INDIVIDUAL IDENTIFICATION: MAMMALS
ANTEATERS, ARMADILLOS, AND SLOTHS
Möcklinghoff L., Schuchmann K. L., Marques M. I. (2018): New non-invasive photo-identification technique for free-ranging giant anteaters (Myrmecophaga tridactyla) facilitates urgently needed field studies. Journal of Natural History 52: 2397-2411.
FULL TEXT
Abstract
Pelage patterns, colouration and other biometric traits are perceived to be uniform in the Neotropical giant anteater (Myrmecophaga tridactyla), a conception precluding the identification of individuals, which is essential for field research on the little known aspects of the species’ ecology and behaviour. Here we present a new, non-invasive technique of matrix photo-identification to identify individual giant anteaters by their natural markings. In a long-term field study in the Brazilian Pantanal, photographs of 475 giant anteater observations (396 = direct sightings, 79 = camera traps) were captured from 2010 to 2015 and considered for our analysis. Photographs were stored in a catalogue and coded in a computerised identification table, with biometric traits being categorised and described for each observed individual in a matrix. In 71% of all photographed giant anteaters, differences in pelage marking patterns, as well as other characteristics such as ear shape and scars, allowed individual recognition. We ensured consistency of the method by conducting a double-blind verification by an experienced researcher and naïve volunteers. This simple, non-invasive method can push the level of information about life history and population structure of giant anteaters, as it applies to a large array of study designs. It can thus enhance future studies, be integrated in ongoing research projects or supply additional information out of older data sets. It is applicable to expand data collection and raise awareness in local communities, and potentially for participatory citizen science methods. Altogether these are important cornerstones for conservation actions on the species which is listed as ‘Vulnerable’ on the International Union for Conservation of Nature Red List of Threatened Species.
Bertassoni A., Bianchi R. D. C., Desbiez A. L. J. (2021): Giant anteater population density estimation and viability analysis through motion‐sensitive camera records. The Journal of Wildlife Management 85: 1554-1562.
FULL TEXT
Abstract
Giant anteater (Myrmecophaga tridactyla) populations are decreasing throughout their range. We tested a methodology for individual identification using camera records and fur patterns to estimate the giant anteater population in a protected area of the Brazilian Cerrado. We identified 9 adult individuals and successfully modeled population abundance and density. Our models estimated an adult population of 16.8 (range = 15–19) giant anteaters through a mark-resight approach and 12.5 individuals (range = 9.7–25.5) through a spatially explicit capture-recapture approach. Density estimates were 0.3–0.4 animals/km2. Using these estimates, we performed a population viability analysis to understand and predict this population’s future. We modeled scenarios without direct effects and models simulating a double carrying capacity and the supplementation and removal of individuals. Even in the more optimistic scenarios, the population is predicted to decreases over time, with ≤8 individuals remaining in 100 years. Given the study area surroundings, realistic models include removals of giant anteaters in the population. Identification of giant anteaters using cameras can inspire conservationists to acquire population data throughout its distribution and obtain population trends to evaluate the species’ conservation status. Individual identification of giant anteaters using a motion-sensitive camera design is feasible, opens new avenues for population analyses, and allows the study of population trends in difficult regions.
Gallo J. A., Abba A. M., Superina M. (2022): Individual identification of armadillos (Mammalia, Cingulata) using a photo-identification software. Mammalian Biology 102: 855-861.
FULL TEXT
Abstract
The scales of the head shield of armadillos form a distinctive pattern which facilitates a reliable identification of individuals. Comparing images of head shields of different individuals by eye is, however, time-consuming, especially when working with large catalogues of photographed animals. Computer-assisted matching of photographs calculates similarity scores between images, thus allowing to limit the visual comparisons to those individuals having the highest scores. The aim of our study was to identify individual patterns of scales on the head shield of ten armadillo species, and test a photo-identification software in identifying individuals based on their head shield pattern. We analyzed 354 frontal photographs of the head shield of 10 species of armadillos taken from different sources. Of those pictures, 153 were different photographs of the same animals. At first, all photographs were compared by eye. Then, we used the pattern extraction and matching software Wild-ID to compare images and find possible matches. None of the individuals on the images were misidentified as other individuals (no false acceptance errors), but the software falsely rejected 26 images. While the most common factors affecting the computer-assisted matching process were the reflection of a flash-light (when flash photography was used) and exposure issues, the flash reflection occurs only when photographing museum collections and is unlikely to affect field photographic surveys unless flash is used in specific conditions, such as for nocturnal armadillos. Our results suggest that the software Wild-ID is a useful tool to individually identify armadillos for capture-recapture studies, including long-term studies.
BATS
Rydell J., Russo D. (2015): Photography as a low-impact method to survey bats. Mammalian Biology 80: 182-184.
FULL TEXT
Abstract
Bats are mammals of chief conservation concern and also represent potentially powerful bio-indicators. Surveying bats is thus an important task but the approaches adopted may either be too invasive (capture) or prone to identification errors (acoustic methods). We here report on the use of a photographic trap to survey bat species richness we tested at two drinking sites in central Italy. The species richness we estimated was similar to that obtained by a previous mist-netting effort at the same sites. We also photographed species often overlooked in acoustic surveys due to their faint echolocation calls. From the photographs we could frequently identify sex, reproductive status, age class and individual marks. Given the relative non-invasiveness of this approach, we strongly recommend it in lieu of capture at sensitive sites or to complement acoustic surveys in order to improve identification rates.
Amelon S. K., Hooper S. E., Womack K. M. (2017): Bat wing biometrics: using collagen–elastin bundles in bat wings as a unique individual identifier. Journal of Mammalogy 98: 744-751.
FULL TEXT
Abstract
The ability to recognize individuals within an animal population is fundamental to conservation and management. Identification of individual bats has relied on artificial marking techniques that may negatively affect the survival and alter the behavior of individuals. Biometric systems use biological characteristics to identify individuals. The field of animal biometrics has expanded to include recognition of individuals based upon various morphologies and phenotypic variations including pelage patterns, tail flukes, and whisker arrangement. Biometric systems use 4 biologic measurement criteria: universality, distinctiveness, permanence, and collectability. Additionally, the system should not violate assumptions of capture–recapture methods that include no increased mortality or alterations of behavior. We evaluated whether individual bats could be uniquely identified based upon the collagen–elastin bundles that are visible with gross examination of their wings. We examined little brown bats (Myotis lucifugus), northern long-eared bats (M. septentrionalis), big brown bats (Eptesicus fuscus), and tricolored bats (Perimyotis subflavus) to determine whether the “wing prints” from the bundle network would satisfy the biologic measurement criteria. We evaluated 1,212 photographs from 230 individual bats comparing week 0 photos with those taken at weeks 3 or 6 and were able to confirm identity of individuals over time. Two blinded evaluators were able to successfully match 170 individuals in hand to photographs taken at weeks 0, 3, and 6. This study suggests that bats can be successfully re-identified using photographs taken at previous times. We suggest further evaluation of this methodology for use in a standardized system that can be shared among bat conservationists.
Kirkpatrick L., Apoznański G., Bruyn L. D., Gyselings R., Kokurewicz T. (2019): Bee markers: A novel method for non-invasive short term marking of bats. Acta Chiropterologica 21: 465-471.
FULL TEXT
Abstract
The ability to accurately identify individuals is essential for both management and research of wild or captive animals. Marks need to be clear enough to be recovered, long-lived enough to provide useful information and not have health implications for the tagged animal which may affect survival. Substantial evidence suggests that more invasive, permanent marking techniques such as wing banding can cause injuries to bats, limiting the species with which such procedures can be used. Furthermore, low recovery rates can mean that the long term impacts of permanent marking on survival cannot be assessed. Here we present a new non-invasive, low cost approach to tag bats which can be carried out with the minimum of handling. Recovery of marks is simple and does not require handling. Our approach repurposes ‘queen bee markers’, small, coloured and numbered plastic disks which are commonly used to mark queen bees; instead they are affixed with superglue directly onto the fur of the bat. We carried out a pilot study at a large hibernaculum in West Poland, home to ca. 35,000 bats, the majority of which are Myotis myotis, the target species for this study. In November 2017 we marked 203 bats during a census of the underground fortification system. We recovered 30% of the originally marked bats over a time period of five months; 27% of the originally marked bats were identified to individual when resighted. Fifteen individuals were recovered in March that were not recovered in January, suggesting that they were either missed by recorders or were not present in the system during the census. Using the colour marking system allowed us to derive information about changes in bat behaviour when identification to individual was not possible, and individual identification revealed differences in hibernation strategies. We conclude that bee markers are an effective, cheap and less invasive approach for short term monitoring of bat populations and will also be useful to monitor whether recovery rates warrant the use of more invasive marking techniques.
Sun C., Zhang C., Lucas J. R., Lin A., Feng J., Jiang T. (2021): Territorial calls of the bat Hipposideros armiger may encode multiple types of information: body mass, dominance rank and individual identity. Animal Cognition 24: 689-702.
FULL TEXT
Abstract
In highly vocal species, territorial aggression is often accompanied using vocalizations. These vocalizations can play a critical role in determining the outcome of male–male agonistic interactions. For this, vocalizations of contestants must contain information that is indicative of each competitor’s fighting ability as well as its identity, and also contestants must be able to perceive information about the physical attributes, quality and identity of the vocalizer. Here, we used adult male Great Himalayan leaf-nosed bats (Hipposideros armiger) to test whether territorial calls encoded honest information about a caller’s physical attributes, quality and individual identity. We did this by exploring the relationships between territorial calls and two potential indices of fighting ability: body mass and dominance rank. Using synchronized audio–video recording, we monitored bat territorial calls and dominance rank of 16 adult male H. armiger in the laboratory. Additionally, habituation–dishabituation playback experiments were performed to test for vocal discrimination. Results showed that body mass was negatively related to minimum frequency and positively related to syllable duration. Dominance score was also negatively related to minimum frequency and positively related to peak frequency. Furthermore, a discriminant function analysis suggested that territorial calls encode an individual signature. Therefore, our data show that males have the ability to utilize this vocal individual signature to discriminate between vocalizing males. In short, territorial calls of male H. armiger contain information about body mass, dominance rank and individual identity, and contestants are probably capable of perceiving this information and may use it to make appropriate decisions during agonistic interactions.
CARNIVORES
Dixon D. R. (2003): A non‐invasive technique for identifying individual badgers Meles meles. Mammal Review 33: 92-94.
FULL TEXT
Abstract
Individual badgers Meles meles can be reliably identified in the field on the basis of variation in the appearance of the tail. Tests of the technique using video surveillance demonstrated that in 95% of instances individuals were identified correctly on the basis of tail patterns. It is possible that tail patterns and posture may be a significant means of communication in this species.
Karanth K. U., Nichols J. D., Kumar N. S., Hines J. E. (2006): Assessing tiger population dynamics using photographic capture–recapture sampling. Ecology 87: 2925-2937.
FULL TEXT
Abstract
Although wide‐ranging, elusive, large carnivore species, such as the tiger, are of scientific and conservation interest, rigorous inferences about their population dynamics are scarce because of methodological problems of sampling populations at the required spatial and temporal scales. We report the application of a rigorous, noninvasive method for assessing tiger population dynamics to test model‐based predictions about population viability. We obtained photographic capture histories for 74 individual tigers during a nine‐year study involving 5725 trap‐nights of effort. These data were modeled under a likelihood‐based, “robust design” capture–recapture analytic framework. We explicitly modeled and estimated ecological parameters such as time‐specific abundance, density, survival, recruitment, temporary emigration, and transience, using models that incorporated effects of factors such as individual heterogeneity, trap‐response, and time on probabilities of photo‐capturing tigers. The model estimated a random temporary emigration parameter of γŷ″ = γŷ′ = 0.10 ± 0.069 (values are estimated mean ± se). When scaled to an annual basis, tiger survival rates were estimated at Ŝ = 0.77 ± 0.051, and the estimated probability that a newly caught animal was a transient was τŷ = 0.18 ± 0.11. During the period when the sampled area was of constant size, the estimated population size Nŷt varied from 17 ± 1.7 to 31 ± 2.1 tigers, with a geometric mean rate of annual population change estimated as λ = 1.03 ± 0.020, representing a 3% annual increase. The estimated recruitment of new animals, B¯t , varied from 0 ± 3.0 to 14 ± 2.9 tigers. Population density estimates, B¯ , ranged from 7.33 ± 0.8 tigers/100 km2 to 21.73 ± 1.7 tigers/100 km2 during the study. Thus, despite substantial annual losses and temporal variation in recruitment, the tiger density remained at relatively high levels in Nagarahole. Our results are consistent with the hypothesis that protected wild tiger populations can remain healthy despite heavy mortalities because of their inherently high reproductive potential. The ability to model the entire photographic capture history data set and incorporate reduced‐parameter models led to estimates of mean annual population change that were sufficiently precise to be useful. This efficient, noninvasive sampling approach can be used to rigorously investigate the population dynamics of tigers and other elusive, rare, wide‐ranging animal species in which individuals can be identified from photographs or other means.
Anderson C. J. R., Roth J. D., Waterman J. M. (2007): Can whisker spot patterns be used to identify individual polar bears? Journal of Zoology 273: 333-339.
FULL TEXT
Abstract
Studies of population dynamics, movement patterns and animal behavior usually require identification of individuals. We evaluated the reliability of using whisker spot patterns to noninvasively identify individual polar bears Ursus maritimus. We obtained the locations of polar bear whisker spots from photographs taken in western Hudson Bay, tested the independence of spot locations, estimated the complexity of each spot pattern in terms of information and determined whether each whisker spot pattern was reliable from its information content. Of the 50 whisker spot patterns analyzed, 98% contained enough information to be reliable, and this result varied little among observers. Photographs taken <50 m from polar bears were most useful. Our results suggest that individual identification of polar bears in the field based on whisker spot pattern variations is reliable. Researchers studying polar bear behavior or estimating population parameters can benefit from this method if proximity to the bears is feasible.
Simcharoen S., Pattanavibool A., Karanth K. U., Nichols J. D., Kumar N. S. (2007): How many tigers Panthera tigris are there in Huai Kha Khaeng Wildlife Sanctuary, Thailand? An estimate using photographic capture-recapture sampling. Oryx 41: 447-453.
FULL TEXT
Abstract
We used capture-recapture analyses to estimate the density of a tiger Panthera tigris population in the tropical forests of Huai Kha Khaeng Wildlife Sanctuary, Thailand, from photographic capture histories of 15 distinct individuals. The closure test results (z = 0.39, P = 0.65) provided some evidence in support of the demographic closure assumption. Fit of eight plausible closed models to the data indicated more support for model Mh, which incorporates individual heterogeneity in capture probabilities. This model generated an average capture probability 0.42 and an abundance estimate of 19 (9.65) tigers. The sampled area of 477.2 (58.24) km2 yielded a density estimate of 3.98 (0.51) tigers per 100 km2. Huai Kha Khaeng Wildlife Sanctuary could therefore hold 113 tigers and the entire Western Forest Complex c. 720 tigers. Although based on field protocols that constrained us to use sub-optimal analyses, this estimated tiger density is comparable to tiger densities in Indian reserves that support moderate prey abundances. However, tiger densities in well-protected Indian reserves with high prey abundances are three times higher. If given adequate protection we believe that the Western Forest Complex of Thailand could potentially harbour >2,000 wild tigers, highlighting its importance for global tiger conservation. The monitoring approaches we recommend here would be useful for managing this tiger population.
Cunningham L. (2009): Using computer-assisted photo-identification and capture-recapture techniques to monitor the conservation status of harbour seals (Phoca vitulina). Aquatic Mammals 35: 319-329.
FULL TEXT
Abstract
Conservation policies require the status of protected species to be monitored. The choice of monitoring methods may be constrained in situations in which there is concern about disturbance or in which sighting individuals is difficult. This study investigated the potential of using a computer- assisted photo-identification method to measure population size in adult harbour seals (Phoca vitulina). Pattern cells or combinations of pattern cells from photographs (i.e., ventral, flank, shoulder, and head) were used for computerized selection of potential matching pairs, and the pelage patterns of those pairs were then checked visually. There was monthly variation in capture-recapture population estimates, with the highest number of adult harbour seals in May (117, CV = 7.2). Around three times more individuals used the sampling area in northwest Scotland between April and October (268, CV = 0.04) than were estimated per month (mean = 86, CV = 0.07). Using computer-assisted photo-identification and capture-recapture methods may be the only practical way of obtaining a measurement of how many seals use a site. This approach has important implications for determining the effectiveness of designated conservation areas for protecting seals and will influence management decisions, including the size of management units.
Anderson C. J., Da Vitoria Lobo N., Roth J. D., Waterman J. M. (2010): Computer-aided photo-identification system with an application to polar bears based on whisker spot patterns. Journal of Mammalogy 91: 1350-1359.
FULL TEXT
Abstract
Ecologists often rely on unique natural markings to identify individual free-ranging animals without disturbing them. We developed a computer-aided photo-identification system for identifying polar bears (Ursus maritimus) based on whisker spot pattern recognition. We automated our system so that the selection of 3 reference points on the input image is the only manual step required during image preprocessing. Our pattern-matching algorithm is unique in that the variability within spot patterns is considered fully rather than representing them as points and applying a point-pattern matching algorithm. We also measured the reliability of our method as probabilities of true positives and false positives using photographs of various qualities taken at different angles. When we excluded photographs of poor quality and angle the probability of true positives was >80% at a false positive probability of 10%. A new photograph could be preprocessed in <1 min and tested against a reference library of 100 individuals in <10 min. Our computer-aided identification system could be extended for use in other species with variable spot patterns, which could be useful in efforts to estimate various population dynamics parameters essential for the study and conservation of wildlife, particularly threatened and endangered species.
Magoun A. J., Long C. D., Schwartz M. K., Pilgrim K. L., Lowell R. E., Valkenburg P. (2011): Integrating motion‐detection cameras and hair snags for wolverine identification. The Journal of Wildlife Management 75: 731-739.
FULL TEXT
Abstract
We developed an integrated system for photographing a wolverine’s (Gulo gulo) ventral pattern while concurrently collecting hair for microsatellite DNA genotyping. Our objectives were to 1) test the system on a wild population of wolverines using an array of camera and hair-snag (C&H) stations in forested habitat where wolverines were known to occur, 2) validate our ability to determine identity (ID) and sex from photographs by comparing photographic data with that from DNA, and 3) encourage researchers and managers to test the system in different wolverine populations and habitats and improve the system design. Of the 18 individuals (10 M, 8 F) for which we obtained genotypes over the 2 years of our study, there was a 100% match between photographs and DNA for both ID and sex. The integrated system made it possible to reduce cost of DNA analysis by >74%. Integrating motion-detection cameras and hair snags provides a cost-effective technique for wildlife managers to monitor wolverine populations in remote habitats and obtain information on important population parameters such as density, survival, productivity, and effective population size.
Higashide D., Miura S., Miguchi H. (2012): Are chest marks unique to Asiatic black bear individuals? Journal of Zoology 288: 199-206.
FULL TEXT
Abstract
Estimating population size based on a capture‐recapture model requires identification of individual animals. We evaluated the reliability of the chest mark to noninvasively identify individual Asiatic black bears Ursus thibetanus. Using image analysis, we collated the chest marks of bears from the photographs taken while the bears were in captivity (Ani Mataginosato Bear Park) to examine the universality, uniqueness and persistence of the marks. Of the 62 bears, 95% had a distinct chest mark by which they could be reliably identified, and the probability of mistakenly identifying two different bears as identical was calculated to be 0.00075. The shape of the mark was found to change slightly from year to year, but this did not hamper individual identification. Thus, individual identification of the bears was highly reliable. A high percentage of correct answers was obtained in a blind test to visually identify individuals based on their chest mark. Considering that it is both an inexpensive and an easy‐to‐use technique, chest mark comparison is suitable for individual identification in order to estimate the abundance of the black bear population.
Zimmermann F., Breitenmoser-Würsten C., Molinari-Jobin A., Breitenmoser U. (2013): Optimizing the size of the area surveyed for monitoring a Eurasian lynx (Lynx lynx) population in the Swiss Alps by means of photographic capture-recapture. Integrative Zoology 8: 232-243.
FULL TEXT
Abstract
We studied the influence of surveyed area size on density estimates by means of camera‐trapping in a low‐density felid population (1–2 individuals/100 km2). We applied non‐spatial capture–recapture (CR) and spatial CR (SCR) models for Eurasian lynx during winter 2005/2006 in the northwestern Swiss Alps by sampling an area divided into 5 nested plots ranging from 65 to 760 km2. CR model density estimates (95% CI) for models M0 and Mh decreased from 2.61 (1.55–3.68) and 3.6 (1.62–5.57) independent lynx/100 km2, respectively, in the smallest to 1.20 (1.04–1.35) and 1.26 (0.89–1.63) independent lynx/100 km2, respectively, in the largest area surveyed. SCR model density estimates also decreased with increasing sampling area but not significantly. High individual range overlaps in relatively small areas (the edge effect) is the most plausible reason for this positive bias in the CR models. Our results confirm that SCR models are much more robust to changes in trap array size than CR models, thus avoiding overestimation of density in smaller areas. However, when a study is concerned with monitoring population changes, large spatial efforts (area surveyed ≥760 km2) are required to obtain reliable and precise density estimates with these population densities and recapture rates.
Alexander J. S., Gopalaswamy A. M., Shi K., Riordan P. (2015): Face value: towards robust estimates of snow leopard densities. Plos One 10: e0134815.
FULL TEXT
Abstract
When densities of large carnivores fall below certain thresholds, dramatic ecological effects can follow, leading to oversimplified ecosystems. Understanding the population status of such species remains a major challenge as they occur in low densities and their ranges are wide. This paper describes the use of non-invasive data collection techniques combined with recent spatial capture-recapture methods to estimate the density of snow leopards Panthera uncia. It also investigates the influence of environmental and human activity indicators on their spatial distribution. A total of 60 camera traps were systematically set up during a three-month period over a 480 km2 study area in Qilianshan National Nature Reserve, Gansu Province, China. We recorded 76 separate snow leopard captures over 2,906 trap-days, representing an average capture success of 2.62 captures/100 trap-days. We identified a total number of 20 unique individuals from photographs and estimated snow leopard density at 3.31 (SE = 1.01) individuals per 100 km2. Results of our simulation exercise indicate that our estimates from the Spatial Capture Recapture models were not optimal to respect to bias and precision (RMSEs for density parameters less or equal to 0.87). Our results underline the critical challenge in achieving sufficient sample sizes of snow leopard captures and recaptures. Possible performance improvements are discussed, principally by optimising effective camera capture and photographic data quality.
Osterrieder S. K., Kent C. S., Anderson C. J. R., Parnum I. M., Robinson R. W. (2015): Whisker spot patterns: a noninvasive method of individual identification of Australian sea lions (Neophoca cinerea). Journal of Mammalogy 96: 988–997.
FULL TEXT
Abstract
Reliable methods for identification of individual animals are advantageous for ecological studies of population demographics and movement patterns. Photographic identification, based on distinguishable patterns, unique shapes, or scars, is an effective technique already used for many species. We tested whether photographs of whisker spot patterns could be used to discriminate among individual Australian sea lion (Neophoca cinerea). Based on images of 53 sea lions, we simulated 5,000 patterns before calculating the probability of duplication in a study population. A total of 99% (± 1.5 SD) of patterns were considered reliable for a population of 50, 98% (± 1.7 SD) for 100, 92% (± 4.7 SD) for 500, and 88% (± 5.7 SD) for 1,000. We tested a semiautomatic approach by matching 16 known individuals at 3 different angles (70°, 90°, and 110°), 2 distances (1 and 2 m), and 6 separate times over a 1-year period. A point-pattern matching algorithm for pairwise comparisons produced 90% correct matches of photographs taken on the same day at 90°. Images of individuals at 1 and 2 m resulted in 89% correct matches, those photographed at different angles and different times (at 90°) resulted in 48% and 73% correct matches, respectively. Our results show that the Chamfer distance transform can effectively be used for individual identification, but only if there is very little variation in photograph angle. This point-pattern recognition application may also work for other otariid species.
Jewell Z. C., Alibhai S. K., Weise F., Munro S., Van Vuuren M., Van Vuuren R. (2016): Spotting cheetahs: identifying individuals by their footprints. Journal of Visualized Experiments 111: e54034.
FULL TEXT
Abstract
The cheetah (Acinonyx jubatus) is Africa’s most endangered large felid and listed as Vulnerable with a declining population trend by the IUCN. It ranges widely over sub-Saharan Africa and in parts of the Middle East. Cheetah conservationists face two major challenges, conflict with landowners over the killing of domestic livestock, and concern over range contraction. Understanding of the latter remains particularly poor. Namibia is believed to support the largest number of cheetahs of any range country, around 30%, but estimates range from 2,905 to 13,520. The disparity is likely a result of the different techniques used in monitoring. Current techniques, including invasive tagging with VHF or satellite/GPS collars, can be costly and unreliable. The footprint identification technique is a new tool accessible to both field scientists and also citizens with smartphones, who could potentially augment data collection. The footprint identification technique analyzes digital images of footprints captured according to a standardized protocol. Images are optimized and measured in data visualization software. Measurements of distances, angles, and areas of the footprint images are analyzed using a robust cross-validated pairwise discriminant analysis based on a customized model. The final output is in the form of a Ward’s cluster dendrogram. A user-friendly graphic user interface (GUI) allows the user immediate access and clear interpretation of classification results. The footprint identification technique algorithms are species specific because each species has a unique anatomy. The technique runs in a data visualization software, using its own scripting language (jsl) that can be customized for the footprint anatomy of any species. An initial classification algorithm is built from a training database of footprints from that species, collected from individuals of known identity. An algorithm derived from a cheetah of known identity is then able to classify free-ranging cheetahs of unknown identity. The footprint identification technique predicts individual cheetah identity with an accuracy of >90%.
Koivuniemi M., Auttila M., Niemi M., Levänen R., Kunnasranta M. (2016): Photo-ID as a tool for studying and monitoring the endangered Saimaa ringed seal. Endangered Species Research 30: 29-36.
FULL TEXT
Abstract
Photo-identification (photo-ID) with camera traps was examined as a non-invasive method for studying and monitoring the endangered Saimaa ringed seal Phoca hispida saimensis. An average of 51 game cameras were set up at shoreline haul-out sites in central Lake Saimaa during the moulting seasons in each of the years from 2010 to 2014. Individuals were identified from their lifelong unique lateral fur patterns. A total of 220000 digital images of seals were obtained from these game cameras and from digital cameras during this period, allowing 164 individuals to be identified, 43% of which were re-sighted in successive years. In the majority of game camera images, both sides of the seal were shown, and it was possible to determine the sex of the seal. The average distance between sightings of individual seals in different years was 1.6 km, suggesting that Saimaa ringed seals exhibit a high degree of moulting site fidelity. In addition, the results support suggestions of natal site fidelity. We propose that photo-ID sampling based on camera traps, and larger-scale photographic survey of the seals, should be implemented as a population monitoring tool of the Saimaa ringed seal. Further application of the photo-ID technique may facilitate mark-recapture population and survival rate estimates for this threatened seal species and provide significant insights into its life history and social behaviour.
Zheng X., Owen M. A., Nie Y., Hu Y., Swaisgood R. R., Yan L., Wei F. (2016): Individual identification of wild giant pandas from camera trap photos – a systematic and hierarchical approach. Journal of Zoology 300: 247-256.
FULL TEXT
Abstract
The scientific and conservation‐management value of using camera traps is greatly enhanced if the identity of individual animals ‘captured’ can be assigned. Identification of individuals is necessary to make unbiased estimates of population parameters, and can allow for the generation of more robust inferences from studies of spatial and behavioral ecology. Here we tested the utility of an approach to individually identify wild giant pandas Ailuropoda melanoleuca from camera trap images, by cataloguing and careful scrutiny of numerous traits. We developed our identification strategy first by analyzing images of known (captive) individuals (N = 7). We then deployed camera traps at 23 control sites and at seven camera trap arrays ‘baited’ with conspecific decoys, in Foping Nature Reserve, China. From a sample of 12 871 photographs, and using the method we developed with known individuals, we were able to identify 11 individual giant pandas. We tested the repeatability of this approach using a double blind test with 12 naïve volunteers, achieving a relatively high inter‐observer agreement of 80%, which was increased to 93% when observers reported a high degree of confidence. We also found that image quality was significantly higher at decoy sites than at control sites. We suggest that this approach will be useful for future field projects, allowing researchers to address a broader array of questions and provide new insights into panda ecology and conservation.
Alibhai S., Jewell Z., Evans J. (2017): The challenge of monitoring elusive large carnivores: an accurate and cost-effective tool to identify and sex pumas (Puma concolor) from footprints. Plos One 12: e0172065.
FULL TEXT
Abstract
Acquiring reliable data on large felid populations is crucial for effective conservation and management. However, large felids, typically solitary, elusive and nocturnal, are difficult to survey. Tagging and following individuals with VHF or GPS technology is the standard approach, but costs are high and these methodologies can compromise animal welfare. Such limitations can restrict the use of these techniques at population or landscape levels. In this paper we describe a robust technique to identify and sex individual pumas from footprints. We used a standardized image collection protocol to collect a reference database of 535 footprints from 35 captive pumas over 10 facilities; 19 females (300 footprints) and 16 males (235 footprints), ranging in age from 1–20 yrs. Images were processed in JMP data visualization software, generating one hundred and twenty three measurements from each footprint. Data were analyzed using a customized model based on a pairwise trail comparison using robust cross-validated discriminant analysis with a Ward’s clustering method. Classification accuracy was consistently > 90% for individuals, and for the correct classification of footprints within trails, and > 99% for sex classification. The technique has the potential to greatly augment the methods available for studying puma and other elusive felids, and is amenable to both citizen-science and opportunistic/local community data collection efforts, particularly as the data collection protocol is inexpensive and intuitive.
Devens C., Tshabalala T., McManus J., Smuts B. (2018): Counting the spots: the use of a spatially explicit capture–recapture technique and GPS data to estimate leopard (Panthera pardus) density in the Eastern and Western Cape, South Africa. African Journal of Ecology 56: 850-859.
FULL TEXT
Abstract
Estimating species population density directly contributes to the conservation of species. As keystone species, carnivores are important to conserve; however, estimating density of wide‐ranging, elusive and solitary carnivores has proven difficult. The leopard (Panthera pardus) is the last large free‐roaming top carnivore in South Africa, and no formal density study had been conducted across the Eastern and Western Cape. We estimated leopard density and abundance using GPS data from 21 collared leopards and a spatially explicit capture–mark–recapture (SECR) method with camera trap survey data. Four regional sites were surveyed using 173 camera trap locations over 15,390 camera trap days, capturing 740 leopard images of which 77 individuals were identified. SECR averaged 0.95 leopards/100 km2 and the two GPS methods averaged 1 and 1.11 leopards/100 km2. Based on predicted available leopard habitat for the region, leopard abundance was estimated between 467 (±112.8) and 553 (±168.8) in the Western Cape and between 365 (±93.2) and 430 (±139.9) in the Eastern Cape. Discrepancies in density estimates can be complex stemming from biological behaviour, anthropogenic factors and prey density. However, our estimates appear to show relatively little variation, suggesting that SECR methods and GPS data capture the population density estimates of the species well.
Li B. V., Alibhai S., Jewell Z., Li D., Zhang H. (2018): Using footprints to identify and sex giant pandas. Biological Conservation 218: 83-90.
FULL TEXT
Abstract
Data on numbers and distribution of free-ranging giant panda are essential to the formulation of effective conservation strategies. There is still no ideal method to identify individuals and sex this species. The traditional bite-size method using bamboo fragments in their feces lacks accuracy. The modern DNA-based estimation is expensive and demands fresh samples. The lack of identifiable individual features on panda pelage and no apparent sexual dimorphism impede reliable estimation from camera trap images. Here, we propose an innovative and non-invasive technique to identify and sex this species using a footprint identification technique (FIT). It is based on a pairwise comparison of trails (unbroken series of footprints) using discriminant analysis, with a Ward’s clustering method. We collected footprints from 30 captive animals to train our algorithm and used another 11 animals for model validation. The accuracy for individual identification was > 90% for individuals with more than six footprints and 89% with fewer footprints per trail. The accuracy for sex discrimination was about 84% using a single footprint and 91% using trails. This cost-effective method provides a promising future for monitoring wild panda populations and understanding their dynamics and especially useful for monitoring reintroduced animals after the detachment of GPS collars. The data collection protocol is straightforward and accessible to citizen scientists and conservation professionals alike.
Dorning J., Harris S. (2019): The challenges of recognising individuals with few distinguishing features: identifying red foxes Vulpes vulpes from camera-trap photos. Plos One 14: e0216531.
FULL TEXT
Abstract
Over the last two decades, camera traps have revolutionised the ability of biologists to undertake faunal surveys and estimate population densities, although identifying individuals of species with subtle markings remains challenging. We conducted a two-year camera-trapping study as part of a long-term study of urban foxes: our objectives were to determine whether red foxes could be identified individually from camera-trap photos, and highlight camera-trapping protocols and techniques to facilitate photo identification of species with few or subtle natural markings. We collected circa 800,000 camera-trap photos over 4945 camera days in suburban gardens in the city of Bristol, UK: 152,134 (19%) included foxes, of which 13,888 (9%) contained more than one fox. These provided 174,063 timestamped capture records of individual foxes; 170,923 were of foxes ≥ 3 months old. Younger foxes were excluded because they have few distinguishing features. We identified the individual (192 different foxes: 110 males, 49 females, 33 of unknown sex) in 168,417 (99%) of these capture records; the remainder could not be identified due to poor image quality or because key identifying feature(s) were not visible. We show that carefully designed survey techniques facilitate individual identification of subtly-marked species. Accuracy is enhanced by camera-trapping techniques that yield large numbers of high resolution, colour images from multiple angles taken under varying environmental conditions. While identifying foxes manually was labour-intensive, currently available automated identification systems are unlikely to achieve the same levels of accuracy, especially since different features were used to identify each fox, the features were often inconspicuous, and their appearance varied with environmental conditions. We discuss how studies based on low numbers of photos, or which fail to identify the individual in a significant proportion of photos, risk losing important biological information, and may come to erroneous conclusions.
Park H., Lim A., Choi T. Y., Baek S. Y., Song E. G., Park Y. C. (2019): Where to spot: individual identification of leopard cats (Prionailurus bengalensis euptilurus) in South Korea. Journal of Ecology and Environment 43: 39.
FULL TEXT
Abstract
Knowledge of abundance, or population size, is fundamental in wildlife conservation and management. Camera-trapping, in combination with capture-recapture methods, has been extensively applied to estimate abundance and density of individually identifiable animals due to the advantages of being non-invasive, effective to survey wide-ranging, elusive, or nocturnal species, operating in inhospitable environment, and taking low labor. We assessed the possibility of using coat patterns from images to identify an individual leopard cat (Prionailurus bengalensis), a Class II endangered species in South Korea. We analyzed leopard cat images taken from Digital Single-Lense Relfex camera (high resolution, 18Mpxl) and camera traps (low resolution, 3.1Mpxl) using HotSpotter, an image matching algorithm. HotSpotter accurately top-ranked an image of the same individual leopard cat with the reference leopard cat image 100% by matching facial and ventral parts. This confirms that facial and ventral fur patterns of the Amur leopard cat are good matching points to be used reliably to identify an individual. We anticipate that the study results will be useful to researchers interested in studying behavior or population parameter estimates of Amur leopard cats based on capture-recapture models.
Sayer S., Allen R., Hawkes L. A., Hockley K., Jarvis D., Witt M. J. (2019): Pinnipeds, people and photo identification: the implications of grey seal movements for effective management of the species. Journal of the Marine Biological Association of the United Kingdom 99: 1221-30.
FULL TEXT
Abstract
Grey seals (Halichoerus grypus) of the North-east Atlantic are protected at designated European Marine Sites (Special Areas of Conservation, SACs) typically during their reproductive periods and in the UK at Sites of Special Scientific Interest (SSSI). As a mobile marine species, grey seals spend other parts of their annual life cycle in non-designated habitat. There is limited information on individual grey seal movements in south-west England. Citizen science photo identification (PID) revealed the movements of 477 grey seals at a regional scale (54 haul-outs up to 230 km apart) for over a decade. Reconstructed movements showed considerable individual variability. Four SACs were linked to up to 18 non-designated sites and two SSSIs in Cornwall were linked to a maximum of 41 non-designated sites. Observations support the value of existing SSSIs at both the well-connected West and North Cornwall sites. Thirteen Marine Protected Areas (MPAs) were visited by grey seals from four SACs and two SSSIs in Cornwall. As a mobile species, grey seals could be included in English MPA management plans. The application of functional linkage from SACs and SSSIs, informed by the movements evidenced in this research, could aid management efforts. This analysis reveals grey seal movements occur across a complex network of interconnected designated and non-designated sites that need to be managed holistically for this species for which the UK has a special responsibility.
Young J. K., Golla J. M., Broman D., Blankenship T., Heilbrun R. (2019): Estimating density of an elusive carnivore in urban areas: use of spatially explicit capture-recapture models for city-dwelling bobcats. Urban Ecosystems 22: 507-512.
FULL TEXT
Abstract
An important first step in managing urban carnivores or the habitat in which they live to reduce risk of conflicts with humans is to understand their basic ecology and population dynamics. Traditional density estimators may be inappropriate in urban areas because of extensive areas of impermeable development but new techniques that include spatial structure could be useful within large urban metropolitan areas. Yet to date, these techniques have largely remained untested. We evaluated whether spatially explicit capture-recapture models (SECR) could provide a reliable density estimate of bobcats (Lynx rufus) in the Dallas Fort-Worth metroplex, Texas, USA. We obtained 1003 photographs of bobcats in an urbanized landscape from June–November 2014, using 41 double camera stations spaced approximately 1.05 km apart. We individually identified bobcats from their distinct pelage patterns and used SECR to predict density throughout the study area. The overall density was at least one bobcat per km2, which calculated to approximately 43 independent-aged bobcats across the entire camera grid, an estimate higher than documented bobcat densities in both rural and peri-urban studies in Texas. Our study revealed a high density of bobcats in an urban landscape despite most assumptions that bobcats require large areas of habitat and are sensitive to fragmentation.
Clapham M., Miller E., Nguyen M., Darimont C. T. (2020): Automated facial recognition for wildlife that lack unique markings: A deep learning approach for brown bears. Ecology and Evolution 10: 12883-12892.
FULL TEXT
Abstract
Emerging technologies support a new era of applied wildlife research, generating data on scales from individuals to populations. Computer vision methods can process large datasets generated through image-based techniques by automating the detection and identification of species and individuals. With the exception of primates, however, there are no objective visual methods of individual identification for species that lack unique and consistent body markings. We apply deep learning approaches of facial recognition using object detection, landmark detection, a similarity comparison network, and an support vector machine-based classifier to identify individuals in a representative species, the brown bear Ursus arctos. Our open-source application, BearID, detects a bear’s face in an image, rotates and extracts the face, creates an “embedding” for the face, and uses the embedding to classify the individual. We trained and tested the application using labeled images of 132 known individuals collected from British Columbia, Canada, and Alaska, USA. Based on 4,674 images, with an 80/20% split for training and testing, respectively, we achieved a facial detection (ability to find a face) average precision of 0.98 and an individual classification (ability to identify the individual) accuracy of 83.9%. BearID and its annotated source code provide a replicable methodology for applying deep learning methods of facial recognition applicable to many other species that lack distinguishing markings. Further analyses of performance should focus on the influence of certain parameters on recognition accuracy, such as age and body size. Combining BearID with camera trapping could facilitate fine-scale behavioral research such as individual spatiotemporal activity patterns, and a cost-effective method of population monitoring through mark–recapture studies, with implications for species and landscape conservation and management. Applications to practical conservation include identifying problem individuals in human–wildlife conflicts, and evaluating the intrapopulation variation in efficacy of conservation strategies, such as wildlife crossings.
Hou J., He Y., Yang H., Connor T., Gao J., Wang Y., Zeng Y., Zhang J., Huang J., Zheng B., Zhou S. (2020): Identification of animal individuals using deep learning: A case study of giant panda. Biological Conservation 242: 108414.
FULL TEXT
Abstract
Giant panda (Ailuropoda melanoleuca) is an iconic species of conservation. However, long-term monitoring of wild giant pandas has been a challenge, largely due to the lack of appropriate method for the identification of target panda individuals. Although there are some traditional methods, such as distance-bamboo stem fragments methods, molecular biological method, and manual visual identification, they all have some limitations that can restrict their application. Therefore, it is urgent to explore a reliable and efficient approach to identify giant panda individuals. Here, we applied the deep learning technology and developed a novel face-identification model based on convolutional neural network to identify giant panda individuals. The model was able to identify 95% of giant panda individuals in the validation dataset. In all simulated field situations where the quality of photo data was degraded, the model still accurately identified more than 90% of panda individuals. The identification accuracy of our model is robust to brightness, small rotation, and cleanness of photos, although large rotation angle (>20°) of photos has significant influence on the identification accuracy of the model (P < 0.01). Our model can be applied in future studies of giant panda such as long-term monitoring, big data analysis for behavior and be adapted for individual identification of other wildlife species.
Prop J., Staverløkk A., Moe B. (2020): Identifying individual polar bears at safe distances: A test with captive animals. Plos One 15: e0228991.
FULL TEXT
Abstract
The need to recognise individuals in population and behavioural studies has stimulated the development of various identification methods. A commonly used method is to employ natural markers to distinguish individuals. In particular, the automated processing of photographs of study animals has gained interest due to the speed of processing and the ability to handle a high volume of records. However, automated processing requires high-quality photographs, which means that they need to be taken from a specific angle or at close distances. Polar bears Ursus maritimus, for example, may be identified by automated analysis of whisker spot patterns. However, to obtain photographs of adequate quality, the animals need to be closer than is usually possible without risk to animal or observer. In this study we tested the accuracy of an alternative method to identify polar bears at further distances. This method is based on distinguishing a set of physiognomic characteristics, which can be recognised from photographs taken in the field at distances of up to 400 m. During five trials, sets of photographs of 15 polar bears from six zoos, with each individual bear portrayed on different dates, were presented for identification to ten test observers. Among observers the repeatability of the assessments was 0.68 (SE 0.011). Observers with previous training in photogrammetric techniques performed better than observers without training. Experience with observing polar bears in the wild did not improve skills to identify individuals on photographs. Among the observers with photogrammetric experience, the rate of erroneous assessment was on average 0.13 (SE 0.020). For the inexperienced group this was 0.72 (SE 0.018). Error rates obtained with automated whisker spot analysis were intermediate (0.26–0.58). We suggest that wildlife studies will benefit from applying several identification techniques to collect data under different conditions.
Russo L. F., Loy A. (2020): Who am I? Testing I3S Contour on the facial mask of the Western polecat (Mustela putorius). Hystrix: the Italian Journal of Mammalogy 31: 83-85.
FULL TEXT
Abstract
Individual recognition of wild animals is a fundamental tool to acquire information about the structure and dynamics of animal populations. Recently, individual identification from camera trapping has been successfully applied to Capture-Mark- Recapture (CMR) studies in various taxa. We collected 281 photos of 48 specimens of Western Polecat (Mustela putorius) from various Italian Museums to test the capabilities of I3S contour software to automatically recognize different individuals from their facial mask. After selecting 52 high quality pictures from different specimens, we obtained a 100% success rate of correct individual identification. This suggested that both facial mask pattern and automatic identification might be successfully applied to the study of this highly elusive species through camera trapping.
Coe S. T., Elmore J. A., Elizondo E. C., Loss S. R. (2021): Free‐ranging domestic cat abundance and sterilization percentage following five years of a trap–neuter–return program. Wildlife Biology 2021: wlb.00799.
FULL TEXT
Abstract
Increasing free-ranging cat populations are a cause of concern for wildlife management and biodiversity conservation. Cats carry and transmit multiple diseases, annually depredate billions of birds and mammals in the mainland United States, and have caused extinctions and declines of wildlife populations worldwide. Trap–neuter–return (TNR) efforts, which entail trapping, sterilizing and releasing unowned free-ranging cats with the goal of reducing populations, have been implemented globally despite limited evidence of their ability to reduce cat numbers. To assess the effectiveness of a TNR program initiated in 2013 in Stillwater, Oklahoma, USA, we used trail cameras at 15 locations to estimate changes in cat abundance and the percentage of ear-tipped (i.e. sterilized) individuals between 2014 and 2018. We reviewed photographs to identify individual cats, and after accounting for detectability with mark–resight analyses, we estimated a non-significant decrease in abundance from 62 to 48 total cats across sampled locations. In 2018, approximately 27% of cats were ear-tipped compared to 0% in 2014, yet this percentage remains far below estimated sterilization levels needed for TNR to reduce unowned cat populations. Although additional long-term monitoring is needed, our results suggest that TNR conducted at its current intensity is unlikely to reduce Stillwater’s cat population. Our research adds further evidence to the growing body of scientific literature indicating that TNR is ineffective in reducing cat populations.
Langley I., Hague E., Civil M. A. (2021): Assessing the performance of open-source, semi-automated pattern recognition software for harbour seal (P. v. vitulina) photo ID. Mammalian Biology 102: 973-982.
FULL TEXT
Abstract
Photographic identification (photo ID) is a well-established, non-invasive, and relatively cost-effective technique to collect longitudinal data from species that can be individually recognised based on natural markings. This method has been improved by computer-assisted pattern recognition software which speed up the processing of large numbers of images. Freely available algorithms exist for a wide range of species, but the choice of software can have significant effects on the accuracy of individual capture histories and derived demographic parameter estimates. We tested the performance of three open source, semi-automated pattern recognition software algorithms for harbour seal (Phoca vitulina vitulina) photo ID: ExtractCompare, I3S Pattern and Wild-ID. Performance was measured as the ability of the software to successfully score matching images higher than non-matching images using the cumulative density function (CDF). The CDF for the top ranked potential match was highest for Wild-ID (CDF1 = 0.34–0.58), followed by ExtractCompare (CDF1 = 0.24–0.36) and I3S pattern (CDF1 = 0.02–0.3). This trend emerged regardless of how many potential matches were inspected. The highest performing aspects in ExtractCompare were left heads, whereas in I3S Pattern and Wild-ID these were front heads. Within each aspect, images collected using a camera and lens performed higher than images taken by a camera and scope. Data processing within ExtractCompare took > 4 × longer than Wild-ID, and > 3 × longer than I3S Pattern. We found that overall, Wild-ID outperformed both ExtractCompare and I3S Pattern under tested scenarios, and we therefore recommend its assistance in harbour seal photo ID.
Clapham M., Miller E., Nguyen M., Van Horn R. C. (2022): Multispecies facial detection for individual identification of wildlife: a case study across ursids. Mammalian Biology 102: 921-933.
FULL TEXT
Abstract
To address biodiversity decline in the era of big data, replicable methods of data processing are needed. Automated methods of individual identification (ID) via computer vision are valuable in conservation research and wildlife management. Rapid and systematic methods of image processing and analysis are fundamental to an ever-growing need for effective conservation research and practice. Bears (ursids) are an interesting test system for examining computer vision techniques for wildlife, as they have variable facial morphology, variable presence of individual markings, and are challenging to research and monitor. We leveraged existing imagery of bears living under human care to develop a multispecies bear face detector, a critical part of individual ID pipelines. We compared its performance across species and on a pre-existing wild brown bear Ursus arctos dataset (BearID), to examine the robustness of convolutional neural networks trained on animals under human care. Using the multispecies bear face detector and retrained sub-applications of BearID, we prototyped an end-to-end individual ID pipeline for the declining Andean bear Tremarctos ornatus. Our multispecies face detector had an average precision of 0.91–1.00 across all eight bear species, was transferable to images of wild brown bears (AP = 0.93), and correctly identified individual Andean bears in 86% of test images. These preliminary results indicate that a multispecies-trained network can detect faces of a single species sufficiently to achieve high-performance individual classification, which could speed-up the transferability and application of automated individual ID to a wider range of taxa.
Nepovinnykh E., Chelak I., Lushpanov A., Eerola T., Kälviäinen H., Chirkova O. (2022): Matching individual Ladoga ringed seals across short-term image sequences. Mammalian Biology 102: 935-950.
FULL TEXT
Abstract
Automated wildlife reidentification has attracted increasing attention in recent years as it provides a non-invasive tool to identify and to track individual wild animals over time. In this paper, the first steps are taken towards the automatic photo-identification of the Ladoga ringed seals (Pusa hispida ladogensis). A method is proposed that takes a sequence of images, each containing multiple individuals as the input, and produces cropped images of seals grouped based on one certain individual per group. The method starts by detecting each seal from the images and proceeds to matching the individual seals between the images. It is shown that high grouping accuracy can be obtained with a general-purpose image retrieval method on an image sequence taken from the same location within a relatively short period of time. Each resulting group contains multiple images of one individual with slightly different variations, for example, in pose and illumination. Utilizing these images simultaneously provides more information for the individual re-identification compared to the traditional approach, i.e., which utilizes just one image at a time. It is further demonstrated that a convolutional neural network based method can be used to extract the unique pelage patterns of the seals despite the low contrast. Finally, a method is proposed and experiments with the novel Ladoga ringed seals data are carried out to provide a proof-of-concept for the individual re-identification.
Strampelli P., Searle C. E., Smit J. B., Henschel P., Mkuburo L., Ikanda D., Macdonald D. W., Dickman A. J. (2022): Camera trapping and spatially explicit capture–recapture for the monitoring and conservation management of lions: Insights from a globally important population in Tanzania. Ecological Solutions and Evidence 3: e12129.
FULL TEXT
Abstract
Accurate and precise estimates of population status are required to inform and evaluate conservation management and policy interventions. Although the lion (Panthera leo) is a charismatic species receiving increased conservation attention, robust status estimates are lacking for most populations. While for many large carnivores population density is often estimated through spatially explicit capture–recapture (SECR) applied to camera trap data, the lack of pelage patterns in lions has limited the application of this technique to the species. Here, we present one of the first applications of this methodology to lion, in Tanzania’s Ruaha-Rungwa landscape, a stronghold for the species for which no empirical estimates of status are available. We deployed four camera trap grids across habitat and land management types, and we identified individual lions through whisker spots, scars and marks, and multiple additional features. Double-blind identification revealed low inter-observer variation in photo identification (92% agreement), due to the use of xenon-flash cameras and consistent framing and angles of photographs. Lion occurred at highest densities in a prey-rich area of Ruaha National Park (6.12 ± SE 0.94 per 100 km2), and at relatively high densities (4.06 ± SE 1.03 per 100 km2) in a community-managed area of similar riparian-grassland habitat. Miombo woodland in both photographic and trophy hunting areas sustained intermediate lion densities (1.75 ± SE 0.62 and 2.25 ± SE 0.52 per 100 km2, respectively). These are the first spatially explicit density estimates for lion in Tanzania, including the first for a trophy hunting and a community-managed area, and also provide some of the first insights into lion status in understudied miombo habitats. We discuss in detail the methodology employed, the potential for scaling-up over larger areas, and its limitations. We suggest that the method can be an important tool for lion monitoring and explore the implications of our findings for lion management.
Alibhai S. K., Gu J., Jewell Z. C., Morgan J., Liu D., Jiang G. (2023): ‘I know the tiger by his paw’: A non-invasive footprint identification technique for monitoring individual Amur tigers (Panthera tigris altaica) in snow. Ecological Informatics 73: 101947.
FULL TEXT
Abstract
Apex predator populations are in decline around the world. Many exist at low density and are elusive, making the acquisition of reliable data on their numbers and distribution a considerable challenge. The Amur tiger (Panthera tigris altaica) is the largest of the five extant sub-species of tiger. The single most significant, contiguous population, an estimated 550 animals, exists in the Russian Far East, with smaller populations on the far eastern Sino-Russian border. For the last few decades, active efforts on the part of Chinese authorities have encouraged the recolonization of these populations back to their former ranges in Northeast China. Reliable data on Amur tiger numbers and distribution are required to assess population recovery at the landscape scale. Footprints, ubiquitous in the snow over range areas, could inform on these baseline data. This paper describes a statistically robust, cost-effective and non-invasive footprint identification technique (FIT) to identify individual tigers from footprints in snow. It is based on a rigorous data collection and data-processing protocol, combined with a cross-validated discriminant analysis method. A Ward’s clustering technique provides a visual output of individual classification. The analytical tools are packaged in a user-friendly analytical interface. Between December 2011 and December 2012, we collected a series of 605 footprint images from 44 captive individual Amur tigers for a reference database from which to derive a classification algorithm. The 23 females and 21 males ranged in age from 3 to 13 years (female mean age 7.95 +/− 0.18; male mean age 8.08 +/− 0.19). 128 measurements (areas, lengths and angles) were taken from each print and analyzed with the FIT add-in to JMP software. The derived classification algorithm was then applied to 21 footprint trails collected from an unknown number of free-ranging Amur tigers during 2012 and 2015/2016. The algorithm predicted 7 Amur tigers at the site surveyed in 2012, and 4 tigers surveyed at two sites in 2015/16. We demonstrate that the footprint identification technique translates traditional tracking methodologies into a statistically robust and objective analytical tool that can be deployed by both scientists and local communities to monitor the recovery of big cat populations.
CETACEANS
Auger-Méthé M., Marcoux M., Whitehead H. (2011): Computer-assisted photo-identification of narwhals. Arctic 64: 342-352.
FULL TEXT
Abstract
Although the narwhal (Monodon monoceros) is economically and culturally important to northern residents, sound management of this species is impaired by large gaps in knowledge. Research on this species has been limited partly by the cost of the methods used, and partly because some of these methods are invasive and therefore condemned by Inuit communities. Photo-identification, a non-invasive, inexpensive, and easy-to-use method recently developed for narwhals, uses photographs of natural marks to identify individuals. Its main drawback is the extended time required to process photographs. We developed a computer program to accelerate the identification process and thus mitigate the main drawback of photo-identification. This program uses the locations of notches on the dorsal ridge to compare a new image to each individual in a catalogue and lists those individuals in decreasing order of similarity. We tested consistency in user assignment of dorsal ridge features and the accuracy of the program by comparing sets of known individuals. While assignment errors were common, the program ranked the true match within the first 10% of the catalogue 78% of the time. The program accelerates the matching process by 1.2 to 4.1 times for catalogues ranging in size from 40 to 500 individuals, and the degree of acceleration increases with the size of the catalogue. This program could also be applied to the beluga whale (Delphinapterus leucas), another important northern species.
Fearnbach H., Durban J., Parsons K., Claridge D. (2012): Photographic mark–recapture analysis of local dynamics within an open population of dolphins. Ecological Applications 22: 1689-1700.
FULL TEXT
Abstract
Identifying demographic changes is important for understanding population dynamics. However, this requires long‐term studies of definable populations of distinct individuals, which can be particularly challenging when studying mobile cetaceans in the marine environment. We collected photo‐identification data from 19 years (1992–2010) to assess the dynamics of a population of bottlenose dolphins (Tursiops truncatus) restricted to the shallow (<7 m) waters of Little Bahama Bank, northern Bahamas. This population was known to range beyond our study area, so we adopted a Bayesian mixture modeling approach to mark–recapture to identify clusters of individuals that used the area to different extents, and we specifically estimated trends in survival, recruitment, and abundance of a “resident” population with high probabilities of identification. There was a high probability (p = 0.97) of a long‐term decrease in the size of this resident population from a maximum of 47 dolphins (95% highest posterior density intervals, HPDI = 29–61) in 1996 to a minimum of just 24 dolphins (95% HPDI = 14–37) in 2009, a decline of 49% (95% HPDI = −5% to −75%). This was driven by low per capita recruitment (average ∼0.02) that could not compensate for relatively low apparent survival rates (average ∼0.94). Notably, there was a significant increase in apparent mortality (∼5 apparent mortalities vs. ∼2 on average) in 1999 when two intense hurricanes passed over the study area, with a high probability (p = 0.83) of a drop below the average survival probability (∼0.91 in 1999; ∼0.94, on average). As such, our mark–recapture approach enabled us to make useful inference about local dynamics within an open population of bottlenose dolphins; this should be applicable to other studies challenged by sampling highly mobile individuals with heterogeneous space use.
Alessi J., Aïssi M., Fiori C. (2014): Photo‐identification of sperm whales in the north‐western Mediterranean Sea: an assessment of natural markings. Aquatic Conservation: Marine and Freshwater Ecosystems 24: 11-22.
FULL TEXT
Abstract
Photo‐identification is an important, non‐invasive tool that can be used to obtain data about cetacean population dynamics that are essential for proper environmental management. A standardized protocol is crucial for obtaining optimal results, particularly for long‐lived and highly migratory species such as sperm whales (Physeter macrocephalus) but also to study resident populations. Photo‐identification of individuals using natural marks has been widely used to study sperm whales in the Mediterranean Sea. However, the prevalence of the marks used for identification is unknown. Thus, the goal of this study was to identify which mark types are most useful for identifying individual sperm whales. A photo‐identification catalogue of sperm whales from the north‐western Mediterranean Sea was produced and examined to determine the most frequent mark types present on whales in the study area. Mark types and their distribution were described, prevalence and the size of each mark type were calculated, variability in visibility was investigated, and gain and loss rates were analysed. Analysis of natural pigmentation may characterize sperm whale flanks in this study area as the best and most convenient element on which matching technique can be based. Indeed, this technique led to the identification of 97% of photographed individuals.
Beck S., Foote A. D., Koetter S., Harries O., Mandleberg L., Stevick P. T., Whooley P., Durban J. W. (2014): Using opportunistic photo-identifications to detect a population decline of killer whales (Orcinus orca) in British and Irish waters. Journal of the Marine Biological Association of the United Kingdom 94: 1327.
FULL TEXT
Abstract
An assemblage of killer whales that has been sighted in waters off the west coast of the British Isles and Ireland has previously been shown to be isolated from other North Atlantic killer whale communities based on association patterns. By applying a Bayesian formulation of the Jolly–Seber mark-recapture model to the photo-identification data compiled from opportunistic photographic encounters with this population of killer whales, we show that such sparse and opportunistically-collected data can still be valuable in estimating population dynamics of small, wide-ranging groups. Good quality photo-identification data was collected from 32 encounters over 19 years. Despite a cumulative total of 77 identifications from these encounters, just ten individuals were identified and the remaining 67 identifications were re-sights of these ten animals. There was no detected recruitment through births during the study and, as a result, the population appears to be in a slight decline. The demography of the population was highly skewed towards older individuals and had an unusually high ratio of adult males, and we suggest that demographic stochasticity due to a small population size may be further impacting the population growth rate. We recommend that this population be managed as a separate conservation unit from neighbouring killer whale populations.
McCordic J. A., Root-Gutteridge H., Cusano D. A., Denes S. L., Parks S. E. (2016): Calls of North Atlantic right whales Eubalaena glacialis contain information on individual identity and age class. Endangered Species Research 30: 157-169.
FULL TEXT
Abstract
Passive acoustic monitoring is a powerful tool that allows remote detection of marine mammals through their vocalizations. While call detection provides information on species presence, additional information may be contained within the vocalizations that could provide more information regarding the demographics and/or number of individuals in a particular area based on passive acoustic detections. The North Atlantic right whale Eubalaena glacialis produces a stereotyped upswept call, termed the upcall, that is thought to function as a long-distance contact call in this species. As such, the call is likely to contain cues providing information about the individual producing it. The goal of this study was to test whether the right whale upcall could potentially encode information related to the identity and age of the caller. Using upcalls recorded from 14 known individuals through non-invasive suction cup archival acoustic tags, we demonstrate that the upcall does contain sufficient information to discriminate individual identity and age class, with average classification levels of 72.6 and 86.1%, respectively. Parameters measured from the fundamental frequency, duration, and formant structure were most important for discrimination among individuals. This study is the first step in demonstrating the feasibility of obtaining additional data from passive acoustic monitoring to aid in the conservation efforts for this highly endangered species.
Carvajal-Gámez B. E., Trejo-Salazar D. B., Gendron D., Gallegos-Funes F. J. (2017): Photo-id of blue whale by means of the dorsal fin using clustering algorithms and color local complexity estimation for mobile devices. EURASIP Journal on Image and Video Processing 2017: 6.
FULL TEXT
Abstract
We present an automatic program of blue whale photo-identification for mobile devices. The proposed technique works in the wavelet domain to reduce the image size and the processing time of the proposed algorithm, and with an edge enhancement filter, the characteristics of the blue whale are preserved. Additionally, an image palette reduction algorithm based on local image complexity estimation is introduced to eliminate redundant colors, thus decreasing the number of pixels that are bad classified in the segmentation process and minimizing the resource consumption of the mobile device. The segmented image is obtained with the FCM (fuzzy C-means) or K-means algorithms incorporating a dynamic filtering which is proposed in this paper to improve the brightness and contrast of the acquired image increasing the performance of the image segmentation. Experimental results show that the proposed approach potentially could provide a real-time solution to photo-id of blue whale images and it can be transportable and portable power for mobile devices. Finally, the proposed methodology is simple, efficient, and feasible for photo-id applications in mobile devices.
Bichell L. M. V., Krzyszcyk E., Patterson E. M., Mann J. (2018): The reliability of pigment pattern‐based identification of wild bottlenose dolphins. Marine Mammal Science 34: 113-124.
FULL TEXT
Abstract
Long‐term studies often rely on natural markings for individual identification across time. The primary method for identification in small cetaceans relies on dorsal fin shape, scars, and other natural markings. However, dorsal fin markings can vary substantially over time and the dorsal fin can become unrecognizable after an encounter with a boat or shark. Although dorsal fins have the advantage in that they always break the water surface when the cetacean breathes, other physical features, such as body scars and pigmentation patterns can supplement. The goal of this study was to explore the use of dorso‐lateral pigment patterns to identify wild bottlenose dolphins. We employed photographic pigment matching tests to determine if pigmentation patterns showed (1) longitudinal consistency and (2) bilateral symmetry using a 30 yr photographic database of bottlenose dolphins (Tursiops aduncus). We compared experienced dolphin researchers and inexperienced undergraduate student subjects in their ability to accurately match images. Both experienced and inexperienced subjects correctly matched dolphin individuals at a rate significantly above chance, even though they only had 10 s to make the match. These results demonstrate that pigment patterns can be used to reliably identify individual wild bottlenose dolphins, and likely other small cetacean species at other sites.
Franklin T., Franklin W., Brooks L., Harrison P., Burns D., Holmberg J., Calambokidis J. (2020): Photo-identification of individual Southern Hemisphere humpback whales (Megaptera novaeangliae) using all available natural marks: managing the potential for misidentification. Journal of Cetacean Research and Management 21: 71-83.
FULL TEXT
Abstract
Misidentification errors in capture-mark recapture studies of humpback whales (Megaptera novaeangliae) related to poor quality of photographs as well as changes in natural marks can seriously affect population dynamics parameter estimates and derived estimates of population size when using sophisticated modelling techniques. In this study we used an innovative photo-identification matching system to investigate and examine the long-term stability and/or changes in natural marks on ventral-tail flukes, dorsal fin shapes and lateral body marks from a sample of 79 individual humpback whales, resighted in 2 to 11 years over timespans ranging from 2 to 21 years. A binary logistic mixed effects model, on a pair-matched sample of the 79 individual whales, found no significant differences in the proportions of ventral-tail fluke marks, dorsal fin shapes and lateral body marks, that displayed changes in primary and/or secondary characteristics over years (F=0.939, df=1/156, p =0.334). The results of this study substantiate the value and reliability of using primary and secondary natural marks on the ventral-tail flukes, in conjunction with dorsal fin shapes and secondary lateral body marks as double-tags. This provides a means of maximising observations of individual humpback whales over years, while minimising and managing misidentification errors in the photo-identification matching process, thus significantly improving modelling outcomes.
Ramos-Arredondo R. I., Carvajal-Gámez B. E., Gendron D., Gallegos-Funes F. J., Mújica-Vargas D., Rosas-Fernández J. B. (2020): PhotoId-Whale: Blue whale dorsal fin classification for mobile devices. Plos One 15: e0237570.
FULL TEXT
Abstract
Photo-identification (photo-id) is a method used in field studies by biologists to monitor animals according to their density, movement patterns and behavior, with the aim of predicting and preventing ecological risks. However, these methods can introduce subjectivity when manually classifying an individual animal, creating uncertainty or inaccuracy in the data as a result of the human criteria involved. One of the main objectives in photo-id is to implement an automated mechanism that is free of biases, portable, and easy to use. The main aim of this work is to develop an autonomous and portable photo-id system through the optimization of image classification algorithms that have high statistical dependence, with the goal of classifying dorsal fin images of the blue whale through offline information processing on a mobile platform. The new proposed methodology is based on the Scale Invariant Feature Transform (SIFT) that, in conjunction with statistical discriminators such as the variance and the standard deviation, fits the extracted data and selects the closest pixels that comprise the edges of the dorsal fin of the blue whale. In this way, we ensure the elimination of the most common external factors that could affect the quality of the image, thus avoiding the elimination of relevant sections of the dorsal fin. The photo-id method presented in this work has been developed using blue whale images collected off the coast of Baja California Sur. The results shown have qualitatively and quantitatively validated the method in terms of its sensitivity, specificity and accuracy on the Jetson Tegra TK1 mobile platform. The solution optimizes classic SIFT, balancing the results obtained with the computational cost, provides a more economical form of processing and obtains a portable system that could be beneficial for field studies through mobile platforms, making it available to scientists, government and the general public.
van der Linde M. L., Eriksson I. K. (2020): An assessment of sperm whale occurrence and social structure off São Miguel Island, Azores using fluke and dorsal identification photographs. Marine Mammal Science 36: 47-65.
FULL TEXT
Abstract
Female and immature sperm whales form stable social units in tropical to subtropical waters. One such area is the Azores archipelago, where details of their year‐round occurrence and social organization are not well known. We used year‐round sightings data collected from whale watching vessels to assess sperm whale occurrence and social structure off São Miguel Island, Azores from 2010 to 2017. Individuals were photo‐identified by their flukes (n = 393) and dorsal pigmentation marks were examined to explore their use in assisting with reidentifications. Of all cataloged whales, 78.8% were sufficiently distinctive to be reidentified from dorsal pigmentation patterns. Associations between individuals were analyzed to determine social structure and delineate social units. We identified 12 units comprising 2–13 members that had stable, and perhaps preferred associations for periods up to the eight years of the study. Preferences between some pairs of units may exist, but more research is required to better understand dynamics within and between units. This local scale study is an important contributor to our knowledge of geographic variation on a global scale. We recommend making use of all available information from the entire archipelago, to further increase our understanding of sperm social organization in the Azores.
Bergler C., Gebhard A., Towers J. R., Butyrev L., Sutton G. J., Shaw T. J., Maier A., Nöth E. (2021): FIN-PRINT a fully-automated multi-stage deep-learning-based framework for the individual recognition of killer whales. Scientific Reports 11: 23480.
FULL TEXT
Abstract
Biometric identification techniques such as photo-identification require an array of unique natural markings to identify individuals. From 1975 to present, Bigg’s killer whales have been photo-identified along the west coast of North America, resulting in one of the largest and longest-running cetacean photo-identification datasets. However, data maintenance and analysis are extremely time and resource consuming. This study transfers the procedure of killer whale image identification into a fully automated, multi-stage, deep learning framework, entitled FIN-PRINT. It is composed of multiple sequentially ordered sub-components. FIN-PRINT is trained and evaluated on a dataset collected over an 8-year period (2011–2018) in the coastal waters off western North America, including 121,000 human-annotated identification images of Bigg’s killer whales. At first, object detection is performed to identify unique killer whale markings, resulting in 94.4% recall, 94.1% precision, and 93.4% mean-average-precision (mAP). Second, all previously identified natural killer whale markings are extracted. The third step introduces a data enhancement mechanism by filtering between valid and invalid markings from previous processing levels, achieving 92.8% recall, 97.5%, precision, and 95.2% accuracy. The fourth and final step involves multi-class individual recognition. When evaluated on the network test set, it achieved an accuracy of 92.5% with 97.2% top-3 unweighted accuracy (TUA) for the 100 most commonly photo-identified killer whales. Additionally, the method achieved an accuracy of 84.5% and a TUA of 92.9% when applied to the entire 2018 image collection of the 100 most common killer whales. The source code of FIN-PRINT can be adapted to other species and will be publicly available.
Blount D., Gero S., Van Oast J., Parham J., Kingen C., Scheiner B., Stere T., Fisher M., Minton G., Khan C., Dulau V. (2022): Flukebook: an open-source AI platform for cetacean photo identification. Mammalian Biology 102: 1005-1023.
FULL TEXT
Abstract
Determining which species are at greatest risk, where they are most vulnerable, and what are the trajectories of their communities and populations is critical for conservation and management. Globally distributed, wide-ranging whales and dolphins present a particular challenge in data collection because no single research team can record data over biologically meaningful areas. Flukebook.org is an open-source web platform that addresses these gaps by providing researchers with the latest computational tools. It integrates photo-identification algorithms with data management, sharing, and privacy infrastructure for whale and dolphin research, enabling the global collaborative study of these global species. With seven automatic identification algorithms trained for 15 different species, resulting in 37 species-specific identification pipelines, Flukebook is an extensible foundation that continually incorporates emerging AI techniques and applies them to cetacean photo identification through continued collaboration between computer vision researchers, software engineers, and biologists. With over 2.0 million photos of over 52,000 identified individual animals submitted by over 250 researchers, the platform enables a comprehensive understanding of cetacean populations, fostering international and cross-institutional collaboration while respecting data ownership and privacy. We outline the technology stack and architecture of Flukebook, its performance on real-world cetacean imagery, and its development as an example of scalable, extensible, and reusable open-source conservation software. Flukebook is a step change in our ability to conduct large-scale research on cetaceans across biologically meaningful geographic ranges, to rapidly iterate population assessments and abundance trajectories, and engage the public in actions to protect them.
Cheeseman T., Southerland K., Park J., Olio M., Flynn K., Calambokidis J., Jones L., Garrigue C., Frisch Jordan A., Howard A., Reade W., Neilson J., Gabriele C., Clapham P. (2022): Advanced image recognition: a fully automated, high-accuracy photo-identification matching system for humpback whales. Mammalian Biology 102: 915-929.
FULL TEXT
Abstract
We describe the development and application of a new convolutional neural network-based photo-identification algorithm for individual humpback whales (Megaptera novaeangliae). The method uses a Densely Connected Convolutional Network (DenseNet) to extract special keypoints of an image of the ventral surface of the fluke and then a separate DenseNet trained to look for features within these keypoints. The extracted features are then compared against those of the reference set of previously known humpback whales for similarity. This offers the potential to successfully automate recognition of individuals in large photographic datasets such as in ocean basin-wide marine mammal studies. The algorithm requires minimal image pre-processing and is capable of accurate, rapid matching of fair to high-quality humpback fluke photographs. In real world testing compared to manual image matching, the algorithm reduces image management time by at least 98% and reduces error rates of missing potential matches from approximately 6–9% to 1–3%. The success of this new system permits automated comparisons to be made for the first time across photo-identification datasets with tens to hundreds of thousands of individually identified encounters, with profound implications for long-term and large population studies of the species.
Cheney B. J., Dale J., Thompson P. M., Quick N. J. (2022): Spy in the sky: a method to identify pregnant small cetaceans. Remote Sensing in Ecology and Conservation 8: 492-505.
FULL TEXT
Abstract
Data on sex ratios, age classes, reproductive success and health status are key metrics to manage populations, yet can be difficult to collect in wild cetacean populations. Long-term individual-based studies provide a unique opportunity to apply unoccupied aerial system (UAS) photogrammetry to non-invasively measure body morphometrics of individuals with known life history information. The aims of this study were (1) to compare length measurements from UAS photogrammetry with laser photogrammetry and (2) to explore whether UAS measurements of body width could be used to remotely determine pregnancy status, sex or age class in a well-studied bottlenose dolphin population in Scotland. We carried out five boat-based surveys in July and August 2017, with concurrent photo-identification, UAS and laser photogrammetry. Photographs were measured using bespoke programmes, MorphMetriX for UAS photos and a Zooniverse project for laser photos. In total 64 dolphins were identified using photo-ID, 54 of which had concurrent UAS body length and 47 with laser body length measurements. We also measured body widths at 10% increments from 10% to 90% of body length for 48 individuals of known sex, age class and/or pregnancy status. There was no significant difference in the length of individuals measured with UAS and laser photogrammetry. Discriminant analyses of the body width–length (WL) ratios expected to change during pregnancy, correctly assigned pregnancy status for 14 of the 15 females of known pregnancy status. Only one pregnant female was incorrectly assigned as not pregnant. However, our results showed that length and body width cannot accurately allocate these bottlenose dolphins to sex or age class using photogrammetry techniques alone. The present study illustrates that UAS and laser photogrammetry measurements are comparable for small cetaceans and demonstrates that UAS measurements of body WL ratio can accurately assign pregnancy status in bottlenose dolphins.
Elliser C. R., van der Linde K., MacIver K. (2022): Adapting photo-identification methods to study poorly marked cetaceans: a case study for common dolphins and harbor porpoises. Mammalian Biology 102: 811-827.
FULL TEXT
Abstract
Historically, traditional photo-identification (photo-ID) methods have been applied to cetaceans with relatively small group sizes, closed and/or small populations, distinctive dorsal fin nicks and/or notches and behavior allowing for photographic capture. Conversely, species with larger group and/or population sizes that often occur in open populations, which have less distinctive natural markings and/or whose behavior is prohibitive for photographic capture, have not been the focus of many photo-ID studies. We review two successful photo-ID studies on species that have less distinctive markings, but differ in habitats and behavior: (1) common dolphins (Delphinus delphis) that live in oceanic environments, have large group sizes, occur in open populations and are easily observed and (2) harbor porpoises (Phocoena phocoena) that live in coastal environments, have small group sizes, but are behaviorally cryptic and elusive. We discuss how traditional photo-ID methods were adapted by incorporating different: (1) identification sides; (2) identification features; (3) levels of photo quality (PQ); (4) distinctiveness and; (5) methods for error checking. Adaptations include: using symmetry of the identification features to determine if both sides of the animal are used, using more than one identification feature, developing a matrix for describing/sorting by the identification features, using three levels of distinctiveness, incorporating PQ and distinctiveness into a flowchart to identify distinctively marked individuals (DMIs) and applying a more rigorous review process to identify possible errors in cataloguing. We discuss how adapting traditional photo-ID methods will improve our ability to use photo-ID for species not traditionally studied using this method.
Grimes C., Brent L. J., Weiss M. N., Franks D. W., Balcomb K. C., Ellifrit D. K., Ellis S., Croft D. P. (2022): The effect of age, sex, and resource abundance on patterns of rake markings in resident killer whales (Orcinus orca). Marine Mammal Science 38: 941-958.
FULL TEXT
Abstract
Fluctuations in aggressive behavior of group-living species can reflect social conflict and competition for resources faced by individuals throughout their lifespan and can negatively impact survival and reproduction. In marine mammals, where social interactions are difficult to observe, tooth rake marks can be used as an indicator of received aggression. Using 38 years of photographic data, we quantified the occurrence of tooth rake marks on wild resident killer whales (Orcinus orca), examining the effects of age, sex, and prey abundance on rake density. Our analysis revealed sex and age effects, with males exhibiting higher rake density than females and rake density declining significantly with age. Contrary to predictions, we observed an increase in rake density across the population as the abundance of their primary food resource, Chinook salmon (Oncorhynchus tshawytscha), increased. These results provide indirect evidence of fluctuations in received aggression from conspecifics across the lifespan of an individual, possibly reflecting changes in patterns of social conflict which may be mediated by resource abundance. Our findings highlight the need for further research to examine the fitness consequences of aggression in killer whales and to understand the proximate mechanisms by which resource abundance influences rates of aggression in the population.
Sarano V., Sarano F., Girardet J., Preud’homme A., Vitry H., Heuzey R., Sarano M., Delfour F., Glotin H., Adam O., Madon B., Jung J. (2022): Underwater photo-identification of sperm whales (Physeter macrocephalus) off Mauritius. Marine Biology Research 18: 131-146.
FULL TEXT
Abstract
The long-term monitoring of long-lived animals often requires individual identification. For cetaceans, this identification may be based on morphological characters observable from a boat such as shape, spots and cuts of the back, fluke and dorsal fins. However, for some species such as the sperm whales (Physeter macrocephalus), this approach may be challenging as individuals display a rather uniform skin pigmentation. They also do not very often show their fluke, complicating individual identification from a boat. Immature sperm whales that usually have an unharmed fluke may be excluded from photo-identification catalogues. Within the framework of the Maubydick project, focusing on the long-term monitoring of sperm-whales in Mauritius, passive underwater observation and video recording were used to identify long-lasting body markers (e.g. sex, ventral white markings, cut-outs of fins). A catalogue of 38 individuals (six adult males, 18 adult females and 14 immatures) enabled observers to record some nearly-daily, and yearly resightings. Advantages and disadvantages of this method are presented here. Such catalogues represent a robust baseline for conducting behavioural, genetic and acoustic studies in social species. Benefits of such newly acquired knowledge are of primary importance to implement relevant conservation plans in the marine realm.
Thompson J. W., Zero V. H., Schwacke L. H., Speakman T. R., Quigley B. M., Morey J. S., McDonald T. L. (2022): finFindR: Automated recognition and identification of marine mammal dorsal fins using residual convolutional neural networks. Marine Mammal Science 38: 139-150.
FULL TEXT
Abstract
Photographic identification is an essential research and management tool for marine mammal scientists. However, manual identification of individuals is time-consuming. To shorten processing times, we developed finFindR, an open-source application that uses a series of neural networks to autonomously locate dorsal fins in unedited field images, quantify an individual’s unique fin characteristics, and match them to an existing photograph catalog. During a blind test comparing manual searching to finFindR for common bottlenose dolphin (Tursiops Tursiops truncatus) photographs, experienced photo-identification technicians achieved similar match rates but examined an order of magnitude fewer photographs using finFindR (an average of 10 required with finFindR versus 124 with manual search). In those tests, the correct identity was ranked in the first position in 88% of cases and was within the top 50 ranked positions in 97% of cases. Our observations suggest that finFindR’s matching capabilities are robust to moderate variation in image quality and fin distinctiveness. Importantly, finFindR allows users to build a catalog of known individuals through time and match an unlimited number of individuals instead of being restricted to a predefined set. finFindR’s convolutional neural networks could be re-trained to identify members of many marine mammal species without altering finFindR’s inherent structure.
Maglietta R., Bussola A., Carlucci R., Fanizza C., Dimauro G. (2023): ARIANNA: a novel deep learning-based system for fin contours analysis in individual recognition of dolphins. Intelligent Systems with Applications 18: 200207.
FULL TEXT
Abstract
Starting from a digital image that represents the dolphin’s body, distinctive features are extracted and used to find the identity of the unknown dolphin in a set of known individuals. This process is called photo identification, used by experts to monitor dolphins, providing relevant data to preserve the environment and its biodiversity. In this work, we show how semantic segmentation can be used to automatically extract a dolphin’s fin contour starting from a cropped photo of the fin, and how this contour can be used for individual identification. A novel contour-based system, called ARIANNA, for the automated cetacean photo identification was designed, developed and tested. The novelty of this system is the adoption of two original modules. The first one, which takes as input a new cropped fin image of unknown dolphin, is devoted to the extraction of a mask that depicts the outline of the unknown fin; the core of this module is a trained neural network, specialized in semantic segmentation of images. The second module is designed to compare the outline of the unknown fin with the outlines of all known dolphins, collected in a referring catalogue, returning a ranked list of the best matches where to search the dolphin identity. The experiments were conducted on images collected between 2013 and 2020 in the Northern Ionian Sea (Central-eastern Mediterranean Sea), which presented cropped fins of Risso’s dolphin Grampus griseus, one of the least-known cetacean species on a global and Mediterranean scale. The results suggest that ARIANNA provides advances over the state-of-the-art methods, can efficiently assist researchers in the photo identification of dolphins and can be a starting point for further studies on the photo identification of different species based on semantic segmentation.
Patton P. T., Cheeseman T., Abe K., Yamaguchi T., Reade W., Southerland K., Howard A., Oleson E. M., Allen J. B., Ashe E., Athayde A., et al. (2023): A deep learning approach to photo–identification demonstrates high performance on two dozen cetacean species. Methods in Ecology and Evolution 14: 2611-2625.
FULL TEXT
Abstract
Researchers can investigate many aspects of animal ecology through noninvasive photo–identification. Photo–identification is becoming more efficient as matching individuals between photos is increasingly automated. However, the convolutional neural network models that have facilitated this change need many training images to generalize well. As a result, they have often been developed for individual species that meet this threshold. These single-species methods might underperform, as they ignore potential similarities in identifying characteristics and the photo–identification process among species. In this paper, we introduce a multi-species photo–identification model based on a state-of-the-art method in human facial recognition, the ArcFace classification head. Our model uses two such heads to jointly classify species and identities, allowing species to share information and parameters within the network. As a demonstration, we trained this model with 50,796 images from 39 catalogues of 24 cetacean species, evaluating its predictive performance on 21,192 test images from the same catalogues. We further evaluated its predictive performance with two external catalogues entirely composed of identities that the model did not see during training. The model achieved a mean average precision (MAP) of 0.869 on the test set. Of these, 10 catalogues representing seven species achieved a MAP score over 0.95. For some species, there was notable variation in performance among catalogues, largely explained by variation in photo quality. Finally, the model appeared to generalize well, with the two external catalogues scoring similarly to their species’ counterparts in the larger test set. From our cetacean application, we provide a list of recommendations for potential users of this model, focusing on those with cetacean photo–identification catalogues. For example, users with high quality images of animals identified by dorsal nicks and notches should expect near optimal performance. Users can expect decreasing performance for catalogues with higher proportions of indistinct individuals or poor quality photos. Finally, we note that this model is currently freely available as code in a GitHub repository and as a graphical user interface, with additional functionality for collaborative data management, via Happywhale.com.
ELEPHANTS
Bedetti A., Greyling C., Paul B., Blondeau J., Clark A., Malin H., Horne J., Makukule R., Wilmot J., Eggeling T., Henley M. (2020): System for Elephant Ear-pattern Knowledge (SEEK) to identify individual African elephants. Pachyderm 61: 63-77.
FULL TEXT
Abstract
Elephant numbers have drastically declined over the past century with illegal killings, habitat fragmentation and human-elephant-conflict representing the greatest threats. Information on estimates of abundance and demographics are important to understand the long-term implications of these threats. Mark-resighting studies can provide valuable insights but depend on the individual identification of numerous elephants within populations across both Africa and Asia. Most photographic elephant identification studies are still reliant on human memory and manual matching of known individuals. A process that is not only labour intensive but also largely dependent on experiential skills that need to be developed over time by researchers. Over the course of almost 25 years, Elephants Alive has developed a unique System of Elephant Ear-pattern Knowledge (SEEK), which makes allowance for rapid individual identification of savannah elephants with reduced observer bias using basic software while also accommodating missing information or changes in identification features over time.
Duporge I., Isupova O., Reece S., Macdonald D. W., Wang T. (2021): Using very‐high‐resolution satellite imagery and deep learning to detect and count African elephants in heterogeneous landscapes. Remote Sensing in Ecology and Conservation 7: 369-381.
FULL TEXT
Abstract
Satellites allow large-scale surveys to be conducted in short time periods with repeat surveys possible at intervals of <24 h. Very-high-resolution satellite imagery has been successfully used to detect and count a number of wildlife species in open, homogeneous landscapes and seascapes where target animals have a strong contrast with their environment. However, no research to date has detected animals in complex heterogeneous environments or detected elephants from space using very-high-resolution satellite imagery and deep learning. In this study, we apply a Convolution Neural Network (CNN) model to automatically detect and count African elephants in a woodland savanna ecosystem in South Africa. We use WorldView-3 and 4 satellite data –the highest resolution satellite imagery commercially available. We train and test the model on 11 images from 2014 to 2019. We compare the performance accuracy of the CNN against human accuracy. Additionally, we apply the model on a coarser resolution satellite image (GeoEye-1) captured in Kenya, without any additional training data, to test if the algorithm can generalize to an elephant population outside of the training area. Our results show that the CNN performs with high accuracy, comparable to human detection capabilities. The detection accuracy (i.e., F2 score) of the CNN models was 0.78 in heterogeneous areas and 0.73 in homogenous areas. This compares with the detection accuracy of the human labels with an averaged F2 score 0.77 in heterogeneous areas and 0.80 in homogenous areas. The CNN model can generalize to detect elephants in a different geographical location and from a lower resolution satellite. Our study demonstrates the feasibility of applying state-of-the-art satellite remote sensing and deep learning technologies for detecting and counting African elephants in heterogeneous landscapes. The study showcases the feasibility of using high resolution satellite imagery as a promising new wildlife surveying technique. Through creation of a customized training dataset and application of a Convolutional Neural Network, we have automated the detection of elephants in satellite imagery with accuracy as high as human detection capabilities. The success of the model to detect elephants outside of the training data site demonstrates the generalizability of the technique.
Wierucka K., Henley M. D., Mumby H. S. (2021): Acoustic cues to individuality in wild male adult African savannah elephants (Loxodonta africana). PeerJ 9: e10736.
FULL TEXT
Abstract
The ability to recognize conspecifics plays a pivotal role in animal communication systems. It is especially important for establishing and maintaining associations among individuals of social, long-lived species, such as elephants. While research on female elephant sociality and communication is prevalent, until recently male elephants have been considered far less social than females. This resulted in a dearth of information about their communication and recognition abilities. With new knowledge about the intricacies of the male elephant social structure come questions regarding the communication basis that allows for social bonds to be established and maintained. By analyzing the acoustic parameters of social rumbles recorded over 1.5 years from wild, mature, male African savanna elephants (Loxodonta africana) we expand current knowledge about the information encoded within these vocalizations and their potential to facilitate individual recognition. We showed that social rumbles are individually distinct and stable over time and therefore provide an acoustic basis for individual recognition. Furthermore, our results revealed that different frequency parameters contribute to individual differences of these vocalizations.
Chan S. C., Chui S. Y., Pretorius Y., Karczmarski L. (2022): Estimating population parameters of African elephants: a photographic mark-recapture application in a South African protected area. Mammalian Biology 102: 1231-1247.
FULL TEXT
Abstract
Accurate estimates of demographic parameters are instrumental in effective management of animal populations. For species with individually distinctive features, photo-identification (photo-ID) provides a reliable means to gather capture–recapture data for population parameter estimation with considerable precision and accuracy. We use a 3-year photo-ID mark-recapture dataset of African savannah elephants (Loxodonta africana) in Pilanesberg National Park (PNP), South Africa, to model their population size and estimate survival rates. All photographed elephants, irrespective of age, were individually identified based on their unique pattern of facial wrinkles. The population currently numbers 385 elephants (95% CI = 380–401), of which nearly half are grown individuals in a sex ratio of 1 male: 1.23 female. Considerable heterogeneity in capture and recapture probabilities, both within and between sex-age classes suggests some form of individual-specific or herd-specific variability, perhaps behavioural or spatio-behavioural dissimilarity within the PNP population. Estimated annual survival rates are high (0.967–0.996) and do not differ between sex-age classes, a likely expression of an extended parental care, low predation pressure, access to rich food and water resources, and absence of targeted killing/poaching. The lack of detectable difference between sexes in adult survival/mortality sets PNP elephants apart from other known African elephant populations and warrants further research attention. Given previous estimates (aerial counts in the early 2000s), the PNP elephant population has grown ~ 5.7% per annum over a 16-year period. This is similar to what is reported in other conservation areas in South Africa, but considerably lower than previously projected. Natural mortality, even if low as 0.4–3.3%, is not negligible and plays a role in moderating population growth. This realisation must be recognised when considering population management measures. It is, therefore, important to obtain and apply the most up-to-date population-specific demographic parameters when making management decisions. Periodic photo-ID surveys with mark-recapture analyses can generate such demographic indicators with a considerable accuracy and should be adopted as a useful tool to inform management decisions, complimentary to direct aerial counts, especially in small-to-medium size fenced conservation areas.
Chui S., Karczmarski L. (2022): Everyone matters: identification with facial wrinkles allows more accurate inference of elephant social dynamics. Mammalian Biology 102: 645-666.
FULL TEXT
Abstract
Reliable identification of individuals plays an important role in behavioural studies of free-ranging animal populations. In field studies of elephants, the naturally acquired markings on their ears, such as notches, tears and holes, are frequently used for individual identification. Although not as easily discernible from a distance as ear markings, the facial wrinkle pattern around the eye, temporal gland and ear on both sides of elephant’s head are individually unique and, with application of high-resolution photographs, can also be used for individual identification. In fact, the wrinkle pattern is highly consistent and reliable as the primary identifiable feature; it changes little over time, facilitates identification of individuals with non-distinctive ear pattern (e.g., calves), and performs well against several practical challenges to the traditional ear-pattern approach. We used data from a 3-year photo-identification study of African elephant population to examine how the two identification methods, one that uses marks on elephant ears and the other using facial wrinkle pattern, affect the results of basic analyses of social dynamics, such as patterns of associations and social preferences, derived from datasets generated with these two identification methods. Comparative analyses demonstrate that by increasing the identifiability of otherwise poorly marked individuals and minimising identification error, the wrinkle-based method reduces substantially the sample bias, enhances the robustness of datasets, and minimises analytical error. While ear-pattern-based distinctiveness is age-dependent, the wrinkle-based method facilitates a more representative sample of the population, with photo-ID data collected non-discriminately across all age classes. This carries further implications, such as enabling more accurate depiction of elephant sociality, long-term population monitoring, calculation of class-specific population parameters, etc. Adopting the facial wrinkle pattern for elephant individual identification is relatively easy, and we encourage future and ongoing studies to consider incorporating the facial wrinkle approach. Given the advantages of wrinkle-based identification and recent advances in machine learning, we recommend it to be considered for the development of automated matching algorithms; such development would benefit long-term socio-behavioural studies and monitoring of elephant populations.
de Silva E. M., Kumarasinghe P., Indrajith K. K., Pushpakumara T. V., Vimukthi R. D., de Zoysa K., Gunawardana K., de Silva S. (2022): Feasibility of using convolutional neural networks for individual-identification of wild Asian elephants. Mammalian Biology 10: 931-941.
FULL TEXT
Abstract
Individual identification is a basic requirement for research in behavior, ecology and conservation. Photographic records are commonly used in situations where individuals are visually distinct. However, keeping track of identities becomes challenging with increasing population sizes and corresponding datasets. There is growing interest in the potential of deep-learning methods for computer vision to assist with automating this task. Here we apply Convolutional Neural Networks, a popular architecture for Artificial Neural Networks used in image classification, to the problem of identifying individual Asian elephants through photographs. We evaluate the performance of five different types of CNN models used in facial recognition (VGG16, ResNet50, InceptionV3, Xception, and Alexnet), on datasets representing three different feature regions (the full body, face, and ears), trained with two techniques (transfer learning vs. training from scratch) for n = 56 elephants. We tested accuracy in matching the top candidate as well as top five candidates. We found that VGG16 trained with the transfer-learning technique outperformed other models on the body and face datasets with accuracies of 21.34% and 42.35%, respectively, in matching the top candidate. Nevertheless, the best performance was achieved by an Xception model trained from the scratch on the ear dataset, with an accuracy of 89.02% for matching the top candidate and 99.27% for including the correct individual among the top five. However, this impressive level of accuracy was obtained with a dataset of 3816 labeled training images of 56 elephants. There are more than 1000 wild elephants in the population under observation, requiring extensive human effort and skill to initially annotate the images used as training data. Therefore, we consider this approach impractical for monitoring large wild populations. Nevertheless this it could be very useful in record keeping and fraud prevention for large captive elephant populations, as well as monitoring animals that have been rehabilitated and released or moved for management purposes.
Montero-De La Torre S., Jacobson S. L., Chodorow M., Yindee M., Plotnik J. M. (2023): Day and night camera trap videos are effective for identifying individual wild Asian elephants. PeerJ 11: e15130.
FULL TEXT
Abstract
Regular monitoring of wild animal populations through the collection of behavioral and demographic data is critical for the conservation of endangered species. Identifying individual Asian elephants (Elephas maximus), for example, can contribute to our understanding of their social dynamics and foraging behavior, as well as to human-elephant conflict mitigation strategies that account for the behavior of specific individuals involved in the conflict. Wild elephants can be distinguished using a variety of different morphological traits—e.g., variations in ear and tail morphology, body scars and tumors, and tusk presence, shape, and length—with previous studies identifying elephants via direct observation or photographs taken from vehicles. When elephants live in dense forests like in Thailand, remote sensing photography can be a productive approach to capturing anatomical and behavioral information about local elephant populations. While camera trapping has been used previously to identify elephants, here we present a detailed methodology for systematic, experimenter differentiation of individual elephants using data captured from remote sensing video camera traps. In this study, we used day and night video footage collected remotely in the Salakpra Wildlife Sanctuary in Thailand and identified 24 morphological characteristics that can be used to recognize individual elephants. A total of 34 camera traps were installed within the sanctuary as well as crop fields along its periphery, and 107 Asian elephants were identified: 72 adults, 11 sub-adults, 20 juveniles, and four infants. We predicted that camera traps would provide enough information such that classified morphological traits would aid in reliably identifying the adult individuals with a low probability of misidentification. The results indicated that there were low probabilities of misidentification between adult elephants in the population using camera traps, similar to probabilities obtained by other researchers using handheld cameras. This study suggests that the use of day and night video camera trapping can be an important tool for the long-term monitoring of wild Asian elephant behavior, especially in habitats where direct observations may be difficult.
HEDGEHOGS, SHREWS, AND DESMANS
Esser D., Schehka S., Zimmermann E. (2008). Species-specificity in communication calls of tree shrews (Tupaia: Scandentia). Journal of Mammalogy 89: 1456-1463.
FULL TEXT
Abstract
Tree shrews are small mammals living in the tropical forest of Southeast Asia. The habitus of species within the genus Tupaia is often quite similar, so that it is difficult to differentiate the species based on their morphology. We applied comparative bioacoustics, a tool successfully used to discriminate cryptic species of nocturnal mammals, to investigate whether species in the diurnal genus Tupaia can be recognized noninvasively on the basis of a conspicuous loud call, the chatter. We studied to what extent the chatter call of 2 tree shrew species, Tupaia glis and T. belangeri, differed in acoustic structure. We also acoustically analyzed the chatter call of T. chinensis, a subspecies or closely related parapatric species of T. belangeri Analyzed acoustic features allowed assigning chatter calls with a probability of more than 73% to the species that produced them. Bioacoustical differences are in line with subtle morphological differences, supporting species status for all 3 studied tree shrew species and corroborating immunodiffusion and genetic data that differentiate T. glis and T. belangeri. Loud calls may offer a reliable noninvasive tool for species diagnosis and discrimination in cryptic species of this diurnal mammalian group.
Mori E., Menchetti M., Bertolino S., Mazza G., Ancillotto L. (2015): Reappraisal of an old cheap method for marking the European hedgehog. Mammal Research 60: 189-193.
FULL TEXT
Abstract
Marking quilled animals for individual recognition may be challenging. This is particularly true for European hedgehogs Erinaceus europaeus, whose defense mechanism involves the concealment of muzzle and underparts. Heat-shrink tubes have been widely used to mark quills, but they do not adapt to the morphological structure of the spines and are rapidly lost, thus reducing method effectiveness. We adapted a cheap and ethical method used to mark crested porcupine quills, which involves the use of colored adhesive tapes applied to quills. The retention period of this marking technique lasts up to 9 months, allowing short-term field studies and possibly dispersal distances measurements. The method could be improved by doubling the number of marked spines and by reapplying adhesive tape at every recapture event. Moreover, the use of a marking code can be obtained by subdividing the body of the hedgehog into six body areas, to increase the number of marking possibilities, through the combination of tape colors and body areas.
MANATEES
Langtimm C. A., Beck C. A., Edwards H. H., Fick‐Child K. J., Ackerman B. B., Barton S. L., Hartley W. C. (2004): Survival estimates for Florida manatees from the photo‐identification of individuals. Marine Mammal Science 20: 438-463.
FULL TEXT
Abstract
We estimated adult survival probabilities for the endangered Florida manatee (Trichechus manatus latirostris) in four regional populations using photoidentification data and open‐population capture‐recapture statistical models. The mean annual adult survival probability over the most recent 10‐yr period of available estimates was as follows: Northwest ‐ 0.956 (SE 0.007), Upper St. Johns River ‐ 0.960 (0.011), Atlantic Coast ‐ 0.937 (0.008), and Southwest ‐ 0.908 (0.019). Estimates of temporal variance independent of sampling error, calculated from the survival estimates, indicated constant survival in the Upper St. Johns River, true temporal variability in the Northwest and Atlantic Coast, and large sampling variability obscuring estimates for the Southwest. Calf and subadult survival probabilities were estimated for the Upper St. Johns River from the only available data for known‐aged individuals: 0.810 (95% CI 0.727–0.873) for 1st year calves, 0.915 (0.827–0.960) for 2nd year calves, and 0.969 (0.946–0.982) for manatee 3 yr or older. These estimates of survival probabilities and temporal variance, in conjunction with estimates of reproduction probabilities from photoidentification data can be used to model manatee population dynamics, estimate population growth rates, and provide an integrated measure of regional status.
Landeo-Yauri S. S., Ramos E. A., Castelblanco-Martínez D. N., Niño-Torres C. A., Searle L. (2020): Using small drones to photo-identify Antillean manatees: a novel method for monitoring an endangered marine mammal in the Caribbean Sea. Endangered Species Research 41: 79-90.
FULL TEXT
Abstract
Population assessments and species monitoring for many endangered marine megafauna are limited by the challenges of identifying and tracking individuals that live underwater in remote and sometimes inaccessible areas. Manatees can acquire scars from watercraft injury and other incidences that can be used to identify individuals. Here we describe a novel method for photo-identification of Antillean manatees Trichechus manatus manatus using aerial imagery captured during flights with a small multirotor drone. Between 2016 and 2017, we conducted 103 flights to detect and observe manatees in Belize, primarily at St. George’s Caye (SGC) near the Belize Barrier Reef. Review of aerial videos from these flights resulted in 279 sightings of manatees (245 adults, 34 calves). High-resolution images of individual manatees were extracted and classified according to image quality and distinctiveness of individual manatees for photo-identification. High-quality images of manatees classified as sufficiently distinctive were used to create a catalog of 17 identifiable individuals. At SGC, 21% of all sighted adult manatees (N = 214) were considered photo-identifiable over time. We suggest that the method can be used for investigating individual site fidelity, habitat use, and behavior of manatee populations. Our photo-identification protocol has the potential to improve long-term monitoring of Antillean manatees in Belize and can be applied throughout clear, shallow waters in the Caribbean and elsewhere.
Beck C. A. (2022): Manatee population traits elucidated through photo-identification. Mammalian Biology 102: 1073-1088.
FULL TEXT
Abstract
Data on the demography and distribution of wildlife populations are important for informing conservation and management decisions; however, determination of life history traits and population trends often are elusive. All four extant species in the order Sirenia are deemed vulnerable to extinction; therefore, determining the demography and distribution for populations worldwide is crucial. Aerial surveys, radio-tagging and tracking, genetic sampling and analyses, health assessments, carcass examination, and photographic documentation are all techniques used to study sirenian populations. A 40 +-year computer-aided catalog of images and demography data collected on Florida manatees enables searches of individuals by descriptions of feature (scar) types and has enabled estimates of annual survival and reproductive rates, documented extra-limital movements, and advanced modeling designs. Photography is discussed as a method for the documentation of unique and acquired features specifically on Florida manatees. By means of these features, individual Florida manatees have been re-identified as far from their established range as Cape Cod, Massachusetts, Houston, Texas, and in Cuba, The Bahamas, and Mexico. The length of gestation (11–13 months) and calf dependency (1–3 years), and potential longevity in the wild (> 50 years), have been verified. To meet the challenge of an increasing number of images collected with the advent of digital photography, there has been an increasing interest and potential for new techniques to assist with individual identification. Several researchers are utilizing drones and artificial intelligence to find, photograph, and streamline the individual identification of sirenians as well as other marine mammal species. New techniques have potential to simplify the photographic identification of Florida manatees. Photographic documentation could be a model for demographic and distribution research of sirenian populations outside of Florida and as a tool to monitor the viability of sirenian populations, particularly as threats emerge due to anthropogenic pressures and global climate change.
PRIMATES
Crouse D., Jacobs R. L., Richardson Z., Klum S., Jain A., Baden A. L., Tecot S. R. (2017): LemurFaceID: a face recognition system to facilitate individual identification of lemurs. BMC Zoology 2: 2.
FULL TEXT
Abstract
Long-term research of known individuals is critical for understanding the demographic and evolutionary processes that influence natural populations. Current methods for individual identification of many animals include capture and tagging techniques and/or researcher knowledge of natural variation in individual phenotypes. These methods can be costly, time-consuming, and may be impractical for larger-scale, population-level studies. Accordingly, for many animal lineages, long-term research projects are often limited to only a few taxa. Lemurs, a mammalian lineage endemic to Madagascar, are no exception. Long-term data needed to address evolutionary questions are lacking for many species. This is, at least in part, due to difficulties collecting consistent data on known individuals over long periods of time. Here, we present a new method for individual identification of lemurs (LemurFaceID). LemurFaceID is a computer-assisted facial recognition system that can be used to identify individual lemurs based on photographs.
Brust C. A., Burghardt T., Groenenberg M., Kading C., Kuhl H. S., Manguette M. L., Denzler J. (2017): Towards automated visual monitoring of individual gorillas in the wild. Proceedings of the IEEE International Conference on Computer Vision Workshops 2017: 2820-2830.
FULL TEXT
Abstract
In this paper we report on the context and evaluation of a system for an automatic interpretation of sightings of individual western lowland gorillas (Gorilla gorilla gorilla) as captured in facial field photography in the wild. This effort aligns with a growing need for effective and integrated monitoring approaches for assessing the status of biodiversity at high spatio-temporal scales. Manual field photography and the utilisation of autonomous camera traps have already transformed the way ecological surveys are conducted. In principle, many environments can now be monitored continuously, and with a higher spatio-temporal resolution than ever before. Yet, the manual effort required to process photographic data to derive relevant information delimits any large scale application of this methodology. The described system applies existing computer vision techniques including deep convolutional neural networks to cover the tasks of detection and localisation, as well as individual identification of gorillas in a practically relevant setup. We evaluate the approach on a relatively large and challenging data corpus of 12,765 field images of 147 individual gorillas with image-level labels (i.e. missing bounding boxes) photographed at Mbeli Bai at the Nouabale-Ndoki National Park, Republic of Congo. Results indicate a facial detection rate of 90.8% AP and an individual identification accuracy for ranking within the Top 5 set of 80.3%. We conclude that, whilst keeping the human in the loop is critical, this result is practically relevant as it exemplifies model transferability and has the potential to assist manual identification efforts. We argue further that there is significant need towards integrating computer vision deeper into ecological sampling methodologies and field practice to move the discipline forward and open up new research horizons.
Schofield D., Nagrani A., Zisserman A., Hayashi M., Matsuzawa T., Biro D., Carvalho S. (2019): Chimpanzee face recognition from videos in the wild using deep learning. Science Advances 5: eaaw0736.
FULL TEXT
Abstract
Video recording is now ubiquitous in the study of animal behavior, but its analysis on a large scale is prohibited by the time and resources needed to manually process large volumes of data. We present a deep convolutional neural network (CNN) approach that provides a fully automated pipeline for face detection, tracking, and recognition of wild chimpanzees from long-term video records. In a 14-year dataset yielding 10 million face images from 23 individuals over 50 hours of footage, we obtained an overall accuracy of 92.5% for identity recognition and 96.2% for sex recognition. Using the identified faces, we generated co-occurrence matrices to trace changes in the social network structure of an aging population. The tools we developed enable easy processing and annotation of video datasets, including those from other species. Such automated analysis unveils the future potential of large-scale longitudinal video archives to address fundamental questions in behavior and conservation.
Guo S., Xu P., Miao Q., Shao G., Chapman C. A., Chen X., He G., Fang D., Zhang H., Sun Y., Shi Z., Li B. (2020): Automatic identification of individual primates with deep learning techniques. iScience 23: 101412.
FULL TEXT
Abstract
The difficulty of obtaining reliable individual identification of animals has limited researcher’s ability to obtain quantitative data to address important ecological, behavioral, and conservation questions. Traditional marking methods placed animals at undue risk. Machine learning approaches for identifying species through analysis of animal images has been proved to be successful. But for many questions, there needs a tool to identify not only species but also individuals. Here, we introduce a system developed specifically for automated face detection and individual identification with deep learning methods using both videos and still-framed images that can be reliably used for multiple species. The system was trained and tested with a dataset containing 102,399 images of 1,040 individuals across 41 primate species whose individual identity was known and 6,562 images of 91 individuals across four carnivore species. For primates, the system correctly identified individuals 94.1% of the time and could process 31 facial images per second.
Schofield D. P., Albery G. F., Firth J. A., Mielke A., Hayashi M., Matsuzawa T., Biro D., Carvalho S. (2023): Automated face recognition using deep neural networks produces robust primate social networks and sociality measures. Methods in Ecology and Evolution 14: 1937-1951.
FULL TEXT
Abstract
Longitudinal video archives of behaviour are crucial for examining how sociality shifts over the lifespan in wild animals. New approaches adopting computer vision technology hold serious potential to capture interactions and associations between individuals in video at large scale; however, such approaches need a priori validation, as methods of sampling and defining edges for social networks can substantially impact results. Here, we apply a deep learning face recognition model to generate association networks of wild chimpanzees using 17 years of a video archive from Bossou, Guinea. Using 7 million detections from 100 h of video footage, we examined how varying the size of fixed temporal windows (i.e. aggregation rates) for defining edges impact individual-level gregariousness scores. The highest and lowest aggregation rates produced divergent values, indicating that different rates of aggregation capture different association patterns. To avoid any potential bias from false positives and negatives from automated detection, an intermediate aggregation rate should be used to reduce error across multiple variables. Individual-level network-derived traits were highly repeatable, indicating strong inter-individual variation in association patterns across years and highlighting the reliability of the method to capture consistent individual-level patterns of sociality over time. We found no reliable effects of age and sex on social behaviour and despite a significant drop in population size over the study period, individual estimates of gregariousness remained stable over time. We believe that our automated framework will be of broad utility to ethology and conservation, enabling the investigation of animal social behaviour from video footage at large scale, low cost and high reproducibility. We explore the implications of our findings for understanding variation in sociality patterns in wild ape populations. Furthermore, we examine the trade-offs involved in using face recognition technology to generate social networks and sociality measures. Finally, we outline the steps for the broader deployment of this technology for analysis of large-scale datasets in ecology and evolution.
RABBITS, HARES, AND PIKAS
Millar C. I., Hickman K. T. (2021): Camera traps provide insights into American pika site occupancy, behavior, thermal relations, and associated wildlife diversity. Western North American Naturalist 81: 141-170.
FULL TEXT
Abstract
Custom camera traps were positioned at American pika (Ochotona princeps) haypiles in 12 warm, low-elevation locations of eastern California over 5 years and during warm and cold seasons. These camera traps detected 26 mammal and 10 bird species, including 4331 pika events as well as visits by 16 sympatric herbivores and 7 pika predators. Camera traps documented pika occupancy at some sites that had been evaluated from field surveys as extirpated, and they also confirmed field assessments of extirpation at other sites. Individual pikas could be distinguished by scars, size, and pelage diagnostics, allowing animals to be followed through sequences of photos and enabling behavioral interpretations and documentation of winter and warm-season activities. Temperature measurements at haypiles and talus interiors corroborated prior findings that rocky interiors are much cooler than surfaces, have highly attenuated daily temperature fluctuations, and offer refuge for pikas from high daytime temperatures. Nocturnal activity was recorded for pikas as well as for many other species, but we found little evidence that night activity of pikas at haypiles increased when prior day temperatures were excessively warm. The capacity of pikas to be active at all times of the day adds to their resilience in the face of predators, foraging needs, and changing climates.
RODENTS
Pollard K. A., Blumstein D. T., Griffin S. C. (2010): Pre‐screening acoustic and other natural signatures for use in noninvasive individual identification. Journal of Applied Ecology 47: 1103-1109.
FULL TEXT
Abstract
Common ecological tasks, such as wildlife monitoring, adaptive management, and behavioural study, often make use of natural signatures (e.g. animal calls or visual markings) to identify individual animals noninvasively. However, there is no accepted method for pre-screening candidate natural signatures to select which signatures are the best-suited for this purpose. In this paper, we suggest a pre-screening checklist and focus on the challenge of assessing a candidate signature’s individuality. Individuality is critical, as the use of low-individuality natural signatures can lead to misidentification of individuals and therefore bias estimation of population parameters and population response to management actions. An information-based metric of individuality could assist researchers with selecting suitable signatures by allowing comparison among candidate signatures and providing an estimate of how many individuals may be reliably discriminated using a particular signature. Before an individuality metric can be used to pre-screen natural signatures, the metric must first be calculated from preliminary sampling and must be robust to typical sampling concerns. We used field-collected animal vocalizations as well as simulations to test how robust the metric is to variation in sampling design. We found that the metric is fairly robust to the number of animals sampled and the number of sessions (e.g. calling bouts) analysed, but that it is sensitive to the number of observations per session. Managers and researchers could save time and energy and improve the accuracy of estimates (such as abundance, survival, or population response) based on individual identification by first pre-screening candidate natural signatures for their individuality. As long as the number of observations per session is controlled, the relative values of the individuality metric can be meaningfully compared. The metric can thus be used as a tool to estimate relative individuality and so facilitates a difficult step in choosing a natural signature for noninvasive individual identification. We include instructions on how to calculate and interpret the individuality metric.
Di Cerbo A. R., Biancardi C. M. (2013): Monitoring small and arboreal mammals by camera traps: effectiveness and applications. Acta Theriologica 58: 279-283.
FULL TEXT
Abstract
Camera trapping has been widely applied to studies of medium to large terrestrial mammals, but its application to small arboreal mammals has hardly been tested. We employed camera trapping and other conventional monitoring methods during a mammal survey in a Site of Community Importance located within the Adda North Regional Park (Lombardy, Italy). Camera trapping was particularly successful for monitoring arboreal mammals, allowing the first detection of presence of the invasive grey squirrel (Sciurus carolinensis) in an area occupied by indigenous red squirrels (Sciurus vulgaris) and the collection of a large amount of data on squirrels and common dormice (Muscardinus avellanarius). When triggered, cameras were set to record short video clips (10 to 40 s). More than 400 events were recorded and analysed, mainly from the autumn and winter months. The daily activity pattern of both species displayed a trend from two to three activity peaks in summer to a unimodal pattern in winter, with the peaks of the two species temporally separated. Camera trapping could be a useful method also when applied to monitoring small mammals, particularly endangered arboreal or invasive alien species. For instance, the monitoring of the spread of S. carolinensis is particularly important, where the early detection of new population can be crucial for the conservation of indigenous European species. Camera trapping can be an effective addition to traditional survey methods. It provides a simple non-invasive technique for collecting a large amount of data per device with limited human effort.
Mills C. A., Godley B. J., Hodgson D. J. (2016): Take only photographs, leave only footprints: novel applications of non-invasive survey methods for rapid detection of small, arboreal animals. Plos One 11: e0146142.
FULL TEXT
Abstract
The development of appropriate wildlife survey techniques is essential to promote effective and efficient monitoring of species of conservation concern. Here, we demonstrate the utility of two rapid-assessment, non-invasive methods to detect the presence of elusive, small, arboreal animals. We use the hazel dormouse, Muscardinus avellanarius, a rodent of conservation concern, as our focal species. Prevailing hazel dormouse survey methods are prolonged (often taking months to years to detect dormice), dependent on season and habitat, and/or have low detection rates. Alternatives would be of great use to ecologists who undertake dormouse surveys, especially those assessing the need for mitigation measures, as legally required for building development projects. Camera traps and footprint tracking are well-established tools for monitoring elusive large terrestrial mammals, but are rarely used for small species such as rodents, or in arboreal habitats. In trials of these adapted methods, hazel dormice visited bait stations and were successfully detected by both camera traps and tracking equipment at each of two woodland study sites, within days to weeks of installation. Camera trap images and footprints were of adequate quality to allow discrimination between two sympatric small mammal species (hazel dormouse and wood mouse, Apodemus sylvaticus). We discuss the relative merits of these methods with respect to research aims, funds, time available and habitat.
Dytkowicz M., Hinds R., Megill W. M., Buttschardt T. K., Rosell F. (2023): A camera trapping method for the targeted capture of Eurasian beaver (Castor fiber) tails for individual scale pattern recognition. European Journal of Wildlife Research 69: 39.
FULL TEXT
Abstract
Camera traps are commonly used to monitor and study wild animals in their natural habitat, with minimal disturbance. Several investigations have shown that the natural markings of animals for some species can be used for individual recognition. However, most commercially available cameras are unable to obtain photos of sufficient quality to highlight these features. Our study further exemplifies the use of applying an external lens to a camera, to obtain higher quality images. We tested various lenses and their ability to record the scale patterns on Eurasian beaver (Castor fiber) tails, for individual identification. We tested eleven different commercially available camera trap models, across six different beaver territories in the Districts of Kleve and Wesel (North Rhine-Westphalia, Germany). The use of an external lens, attached to the camera, produced the best quality pictures for reliable identification of individual beavers based on the scale patterns on their tales. These results further exemplify the application of external lenses for improving image quality for individual recognition which has potential applications for other species.
Hinds R., Dytkowicz M., Tania M., Megill W. M., Rosell F. (2023): A tale of tails: the use of Eurasian beaver (Castor fiber) tails for ageing and individual identification. European Journal of Wildlife Research 69: 88.
FULL TEXT
Abstract
With increasing technology and knowledge, the range of methods used to monitor wildlife is growing. As many invasive techniques have been shown to negatively impact study populations, the use of non-invasive methods is increasing. With Eurasian beaver (Castor fiber) reintroductions occurring across much of Europe, monitoring of beavers is becoming increasingly important; however, some frequently used techniques are invasive. We therefore aimed to examine potentially non-invasive methods of identifying and ageing them from the tail. Tails from previously deceased beavers were photographed with a Nikon D3500 DSLR camera across 3 distances: ‘close’, ‘medium’ and ‘far’, and the pattern of the scales were examined by eye to determine accuracy of individual identification. Photographs including a grey standard were used to determine the accuracy of ageing beavers from the colour of the tail. The accuracy of individual identification was 100% across all distances; however, the results from ageing showed the method to be inaccurate. The success of the individual identification shows that this method could be effectively used as a non-invasive method for monitoring beaver populations, especially in captivity.
UNGULATES
Alibhai S. K., Jewell Z. C., Law P. R. (2008): A footprint technique to identify white rhino Ceratotherium simum at individual and species levels. Endangered Species Research 4: 205-218.
FULL TEXT
Abstract
A non-invasive and cost-effective footprint identification technique (FIT) is presented, which can aid the identification of individual white rhino Ceratotherium simum and the differentiation of this species from black rhino Diceros bicornis. FIT is an adaptation of a traditional tracking identification technique and is a useful censusing and monitoring tool for wildlife conservation. We implemented FIT to identify 40 white rhino. Geometric profiles were extracted from digital images of footprints, and subjected to an algorithm based on multivariate statistical analyses. FIT’s classification rules were tested using a dataset of 1276 footprints from 159 tracks of 40 white rhino from a fenced wild population in Namibia. Using 2 different test models for pairwise track matching, FIT gave accuracies of 91 and 95% for population estimate (census) prediction. In a monitoring scenario (matching a ‘test’ track to one of the footprint sets of known individuals) the accuracies for 2 test models were 97 and 99%. For species discrimination, we used a dataset of 1636 footprints, with 218 tracks of which 59 were from black and 159 from white rhino. FIT gave a species discrimination accuracy for tracks of 98 to 99% using 3 different test models. We outline how the underlying FIT has been adapted for white rhino and detail work in progress to extend the method to other species. We anticipate that the technique will offer an objective and accurate tool for monitoring and censusing, with flexibility as regards target species and locale. Data collection for FIT is intuitive for skilled trackers and thus local expertise can be employed. The technique promises to be an effective tool for management and ecological studies, especially for nocturnal or otherwise elusive species, and is expected to be effective as a complementary tool for other monitoring techniques, such as mark-recapture or camera-trapping.
Lubow B. C., Ransom J. I. (2009): Validating aerial photographic mark‐recapture for naturally marked feral horses. The Journal of Wildlife Management 73: 1420-1429.
FULL TEXT
Abstract
Accurately estimating large mammal populations is a difficult challenge because species of interest often occupy vast areas and exhibit low and heterogeneous visibility. Population estimation techniques using aerial surveys and statistical design and analysis methods provide a means for meeting this challenge, yet they have only rarely been validated because wild populations of known size suitable for field tests are rare. Our study presents field validations of a photographic aerial mark‐recapture technique that takes advantage of the recognizable natural markings on free‐roaming feral horses (Equus caballus) to accurately identify individual animals and groups of animals sighted on multiple occasions. The 3 small populations of feral horses (<400 animals each) in the western United States used in the study were all closely monitored on a weekly basis by local researchers, thus providing test populations of known size. We were able to accurately estimate these population sizes with aerial surveys, despite rugged terrain and dense vegetation that created substantial heterogeneity of sighting probability among horse groups. Our best estimates at the 3 sites were within −6.7%, 2.6%, and −8.6% of known truth (‐4.2% mean error, 6.0% mean absolute error). In contrast, we found undercount bias as large as 32% before any statistical corrections. The necessary corrections varied both temporally and spatially, in response to previous sighting history (behavioral response), and by the number of horses in a group. Despite modeling some of the differences in horse‐group visibility with sighting covariates, we found substantial residual unmodeled heterogeneity that contributed to underestimation of the true population by as much as 22.7% when we used models that did not fully account for these unmeasured sources. We also found that the cost of the accurate and validated methods presented here is comparable to that of raw count (so called, census) methods commonly employed across feral horse ranges in 10 western states. We believe this technique can assist managers in accurately estimating many feral horse populations and could be applied to other species with sufficiently diverse and distinguishable visible markings.
Patton F., Jones M. (2010): Determining the suitability of using eye wrinkle patterns for the accurate identification of individual black rhinos. Pachyderm 48: 18-23.
FULL TEXT
Abstract
Photograhic identification is used to identify individual black rhinos (Diceros bicornis) with the distinguishing features of sex, the size and shape of the anterior and posterior horns, peculiarities of the ears, the pattern of wrinkle contours on the snout, prominent scars and sores on the body, the state of the tail, body size including the size of a calf in relation to the mother and skin folds. Eye wrinkle patterns have received little attention but were found to be useful when separating a large number of photographs of 19 captive rhinos particularly for distinguishing individual sub-adults where other features such as horn length and shape were very similar. Each rhino was found to have unique eye wrinkle patterns which remained consistent when the eyes were open or closed. By developing and applying a series of tests, judgement errors that occur when reviewing eye wrinkle photographs were determined and are reported. Results show that individual black rhinos can be accurately identified from suitable photographs but, even for the best of the judges, using eye wrinkles alone to identify individual rhinos was not completely reliable.
Stein A., Erckie B., Fuller T., Marker L. (2010): Camera trapping as a method for monitoring rhino populations within the Waterberg Plateau Park, Namibia. Pachyderm 48: 67-70.
FULL TEXT
Abstract
For species with unique markings, camera trapping has been used as a non-invasive method for generating population estimates and monitoring the fate of particular individuals. Rhinos – both black (Diceros bicornis) and white (Ceratotherium simum) – have unique horn sizes, shapes and scarring, making camera trapping a monitoring technique that could be useful. Over a 7-week period during 2006 in the Waterberg Plateau Park (WPP) in Namibia, we obtained 125 photos of rhinos from 11 camera stations during 545 camera nights, about half of which were useful in identifying 18 individual black rhinos and 13 white rhinos. Additional coverage of the Park could lead to more complete counts that would complement ongoing monitoring efforts.
Law P. R., Jewell Z., Alibhai S. (2013): Using shape and size to quantify variation in footprints for individual identification: Case study with white rhinoceros (Ceratotherium simum). Wildlife Society Bulletin 37: 433-438.
FULL TEXT
Abstract
For those vertebrate species that create sufficiently complex footprints, identifying individuals from their footprints promises to be a noninvasive technique of great potential for wildlife studies and conservation, but with statistical challenges. Various approaches to employing footprints for identification appear in the literature, but doubt often remains as to the information contained in the footprints and therefore of the reliability of the procedures. For footprints represented by landmarks, we propose using pre‐assigned measures of shape and size of configurations of landmarks to quantify the variation in footprints amongst individuals relative to the variation in each individual’s footprints. Our method provides a relatively simple means of assessing when footprints (represented by landmarks) from individuals of a population will be useful for identifying individuals, independent of any particular identification algorithm, and is also a tool for exploring footprint landmark data to aid development of discrimination routines. We illustrate the method using footprints collected from a population of white rhinoceros (Ceratotherium simum) at Otjiwa Game Ranch, Namibia, during late 1999.
Merkle J. A., Fortin D. (2014): Likelihood‐based photograph identification: Application with photographs of free‐ranging bison. Wildlife Society Bulletin 38: 196-204.
FULL TEXT
Abstract
Using photographs to identify individual animals and monitor populations is becoming more common. However, photographic identification methods where measurements of morphological traits (e.g., horn length) are compared have received little attention. We present an approach for aiding with the identification of individual animals from photographs. The approach incorporates measurement data, metadata from photographs, and multiple sources of error, and calculates a matching score between pairs of photographs using a likelihood‐based algorithm. We tested and identified the false‐rejection error rate using 91 photographs, representing 33 known free‐ranging bison (Bison bison), and 117 simulated data sets with varying numbers of unique individuals, morphological measurements, and photograph error. We then used the approach to estimate the adult population size of bison in Prince Albert National Park, Canada, in 2011. For bison, the false‐rejection rate of our approach was 0.055. Using a Huggins closed population model with misidentification, we estimated 103 (95% CI = 82–130) and 46 (95% CI = 37–58) adult female and male bison, respectively. After incorporating field‐based calf‐ and juvenile‐to‐female ratios, we estimated 202 (95% CI = 171.6–231.4) bison. We found this estimate to be plausible using 2 minimum‐count aerial surveys conducted in March 2011 and 2012 for comparison. With our approach, researchers and managers can build capture histories of individuals, which can be used for studies of population dynamics and habitat selection. This approach can incorporate any morphological measurements extracted from photographs (e.g., coat color), making it robust to a variety of species and study systems.
Gibbon G. E., Bindemann M., Roberts D. L. (2015): Factors affecting the identification of individual mountain bongo antelope. PeerJ 3: e1303.
FULL TEXT
Abstract
The recognition of individuals forms the basis of many endangered species monitoring protocols. This process typically relies on manual recognition techniques. This study aimed to calculate a measure of the error rates inherent within the manual technique and also sought to identify visual traits that aid identification, using the critically endangered mountain bongo, Tragelaphus eurycerus isaaci, as a model system. Identification accuracy was assessed with a matching task that required same/different decisions to side-by-side pairings of individual bongos. Error rates were lowest when only the flanks of bongos were shown, suggesting that the inclusion of other visual traits confounded accuracy. Accuracy was also higher for photographs of captive animals than camera-trap images, and in observers experienced in working with mountain bongos, than those unfamiliar with the sub-species. These results suggest that the removal of non-essential morphological traits from photographs of bongos, the use of high-quality images, and relevant expertise all help increase identification accuracy. Finally, given the rise in automated identification and the use of citizen science, something our results would suggest is applicable within the context of the mountain bongo, this study provides a framework for assessing their accuracy in individual as well as species identification.
Zaumyslova O. Y., Bondarchuk S. N. (2015): The use of camera traps for monitoring the population of long-tailed gorals. Achievements in the Life Sciences 9: 15-21.
FULL TEXT
Abstract
The long-tailed goral (Nemorhaedus caudatus) is a rare montane ungulate species with a patchy distribution. In the Sikhote-Alin Reserve, gorals occupy the northern part of their range, concentrated primarily in a small coastal area (6.4 km2) in Abrek Urochishe. Our pilot study tested the feasibility of individual photo-identification of gorals and population size estimation using the capture–recapture method. We used 10 camera traps spaced 0.6–2 km apart on coastal slopes to monitor the gorals. Four additional cameras were placed at the Reserve boundaries, mainly for law enforcement purposes, such as documenting trespassers. Between June 1 and December 31, 2013, we collected nearly 3000 photographs of gorals, 500 photographs of other wildlife, and 12 images of illegal activities within the Reserve. The total sampling effort was 1870 camera days. Photo data showed that goral horns are reliable biometric identifiers, distinguishable by size, shape, pattern, and the number of rings. The proportion of individually identified gorals in our photos was 0.64 (SE = 0.05). Most individuals (45) were marked (i.e., first detected on camera) in the fall; therefore, preliminary estimates of the goral population size were made between October 11 and December 20, 2013. A closure test confirmed that the population was, in fact, closed (z = − 2.670, P = 0.004). The best-fit closed population multiple recapture model for our data was the heterogeneity model Mh (programme CAPTURE), which assumes an unequal capture probability (χ2 = 112.19; d.f. = 9; P = 0.000). The average goral capture probability was 0.16, and the corresponding population size was estimated at 90 individuals (SE = 6.91; 95% CI: 77–125 individuals). The average goral population density in a 3.5 km2 effective sampled area (56% of the entire plot area) was 25 individuals/km2 (SE = 5.62). Extrapolation to locations that lacked data suggests that Abrek Urochishe supports a goral population of 160 individuals. Our results demonstrate that camera trap data can be used for photographic capture–recapture sampling of goral populations. This approach may be more effective than traditional visual surveys of montane ungulates that tend to underestimate the population abundance. The use of camera traps will undoubtedly enhance goral monitoring efforts, aiding in the conservation of this rare species.
Patton F. (2017): The use of ear tufts to assist in the identification of individual black rhinos. Pachyderm 58: 148-151.
FULL TEXT
Excerpt
In Kenya, one objective set for black rhino monitoring is to “maintain at least 60% of independent animals in each population, identifiable by all trained observers” (KWS 2012). For an individual rhino to be considered ‘Identifiable’, its distinguishing features must be obvious and of such a nature that all observers can record them. It should also be recognised as the same animal when seen at different times. Therefore, the more permanent an identification feature, the more useful it is. Figures 1-4 show photographs of four rhinos from a population taken at different times over several years where ear tufts have been used to assist in individual identification. It can be seen that tufts can be a reliable identification feature.
Gardner P. C., Vaughan I. P., Liew L. P., Goossens B. (2019): Using natural marks in a spatially explicit capture-recapture framework to estimate preliminary population density of cryptic endangered wild cattle in Borneo. Global Ecology and Conservation 20: e00748.
FULL TEXT
Abstract
The behaviour of cryptic tropical forest ungulates that are not identifiable from unique coat colour and patterns often impedes detectability and investigations of population density, which underpin conservation plans. The shy and endangered Bornean banteng has a declining trend, but quantifying this requires sufficient detections to estimate robust population parameters, which are currently unavailable. Using intensive camera trapping and individual identification from natural marks by two observers, we estimated the baseline population density of Bornean bantengs in Malua and Tabin forests in Sabah (Malaysian Borneo) using a spatially explicit capture-recapture framework. We also investigated the efficacy of two commonly-used survey methods (camera trapping and signs) that have previously failed to detect the species, by contrasting capture frequencies to estimate the probability of odds of capture. Density estimates and simulated 95% confidence limits were exceptionally low in both forests and with negligible differences arising from small disparities in the interpretation of natural marks. Density in Malua ranged from 0.5 individuals per 100 km2 (0.21–1.48) to 0.56 (0.15–2.09), and in Tabin between 0.61 (0.32–1.16) to 0.95 (0.54–1.66). The capture odds were significantly greater for camera traps (X2 = 20, p < 0.001, and OR = > 4, p < 0.001); sign survey efficacy declined at higher elevations and under dense canopy. Using natural marks for individual identification was resource-demanding, but provided robust population density parameters for an otherwise challenging species to detect. Extremely low-density estimates of Bornean bantengs highlights the urgency for greater control of poaching, which is almost certainly decimating the population. Rapid implementation of actions to mitigate against further losses are essential for halting the declining trend. The estimation of density parameters in other forests in Sabah that contain bantengs would set the context for our density estimates. It would additionally provide a basis for long-term population monitoring, and facilitate investigations into the effectiveness of enforcement strategies.
Macaulay L. T., Sollmann R., Barrett R. H. (2020): Estimating deer populations using camera traps and natural marks. The Journal of Wildlife Management 84: 301-310.
FULL TEXT
Abstract
Despite the ubiquity of camera traps in wildlife monitoring projects, the data gathered are rarely used to estimate wildlife population demographics, a critical step in detecting declines, managing populations, and understanding ecosystem health. In contrast to abundant white‐tailed deer (Odocoileus virginianus) in the eastern United States, black‐tailed deer (Odocoileus hemionus columbianus) in the western United States have declined over the past several decades. We tested whether passively operating camera traps can be used to quantify population characteristics for black‐tailed deer. We used images of naturally occurring physical characteristics of deer to develop movement and activity data and inform a Bayesian spatial mark‐resight model that estimates deer abundance, density, sex ratio, ratio of fawns to adult females, and home range size. We developed the model to account for the effect of attractants (bait) on encounter rate. We placed 13 cameras on all known water sources of a private ranch in California and provided bait once a month in front of each camera. Over 9,000 visits occurred between 24 May 2012 and 21 January 2013, and we identified 50 individual deer from ear notches or antler characteristics. We estimated density at 7.7 deer/km2 in summer and 8.6 deer/km2 in fall. In the summer, home ranges were 2.3 km2 for females and fawns and 16.8 km2 for males. Home ranges constricted slightly in fall. We estimated a sex ratio of 12.5 males/100 females, and a ratio of 47.0 fawns/100 adult females. Bait increased baseline encounter rates (visits/week) by 3.7 times in summer and 4.95 times in fall. We found slightly higher densities of deer in our study area compared to other recent studies in more mountainous areas of California, and lower male:female sex ratios. This approach shows that commonly deployed camera traps can be used to quantify population characteristics, monitor populations, and inform harvest or habitat management decisions.
Koopmans M., Stokes E. J., Opepa C. K., Mouele A. M., Abea G., Strindberg S., Brncic T. M. (2021): Wild bongo density estimation and population viability analysis improves conservation management. Global Ecology and Conservation 28: e01661.
FULL TEXT
Abstract
Bongo (Tragelaphus eurycerus eurycerus) are rare, mostly nocturnal forest antelopes in tropical Africa that are prized by trophy hunters for their large size, curved horns, and colorful pelage. Estimating bongo population size is difficult because of their low densities and forested habitat, and therefore management decisions have not previously been based on realistic models of population dynamics. In this study, we estimate bongo density and population size in a safari hunting concession in northern Republic of Congo using spatially explicit capture-recapture (SECR) models based on camera trap data. The SECR bongo population estimates were used to model population dynamics under different scenarios of regulated and unregulated harvest and catastrophic natural events. Bongo density in the safari hunting concession was 8.77 bongo 100 km−2 (4.78–15.58). Vortex modelling showed that trophy hunting quotas granted previously would, if achieved, have led to rapid local extinction of bongo. As a result of this study, quotas were reduced to three bongo per year, which is only sustainable if current optimal conditions are maintained.
Prinsloo, N. D., Postma, M., & de Bruyn, P. N. (2021). How unique is unique? Quantifying geometric differences in stripe patterns of Cape mountain zebra, Equus zebra zebra (Perissodactyla: Equidae). Zoological Journal of the Linnean Society 191: 612-625.
FULL TEXT
Abstract
Quantified coat pattern dissimilarity provides a visible surface for individual animal traceability to populations. We determined the feasibility in quantifying uniqueness of stripe patterns of Cape mountain zebra (CMZ; Equus zebra zebra) using geometric morphometrics. We photogrammetrically created dense surface models of CMZ (N = 56). Stripe edges were landmarked, superimposed and compared for shape variation across replicates and the population. Significant allometry in stripe patterns prompted allometric correction to remove increased curvature of stripes at the rump, belly and back with larger adult individuals, to facilitate equilibrated comparison between individuals. Re-landmarked replicates showed lower dissimilarity (Di) than non-replicates (Dp), representing minimal landmarking error. Individuals were 78.07 ± 1.79% unique (U=1−Di/Dp×100%) relative to the study population. Size, the number of torso stripes and degree of branching in four rear torso stripes described the most shape variation (36.79%) but a significant portion could only be distinguished with geometric morphometrics (41.82%). This is the first known use of geometric morphometrics to quantify coat pattern uniqueness, using a model species to provide baseline individual morphological variation. Measures of coat pattern similarity have a place in phenotypic monitoring and identification.
Prinsloo, N. D., Postma, M., & de Bruyn, P. N. (2021). How unique is unique? Quantifying geometric differences in stripe patterns of Cape mountain zebra, Equus zebra zebra (Perissodactyla: Equidae). Zoological Journal of the Linnean Society 191: 612-625.
FULL TEXT
Abstract
Quantified coat pattern dissimilarity provides a visible surface for individual animal traceability to populations. We determined the feasibility in quantifying uniqueness of stripe patterns of Cape mountain zebra (CMZ; Equus zebra zebra) using geometric morphometrics. We photogrammetrically created dense surface models of CMZ (N = 56). Stripe edges were landmarked, superimposed and compared for shape variation across replicates and the population. Significant allometry in stripe patterns prompted allometric correction to remove increased curvature of stripes at the rump, belly and back with larger adult individuals, to facilitate equilibrated comparison between individuals. Re-landmarked replicates showed lower dissimilarity (Di) than non-replicates (Dp), representing minimal landmarking error. Individuals were 78.07 ± 1.79% unique (U=1−Di/Dp×100%) relative to the study population. Size, the number of torso stripes and degree of branching in four rear torso stripes described the most shape variation (36.79%) but a significant portion could only be distinguished with geometric morphometrics (41.82%). This is the first known use of geometric morphometrics to quantify coat pattern uniqueness, using a model species to provide baseline individual morphological variation. Measures of coat pattern similarity have a place in phenotypic monitoring and identification.
Lee D. E., Lohay G. G., Cavener D. R., Bond M. L. (2022): Using spot pattern recognition to examine population biology, evolutionary ecology, sociality, and movements of giraffes: a 70-year retrospective. Mammalian Biology 102: 1055-1071.
FULL TEXT
Abstract
Individual-based studies where animals are monitored through space and time enable explorations of ecology, demography, evolutionary biology, movements, and behavior. Here, we review 70 years of research on an endangered African herbivore, the giraffe, based on individual spot pattern recognition, and profile an example of a long-term photographic mark–recapture study of Masai giraffes in Tanzania. We illustrate how individual-based data can be used to examine the fitness consequences (variation in survival and reproduction) of extrinsic environmental factors or intrinsic traits in an evolutionary ecology framework. These data also allow the study of social structure, space use, life histories, and health. The giraffe offers an excellent opportunity to study dynamics of an ungulate species with a highly fission–fusion social system using spot pattern recognition.
Ness I. F., Jung T. S., Schmiegelow F. K. (2022): Evaluating likelihood-based photogrammetry for individual recognition of four species of northern ungulates. Mammalian Biology 102: 701-718.
FULL TEXT
Abstract
Estimating abundance is a key component of wildlife management and capture–mark–recapture (CMR) methods are commonly used. Photography has been widely explored as a method of ‘marking’ individuals for CMR. However, it is most often used on species with unique markings, and application to species without obvious marks remains challenging. We tested a likelihood-based photogrammetric method to identify individuals from four species of ungulates: muskox (Ovibos moschatus; 31 photos of 16 individuals), thinhorn sheep (Ovis dalli; 110 photos of 32 individuals), mountain goats (Oreamnos americanus; 47 photos of 19 individuals) and mule deer (Odocoileus hemionus; 60 photos of 30 individuals). Misidentification rates (false rejection [FRR] and false acceptance [FAR] rates) varied widely between species (FRR = < 1–13%, FAR = 2–22%), as did matching success rates (48–96%). Muskox and sheep had the highest matching success rates (96% and 88%, respectively), while those for goats (80%) and deer (58%) were lower. Automated matching success rate for deer and sheep, calculated based on the top-ranked photograph, was compared to a user-generated matching success rate. The latter was significantly higher in all cases, indicating that the final subjective step was important. Significant observer bias was found for deer, but not sheep. The main findings of our study were that a photogrammetric approach to individual recognition of ungulates was species specific and sensitive to the inclusion of horn measurements, season, photo quality, and final selection from top choices by the user. While matching success rates of photogrammetry was reasonably good (> 90%) for muskox, it was only moderately accurate 80–90% for sheep and goats, and poor (< 80%) for deer. Future work should address the temporal stability of horn measurements, seasonal effects, feasibility of capturing useable photographs from remote cameras, and application to other species.
MULTIPLE SPECIES
Gracanin A., Mikac K. M. (2022): The use of selfie camera traps to estimate home range and movement patterns of small mammals in a fragmented landscape. Animals 12: 912.
FULL TEXT
Abstract
The use of camera traps to track individual mammals to estimate home range and movement patterns, has not been previously applied to small mammal species. Our aim was to evaluate the use of camera trapping, using the selfie trap method, to record movements of small mammals within and between fragments of habitat. In a fragmented landscape, 164 cameras were set up across four survey areas, with cameras left to record continuously for 28 nights. Live trapping was performed prior to ear mark animals to facilitate individual identification on camera. Four small mammal species (sugar glider; Petaurus breviceps; brown antechinus; Antechinus stuartii, bush rat; Rattus fuscipes, and brown rat; Rattus norvigecus) were recorded on camera (N = 284 individuals). The maximum distance travelled by an individual sugar glider was 14.66 km, antechinus 4.24 km; bush rat 1.90 km and brown rat 1.28 km. Movements of both female and male sugar gliders in linear fragments were recorded at much higher rates than in larger patches of forest sampled in grids. Short term core homes ranges (50% KDE) of 34 sugar gliders ranged from 0.3 ha to 4.2 ha. Sugar glider core home ranges were on average 1.2 ha (±0.17) for females and 2.4 ha (±0.28) for males. The selfie trap is an efficient camera trapping method for estimating home ranges and movements due to its ability to obtain high recapture rates for multiple species and individuals. In our study landscape, linear strips of habitat were readily utilised by all small mammals, highlighting their importance as wildlife corridors in a fragmented landscape.