The Relevance of Ecological Transitions to Intelligence in Marine Mammals

 The Relevance of Ecological Transitions to Intelligence in Marine Mammals Gordon B. Bauer1,2, Peter F. Cook1,2 and Heidi E. Harley1,2,3 1Division of Social Sciences, New College of Florida, Sarasota, FL, United States 2Mote Marine Laboratory, Sarasota, FL, United States 3The Seas, Epcot, Walt Disney World Resorts, Lake Buena Vista, FL, United States Macphails comparative approach to intelligence focused on associative processes, an orientation inconsistent with more multifaceted lay and scientific understandings of the term. His ultimate emphasis on associative processes indicated few differences in intelligence among vertebrates. We explore options more attuned to common definitions by considering intelligence in terms of richness of representations of the world, the interconnectivity of those representations, the ability to flexibly change those connections, and knowledge. We focus on marine mammals, represented by the amphibious pinnipeds and the aquatic cetaceans and sirenians, as animals that transitioned from a terrestrial existence to an aquatic one, experiencing major changes in ecological pressures. They adapted with morphological transformations related to streamlining the body, physiological changes in respiration and thermoregulation, and sensoryperceptual changes, including echolocation capabilities and diminished olfaction in many cetaceans, both in-air and underwater visual focus, and enhanced senses of touch in pinnipeds and sirenians. Having a terrestrial foundation on which aquatic capacities were overlaid likely affected their cognitive abilities, especially as a new reliance on sound and touch, and the need to surface to breath changed their interactions with the world. Vocal and behavioral observational learning capabilities in the wild and in laboratory experiments suggest versatility in group coordination. Empirical reports on aspects of intelligent behavior like problem-solving, spatial learning, and concept learning by various species of cetaceans and pinnipeds suggest rich cognitive abilities. The high energy demands of the brain suggest that brain-intelligence relationships might be fruitful areas for study when specific hypotheses are considered, e. g. , brain mapping indicates hypertrophy of specific sensory areas in marine mammals. Modern neuroimaging techniques provide ways to study neural connectivity, and the patterns of connections between sensory, motor, and other cortical regions provide a biological framework for exploring how animals represent and flexibly use information in navigating and learning about their environment. At this stage of marine mammal research, it would still be prudent to follow Macphails caution that it is premature to make strong comparative statements without more empirical evidence, but an approach that includes learning more about how animals flexibly link information across multiple representations could be a productive way of comparing species by allowing them to use their specific strengths within comparative tasks. The Relevance of Ecological Transitions to Intelligence in Marine Mammals Since the birth of psychology, scientists have debated the power of associationism as the central mechanism behind the science of mental life James, 18901952, p. 1. James began his seminal psychological work by contrasting three different possibilities for describing the human mind soul, associative processes dictated by experience, and the innate and developed structure of the human mind as a framework that constrains how we process information. Macphail 1982, 1987, after a comprehensive review, concluded that an objective assessment of the vertebrate literature indicated that learning only occurred through a limited set of processes, primarily associative i. e. , habituation, classical, and instrumental conditioning. He further indicated that, restricting comparisons to associative processes, there were no differences in intelligence among vertebrates with the exception of humans, and he speculated that difference might be attributable to language, essentially discounting specific cognitive adaptations to distinct niches. Of course, the circumstances in which animals perform and the underlying mechanisms they use are actually the province of psychology, and Macphails assertion does not inspire a productive comparative research agenda going forward. Here, we explore a broader framework for interpreting intelligent behavior in animals using widely studied marine mammals bottlenose dolphins, sea lions, harbor seals, and West Indian manatees as examples. Their evolutionary history, notably the transition of terrestrial mammals into marine species, forced many adaptations including unique sensory systems, complex social organization, and neurobiological extremes. In addition, many of these animals show flexible cognition, at least, comparable to what has been observed in primates. Macphails definition of intelligence, limited to associative processes, is not consistent with common conceptualizations of human intelligence by experts or the lay public. Expert conceptualizations of human intelligence are multifaceted and include adaptation to the environment, mental processes, and higher order thinking e. g. , reasoning, problem solving, decision-making, and metacognition review in Sternberg, 2003. Studies of lay ideas of intelligence in the United States identify factors such as speed of processing, practical problem solving, verbal ability, non-verbal reasoning, numerical reasoning, and social competence Sternberg et al. , 1981 Chen and Chen, 1988. Unfortunately, these definitional factors do not hold up well cross-culturally Nagoshi, 1987, and there is no strong consensus among psychologists on what the components of intelligence should be, illustrating the problem of generalizing from any single human tradition. The origin of the concept of human intelligence and practical applications in individual differences Binet and Simon, 1916 McNemar, 1964 suggest questionable utility for interspecies comparisons, unless we use individual variability itself, which may be a hallmark of an intelligent species, as a comparative measure. In addition, Mackintosh 1998 notes that associative learning as described by Macphail bears striking similarities to human implicit learning e. g. , Reber, 1993, an area typically not addressed on intelligence tests. This sets a conundrum for comparative psychologists because intelligence defined for humans excludes implicit associative processes, and so intelligence would then appear to lie outside the realm of comparative psychology and Darwinian evolution. Comparative researchers responding to Macphails null hypothesis of no species differences in intelligence emphasized that intelligence consists of multiple facets including sensory and perceptual processes, memory, spatial relations, concept formation, rule learning, and tool use Goldman-Rakic and Preuss 1987 Hodos, 1987 Shettleworth, 1987 Rilling, 1990 Walker, 1990. Bullock 1986 suggested that candidates for investigation might include flexibility in interacting with the environment, social interactions, communication, and difficult, higher forms of cognition, plus problem solving across all the categories. In addition, he considered acquired knowledge essential to considerations of intelligence. Goldman-Rakic and Preuss 1987 and Vauclair 1990 among others also suggested that representation rather than association might be a more productive approach. Representations, based on the environmental information that animals extract through their sensory-motor systems and then organize perceptually and cognitively, vary widely across species and facilitate intelligent behavior. Associations between representations and the breadth and flexibility of those representations may be especially relevant for marine mammals, who become interesting due to their operating so effectively in two vastly different perceptual media water and air where they must recruit sensory-motor systems developed differentially for this split life. More recent approaches to animal intelligence retain a multifaceted approach e. g. , Roth and Dicke, 2017. In considering the intelligence of marine mammals, we start with the assumption that marine mammals have the basic associative processes identified by Macphail 1982, 1987 and demonstrated ubiquitously in marine mammal training Pepper and Defran, 1975, entertainments at commercial oceanaria, and numerous studies. We take an evolutionary stance that the transition from a terrestrial to an aquatic environment modified sensory and perceptual processes, as well as the flexibility and processing speed of other cognitive processes contributing to intelligence. We also investigate the implications of marine mammal neurobiology in the manifestation of intelligent behavior. For us, intelligence is the effectiveness by which one deploys cognitive processes including sensation and perception, instantiated in the central and peripheral nervous systems, and studied through investigations of behavior. Marine Mammals Marine mammals can be characterized as the mammals that depend primarily on the marine environment for survival Rice, 1998. This list could include marine otters Lontra felina, polar bears Ursus maritimus, Arctic foxes Vulpes lagopus, and fishing bats Noctilio leporinus, which feed on marine prey, but much more is known about the senses and cognition of cetaceans whales and dolphins, pinnipeds seals, sea lions, and walruses, and sirenians manatees and dugongs, so our focus will be on these orders. There are species within these orders that are exclusively freshwater such as the river dolphins, family Iniidae and Platinistidae, and some of the manatee species such as Amazonian manatees Trichechus inunguis, but by and large these orders are marine. Within each order only a few species have been studied, so some caution needs to be observed in generalizing across species, but these few species serve to provide a working base with which to compare other species. In addition, sample sizes for laboratory experimentation are small, frequently only one or two subjects, so conclusions are likely to be modified as more subjects are studied. The ancestors of cetaceans whales and dolphins and sirenians sea cows made a major transition from a terrestrial to an aquatic environment during the Eocene 50 million years ago. The pinnipeds seals, sea lions, and walruses made a partial transition more recently during the late Oligocene 2623 million years ago and remain amphibious, feeding at sea but reproducing on land. These three orders, which constitute the most studied groups of marine mammals, responded to new ecological pressures with numerous adaptive changes in morphology, physiology, behavior, and sensoryperceptual processes, thereby shifting the information they could gain about the world and thus their representations of it. Morphological changes included streamlining the body to reduce drag, including loss or reduction in hind limbs and modification of forelimbs, various other skeletal modifications, loss, reduction or modification in pelage, and internalization of male reproductive organs. Respiratory mechanisms had to meet simultaneous demands for the combination of in-air breathing with diving, often to great depths under great physical pressure. Circulatory systems were modified to maintain warm body temperature in cold water environments. Many species developed group social structures and cooperative systems for foraging and defense in an environment with few places to hide either for purposes of prey ambush or predator avoidance, especially for animals coupled to the surface for respiration. Communication systems emphasized auditory and tactile channels, while de-emphasizing or modifying visual systems, which were limited by low light and turbidity underwater and sharp transitions in brightness at the surface. Olfactory systems that evolved on land had limited utility underwater. Novel sensory processes, such as echolocation and exquisite senses of touch, were shaped by natural selection to facilitate foraging and orientation in a marine environment. An evolutionary perspective suggests that adaptation to the aquatic realm overlain on a terrestrial foundation likely affected an array of behavioral and cognitive intellectual processes, preserving some attributes while modifying others. Bullock 1986 provides an entry to a comparative assessment of intelligence beyond associationism by presenting a broad palette of candidate domains for the investigation of animal intelligence. We have selected from that palette to emphasize flexibility in problem solving, the neural plasticity that underlies flexibility, and knowledge. Curiously, Macphail 1987 also emphasized the generality of human intelligence, as well as its dependence on experience. Knowledge is a little studied topic in marine mammal science, but we can identify mechanisms that would allow the accumulation of knowledge the resolution perceptual detail of sensory systems, the speed of information transfer by imitation, retention over long time periods, and facility at problem solving. We have organized this information into four categories sensation and perception, social learning, flexibility of cognitive processes, and the brain. Sensation and Perception Sensory-motor experiences provide a foundation for intelligent thought by providing insight to the quality, range, and resolution of animal worlds or Umwelten von Uexkll, 19341957, the detail creating the representations operated on during cognitive processing to produce intelligent behavior. Early researchers Galton, 1883 Cattell, 1890 Spearman, 1904 considered sensory discrimination as integral to human intelligence, but their view failed to gain traction in mental measurement Deary, 1994 Sternberg, 2003. Nevertheless, subsequent research provided support for this sensory hypothesis for example, Deary et al. 2004 reported a high correlation between general sensory discrimination representing shared variance across several modalities and fluid intelligence Cattell, 1963, which is closely related to working memory Kyllonen and Christal, 1990 Salthouse and Pink, 2008. Intelligence differences between animal species may even more strongly reflect sensory processing, since sensory differences between species are more likely to be greater than differences within a single species, such as humans. The marine environment places specific demands on sensory perception. The slow rate of diffusion of chemical compounds in water limits their utility to marine mammals compared to terrestrial mammals living where olfactants are rapidly dispersed. The olfactory systems of terrestrial mammals, designed to detect and discriminate airborne compounds, are of reduced importance to animals spending substantial time underwater. Taste may be relevant but the overall sense of flavor i. e. , combined effect of taste and smell is probably lessened to the extent that olfaction is unavailable. Underwater vision is constrained by the limits of photic transmission in water, and it loses much of its relevance at depth or in turbid environments where light is limited. Touch provides advantages underwater for sensing hydrodynamic movement caused by currents or distortions in water flowing past objects, as well as for close contact investigation of items. Sound in water travels close to 4. 5 times as fast as sound in air and can be conveyed with fidelity over great distances. The long wavelengths of lower frequency sounds allow them to pass around objects that block light transmission, and high frequency sounds are capable of transmitting detailed information over shorter distances. Adaptations for enhanced acoustic and tactile processing required for life underwater not only fostered new sensory mechanisms for gaining important information but also pushed speed and range of processing to new heights due to the physics of sound transmission and pressure changes in water. Cetaceans There are over 80 species of odontocetes toothed whales living in diverse environments. The river dolphins, who live in muddy waters thick with particulate matter, have extremely poor eyesight. For example, Platanistidae, the Southeast Asian river dolphins, are probably capable of seeing only degrees of brightness and the Inia, the South American river dolphins, have visual acuity of over 40 arc min Mass and Supin, 1989. Bottlenose dolphins Tursiops truncatus have considerably better resolution, useful in the frequently more transparent water of coastal regions. Underwater visual acuity for bottlenose dolphins is about 8. 5 min and in-air is 12. 5 min Herman et al. , 1975. This reasonably good acuity underwater and in-air is surprising because of the differential role the cornea plays in refraction underwater practically none and in air where it is the primary refractive component of the eye. An eye adapted for vision underwater should not be able to focus in air and vice-versa, without specific adaptive mechanisms, which dolphins have Herman et al. , 1975. Dolphins, like other marine mammals tested, are monochromats Ahnelt and Kolb, 2000 who see the world in shades of gray Madsen and Herman, 1980, although there is evidence that they may have some color perception, presumably mediated by the differential sensitivity of rods and the single cone-type Griebel and Schmid, 2002. Dolphins depend on vision to build their representations of the world, their Umwelten. For example, they integrate information from multiple sensory systems, like vision and echolocation, to represent objects e. g. , Harley et al. , 1996, and they can discriminate among photographs and video of fish underwater using vision alone, likely an important ability for stealthy foraging Harley et al. , 2019. Dolphin hearing is exceptional reviews in Supin et al. , 2001 Au and Hastings, 2008 ranging from 0. 15 to 200 kHz, an upper limit over three octaves higher than that of humans. They are excellent at sound localization with 0. 54 degrees of resolution. They have a temporal processing rate, the ability of the nervous system to map sound intervals, as measured by auditory brainstem responses, over 1,500 Hz for amplitude modulated sounds compared to a rate of 50 Hz for humans. These evoked potential measures provide only indirect measures of temporal processing. Behavioral tests, direct tests of the ability of dolphins to discriminate sound intervals, indicate that they have a temporal integration time of an order of magnitude less than humans do Supin et al. , 2001. Dolphins are also active echolocators that have the ability to make subtle distinctions among object characteristics, e. g. , they can discriminate cylinders that vary in wall thickness by less than a millimeter Au and Pawloski, 1992. They can also recognize an equivalence between their visual and echoic experiences of objects Harley et al. , 2003 and share echoic information with nearby eavesdropping dolphins about object identity Xitco and Roitblat, 1996. Clearly, their representations of objects are fine-tuned and flexible. The sense of touch in dolphins has not been investigated to the same extent as in other marine mammals but electrophysiological measures of skin response show greatest sensitivity around the head Ridgway and Carder, 1990, 1993 with sensitivity comparable to human lips and fingers, sufficient to detect underwater turbulence Kolchin and Belkovich, 1973. Hair, important for touch in other marine mammals, has not been investigated well in cetaceans, probably because of its infrequent appearance among odontocetes, where it is found only on the rostrum of river dolphins and some neonates of other species. Sensory hairs are found on the rostrums of baleen whales, but they are difficult to study in these large, pelagic animals, although the structure of hairs of right whales appear to be adapted for detection of small prey such as plankton Murphy et al. , 2015. Cetaceans have missing or greatly reduced olfactory bulbs and ethmoturbinates. Their taste buds are few. Nevertheless, they have low thresholds for sour citric acid and bitter Nachtigall and Hall, 1984 Friedl et al. , 1990 Kuznetsov, 1990. They also can detect salt. Pinnipeds Although a large number of species comprise the pinnipeds seals, sea lions, and walruses, most sensory research has been conducted on the California sea lion Zalophus californianis and harbor seal Phoca vitulina. The vision of the pinnipeds may be most notable for relatively high acuity both in air and underwater. The large, curved orbit of the lens focuses light on the retina underwater. This would lead to myopic near-sighted vision in air, except the cornea in pinnipeds contains a flattened area over the pupil reducing or eliminating corneal refraction in air West et al. , 1991 Miller et al. , 2010. The underwater and in-air acuity of the sea lion are equivalent at moderate and high brightness at 4. 77 arc min, but underwater vision is better under dim light conditions. Seal vision is similar at 58 min. Pinnipeds are monochromats, and therefore, do not have dichromatic color vision as do most terrestrial mammals Ahnelt and Kolb, 2000, although a weak form of mesopic color vision in seals has also been reported Oppermann et al. , 2016. These reports of a weak form of color vision based on rod-cone spectral sensitivity differences Griebel and Peichel, 2003 have been challenged Scholtyssek et al. , 2015. Audiograms for pinnipeds tend to have considerable variability among studies, perhaps attributable to small sample sizes frequently just one animal and individual differences, but in general, the frequency range for harbor seals is about 0. 272 kHz Kastelein et al. , 2009, with sea lions having a somewhat lower upper limit. Early reports of hearing by pinnipeds suggested that underwater hearing was superior, but more recent evidence suggests that they are similar with both having low threshold levels Reichmuth et al. , 2013, a more understandable relationship given pinnipeds amphibious existence and terrestrial ancestry. Pinnipeds demonstrate sensitive mechanoreception both in the active haptic and passive modes, which they use for detecting hydrodynamic stimuli. They can discriminate size and shape by active touch Dehnhardt, 1994 Dehnhardt and Dcker, 1996 and detect water movement at detection thresholds under a micron of particle displacement Dehnhardt et al. , 1998 Dehnhardt and Mauck, 2008. Their high sensitivity to hydrodynamic stimuli allows both seals and sea lions to track fish by the turbulence they generate in swimming. While sea lion vibrissae appear to be more sensitive than those of phocids to relatively low frequency vibrations in the water, harbor seals have shown greater ability at following complex wakes over longer periods of time Glser et al. , 2011, perhaps due to differences in vibrissal structure Hanke et al. , 2010 Witte, 2012. The olfactory bulbs of pinnipeds are reduced in size, and there are fewer nasal turbinates. Nonetheless, scent recognition is a demonstrated feature of individual recognition in pinnipeds, particularly well-studied in mother-pup identification, and likely relevant for reproductive behavior in some species Pitcher et al. , 2011. Gustation has hardly been studied. There are taste buds on the tongue, albeit fewer than among terrestrial mammals. Despite the unimpressive anatomy associated with the chemical senses, sea lions detect sour, bitter, and salt Friedl et al. , 1990. They also have low discrimination thresholds for saline solutions Sticken and Dehnhardt, 2000. Sirenians West Indian manatees have modest visual acuity of approximately 20 arc min Mass et al. , 1997 Bauer et al. , 2003 and probably limited visual tracking capabilities Samuelson et al. , 2012. Unlike many of the cetaceans and pinnipeds studied, they lack a tapetum lucidum for enhancing light sensitivity, but also unlike them have dichromatic color vision Griebel and Schmid, 1996 Ahnelt and Kolb, 2000 Newman and Robinson, 2006. Preliminary evidence from streak retinoscopy indicates emmetropic to hyperopic vision both underwater and in-air Samuelson et al. , 2012. They lack a vomeronasal organ and their neurophysiology suggests modest olfaction review in Reep and Bonde, 2006. They have a higher density of taste receptors than dolphins Yamasaki et al. , 1980, but the psychophysics of taste and other chemical senses has not been investigated. Auditory capabilities include about an eight-octave frequency range extending from about 0. 25 kHz into the ultrasonic range over 70 kHz Gaspard et al.