With other ingenuity and interests it can be seen how this kind of psychophysical approach can be used with sounds rather than light. In these kinds of studies birds are not trained to respond when they see something but when they hear it. Systematic alterations of sound levels and frequency can provide insights into a bird’s absolute sensitivity to sound, its relative sensitivity to sound of different frequencies (audiograms), and how finely the birds can discriminate between sounds of different frequencies. It would also be possible to introduce sounds with different time patterns, trills etc., and investigations with these can give insight into the ability of birds to detect differences in bird songs.
Some investigators have even been able to modify the overall procedure to work with smell and taste to determine which chemical stimuli are more salient and can be detected at low concentrations. However, controlling the smell or taste compounds and presenting them in a uniform way can be very difficult, and perhaps only a few trials can be done each session.
Who, how, and what to measure?
There are many ways in which such training techniques can be used to investigate the senses of birds in a robust way. These allow comparisons between species. However, not all species will be readily amenable to such training, or investigators may not be able to invest the time necessary for the use of these training procedures in full. Some investigations have been able to use the basic principles of these training procedures to gain insights into birds’ senses albeit in a rather limited way compared with the full descriptions that might be aspired to. For example, the first investigations of whether birds could detect ultraviolet light was done using a modified procedure with hummingbirds, and some early work on the sense of smell in doves was carried out using modifications of this kind of approach.
Much depends on the time available to do the work and the motivation of the investigators to work with their species. Unfortunately, modern ways of funding science and pressure to produce papers often preclude long-term training projects of the kinds described here, but work is still being done. For example, recently there have been some valuable investigations of vision in diurnal birds of prey and in parrots, using just these kinds of procedures. Valuable insights into what diving birds might see underwater have come from training investigations of these kinds. All of these studies took many months and required the accumulation of many thousands of trials with the birds.
There remain plenty of opportunities to pose questions about the sensory capacities of most bird species. As the comparative database of species’ sensory thresholds grow, the results become of ever greater value.
Vision first emerged on Earth about 540 million years ago. The first eyes were simple structures able only to register changes in the amount of light falling upon them. However, these simple structures, and the limited information they provided, had a profound influence on the evolution of animals. From a modern perspective these first eyes were game changers, the equivalent of today’s disruptive technologies. It is no exaggeration to suggest that the emergence of eyes changed forever the type, quality and quantity of information that underlies the interactions of animals with their environments, which, of course, includes their interactions with other animals. Ever since the simplest eyes evolved, the gaining of information about the world, its interpretation, and the uses to which it is put, has become increasingly complex and subtle.
The evolution of vision
Prior to the evolution of eyes, the ways that animals were able to gain information about their worlds was both slow and intimate. It depended primarily upon the transmission of chemicals through air or water. This information either took some time to arrive or it concerned objects that were very close to or, more likely, in direct contact with an animal’s body. The key advantage that even simple eyes brought was that information about events remote from the animals were received immediately. This advantage was so dramatic that the first simple light-sensitive structures, which could detect just the presence and intensity of light, rapidly evolved into something far more sophisticated.
This was the birth of ‘spatial vision’. This is the ability to determine not just that light is present and changing in intensity, but also the direction from which light is coming. The refinements of eyes and vision over the past 500 million years have been concerned primarily with elaborating the accuracy of spatial vision. This elaboration has been concerned with extending the degree of detail that can be extracted about light sources at various distances, extending the range of light levels over which this can be achieved, and increasing the volume of space about an animal from which information can be obtained at any instant.
The high utility of such information is indicated by the rapidity with which eyes evolved. It took probably less than 2 million years for eyes that simply registered the presence of light to evolve into ‘camera eyes’. These are eyes that show all of the main features of the eyes that we recognise in species of the present day, including in ourselves. It has been argued that the evolution of sophisticated eyes, showing all of the key features of modern eyes, could have involved only about 400,000 generations of change (Figure 3.1).
FIGURE 3.1 The evolution of well-focused camera eyes, starting from a light-sensitive patch on the surface of an animal. The diagram shows a theoretical model based on conservative assumptions about selection pressure and the amount of variation in natural populations. This model, proposed by Nilsson and Pelger from the University of Lund, suggests that an eye could have evolved very fast, in fewer than 400,000 generations. The starting point is a flat piece of light-sensitive skin (shown in blue) with a transparent protective layer over it, and below the receptor cells a layer of pigmented cells (shown in black). These absorb light not caught by the receptors and help provide integrity of the whole structure. The emerging chamber is filled with a clear fluid and eventually by a lens, while the original protective layer becomes curved and eventually takes on an optical role as the cornea. This evolutionary pathway is not just theoretical, it is informed by the fossil record. (Redrawn from the original scheme proposed by Nilsson and Pelger in 1994.)
The newly evolved ability of spatial vision provided information not only about objects that were close by, but also about those that were far away from the animal. By so doing it established vision as the primary source of information used to guide behaviour. This primary reliance upon vision is found today in nearly all animal taxa including, of course, birds.
Camera eyes (their basic design and functional divisions are described below) had been around for over 350 million years before the first birds appeared on the planet, about 150 million years ago. This means that the first birds, and their dinosaur ancestors, were highly likely to have had elaborate visual capacities. These capacities had been honed through the process of natural selection in response to the challenges of extracting information from the many different environments that had occurred on the earth over a number of geological eras. As sophisticated as these first bird eyes might have been, their evolution has continued. Changes in vision have occurred in response to the new environmental challenges that emerged as bird lineages diversified to exploit the wide range of habitats in which birds now exist. Across today’s 11,000 bird species, eyes exhibit both major and subtle differences in design and function.
It