Exploration of Rapid Evaporative Ionisation Mass Spectrometry as a novel tool for insect identification and characterisation

Wagner, Iris ORCID: 0000-0001-9198-4532
(2021) Exploration of Rapid Evaporative Ionisation Mass Spectrometry as a novel tool for insect identification and characterisation. PhD thesis, University of Liverpool.

[thumbnail of 201235558_Sep2021.pdf] Text
201235558_Sep2021.pdf - Unspecified

Download (22MB) | Preview


Insect identification and monitoring are essential to a number of diverse fields and settings, seeking to identify and study insect populations to learn more about their place in ecosystems as well as their impact on the environment and other species. The long-established approach to identifying insects is by morphological taxonomy, which utilises taxonomic keys and requires or at least greatly benefits from experience. However, identification based on morphological characteristics can be difficult when facing morphologically indistinguishable species, immature life stages or damaged specimens. Additionally, there is the challenge of processing large sample numbers that are being collected for analysis. New, easy-to-use high-throughput tools, capable of handling a variety of samples in vast amounts and with minimal sample preparation, are still needed and could provide much needed support in the wide array of fields requiring rapid insect identification. This PhD project explored the capabilities of Rapid Evaporative Ionisation Mass Spectrometry (REIMS) and its potential as a new tool for insect identification and characterisation. REIMS utilises an ambient ionisation source, specifically designed to analyse aerosols resulting from thermal disintegration caused by the passage of electricity through the sample of interest. The electric current is applied through diathermy tools and the resulting aerosol evacuated through a tube to the source and subsequently the mass spectrometer. Instead of focusing on the identification of single molecules, pattern recognition is applied to identify unique mass patterns that facilitate classification and consequently sample identification. After the first test using a mixture of wild-trapped arthropod species, successfully generating informative mass spectra, a larger proof-of principle study was conducted, based on 800 adult specimens of different Drosophila species. By analysing the REIMS data using random forest analysis, in addition to principal component and linear discriminant analysis (PCA-LDA), high classification rates were achieved when using test data sets. The results demonstrated the ability of REIMS to distinguish species, even closely related ones, as well as discriminate males and females. Further, the same approach correctly discriminated Drosophila species at the larval stage, where specimens are morphologically highly similar or identical. The next stages of the project focussed on mosquitoes and the use of REIMS to help with population characterisation – using both laboratory reared and semi-wild/trapped specimens. Laboratory reared Anopheles mosquitoes from three sibling species, usually requiring DNA analysis to be distinguished, extended the species separation challenge. Furthermore, the ability of REIMS to separate sample groups according to their age was investigated. Establishing the age profile of a mosquito population is challenging, but potentially useful as it allows prediction of disease transmission intensities and evaluation of disease vector control actions. The resulting models allowed for clear distinction between age groups separated by only 24 h and high classification rates when leaving more distinct gaps between age groups. Further, REIMS analyses of local mosquito specimens were completed, testing the system using wild-caught mosquitoes as well as semi-wild specimens, which had been collected as larvae in the field and raised under changing conditions. While the focus remained on species and age, these data sets possessed far more variability and confounding factors than those based on specimens reared under controlled laboratory conditions. The increased variance provided a powerful way to gauge REIMS suitability as identification device and helped underpin results and findings obtained with laboratory reared insects. The species of over 180 unknown specimens, part of a blinded sample set, were correctly identified at a rate of 94 % using a pre-built model and recognition software. The exploration of the potential of REIMS concluded with preliminary proof-of-principle experiments that focussed not on the insects themselves, but the frass (droppings) they produce. Successful separation of different cricket species using only their faecal matter proved that REIMS could have the potential for insect identification and population monitoring on various levels, whether its adult specimens, immature forms or ‘calling cards’ left behind. Without the need for sample preparation, entomological expertise or perfectly preserved specimens, REIMS offers a novel approach to insect typing and analysis and has considerable potential as a new tool for the field biologist.

Item Type: Thesis (PhD)
Divisions: Faculty of Health and Life Sciences
Faculty of Health and Life Sciences > Institute of Infection, Veterinary and Ecological Sciences
Depositing User: Symplectic Admin
Date Deposited: 18 Nov 2021 11:08
Last Modified: 18 Jan 2023 21:27
DOI: 10.17638/03140059
URI: https://livrepository.liverpool.ac.uk/id/eprint/3140059