Fatal injuries to birds from collisions with aircraft reveal anti‐predator behaviours

Abstract

Between 1990 and 2008, more than 87 000 bird–aircraftcollisions (hereafter, bird strikes) were reported to theUS Federal Aviation Administration (FAA) and repre-sented more than US$600 million in direct and indirectcosts to US civil aviation annually (Dolbeer et al. 2009).Worldwide, the costs to civil aviation associated withbird strikes exceed US$1.2 billion annually (Allan Oas a consequence, the authors suggest that a genericavoidance response was initiated prior to impact.Clearly, an aircraft poses a hazard to birds (Blackwellet al. 2009a, Dolbeer et al. 2009) and there is empiricalevidence that birds utilize anti-predator strategies inresponse to human disturbance similar to strategies usedwhen encountering a predator (e.g. Frid & Dill 2002,Moller et al. 2008, Blackwell et al. 2009b). Thus, in anapplied context, anti-predator behaviours can help us tounderstand the mechanisms behind the responses ofwildlife to different types of human activities (Fernan-dez-Juricic et al.2001, Blackwell 2009b).We questioned, therefore, whether informationobtained via necropsy of struck birds would indicate notonly a behavioural response to the aircraft, but distinctanti-predator behaviours (e.g. Lima 1993, Hedenstrom& Rosen 2001, Blackwell et al. 2009b). If so, necropsiesof struck birds could provide information useful inunderstanding avian response to aircraft approach and,potentially, in the development of predictive methodsintended to reduce the frequency of bird strikes. Ourobjectives were to determine whether injuries associatedwith a strike were discernible from those incurred dueto impact with the ground, and whether injuries to birdswithin phylogenetic groups and foraging guilds (cohorts)were distinctive, thus indicating cohort-specific responsebehaviours. We assumed that if a fatal injury occurredbecause of a strike (i.e. a strike not involving engineingestion), but the carcass received further damage dueto impact with the ground or crushing by ground vehi-cles, the location of injuries would be randomly distrib-uted. In contrast, the location of strike injuries alone islikely to be governed by either a generic avoidanceresponse (e.g. Sheehy et al. 2005) or species-specificanti-predator behaviours, and therefore be clumped.Across all birds examined in our study, fatal injurylocations were generally posterior, ventral and on the leftside. Because of the predominant ventral distribution ofinjuries we conclude that the birds had taken evasiveaction in response to the aircraft, reflecting knownaspects of anti-predator behaviour.METHODSCarcass recoveryDuring 2000 and 2001, staff at John F. Kennedy Interna-tional Airport (JFK), New York, USA, recovered car-casses of birds reported to the FAA as killed in collisionswith aircraft at the airfield, as well as those presumed sogiven the location at recovery. We obtained the FAAstrike reports for those birds recovered as a result of areported strike. The carcasses were recovered during fiveto 10 daily sweeps of the runways, taxiways and groundwithin 250 feet of the runways. Upon recovery, the car-casses were frozen and sent to the US Department ofAgriculture’s (USDA) National Wildlife Research Cen-ter, Ohio Field Station, and assigned a unique numberbefore examination. We excluded from our sample ofbirds those which were severely desiccated, partly