Metabarcoding avian diets at airports: implications for birdstrike hazard management planning

Background: Wildlife collisions with aircraft cost the airline industry billions ofdollars per annum and represent a public safety risk. Clearly, adaptingaerodrome habitats to become less attractive to hazardous wildlife willreduce the incidence of collisions. Formulating effective habitat managementstrategies relies on accurate species identification of high-risk species.This can be successfully achieved for all strikes either through morphologyand/or DNA-based identifications. Beyond species identification, dietaryanalysis of birdstrike gut contents can provide valuable intelligence forairport hazard management practices in regards to what food is attractingwhich species to aerodromes. Here, we present birdstrike identification anddietary data from Perth Airport, Western Australia, an aerodrome that sawapproximately 140,000 aircraft movements in 2012. Next-generation highthroughput DNA sequencing was employed to investigate 77 carcasses from 16bird species collected over a 12-month period. Five DNA markers, whichbroadly characterize vertebrates, invertebrates and plants, were used totarget three animal mitochondrial genes (12S rRNA, 16S rRNA, and COI) and aplastid gene (trnL) from DNA extracted from birdstrike carcassgastrointestinal tracts.Results: Over 151,000 DNA sequences were generated, filtered and analyzed by afusion-tag amplicon sequencing approach. Across the 77 carcasses, the mostcommonly identified vertebrate was Mus musculus (house mouse).Acrididae (grasshoppers) was the most common invertebrate family identified,and Poaceae (grasses) the most commonly identified plant family. TheDNA-based dietary data has the potential to provide some key insights intofeeding ecologies within and around the aerodrome.Conclusions: The data generated here, together with the methodological approach, willgreatly assist in the development of hazard management plans and, incombination with existing observational studies, provide an improved way tomonitor the effectiveness o