Assessing kea perception of cereal baits using modelling of spectral reflectance

Abstract

Kea (Nestor notabilis) are highly inquisitive parrots endemic to Aotearoa/New Zealand that often interact with novel items in their environment. To help reduce the risk of by-kill of kea during aerial 1080 pest-control operations, we investigated how kea perceive the different types of cereal baits typically used in such pest control. We measured the spectral reflectance of a range of baits including baits that incorporated different levels of a bird-repellent (anthraquinone) or a UV-reflecting biomarker (pyranine, used in conservation to determine whether a bird has interacted with bait), non-toxic prefeed baits, and green-dyed, toxic bait. In the absence of information about kea vision, we constructed a model of parrot vision by averaging spectral sensitivities of three parrot species. We found that kea are unlikely to visually distinguish between two baits that contain different concentrations of bird-repellent when they are dyed similarly (both green-dyed or both undyed). Additionally, kea are likely to visually distinguish between green-dyed and undyed baits, and baits with or without the biomarker pyranine. Our findings support the addition of bird repellent to pre-feed baits intended to establish learned avoidance behaviour. However, the addition of pyranine might inadvertently influence kea perception and interactions with bait. This research highlights the importance of considering the visual perception of a vulnerable non-target species to reduce risk during pest control efforts.

Genomic evidence of a functional RH2 opsin in New Zealand parrots and implications for pest control

“…The models employed by Brunton-Martin et al 2021 based on "average parrot vision" seem to remain a reasonable proxy for kea vision. We concur with(Brunton-Martin et al 2021) that more research is required to determine keas' behaviour towards green-dyed and undyed types of baits used in aerial predator control operations. 78375…”

“…Aerial poison drops using green toxic bait are frequently preceded by drops of undyed or green-dyed non-toxic pre-feed bait to increase subsequent consumption of toxic baits by neophobic predators. The effects of using either undyed or green pre-feed bait on non-target avian species remains unknown (Cowan and Crowell 2017;Brunton-Martin et al 2021).…”

“…They may directly feed on poison bait, especially if they have learnt that non-toxic pre-feed bait is an acceptable food source (Orr-Walker and Roberts 2009). Brunton-Martin et al (2021) modelled the appearance of different predator control baits based on "average parrot vision", to gauge the ability of kea to discern bait from different backgrounds and between dyed and undyed bait. The study's findings suggest that kea are likely able to distinguish between green-dyed and undyed baits in well-lit environments and that green-dying baits likely had a camouflage effect (Brunton-Martin et al 2021).…”

“…Brunton-Martin et al (2021) modelled the appearance of different predator control baits based on "average parrot vision", to gauge the ability of kea to discern bait from different backgrounds and between dyed and undyed bait. The study's findings suggest that kea are likely able to distinguish between green-dyed and undyed baits in well-lit environments and that green-dying baits likely had a camouflage effect (Brunton-Martin et al 2021). The authors highlighted the need to review the current practice of using undyed pre-feed and green toxic baits in the light of their findings but they also acknowledged that their model based on average parrot vision might not reflect the true visual capabilities of kea.…”