Evidence of learning in animals has been obtained with respect to foraging, territoriality, predator avoidance, dispersal, migration, thermoregulation, communication, mating behavior, parental care, kin recognition, and other social interactions. In these contexts, animals learn relevant features of their environment. Animals can learn the caloric content, nutrient profile, and handling time associated with food types, and the genetic and direct benefits offered by different mates. They learn the quality distribution of essential resources such as food, breeding sites, and mates. In other words, information about virtually every aspect of an animal's ecology may be acquired through learning. Of course, not every animal of every species or population is expected to learn all possible behavior in all ecological contexts. What, if anything, is learned should depend on the net value of learning to the individual. Two observations support this statement. First, animals frequently show biases in what stimuli they are prepared to learn. Second, animals that differ in their ecological requirement for a particular kind of learning differ in how much they invest in such learning.
Learning biases. Animals do not learn all environmental stimuli in all contexts equally well. A classic example of learning bias has been described for food aversion learning. Food aversion learning, best studied in birds, rodents, and insects, involves avoidance of food stimuli associated with a digestive malaise. The sensory properties of food aversion learning in rats are intriguing. Rats readily avoid a flavor associated with an illness but not sound or light. This is not because sounds and light are generally more difficult to learn than flavors: rats more easily learn to associate sound and light with an electric shock (a rapidly acting punishment) than they do a flavor. This pattern most likely reflects an evolutionary history in which foods and their flavors are sometimes associated with delayed illness, favoring a learned aversion toward them, but such illnesses are very rarely associated with sound or light. Similar learning biases are known for song learning in birds. In playback experiments, for example, sparrows prefer to copy their own species-typical song over that of other sparrow species.
Patterns of investment in learning. What information animals learn should depend on the relative costs and benefits of learning that should in turn depend on a species' or population's particular ecological demands. Spatial learning is critical to navigation (including homing) in many animals, as well as to formation of territories and home ranges. Such learning necessarily entails acquiring, processing, and storing large amounts of information. As such, spatial memory can be expected to be very costly and, in fact, large areas in metabolically expensive brains (e.g., the hippocampus in birds and mammals) are devoted to it. Not surprisingly, in view of its high costs, some of the strongest evidence of species differences in learning occurs in relation to spatial learning. In birds and mammals, for example, spatial learning and associated areas of the brain, like the hippocampus, are better developed in species, populations, or sexes in which spatial learning is relatively important. Home range size, for example, is larger in meadow voles than pine voles and thus the former species has correspondingly better spatial memory and larger hippocampi. In meadow voles, the male's home range is much larger than the female's, a difference associated with superior spatial learning and larger hippocampi in males. Food-storing behavior is correlated with similar species differences in corvids. Nutcrackers, which depend strongly on food caches to survive the winter, have better spatial memory ability than scrub jays or pinyon jays, each of which depend less on caches. Significantly, nutcrackers are no better at learning nonspatial tasks than these other jay species.
Was this article helpful?