Predator learning of camouflage types

Lead Research Organisation: University of Exeter
Department Name: Biosciences

Abstract

Most animals face a persistent and significant risk of predation. Correspondingly, prey species have evolved a wide variety of defensive strategies, often including protective coloration. Perhaps the most widespread anti-predator defence is animal camouflage. Camouflage works by preventing detection or recognition of an object, and is also valuable to human applications and recreation. It has long been a key textbook example of natural selection, but in the last decade there has been a substantial increase in research investigating what types of camouflage exist, how they work, and the survival benefit conferred. This work has been undertaken by evolutionary and behavioural biologists, computer scientists, experimental psychologists, art historians, and the military, and represents a highly interdisciplinary subject. Several types of camouflage have been studied, including background matching (resembling the colour and pattern of the background environment), disruptive coloration (high contrast markings at the body edges that break up the shape of the body), and distractive markings (high contrast isolated markings that may draw the predator's attention towards them and away from the body edge).
Despite numerous advances most work has focussed on how different camouflage types prevent initial detection in predators that have not seen the prey before. However, in reality, most predators will encounter similar prey types repeatedly over a period of hours and days, allowing them the opportunity to learn to discriminate prey from the background, and giving them the opportunity to form 'search images' (by selectively paying attention to particular features of the prey's appearance). These search images temporarily increase a predator's ability to find one prey type but decrease its ability to find other prey types. Currently, although learning and search image effects have been investigated in some camouflage studies, almost no work has specifically tested how different types of camouflage and specific features of prey patterns facilitate and affect predator learning and search image formation. Because the success of a camouflage strategy will be affected by how it prevents both initial detection as well as learning and search image formation, there is currently a major gap in our understanding of how camouflage works and the value of different camouflage types.
Here, we use humans searching for hidden targets on touch screens, and domestic chicks searching for hidden artificial prey in arenas, as model predator systems to test how the value of different types and features of camouflage are affected by predators' experience of different prey types. These systems allow us to precisely control prey features and the experience of the subjects acting as predators. Specifically, we will: 1) Determine whether some camouflage types are easier to learn than others by presenting human and chicks with prey types and measuring how effectively detection times decrease over repeated encounters. 2) Determine whether some camouflage types are more resistant to predator search images than others by presenting humans and birds with sequences of the same prey type to test how quickly they form search images, and whether this affects their ability to find prey of other types later. 3) Determine what specific prey features (e.g. pattern contrast, pattern size, and disruptiveness) facilitate search image formation. 4) Test whether findings are consistent between humans and birds by using comparable experiments with birds and humans to determine how the principles about camouflage and visual search and detection can be applied across species.
This project would answer key questions in camouflage research, including how different types of concealment affect learning and search image formation, the nature of search images and which prey features predators attend to, and whether the principles are consistent among taxonomic groups.

Technical Summary

Most animals face a persistent risk of predation. Correspondingly, prey species have evolved a variety of defensive adaptations, with perhaps the most widespread being camouflage. Camouflage works by preventing detection or recognition of an object. It has long been a key system to understand fundamental questions regarding the evolution, tuning, and mechanistic basis of adaptive traits in nature, and it is a textbook example of natural selection. In the last decade there has been a substantial increase in research effort investigating what types of camouflage exist (e.g. background matching, disruptive coloration), how camouflage works, and the survival benefit that it may confer. This work has been undertaken by evolutionary and behavioural biologists, computer scientists, experimental psychologists, art historians, and the military, and represents a highly interdisciplinary subject. Despite numerous advances, most work has been focussed on how different camouflage types prevent initial detection. However, in reality, most predators will encounter the same or similar prey types frequently over a period of hours and days, allowing them the opportunity to learn to discriminate prey from the background and subsequently form search images. Because the success of a camouflage strategy will be affected by how effectively it prevents initial detection, learning and search image formation, there is currently a major gap in our understanding of how camouflage works.
Here, we use humans searching for hidden targets on touch screens, and domestic chicks searing for artificial prey in arenas, as model predator systems. We will determine whether different camouflage types are easier to learn than others or more resistant to predator search images. We will also test what specific prey features (e.g. pattern contrast, spatial frequency, and disruptiveness) facilitate or hinder search image formation, and how findings are consistent between humans and birds.

Planned Impact

1) Public outreach, education, and widening participation.
Predator-prey relationships and animal coloration are subjects with a long history of captivating public interest and attention, and have great potential to inspire future generations about science and learning. The general concepts are readily understandable and easy to demonstrate, as well as being striking and highly visual. Camouflage has long been a key textbook example of natural selection and is frequently covered in the popular media. This interest has significantly increased in recent years, with the subject regularly being covered in television and radio programmes, art and design exhibitions, and several recent popular science books. The PI has a wide range of experience dealing with media interest (both reactive and proactive), setting up websites on our research, and corresponding Twitter and Facebook accounts for disseminating our work in social media (see Pathways to Impact). A key aim of ours is to communicate with and inspire the general public, especially school children and young adults about the subject and science in general. To this end, we will: i) coordinate with press offices about press releases from our papers and communicating our research in general, ii) apply to present a display at the Royal Society Summer Science Exhibition, iii) make an interactive website explaining our work to a broad audience, as well as computer games demonstrating the key concepts of our work, alongside a downloadable phone/tablet app, iv) give talks at various public/outreach events, in particular to local schools, and v) set up a Facebook and Twitter account to disseminate our work and interesting materials, and the subject area in general, to a wide audience.
2) Economic and commercial.
Camouflage has also long had significant importance in human application, such as the military, and recreation and commerce, including art, wildlife watching, and hunting. In addition, it is increasingly used in an urban environment to disguise functional but unattractive objects such as mobile phone transmission masts and satellite dishes. The PI currently co-supervises a CASE studentship part funded by the MoD on motion camouflage. Our work will provide information about how objects can be camouflaged in complex real world environments to different receivers (birds and humans). In addition, our work may also be relevant to advertising and packaging of products. We will investigate how colour patterns can hide objects and prevent observers from learning about them. By turning these principles around, we can understand how to make objects maximally conspicuous and easy to remember. When we have obtained sufficient results to make early findings we will coordinate with relevant departments, the MoD, and industry about the application and exploitation of our work and its potential relevance to products.

Publications

10 25 50
publication icon
Galloway JAM (2020) Finding a signal hidden among noise: how can predators overcome camouflage strategies? in Philosophical transactions of the Royal Society of London. Series B, Biological sciences

publication icon
Skelhorn J (2016) Cognition and the evolution of camouflage. in Proceedings. Biological sciences

publication icon
Skelhorn J (2018) Avoiding death by feigning death. in Current biology : CB

publication icon
Stevens M (2018) The key role of behaviour in animal camouflage. in Biological reviews of the Cambridge Philosophical Society

publication icon
Troscianko J (2017) Quantifying camouflage: how to predict detectability from appearance. in BMC evolutionary biology

publication icon
Troscianko J (2017) Relative advantages of dichromatic and trichromatic color vision in camouflage breaking. in Behavioral ecology : official journal of the International Society for Behavioral Ecology

 
Description Key work shows how camouflage types are learnt differently, with disruptive markings being learnt differently to other types, such as distractive markings, and certain types of camouflage interfere with search image formation. Other key findings involve showing how different visual metrics allow better camouflage breaking and testing, illustrating how camouflage types work and how best to measure types of camouflage in terms of predicting detection. Overall, disruptive camouflage seems resistant to learning effects while distractive camouflage encourages learning. Results are broadly similar between humans and birds. Naturalistic camouflage patterns also interfere with learning, especially when in a disruptive style.
Exploitation Route Further research and funding. Potential applied benefits in camouflage design and breaking. More work on how visual information affects learning.
Sectors Aerospace, Defence and Marine,Digital/Communication/Information Technologies (including Software),Education

 
Description Not specifically the findings related to learning, but the work on camouflage and visual detection has attracted a lot of attention and has been used in museum displays and exhibitions. Methods partly developed in this project to image animal vision have also been taken up for displays by museums such as the Natural History Museum and the National Media Museum.
First Year Of Impact 2016
Sector Education
Impact Types Cultural,Societal

 
Description How to optimise imperfect camouflage
Amount £376,743 (GBP)
Funding ID BB/P018319/1 
Organisation Biotechnology and Biological Sciences Research Council (BBSRC) 
Sector Public
Country United Kingdom
Start 10/2017 
End 09/2020
 
Description Imaging animal vision
Amount £10,842 (GBP)
Funding ID BB/P017339/1 
Organisation Biotechnology and Biological Sciences Research Council (BBSRC) 
Sector Public
Country United Kingdom
Start 12/2016 
End 05/2017
 
Title Multispectral Image Calibration and Analysis Toolbox 
Description Digital cameras can be powerful tools for measuring colours and patterns in a huge range of disciplines. However, in normal 'uncalibrated' digital photographs the pixel values do not scale linearly with the amount of light measured by the sensor. This means that pixel values cannot be reliably compared between different photos or even regions within the same photo unless the images are calibrated to be linear and have any lighting changes controlled for. Some scientists are aware of these issues, but lack convenient, user-friendly software to work with calibrated images, while many others continue to measure uncalibrated images. We have developed a toolbox that can calibrate images using many common consumer digital cameras, and for some cameras the images can be converted to "animal vision", to measure how the scene might look to non-humans. Many animals can see down into the ultraviolet (UV) spectrum, such as most insects, birds, reptiles, amphibians, some fish and some mammals, so it is important to measure UV when working with these animals. Our toolbox can combine photographs taken through multiple colour filters, for example allowing you to combine normal photographs with UV photographs and convert to animal vision across their whole range of sensitivities. The toolbox is free, open source and open access and works with the free programme Image J. It is linked to this paper: Troscianko, J. & Stevens, M. 2015. Image calibration and analysis toolbox - a free software suite for objectively measuring reflectance, colour and pattern. Methods in Ecology and Evolution. 6: 1320-1331. 
Type Of Technology New/Improved Technique/Technology 
Year Produced 2015 
Impact Still early stage. Being used by wide range of scientists already. 
URL http://www.sensoryecology.com/image-analysis-tools/
 
Description Grand National Outreach Event 
Form Of Engagement Activity Participation in an activity, workshop or similar
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Public/other audiences
Results and Impact Grand National at Aintree Racecourse. I am collaborating with Aintree Racecourse to make a visual display (a slideshow) of how the Grand National appears to horse vision. This will form part of a course walk and other events, and is likely to be seen by thousands of visitors each day.
Year(s) Of Engagement Activity 2017
 
Description NHM Exhibition 
Form Of Engagement Activity Participation in an activity, workshop or similar
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Public/other audiences
Results and Impact NHM London 'Colour and Vision' exhibition. We collaborated with the NHM to make a new interactive citizen science game on camouflage (in shore crabs) and predator vision (played by over 6200 people), and to make a series of interactive image displays that show how a range of other animals see the world, from nautilus to dogs. The exhibition was visited by over 40,000 people and both exhibits were commented on by the public as some of the most engaging and enjoyable aspects of the exhibition.
Year(s) Of Engagement Activity 2016
 
Description NHM London 'Colour and Vision' exhibition 
Form Of Engagement Activity Participation in an activity, workshop or similar
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Public/other audiences
Results and Impact I collaborated with the NHM to make a new interactive citizen science game on camouflage (in shore crabs) and predator vision (played by over 6200 people), and to make a series of interactive image displays that show how a range of other animals see the world, from nautilus to dogs. The exhibition was visited by over 40,000 people and both exhibits were commented on by the public as some of the most engaging and enjoyable aspects of the exhibition.
Year(s) Of Engagement Activity 2016
 
Description National Media Museum display 
Form Of Engagement Activity Participation in an activity, workshop or similar
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Public/other audiences
Results and Impact Supersenses exhibition at the National Science and Media Museum - we made images of animal vision to show to the public as part of this exhibition.
Year(s) Of Engagement Activity 2017
 
Description Press coverage of paper 
Form Of Engagement Activity A magazine, newsletter or online publication
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Public/other audiences
Results and Impact Media coverage of paper:
Troscianko, J. & Stevens, M. 2015. Image calibration and analysis toolbox - a free software suite for objectively measuring reflectance, colour and pattern. Methods in Ecology and Evolution. 6: 1320-1331.
Covered by: BBC Cornwall (11 August 2015), Daily Mail (6 August 2015), the Independent, the Washington Post (7 August 2015), IFL Science (8 August 2015).
Year(s) Of Engagement Activity 2015
 
Description Two day conference workshop on anti-predator coloration 
Form Of Engagement Activity Participation in an activity, workshop or similar
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Study participants or study members
Results and Impact 4-5 August 2016: A two-day conference on anti-predator coloration at the Centre for Ecology and Conservation. This was attended by approximately 60 of the leading researches in this subject from all over the world, from China and Australia to North America and Europe.
Year(s) Of Engagement Activity 2016
 
Description Wiko Advanced Studies workshop on animal coloration 
Form Of Engagement Activity Participation in an activity, workshop or similar
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Study participants or study members
Results and Impact Workshop on animal coloration at Wiko in Berlin, attended by researchers from around the world. Outcomes: special issue of Phil Trans B, and maybe an additional high profile multi-authored paper.
Year(s) Of Engagement Activity 2016