Answers to such questions help put broad limits on what an animal can detect and therefore what can influence its behaviour. Unfortunately, even these questions have been answered in relatively few species, and rarely are answers available to all such questions in a single species. However, we do know enough for some general principles to have been established and some rules of thumb to have been generated.
Describing sensory performance
How can sensory performance be quantified? In essence we want to ask some specific questions. We want to ask a Barn Owl and a Starling what it can see and hear, a Sanderling what it can feel at its bill tip and taste on its tongue, a Storm Petrel what it can smell (Figure 2.4).
FIGURE 2.4 Four bird species which differ markedly in their senses and how they use them to gain information that is used to guide their behaviour, especially their foraging. Clockwise from top left: Common Starling Sturnus vulgaris, Barn owl Tyto alba, Leach’s Storm Petrel Oceanodroma leucorhoa, Sanderling Calidris alba. These birds differ in their vision, hearing, sense of smell, and sense of touch. Furthermore, each species relies upon a different primary sense to guide its behaviour. They also differ in how they combine and complement information gained through different senses. In short, each bird lives in a different sensory world. (Photo of Starling by Pam P. Parsons [West Bay Dorset, via Flickr as Pam P Photos], Barn Owl by Graham White [CC BY-NC-SA 2.0], Storm Petrel by C. Schlawe [public domain], Sanderling by J. J. Harrison [https://www.jjharrison.com.au, CC BY-SA 3.0].)
Sensory science is mainly concerned with these kinds of questions, with revealing sensory capacities and discovering the mechanisms that underpin them. Sensory ecology, however, takes this information further and is more interested in revealing what sensory capacities are in play in a particular situation, or in finding out how an animal uses the information that it has available to guide its key behaviours.
The aim of sensory scientists has been to manipulate just one or two parameters of a stimulus at a time, and if possible to use those manipulations to determine the limits of sensory performance. If this can be done in a comparable way across a range of species then we should be able to say with some confidence that this species is more sensitive than that one, or this species is able to gain information over a wider range of parameters. These are often the kinds of things that birdwatchers, journalists, and TV documentary makers are keen to know about.
I am often asked whether this or that species is ‘better’ or ‘worse’ than humans. But the biologically relevant question is whether the species is better or worse than the species that it is competing with. What is important to a sensory ecologist is using this information to understand what information an animal has available to guide its behaviour in real-world tasks, and understanding how information from different senses might be integrated.
It does not matter whether an animal’s senses are ‘better’ or ‘worse’ than humans. Humans are, after all, just one species, adapted to living in certain types of environments through the conduct of particular behaviours, so comparison with humans may not be important. The desire to compare another species’ sensory performance with humans is born of an anthropocentric view of the world. It is a viewpoint which provides a strong pull. I try to resist it, but we shall not be able to escape from it completely in this book.
Measuring senses in a similar way in different species does have great value for comparative studies and also for helping to understand the basic mechanisms that underlie a sense. For example, if differences are found between species in their ability to see detail, and systematic differences are also found in the structure of the eyes’ optics or retinas, then it is possible to start piecing together an understanding of basic mechanisms. This in turn allows the possibility of being able to predict what another species might be able to see just from knowledge of its eye structure.
Sensory thresholds
Limits of sensory performance are defined as absolute thresholds. These are measures of the minimum amount of a particular stimulus that can be detected. In the case of sound, this will be the smallest amount of air disturbance that the hearing system can detect. In vision, it is the lowest number of photons received per unit time that can be detected. In the chemical senses (taste and olfaction), it is the lowest concentration of molecules of a particular substance that can be detected. In mechanical senses (for example, touch sensitivity of a bird’s bill) it is the smallest displacement of the detecting surface.
Absolute thresholds are particularly interesting because they allow simple and direct questions to be posed. Which species has the most sensitive hearing? Which species has the most sensitive eyes? However, comparing thresholds is not as straightforward as one might hope, and answers are rarely clear-cut. This is because thresholds differ not just between species but between individuals.
Like all measures of performance, absolute sensory thresholds tend to be normally distributed when the results for individuals are collected together. This means that the absolute threshold for a species is best described statistically as a mean value with variation around it based on a sample of individuals. Between-individual differences are often particularly notable if age comes into play. A change in sensitivity is usually part of the ageing process in birds, as much as in humans. Therefore, to compare species it is necessary to refer to mean differences, and to be aware that individuals will have sensitivities below or above this mean.
A simple analogy is to consider what the answer might be if you asked how fast humans can run. Every able-bodied person can run, but there is no surprise that the running speeds of people will differ, often markedly so. The answer is that there will be a mean speed for human running performance but around the mean there will be a wide range of speeds. It would then make sense to compare the average running speeds of humans with another animal species, and while there might be a difference in the averages there may well be some overlap in the performances of individuals of the two species.
It is the same with absolute sensory thresholds: differences between individuals do occur. Also, an individual’s thresholds can change owing to a range of factors, including their motivation to participate in the investigation. So, although absolute thresholds are both important and interesting, and will feature a number of times in this book, it will always be necessary to express caution, or at least bear in mind that sensory thresholds are mean values based on a sample of individuals representing a particular species.
Relative sensitivity within a sense
Clearly, the sound and light that an animal is able to detect can vary in their total energy. We can see very bright as well as dim lights, hear very loud sounds as well as very quiet sounds. The ability to detect these stimuli indicates that there is a wide dynamic range to sensory performance. Absolute thresholds measure only the ability to detect the lowest amount of energy for a particular type of stimulus or sensory dimension.
As well as the wide dynamic range, most types of natural stimuli can also vary along a number of dimensions. For example, the vibrations of air molecules that we detect as sounds can have different frequencies, and it is these different frequencies that humans describe as sounds of different pitch. Light also varies in frequency, though we more often describe it in terms of its wavelength. Our visual system detects the different wavelengths as lights of different colours.
In each animal species there are likely to be different absolute thresholds for each frequency of sound or each wavelength of light. This means that a complete description of a bird’s vision or hearing requires knowledge of thresholds across a wide range of stimulus frequencies. For light these are presented as spectral sensitivity functions,