High definition video loggers provide new insights into behaviour, physiology, and the oceanic habitat of a marine predator, the yellow-eyed penguin

A method to estimate prey density from single-camera images: A case study with chinstrap penguins and Antarctic krill

Estimating the densities of marine prey observed in animal-borne video loggers when encountered by foraging predators represents an important challenge for understanding predator-prey interactions in the marine environment. We used video images collected during the foraging trip of one chinstrap penguin (Pygoscelis antarcticus) from Cape Shirreff, Livingston Island, Antarctica to develop a novel approach for estimating the density of Antarctic krill (Euphausia superba) encountered during foraging activities. Using the open-source Video and Image Analytics for a Marine Environment (VIAME), we trained a neural network model to identify video frames containing krill. Our image classifier has an overall accuracy of 73%, with a positive predictive value of 83% for prediction of frames containing krill. We then developed a method to estimate the volume of water imaged, thus the density (N·m-3) of krill, in the 2-dimensional images. The method is based on the maximum range from the camera where krill remain visibly resolvable and assumes that mean krill length is known, and that the distribution of orientation angles of krill is uniform. From 1,932 images identified as containing krill, we manually identified a subset of 124 images from across the video record that contained resolvable and unresolvable krill necessary to estimate the resolvable range and imaged volume for the video sensor. Krill swarm density encountered by the penguins ranged from 2 to 307 krill·m-3 and mean density of krill was 48 krill·m-3 (sd = 61 krill·m-3). Mean krill biomass density was 25 g·m-3. Our frame-level image classifier model and krill density estimation method provide a new approach to efficiently process video-logger data and estimate krill density from 2D imagery, providing key information on prey aggregations that may affect predator foraging performance. The approach should be directly applicable to other marine predators feeding on aggregations of prey.

An Overview of the Penguin Visual System

Penguins require vision that is adequate for both subaerial and submarine environments under a wide range of illumination. Here we provide a structured overview of what is known about their visual system with an emphasis on how and how well they achieve these goals. Amphibious vision is facilitated by a relatively flat cornea, the power in air varying from 10.2 dioptres (D) to 41.3 D depending on the species, and there is good evidence for emmetropia both above and below water. All penguins are trichromats with loss of rhodopsin 2, a nocturnal feature, but only deeper diving penguins have been noted to have pale oil droplets and a preponderance of rods. Conversely, the diurnal, shallow-diving little penguin has a higher ganglion cell density (28,867 cells/mm²) and f-number (3.5) than those that operate in dimmer light. In most species studied, there is some binocular overlap, but this reduces upon submergence. However, gaps in our knowledge remain, particularly with regard to the mechanism of accommodation, spectral transmission, behavioural measurements of visual function in low light, and neural adaptations to low light. The rarer species also deserve more attention.

Consistent Site-Specific Foraging Behaviours of Yellow-eyed Penguins/Hoiho Breeding on Stewart Island, New Zealand

Simple Summary

The yellow-eyed penguin (Megadyptes antipodes) is endemic to New Zealand and has declined c. 72% since 2008/09 within its mainland range. Population monitoring suggests yellow-eyed penguins are tracking below even the most pessimistic scenario modelled, indicating stressors may not be accurately quantified or underestimated. Fisheries-related bycatch, particularly in gillnets, has been identified as a significant contributor to the species’ decline. Mortality mitigation measures exist for penguins breeding on South Island, with a four nautical mile gillnet exclusion zone in place. Penguins breeding on Stewart Island have no protection, leaving them vulnerable to capture and drowning in gillnets. We use GPS-TDR loggers attached to adult breeding penguins from three sites across Stewart Island to track their at-sea activity, diving behaviour, and investigate the degree of foraging plasticity displayed across this range. Penguins from each site showed significant differences in their preferred habitat use and were consistent between trips and years. Results here show that foraging locations at one site cannot be used to assess habitat use by penguins at other sites. The intra-site and inter-annual consistency in preferred foraging locations observed in Stewart Island penguins reveal that implementation of marine protection may be effective in eliminating fisheries-related mortality and reduce the risk of local extinction.

Abstract

The endangered yellow-eyed penguin/hoiho (Megadyptes antipodes) predominantly forages benthically within its mainland range and shows high foraging site fidelity. Identifying consistencies in foraging locations can allow effective conservation, especially when managing bycatch risk. This study investigated the at-sea distribution of penguins breeding on Stewart Island to explore site-specific foraging strategies and inform fisheries management. During the 2020/21 season, 19 adult breeding yellow-eyed penguins from Port Pegasus, Paterson Inlet, and Codfish Island were fitted with GPS-TDR dive loggers to track their movements and diving behaviours. A total of 25,696 dives were recorded across 91 foraging trips. Birds from Port Pegasus reached significantly greater depths, spent longer at the seafloor, and performed longer dives. They also had the smallest foraging distribution, with most activity concentrated inshore. Compared to Port Pegasus, foraging radii and trip lengths were twice as large for Paterson Inlet and four times larger at Codfish Island. Despite differences in available foraging habitat, considerable individual and intra-site consistency for preferred foraging locations was observed. Localised behaviour and inter-site differences in dive metrics suggest significant plasticity in foraging ecology across their mainland range; however, individual behaviour and preferred foraging locations were extremely predictable. Thus, risk of mortality from fisheries can be quantified and managed accordingly.

Critical summer foraging tradeoffs in a subarctic ungulate

Summer diets are crucial for large herbivores in the subarctic and are affected by weather, harassment from insects and a variety of environmental changes linked to climate. Yet, understanding foraging behavior and diet of large herbivores is challenging in the subarctic because of their remote ranges. We used GPS video-camera collars to observe behaviors and summer diets of the migratory Fortymile Caribou Herd (Rangifer tarandus granti) across Alaska, USA and the Yukon, Canada. First, we characterized caribou behavior. Second, we tested if videos could be used to quantify changes in the probability of eating events. Third, we estimated summer diets at the finest taxonomic resolution possible through videos. Finally, we compared summer diet estimates from video collars to microhistological analysis of fecal pellets. We classified 18,134 videos from 30 female caribou over two summers (2018 and 2019). Caribou behaviors included eating (mean = 43.5%), ruminating (25.6%), travelling (14.0%), stationary awake (11.3%) and napping (5.1%). Eating was restricted by insect harassment. We classified forage(s) consumed in 5,549 videos where diet composition (monthly) highlighted a strong tradeoff between lichens and shrubs; shrubs dominated diets in June and July when lichen use declined. We identified 63 species, 70 genus and 33 family groups of summer forages from videos. After adjusting for digestibility, monthly estimates of diet composition were strongly correlated at the scale of the forage functional type (i.e., forage groups composed of forbs, graminoids, mosses, shrubs and lichens; r = 0.79, p < .01). Using video collars, we identified (1) a pronounced tradeoff in summer foraging between lichens and shrubs and (2) the costs of insect harassment on eating. Understanding caribou foraging ecology is needed to plan for their long-term conservation across the circumpolar north, and video collars can provide a powerful approach across remote regions.

Beyond the Chicken: Alternative Avian Models for Developmental Physiological Research

Biomedical research focusing on physiological, morphological, behavioral, and other aspects of development has long depended upon the chicken (Gallus gallus domesticus) as a key animal model that is presumed to be typical of birds and generally applicable to mammals. Yet, the modern chicken in its many forms is the result of artificial selection more intense than almost any other domesticated animal. A consequence of great variation in genotype and phenotype is that some breeds have inherent aberrant physiological and morphological traits that may show up relatively early in development (e.g., hypertension, hyperglycemia, and limb defects in the broiler chickens). While such traits can be useful as models of specific diseases, this high degree of specialization can color general experimental results and affect their translational value. Against this background, in this review we first consider the characteristics that make an animal model attractive for developmental research (e.g., accessibility, ease of rearing, size, fecundity, development rates, genetic variation, etc.). We then explore opportunities presented by the embryo to adult continuum of alternative bird models, including quail, ratites, songbirds, birds of prey, and corvids. We conclude by indicating that expanding developmental studies beyond the chicken model to include additional avian groups will both validate the chicken model as well as potentially identify even more suitable avian models for answering questions applicable to both basic biology and the human condition.

Species in the faeces: DNA metabarcoding as a method to determine the diet of the endangered yellow-eyed penguin

Abstract Context. Diet variability is a significant driver of seabird decline; however, data on seabird diet composition and trends have been affected by changes in precision and resolution owing to the evolution of different sampling methods over time. We investigated the effectiveness of applying a passive molecular diet method using faeces obtained from the endangered yellow-eyed penguin. Aims. To assess the feasibility of applying DNA metabarcoding methods to yellow-eyed penguin faeces to evaluate diet, and to compare the reliability of diet results derived from adults and chicks, and from latrine versus fresh faecal samples. Methods. We collected 313 faecal samples from yellow-eyed penguins resident on the Otago coast of New Zealand from October 2016 to August 2017. We used polymerase chain reaction (PCR) with mitochondrial 16S cephalopod and chordate primers to amplify prey DNA present in the faecal samples, and tested the completeness of our assembled reference databases based on previous diet research. Amplified prey DNA sequences were then assigned to taxa from our reference databases by using QIIME2. Key results. Mitochondrial 16S chordate PCR primers were effective at identifying 29 fish taxa, with 98.3% of amplified sequences being identified to species or genus level in 193 samples (61.7% collected). There was no significant difference in the number, occurrence or proportion of ray-finned fish prey DNA sequences derived from fresh samples or latrines. Mitochondrial 16S cephalopod PCR primers classified 1.98% of amplified DNA sequences as targets, with 96.5% of these target sequences being identified to species or genus level in 48 samples (15.3% collected), and five taxa identified. Conclusions. We recommend the collection of latrine samples to enable long-term monitoring of the diet of yellow-eyed penguins, which will optimise the trade-off between wildlife disturbance and dietary resolution. Further refinement is needed to identify cephalopod dietary components for yellow-eyed penguins, because our cephalopod primers were not as specific as those used for ray-finned fishes, amplifying a large number (&gt;98%) of non-cephalopod species. Implications. DNA metabarcoding offers a robust and comprehensive alternative to other, more intrusive, seabird diet-assessment methods, but still requires parallel studies to provide critical information on prey size, true diet composition and diet quality.

Lateralization (handedness) in Magellanic penguins

Lateralization, or asymmetry in form and/or function, is found in many animal species. Brain lateralization is considered adaptive for an individual, and often results in “handedness,” “footedness,” or a side preference, manifest in behavior and morphology. We tested for lateralization in several behaviors in a wild population of Magellanic penguins Spheniscus magellanicus breeding at Punta Tombo, Argentina. We found no preferred foot in the population (each penguin observed once) in stepping up onto an obstacle: 53% stepped up with the right foot, 47% with the left foot (n = 300, binomial test p = 0.27). We found mixed evidence for a dominant foot when a penguin extended a foot for thermoregulation, possibly depending on the ambient temperature (each penguin observed once). Penguins extended the right foot twice as often as the left foot (n = 121, p < 0.0005) in 2 years when we concentrated our effort during the heat of the day. In a third year when we observed penguins early and late in the day, there was no preference (n = 232, p = 0.59). Penguins use their flippers for swimming, including searching for and chasing prey. We found morphological evidence of a dominant flipper in individual adults: 60.5% of sternum keels curved one direction or the other (n = 76 sterna from carcasses), and 11% of penguins had more feather wear on one flipper than the other (n = 1217). Right-flippered and left-flippered penguins were equally likely in both samples (keels: p = 0.88, feather wear: p = 0.26), indicating individual but not population lateralization. In fights, aggressive penguins used their left eyes preferentially, consistent with the right side of the brain controlling aggression. Penguins that recently fought (each penguin observed once) were twice as likely to have blood only on the right side of the face (69%) as only on the left side (31%, n = 175, p < 0.001). The proportion of penguins with blood only on the right side increased with the amount of blood. In most fights, the more aggressive penguin used its left eye and attacked the other penguin’s right side. Lateralization depended on the behavior tested and, in thermoregulation, likely on the temperature. We found no lateralization or mixed results in the population of Magellanic penguins in three individual behaviors, stepping up, swimming, and thermoregulation. We found lateralization in the population in the social behavior fighting.