduate projects I found that freshwater mussel abundance can be used as a predictor of the number of other invertebrate species present at a site in UK rivers (Aldridge et al. 2007), making mussels a potentially useful indicator group. During a second project (also undergrad), involving light trapping UK moths, I found that the use of a black-light only bulb (often used in heavily populated areas to reduce light pollution) was found to decrease catch size, and number of species relative to a standard mercury vapour bulb (Fayle et al. 2007). Recently, in collaboration with Kalsum Mohd Yusah, I was involved in the development of a novel, and highly efficient bait-based trapping method for ants in the high canopy of tropical rainforest, which we demonstrated was the best method to use, in conjunction with fogging, for rapid biodiversity assessment of canopy ants through analysis using a “greedy algorithm” approach (Yusah et al. 2012). This analytical method allows the identification of the optimal combinations of sampling methods in order to maximise discovery of new species per unit time.
In addition to optimising methods for assessing numbers and identities of species, I am also interested in methods used to detect species interactions, both statistically from large datasets, and using molecular methods on specimens to infer trophic interactions. Null models of species co-occurrence can be used for this purpose, in which a selected metric relating to association (or disassociation) between species is calculated for the observed dataset, and then for multiple randomised datasets. The distribution of the metric under the null hypothesis of there being no interactions can then be compared to the observed value, to understand how likely such a value would be if species were interacting at random. We have used this approach to make predictive models of competition in relation to body size for ants living in epiphytic ferns (Fayle et al. 2015a). I developed a new statistical framework that allows the determination of the particular range of body sizes (or any other trait) over which species segregation occurs. It is then possible to use this information to build "non-null" species assembly models, and to use them to simulate community assembly, with the accuracy of the model being assessed by comparing simulated communities to real ones. I have also used a simulation-based approach to investigate statistical error rates of one of the most commonly used metrics, the C-score. We found that, when used in conjunction with a fixed-fixed randomisation algorithm (the standard algorithm for these analyses), the numbers of simulations usually carried out is likely to result in type 1 errors (false positives). We showed that increasing the number of randomisations effectively solves this problem, with more randomisations being needed for larger data matrices (Fayle & Manica 2010, Fayle & Manica 2011).
Molecular methods for detecting species interactions can be useful when it is challenging to directly observe interacting individuals. This is the case for the predatory interaction between tropical soil ants and their termite prey. We recently developed a molecular protocol for assessing ant predation on termites by screening for termite mitochondrial COII sequences in ant guts (Fayle et al. 2015b). We are now in the process of using this method to understand how ant-termite predation changes with logging of rain forest and conversion to oil palm plantation.
Fayle T.M., Eggleton P., Manica A., Yusah K.M. & Foster W.A. (2015a) Experimentally testing and assessing the predictive power of species assembly rules for tropical canopy ants. Ecology Letters 18: 254-262 [PDF]
Fayle T.M., Scholtz O., Dumbrell A.J., Russell S., Segar S.T. & Eggleton P. (2015b) Detection of mitochondrial COII DNA sequences in ant guts as a method for assessing termite predation by ants. PLoS ONE 10 (4): e0122533 [PDF]
Yusah K.M., Fayle T.M., Harris G. & Foster W.A. (2012). Optimising diversity assessment protocols for high canopy ants in tropical rain forest. Biotropica 44: 73-81. [PDF]
Fayle T.M. & Manica A. (2011). Bias in null model analyses of species co-occurrence: a response to Gotelli and Ulrich (2010). Ecological Modelling 222: 1340-1341. [PDF]
Fayle T.M. & Manica A. (2010). Reducing over-reporting of deterministic co-occurrence patterns in biotic communities. Ecological Modelling 221: 2237-2242. [PDF]
Aldridge D.C., Fayle T.M. & Jackson N. (2007). Freshwater mussel abundance predicts biodiversity in UK lowland rivers. Aquatic Conservation: Marine and Freshwater Ecosystems 17: 554-564. [PDF]
Fayle T.M., Sharp R.E. & Majerus M.E.N. (2007). The effect of moth trap type on size and composition in British Lepidoptera. British Journal of Entomology and Natural History 20: 221-232. [PDF]