Neuroscience & BiobehavioralReviews, Vol. 14, pp. 273-294. o Pergamon Press plc, 1990. Printed in the U.S.A. 0149-7634/90 $3.00 + .00

Behavioral Adaptations to Pathogens and Parasites: Five Strategies

BENJAMIN L. HART

Department of Physiological Sciences, School of Veterinary Medicine University of California, Davis, Davis, CA 95616

Received 21 December 1988

HART, B. L. Behavioral adaptations to pathogens and parasites: Five strategies. NEUROSCI BIOBEHAV REV 14(3) 273-294, 1990.--The ever present threat of viral, bacterial, protozoan and metazoan parasites in the environment of wild animals is viewed as responsible for the natural selection of a variety of behavioral patterns that enable animals to survive and reproduce in this type of environment. Several lines of research, some quite recent, point to five behavioral strategies that vertebrates utilize to increase their personal or inclusive fitness in the face of parasites (broadly defined to include pathogens). These are: 1) avoidance of parasites; 2) controlled exposure to parasites to potentiate the immune system; 3) behavior of sick animals including anorexia and depression to overcome systemic febrile infections; 4) helping sick animals; 5) sexual selection for mating partners with the genetic endowment for resistance to parasites. The point is made that to consider a behavioral pattern as having evolved to serve a parasite control function the parasite or causative agent should be shown to adversely impact the animal's fitness and the behavior in question must be shown to help animals, or their offspring or group mates, in combating their exposure, or reducing their vulnerability, to the parasite.

Parasites Pathogens Evolution Feeding behavior Sexual behavior Maternal behavior Grooming

TO many the most profound theme in our current understanding of examine many examples of behavioral patterns that appear to animal behavior is the influence of natural selection shaping reflect adaptations to the threat of parasites and pathogens. The behavioral patterns. Whether looking at experiential influences or different types of behavioral patterns that animals use to avoid, those reflecting a more direct genetic endowment, there is a focus control or eliminate parasites from their bodies can be divided into on the strategies or coping mechanisms, such as those used in five strategies or categories. The first strategy comprises behav- predator avoidance or resource utilization, which enable animals ioral patterns that enable animals to avoid or minimize their to survive into reproductive adulthood and assure survival of their exposure to parasites. The second strategy is controlled exposure offspring. The fact that animals must avoid predators and acquire by which animals may expose themselves (or offspring) to small resources in order to survive to reproductive age and successfully samples of particular parasites or pathogens to facilitate develop- rear young, shapes various aspects of social, feeding, and repro- ment of the body's immunological competence. The third strategy ductive behavior. The concept presented in this paper is that the relates to the behavioral patterns of anorexia and depression that existence of disease-causing viruses and bacteria, as well as animals show when they are febrile with an infectious disease. The external and internal parasites, represent major forces shaping behavior of sick animals can be viewed as facilitating the fever behavior that are perhaps as profound as the forces having to do response in suppressing microbial infections and increasing the with predation or resource utilization. Wildlife biologists are animal's chance of recovery from the illness. A fourth strategy is familiar with a variety of naturally occurring disease epidemics the behavior of some animals in helping sick group mates or kin. that have had devastating effects on animal populations (57, 58, Finally, as the fifth strategy, animals may select mates on the basis 215). Although relatively rare, these phenomena provide powerful of providing offspring with the genetic basis for effective parasite examples of potential costs of disease-causing organisms when and disease control. behavioral and immunological responses do not adequately protect One can apply the general rules of inclusive fitness (I00) to the animals. Anderson and May (6) point to accumulating evidence behavioral adaptations related to pathogens and parasites just as in that parasites (broadly defined to include viruses, bacteria, proto- other behavioral systems. The efforts an animal may expend in zoans, helminths and arthropods) play a role analogous to that of preserving its own life and reproductive potential will extend to the predators and resource limitation in constraining the growth of offspring, in protecting them from parasites and in possibly animal populations. In studying and observing animals that live in helping sick or injured kin to survive. The concept of parasite relatively clean laboratories, field stations and domestic environ- control strategies focuses on the welfare of individual animals ments, and that are vaccinated against diseases and medically sufficient to assure survival of their offspring to reproductive age. treated when sick, it is easy to forget that animals evolved and This does not imply that individuals remain parasite- or disease- thrived in environments with an array of parasites long before free; it is clear that animals do sometimes succumb to disease and human protective measures were available. parasitism. In some instances parasites may be carded at a load In this paper I will address theoretical issues dealing with the which does not cause any noticeable decrement in health, but role of animal behavior in the control of parasitism or disease and might play a role when there are extreme demands on an animal's

273 274 HART

resources such as in escaping from predators or undergoing to the differential effects of microparasites, macroparasites, para- nutritional stress. As in predator avoidance strategy, the fact that sitoids and predators on the stability of their host populations and an animal may eventually succumb to parasites does not mean that influences in regulating the population dynamics of hosts. the parasite control strategies are not adaptive. Similarly, the There are two requirements for accepting a particular behavior display of such behavioral strategies does not require the parasites as having a parasite control function. One is that the parasites in be always present. question should be shown to have a detrimental effect on the host's After going into some general issues about relations between fitness. The cost in fitness from an infectious, microparasitic parasites or pathogens and their hosts, I will discuss parasite- disease that kills an animal, or makes it susceptible to predation, related behavioral adaptations under each of the five strategies. is self-evident. As for macroparasites, a few lice or fleas may not This paper is written primarily with the student of animal behavior be harmful, but an overwhelming number may kill an animal. An in mind. However, an equal emphasis could be directed to intermediate parasite load may affect growth rate. Some macro- investigators interested in ecology or animal disease epidemiol- parasites may severely affect an animal only in times of nutrition ogy. If parasites are an unappreciated force from the standpoint of or socially related stress or in conjunction with other demands such behavioral science, then from the standpoint of ecology, epidemi- as lactating, fighting with a conspecific, or escaping from a ology, microbiology, and parasitology, behavior is an equally predator. Not to be overlooked is the cumulative effect of several unappreciated factor in how animals cope in an environment types of macroparasites, none of which alone may present a teeming with pathogenic viruses, bacteria and parasites. measurable cost to fitness. A parasite load in the laboratory may be studied in isolation from other parasites and thus not be represen- tative of the actual costs to an animal in the wild. In the section on GENERAL ISSUES OF HOST-PARASITEINTERACTIONS the avoidance strategy experimental findings on the costs of Before examining the specific strategies of behavior and macroparasites are reviewed. control of parasites, some general issues should be discussed. One The second requirement for accepting a behavior as having a of these is the coevolution of host-parasite relationships. In some parasite control function, and one which is basically the theme of instances parasites appear to have evolved behavioral strategies this review, is that the behavior in question should be shown to be themselves to counteract behavioral patterns that hosts use to avoid effective in helping an animal, or its offspring or group mates, to parasites. The predator-prey model is quite applicable here (266). avoid or remove the parasites or decrease their vulnerability to For example, fleas can jump with impressive speed and are thus parasites. Virtually any parasite control behavior will represent less likely to be dislodged by the grooming or scratching of the some cost in fitness to an animal as reflected in energy utilized, host which functions to eliminate fleas. Nasal bot flies that deposit distraction from predator vigilance or loss of feeding time. larvae in the nasal cavities of deer display ambush-like behavior Theoretically, the greater the cost of the parasite control behavior, that has the effect of counteracting the deer's avoidance behavior the more beneficial one would expect the behavior to be. (4). In other instances one can point to parasites that cause There has been a prevailing viewpoint that well-adapted physiological changes that alter the hosts' behavior to the para- parasites, with long-term evolutionary histories with their hosts, sites' advantage. Examples of this latter phenomenon are ad- will have no ill effects on the host (67,127). The devastating dressed below. Another topic that needs to be covered under effects of parasites on animals, such as in the infamous rinderpest general issues is the way that the interaction between animals epidemic of East Africa and myxomatosis in rabbits in Australia, and parasites is sometimes markedly influenced by human inter- it is argued, come from the introduction of novel parasites to hosts vention. to which parasites have not adapted. This peaceful coexistence or commensalistic argument is based on the notion that the parasite Coevolution of Animals and Parasites ought not to destroy its host lest the parasite species become extinct. If this viewpoint were accurate, there would be no basis It is common in the more recent literature on host-parasite for animals to evolve parasite control strategies (and hence no interactions to see reference to the terms microparasites and reason for this review). The commensalistic viewpoint is based macroparasites, a concept introduced by Anderson and May (6). upon unsound group selection arguments (183) and there are, in Microparasites are primarily viruses, bacteria, and protozoa. fact, both theoretical arguments and empirical evidence that many These pathogens usually have a short generation time and multiply parasite species evolve increasing virulence over long-term asso- directly within the host. Their effect on the host generally is rather ciation with their hosts. Anderson and May (7) point out that there transient, inducing either immunity or death in a relatively short are many instances where microparasite transmission is linked to time. Hosts can be classified as sick or nonsick and, if nonsick, its reproductive rate, and where in turn the parasite's reproductive either susceptible, recovered or immune. Macroparasites are rate is linked to pathogenicity within the host in the form of primarily helminths (internal worm-like parasites) or arthropods depletion of resources, damage to host tissues, production of (external parasites). These parasites have a long generation time toxins and suppression of the host's immune system. Certainly and usually go through some part of their life cycle outside the with regard to microparasites, competition among parasite clones hosts. Hosts are continuously reinfected with macroparasites and within an animal's body would often favor the most virulent forms immunity, if it occurs, depends upon the number of parasites and (91). In a survey of mortality of different diseases, Ewald (76) is not long lasting. Hosts can be classified as having a range of provided empirical evidence that vector-born parasites, in which infestation along a continuum from low-level, with little notice- debilitation or immobilization of the vertebrate host facilitates the able impact, to a high-level with major adverse effects on health. likelihood of a vector acquiring the parasite, tend to evolve In this review the term parasites will often be used in the general towards increasing virulence in the host. sense to represent both microparasites and macroparasites. Host- There are instances where a lack of pathogenicity could be parasite interactions may be looked at as involving a continuum selected. One is by vertical transmission in which there is a direct with, at one extreme, microparasites that can live only within or transmission of the infection from parent to progeny and where the upon their hosts and, at the other extreme, predator-like parasites reproductive success of the parasite is directly linked to that of the of invertebrates referred to as parasitoids, which live within a host host (77). A second is where there is a prolonged intrahost stage of in the larval stage and are otherwise free living (77). Toft (269) has the life cycle in which transmission to the next host is impossible discussed host-parasite and prey-predator interactions with regard until the host engages in a new activity. Such benignity would be BEHAVIOR AND PARASITES 275

selected where host migration occurs during the year and the descending back down peripheral nerves including those innervat- parasite must wait until the host returns to the original site for ing the salivary glands. The virus enters the salivary gland tissue breeding before the parasite can be transmitted (77,91). Another and saliva (203). The nerves supplying the laryngeal and pharyn- possibility for selection of low virulence is by interdemal selection geal muscles are often affected as well and the ensuing paralysis of where hosts harboring the most virulent forms of a parasite are these muscles makes swallowing difficult. This causes pooling of killed off so quickly that parasite transmission is less than that of saliva, and when the infected animal bites a new victim the pooled competing conspeciflc parasites with milder virulence. This is saliva is a reservoir of virus to be injected into the new victim. given as an explanation of the evolution of reduced virulence of When transmission of the rabies virus does not depend upon the myxoma virus in Australian rabbits (91). abnormal viciousness for transmission, such as in vampire bats There are many instances where the parasite-host relationship where the animals naturally bite each other rather frequently (50), looks deceptively harmless but where the parasite has the potential or where the virus may be transmitted by the airborne route, to be very harmful. The host's behavior and immune system may abnormal viciousness for transmission from neurological alteration be effective in protecting it from the parasite, leading to a standoff does not occur. In vampire bats the rabies virus reportedly does not between parasite virulence and host resistance. Any letup in the necessarily produce acute death and may not be fatal (50). host's defenses, including behavior, could be disastrous. Thus, Alteration of intermediate hosts by macroparasites. A number parasite avoidance behavior could be important to the host's of intestinal parasites which use vertebrates as definitive hosts survival albeit the threat of serious disease or parasitism may not have invertebrate or vertebrate intermediate hosts and are trans- be obvious. Additionally, in looking at an animal population in ferred to the definitive host when the definitive host preys upon the general, a parasitic threat may not be evident because parasites intermediate host. If the parasite can bring about a change in commonly are unevenly distributed within a population with most behavior of the intermediate host that makes it more likely to be parasites aggregated on or within a few animals (5). preyed upon, the parasite is more likely to secure its def'mitive host. Parasitized intermediate hosts may become more vulnerable prey through reduced stamina, increased conspicuousness, disori- Effects of Parasites on Host Behavior entation and aberrant responses to environmental stimuli (64, 128, 194, 195). The effects of thorny-headed worms on the behavior of Although not the focus of this review, the topic of parasite- pill bugs, making them more likely to be eaten by birds such as induced changes in host behavior has received attention recently starlings which are definitive hosts of the worms, have been and is becoming an area of increasing investigation. It has been examined by Moore in considerable detail (193). Normally, pill mentioned that the alteration of host physiological responses by bugs seek hidden places of high humidity that are sheltered and of parasites may be adaptive to the parasite. Moore and Gotteli (197) dark color. In laboratory experiments pill bugs infested with the caution that whereas there are clear instances in which alteration of worms were much less likely than uninfected bugs to seek high host behavior is advantageous to parasites, available evidence also humidity, shelter and a camouflaging dark substrate. Captive shows that there are instances where alteration of host behavior is starlings were found to differentially take their prey from bugs on not advantageous to the parasites. Ewald (75) has emphasized that the white sand thereby eating more infested than uninfested bugs changes in the host induced by a parasite may be of some (193). In field experiments, starlings were also found taking a advantage to the host (especially if prolonging the life of the host disproportionately greater number of infested than uninfested pill enhances transmission of the parasite), or the changes may be of bugs. advantage to neither host nor parasite but simply reflect a Two other examples serve to illustrate that parasitic worms can by-product of the infection. The following are some examples of make intermediate hosts more vulnerable to predation. Flukes of alteration of host behavior, where vertebrates are part of the fish-eating birds commonly use fish as intermediate hosts. When system of parasite transmission and in which behavioral changes in the fish (dace) are infested with a parasitic eye-fluke their vision is the host are detrimental to the host and of advantage to the impaired and they consequently spend a greater proportion of time parasite. feeding in better illuminated water (54). As a result the dace are Effects of rabies virus on mammals. This is probably the most more likely to be preyed upon by the definitive avian hosts. Moose well-known example of a virus altering animal behavior. The virus serve as an intermediate host of a tapeworm for which wolves are infects domestic dogs and other canids and can turn a calm and the definitive host. Presence of the encysted tapeworm in moose placid animal into a wandering and vicious animal that attacks lungs makes the animals more sluggish and more likely to be almost any target in sight. The rabies virus cannot survive except preyed upon (235). Whether these examples in altering of host in living organisms, yet it is virtually always lethal in man, canids behavior actually increases the likelihood of transmission of the and many other hosts. In these instances, to be propagated the parasite (and hence could be viewed as an adaptive manipulation virus must be transferred to another animal before the host dies. In of host behavior) remains to be tested. the furious or vicious form of rabies the virus usually enters a Alteration of mouse behavior by malarial parasites. Mice wound through a bite by an infected animal. The virus, which is in (69,70), rabbits (277), and birds (284) have highly effective the attacker's saliva, enters the open wound created by the bite, anti-mosquito behaviors, such as head shaking, ear flipping, face moves into exposed peripheral nerves and then travels up inside rubbing, foot stamping and tail biting, that render them quite peripheral nerves to the brain (203). The virus then multiplies in inaccessible to mosquitoes. However, mosquito-borne diseases, several areas of the brain of the new host, but particularly in limbic including malaria, are widespread in certain rodent populations, system structures such as the amygdala, hippocampus and pyri- indicating that some aspect of mosquito-borne diseases makes form lobes. The limbic system is known to control emotional rodents susceptible to mosquito feeding. It was found that infec- behavior, including aggressive and feeding behavior and damage tion with malarial organisms results in a marked reduction in mean or experimental electrical stimulation of this area can lead to rage daily activity level of mice and inhibits anti-mosquito defenses just and undirected, unprovoked attacks. The greater localization and at the time when the malarial gametocytes are most infective to the multiplication of the virus in the limbic system compared with the mosquito vector (60). The reduction in murine anti-mosquito neocortex is apparently what causes abnormal behavior, including behavior is of vital importance to a parasite that must be trans- wandering aimlessly and attacking and biting anything encoun- ferred by mosquito vector. tered (142). While the virus is multiplying in the brain, it is also 276 HART

Impact of Human Intervention efficiency (89). For parasites that invade tissue such as flukes, lung worms, ascarids, and strongyles, there can be actual tissue An important perspective on behavioral adaptations related to damage to major organ systems. Parasites that erode the intestinal parasites is recognizing the widespread effects of human interven- mucosa, such as nematodes and trichostrongyles, can result in tion in hunting animals, transporting them, introducing chemo- diarrhea, leakage of plasma proteins into the gut and decreased therapeutic agents for treating disease, and using immunization nutrient absorption (89). Levels of intestinal parasites that do not programs. For the domesticated species, these programs have result in clinical signs of disease may still impair growth rate and undoubtedly altered the host-pathogen relationship. For example, reproductive efficiency. This was dramatically illustrated in a with the treatment of domestic animals for intestinal parasites series of experiments, where lambs were continuously infested which are transmitted by fecal contamination of foodstuffs, the with gastrointestinal nematodes at a level of infection low enough selection pressures are relaxed for sanitary elimination behavior that no clinical signs could be used to separate the infested from with the apparent result of individual variability in sanitary the clean sheep; even parasitic eggs in feces were low enough to be eliminative behavior (108). Vaccination programs might do the considered clinically insignificant. After 14 weeks of worm same for certain infectious viral and bacterial diseases. infestation, the growth rate of infested lambs was about 50 percent Wild populations of animals are less influenced by our use of of controls and protein deposition and calcium deposition less than vaccinations or chemotherapeutic agents, but these animals could 50 percent (262). be affected by relocation schemes. Existing wild populations of animals that have developed immunological competence and Effects on social behavior may result from moderate parasite parasite control behavioral strategies may be exposed to animals loads. For example, in an experimental infestation with an intestinal nematode, mice with lighter doses of nematode larvae that carry alien parasites with which the existing population has no past evolutionary or immunologic history. This may subsequently became dominant more frequently than those with heavier doses result in massive kill-offs of the resident animals. A prime despite the fact that the heavy dose did not affect weight (85). Similar relationships between dominance status in mice and example was the introduction of rinderpest-carrying Asian cattle to East Africa (256). parasite load were found with experimentally induced TrichineUa infections (233,234). There is also the influence of hunting or harvesting of wild populations of animals. In many instances, predators will usually External parasites commonly act as vectors of disease as well capture the most incapacitated, weakened, or young animals. Most as consuming body resources of their hosts. However, since wild infectious diseases are not transmissible from prey to predator, so animals are often immune to vector-borne microparasitic diseases, when a predator kills and consumes an animal sick with a disease, the direct costs of vector-borne disease would be difficult to assess it reduces the reservoir of disease for other prey conspecifics. In so these will be ignored in favor of discussing only direct costs of our hunting practices we often take the strongest and healthiest macroparasites. Indicative of such direct costs is the blood animals. This can alter a long-established host-parasite rela- removed by blood-sucking flies. In one study of tabanid flies in tionship. upstate New York, investigators estimated that on a typical day, a horse could be bitten by as many as 4000 flies (264). Depending upon species, feeding by a fly was estimated to take 0.08-0.3 ml Human Ethology and Disease Control of blood. According to the author's calculations, a horse could Although this paper focuses on relationships of animal behav- typically lose 0.5 liters of blood per day from fly bites. In cattle the ior to parasite control the same concepts may apply, at some level, presence of flies is correlated with reduced weight gains (117,251) to human behavior. The field of human ethology emphasizes the and a reduction in milk production (106, 134, 200). Studies on genetic influences over human behavior that reflect selection reindeer attribute biting flies with contributing to death of the pressures in our past evolutionary history and certainly parasite animals when they are faced with food shortages or environmental control behavioral patterns would have influenced human repro- stress (123). Warble (oestrid or hypodermal) flies deposit eggs on ductive success before the days of chemotherapeutic drugs and the hairs of cattle and cervids. Within a few days, larvae emerge vaccinations. Some possibilities which come to mind are our from the eggs, enter the skin and burrow through subcutaneous almost universal predilections for sanitary behavior and a tendency tissues of the host for 4-5 months. The larvae eventually encyst in to ostracize chronically sick people. The behavior of people when subcutaneous tissues on the backs where they remain for another they are sick closely matches that of sick animals that are anorexic month or so before emerging, falling to the ground and pupating and depressed (109). As discussed later, these behaviors can be (8). Typical decrements in weight gain of growing cattle attributed viewed as adaptive in helping animals and people overcome acute to these flies range from an estimated 20 to 70 kg per year (40,98). illness. Surface dwelling ectoparasites such as ticks and lice may cause death from anemia and weight loss when the parasite burdens are STRATEGY t: THE AVOIDANCESTRATEGY high (89,210). Some information about the costs of light or moderate parasite loads is available from experimental studies of This is the strategy for which documentation of behavioral tick infestation in cattle. One study found that a tick load of 48 adaptations to the threat of parasitism or disease is the most engorging ticks retarded growth at a mean rate per animal of 29 kg available. As mentioned previously, the costs to fitness from per year or 0.6 kg per engorging tick on an annualized basis (170). microparasites, which can cause acute and severe illness, is Similar results regarding effects on growth rate have been obtained self-evident. The costs of macroparasites, which are more subtle, by other investigators (253, 273, 291, 292). As much as 65 is discussed first in this section. The section then moves on to percent of the decrement in weight gain has been attributed to behavioral patterns related to avoiding macroparasites. appetite loss, presumably from toxins in tick saliva (253). To slender gazelle or impala that face the risk of predation daily, Costs of Macroparasites where intermale competition for territory depends on size and strength, and where lactating females must put out extra nutrients Most of the indications of the effects of large numbers of for nursing their young, the cost of even as few as half a dozen intestinal parasites come from observations on domestic animals engorging ticks could result in failure to escape from a predator, where weight loss and anemia are fairly common along with win a territory, or wean healthy offspring. Even lice can take their diminished efficiency in feed conversion and reduced reproductive toll in body weight gain. In growing cattle infestation of approx- BEHAVIOR AND PARASITES 277

imately 1.6 lice per sq cm retarded growth at the rate of 32 kg per nests or den areas include felids and canids, which would be at year (90). The nests of birds and burrows of rodents are an particular risk of reinfecting themselves with fecal-borne parasitic excellent habitat for ectoparasites. The environment is dark, larvae as they bring kills back to the nest or den to feed upon moist, and warm. Many parasites live on the blood, skin or (108). Since young are the most severely affected by internal feathers of animals causing problems for young and old inhabitants parasites, protection of young from parasitic larval exposure of the nests and burrows. It is from the relatively stable environ- around the nest or den would be particularly important. Owners of ments of nests and burrows, with abundant food, that many pet dogs and cats are usually aware of this native predisposition ectoparasitic arthropods are thought to have evolved (181,276). towards sanitary behavior in their pets and take advantage of this Little is known about the costs of mammalian ectoparasites behavior in housetralning their pets (110). The predisposition of maintained in burrows. There are a number of reports alluding to puppies to eliminate away from the sleeping area is evident as soon the costs of nest-maintained ectoparasites in birds including failure as the pups are capable of walking (239,245). Sows will typically of second broods (233), reduction of clutch size (201) and eliminate far from sleeping areas and piglets eliminate away from desertion of well-incubated eggs (78). the sleeping area when they are 4-5 days old (36,286). Probably the most clear effects of parasites on the fitness of Nest sanitation would not generally apply to primate species, birds is the work of M¢ller (188) on the barn swallow in which many of which roost in trees and drop feces from the trees. With exposure of birds to the hematophagus tropical fowl mite was yellow baboons the collection of feces beneath roosting trees experimentally manipulated by treating nests with a pyrethrin provides a source of endoparasitic infection (121). The area spray to kill mites or inoculating nests with mites. The addition of beneath roosting trees is moist, and has less sun and wind than mites to nests decreased reproductive success as indicated by other areas, so parasitic larvae are not rapidly desiccated. Baboons number of independent fledglings of both first and second clutches. tend to socialize and rest beneath such trees exposing them to Swallow nestlings from inoculated nests had lower body mass (but intestinal nematode larvae. Although baboons rotate roost sites, similar size) than those from sprayed nests. Interclutch intervals such sites are limited within a territorial range and eventually the were shorter in the birds in sprayed nests because more of their troop must return to old roosts. A correlation between baboon nests were reused for the second clutch and because adult troop movement and life cycle intervals of parasite larvae provided swallows apparently suffered less from mite infestations. The barn evidence that the baboon roost-changing behavior may be a swallow is one of the avian species that nests in holes or protected parasite control strategy (121). Most parasitic larvae will have died areas, presumably as a defense against predators. The importance off or been consumed by dung beetles 8-11 days after fecal of parasitism in avian species that nest in holes exceeds that of deposition. Baboons typically stay at a roost site for only 2 nights predation as a selection pressure and it is likely that the cost of and wait about 9-11 days before returning to a roosting grove to parasitism is what hole nesters pay for safety from nest predation use it again. Thus, the pattern of moving roosting sites has the (186,190). effect of minimizing exposure to infectious larvae. Mangabey monkeys of rain forest areas are completely arbo- Avoidance of Macroparasites real, living on fruits and leaves and do not (or cannot) avoid defecating on foliage that they will handle or eat. Rain washes off The variety of behavioral patterns involved in avoidance of leaves in 24 hours but with no rain, feces stay on the vegetation internal and external macroparasites includes feeding, eliminative, longer. The monkeys were found to cover'a greater distance during grooming, preening, social, sexual, maternal and sleeping behav- no rain days and to avoid areas foraged in the previous day, the ior. Many studies are quantitative and creatively designed, others effect being to reduce contact with contaminated foliage (83). are based upon opportunistic field observations. Control of ectoparasites by grooming and preening. For many Foraging and eliminative behavior to avoid internal parasites. rodents, felids, ruminants and primates grooming is one of the Intestinal parasites characteristically lay eggs which pass out with most frequently and regularly performed behavioral patterns. It is the feces and develop into larvae which are then infectious to estimated that rats, for example, can spend up to one-third of their animals that consume them. Animals have behavioral patterns that waking time grooming (24). Observations of female antelope appear to reduce their exposure to infectious stages of internal (gazelles, impala) on the East African plains indicate they orally parasites. This is especially evident in the foraging patterns of self-groom the body over 1000 times and deliver almost as many grazing ungulates. Sheep (265), horses (217), cattle (184), and scratches during a typical 12-hour period of daylight (114). presumably other wild and domestic ungulates, tend to avoid Grooming behavior entails some costs to animals (depending upon consuming forage that is adjacent to recently dropped feces, even species) including energy utilized, distraction from vigilance for though the grass may be quite lush. Michel (184) estimated that predators and loss of water and electrolytes from saliva used in cattle can reduce their intake of infectious larvae to 25 percent of oral grooming. Studies estimate that from 8 to 30 percent of a rat's that which would be consumed on a random basis in a pasture by evaporative water loss in the absence of heat stress is from normal such selective grazing. Experimental observations on horses have grooming (178,240). Preening is very common among avian revealed that animals cue off the odor of fecal material in avoiding species, but the behavior has received less analytical attention as a contaminated forage (217). The rejection of forage by cattle is parasite control mechanism than grooming. Where examined both evident especially when manure is spread over pasture as a grooming and preening have proven to be effective in removing fertilizing slurry. Even 7 weeks after slurry application cattle ectoparasites. Conditioning the pelage and plumage by removing prefer pasture with no slurry (222); this discrimination largely dirty and stale oil is also a commonly mentioned function of disappears by 13 weeks (32). grooming and preening. While selective foraging is one method of reducing intestinal Self-grooming. The best evidence that grooming is important in parasite infestation, selective elimination behavior is another. the removal of ectoparasites is from experiments in which animals Horses, especially those that are corralled, use dunging areas and are prevented from grooming. Mice sustaining limb paralysis or those in pasture commonly defecate in areas that are seldom limb amputation, that prevented scratching, became infested with grazed (217). Sheep on pasture have been reported to selectively lice to about 30 times the normal level when exposed to a seed eliminate at night away from areas that they will graze upon during population (17). Excessive louse infestation of about 60 times the the day (53). normal level followed prevention of oral grooming in mice Species that are recognized as fastidious in not soiling their (205,206). Mite infestation also increases when mice are pre- 278 HART

vented from grooming (288). A positive relationship between investigators as being remarkably free of ectoparasites (99, 266, mean rate of self-grooming in vampire bats and estimated number 281). This may well be due to the effectiveness of their grooming of ectoparasitic flies in their particular roost trees was noted by technique. One investigator took a tame redtail monkey on walks Wilkinson (290), suggesting that as exposure to ectoparasitic flies in the forest to estimate the rate of tick acquisition and found it increased, self-grooming rate increased. collected 8.2 ticks per hour (84). In a study of the effects of grooming on tick infestation in Allogrooming in adult primates has been related to a number of cattle, cows were fitted with harnesses that prevented oral groom- social functions such as reinforcing social bonds, diverting possi- ing and compared with cows which were not harnessed. Subse- ble attacks, and facilitating sexual interactions. However, obser- quently, 40,000 to 50,000 larval or nymphal stage ticks (Boophilus vations on wild primates provide evidence that allogrooming is microplus) were applied to the animals' bodies (257). Of the ticks also important in ectoparasite removal. For example, an adult male applied to harnessed cows 33 percent survived to engorgement vervet monkey which had disappeared from its group for a few whereas only 9 percent of the applied ticks survived on unre- days acquired such a heavy load of ticks that even at some distance strained cows. Similar results were found when cattle were its right ear could be seen to be covered with ticks. Shortly after infested daily with larval or nymphal ticks (18). Experiments on rejoining the group the ticks on the ear (and presumably else- the effectiveness of oral grooming in controlling cattle lice found where) had been removed by other monkeys through aUogrooming that louse population rapidly built up when caRRie were stanchioned (266). Animals without access to allogrooming, including solitary to prevent oral grooming (166). When the cattle were allowed to male baboons (247) and langurs (175), are reported to harbor groom, the louse load was eliminated or markedly reduced within heavier tick loads than those with access to allogrooming. Since 3 days. When licking was allowed on just part of the body, lice primate allogrooming is concentrated on areas inaccessible or were still reduced in noniicked areas apparently because of the invisible to an animal being groomed, this has been used by some movement of lice into licked areas. primatologists to argue that allogrooming has primarily a utilitar- Grooming of young. The effectiveness of grooming in removing ian function (13,136). With this perspective, the social functions ectoparasites could be especially important in newborn and very could be viewed as important but evolutionarily secondary to the young animals in which ectoparasites can be particularly detrimen- utilitarian function. tal. Newborn rodents and carnivores cannot groom themselves and Both mule deer of North America and impala of Africa the nest or den provides an excellent habitat for ectoparasites. frequent brushy areas that presumably place the animalz more at Mothers provide extensive licking to the newborn of these species risk for acquiring ticks than does grassland habitat (171, 213, (239, 243,252). Furthermore, there is an increase in grooming of 241). In both species allogrooming is common among adults and the mammary gland and nipples in mother rats prior to pamLrition is directed to body areas that animals cannot reach by self oral (246). This provides protection to the young because the mam- grooming (114,199). Tick removal is a likely function of this mary area is a major point of contact between the mother's body behavior. In fact impala exhibit a highly reciprocal tit-for-tat type and the young and a potential site for the spread of ectoparasites of allogrooming which is unique among East African antelope and from mother to young. In puppies the initial site of mite location perhaps all ungulates (114). Interestingly, impala also attract tick is on the lips; this has been attributed to the fact that it is the lips birds (oxpeckers) which other small plains-dwelling antelope do that constitute the site of mite transference from mother to infant not (114). during nursing (214). Self-preening and aUopreening in birds. Limited studies indicate In most wild ungulates grooming of the young offspring by the that preening in birds is an effective behavior for controlling mother is common. Immediately after birth grooming of the ectoparasites. In a study of chickens, the birds were debeaked in newborn has been reported to clean and dry the pelage (185), the fashion customary for group housing on farms and each facilitate olfactory recognition of the neonate by the mother chicken infested with a seed population of 50 lice (34). When (97,164), and reduce fawns' odors minimizing the likelihood that examined 33 days later the debeaked animals had a mean of about fawns will be detected by predators (I0, 164, 165). Beyond the 1600 lice compared with a mean of less than 60 lice for normal first couple of weeks after birth these functions diminish in beaked chickens. Louse-infested chickens have been found to importance but grooming of the young by the mother continues. preen at about 5-10 times the rate of noninfected chickens (35). Certainly for ungulates, such as mule deer, that inhabit the brushy, Allopreening, which has been sometimes observed among tick-infested wooded areas, removal of ticks would be an impor- mated pairs of birds (107), has been investigated with regard to tant function of grooming of the young by the mother. Male fawns tick removal among two species of eudypid penguins (31). In of mule deer were noted to solicit and receive more grooming by February and March penguins come ashore to pair and mate, mothers than female fawns (199), a finding consistent with the although some remain unpaired. After leaving the sea and coining notion that removal of ticks prior to engorgement helps male ashore is when they are likely to acquire ticks. Only the mated offspring preferentially achieve the rapid growth and large size pairs engage in allopreening. Despite the fact that unpaired which is needed for consequent mating success (47). penguins self-preened as much as paired penguins (8-12 min per Mutual grooming between adults. By engaging in scratch groom- hr) the unpaired birds had 2-3 times more ticks. ing animals are able to cover areas of the body that cannot be Anting in birds. Another behavior engaged in by birds, and reached by the mouth. Mutual or allogrooming is also a way that which appears to be effective in removal of ectoparasites, is anting animals can obtain grooming coverage and for areas that cannot be in which a bird (passerine) places itself prostrate on an anthill. reached by oral grooming. Mice that are prevented from self This behavior stimulates ants to swarm out in large numbers and scratching, but are housed together and able to mutually groom, spray formic acid and terpenoids as if the bird was attacking their maintain levels of lice lower than comparably treated mice housed nest. Birds customarily flutter their wings while hovering over the singly (15). anthill and occasionally dab ants over their feathers with their Both self-grooming and allogrooming is a common behavior in beaks (254). Anting in birds has been related to a number of primates. Primates groom by parting the fur and removing functions such as autoerotism (287), soothing the skin initiated by particles with the digits or by placing their mouth on the skin to feather growth (225), and removing stale lipids (255). The fact remove parasites or particles. They also scratch with brisk raking that formic acid and terpenoids are toxic to external parasites has movements. The fur parting method is used in allogrooming (84). led to the ectoparasite removal hypothesis (254,255). Probably the When wild primates have been examined they have impressed strongest support for this function comes from the observations of BEHAVIOR AND PARASITES 279

a former eminent Russian parasitologist, Dubinin, whose obser- ticks (174). Other reports include observations of birds removing vations were published in Russian and not readily available to ticks from marine iguanas (3,41) or leeches from turtles (275). Western ornithologists until the translation and discussion of However, these latter reports do not mention any behavior on the pertinent passages by Kelso and Nice (146). Dubinin conducted part of the animal being cleaned such as seeking or facilitating the necropsies of Steppe pipits, 4 of which had been anting and 4 of cleaners' actions. Without clear evidence of the host animal which were collected near their nests and had not been mating. He seeking or accommodating the cleaners' actions, it would be collected 732 mites from the mating pipits of which 90 were dead premature to generally relate tick removal by birds to parasite and 163 died within 12 hours. From 758 mites collected from avoidance or control behavior on the part of the parasite host. nonanting pipits, none were dead and only 5 died within 12 hours. Interestingly, when cleaning activity is sought or facilitated by Anting may have a role in control of microparasites as well as host animals such as impala, fish and Gal~pagus tortoises (as maeroparasites. Recent work on the secretion of metapleural discussed above), it is not possible for the host animal to see the glands of ants reveals that fungus- and bacteria-inhibiting sub- parasites that the cleaners remove. stances (plant auxins, 13-indoleacetic acid and phenylacetic acid) Ticks and other ectoparasites can be controlled by the immune are produced and spread by ants over their bodies (14). It has been system (208, 210, 274, 289). Some of the initial evidence for suggested that mating may involve the acquisition of antibiotic immune system control of parasites comes from work with cattle secretions from ants which would help explain the frequent showing that when grooming is prevented, tick load initially correlation of mating with conditions of high humidity which increases but then may decrease several weeks later presumably promote bacterial and fungal growth on the skin of birds (71). because the immune response is activated with an excessive tick Heterospecific cleaning. Removal of ectoparasites may be load (16,18). One issue that has not received much attention is the performed by heterospecific animals as in the well-known phe- interaction between the physical removal of parasites through nomenon of aquatic cleaning symbiosis in which smaller fish or grooming with the immunological inhibition of feeding by para- shrimp remove ectoparasites, fungi and bacteria from larger fish sites. As implied by Nelsen et al. (210), grooming and preening (79,169). The parasites are food for smaller fish and shrimp. Large are an effective first line of defense against ectoparasites and the fish visit the territories or cleaning stations of cleaner fish or immune response may not become evident unless parasite numbers shrimp, sometimes traveling some distance. The fish seeking become excessive. grooming often present certain body parts to be cleaned, including Animal movement to avoid ticks. The observation that cattle opening their mouths for the cleaner animals to remove and ingest sometimes refused to.graze in small experimental paddocks which the parasites around the gills and mouth surfaces. In reviewing the had been seeded with large numbers of tick larvae, suggested that published information available at the time, as well as citing the cattle may actually avoid coming close to clumps of ticks (125). extensive unpublished research notes of the then late C. Lim- Experiments revealed that the number of ticks picked up on the baugh, Feder (79) noted cleaning symbiosis was apparently a bodies of cattle fell below that expected from a direct proportional world wide occurrence in the marine environment. In many relationship to tick density in the paddocks (261). Close observa- instances the species of cleaner animals resemble normal prey of tion of cattle revealed obvious avoidance behavior as indicated by the fish that they clean but are not attacked by the species they turning away from the tick-infested patches. If tick avoidance customarily clean (79). As of Feder's (79) comprehensive review, occurs in cattle, it is reasonable to expectsimilar behavior in wild 45 species of small fish and 6 species of shrimp had been identified ungulates such as cervids and antelope that are more vulnerable to as cleaners. Among the many opportunistic observations pointing the resource drain of ticks. Because the distribution of ticks in the to the importance of cleaners for parasite removal is an observation environment is patchy (engorged ticks generally drop off when the on a corral reef where all cleaners had been removed. Within two animals rest overnight), there is an opportunity for animals to weeks almost all nonterritorial fish had left and those remaining avoid clumps by visual detection. For the young of cervids and had developed visible white fuzzy blotches, swellings, ulcerated antelope, where newborn animals hide out for the fast week or two sores and/or frayed t-ms indicative of infections and excessive after birth, visual avoidance of tick clumps should be highly parasitism (168). Feder (79) concludes that some species of fish advantageous. The possibility that adults and young of wild would not be able to survive the ravages of ectoparasites, bacteria ungulates may visually avoid clumps of ticks should be investi- or fungi without cleaners. gated. Another type of interspecific cleaning behavior that functions Animal movement to avoid flying parasites. A wide variety of to remove some of the larger ectoparasites from the bodies of behavioral patterns are utilized by various animal species to avoid mammals and reptiles is foraging by tick birds. The foraging by or reduce the intensity of attacks by flies. Some studies reveal the oxpeckers on rhinoceros, buffalo, zebra, giraffe and eland is a effectiveness of the behaviors in reducing fly attacks, others common site in Africa. Stomach analyses found the average consist of correlating prevalence, or rate of the behavior, with fly free-living oxpecker had 400 ticks in its stomach (21). An intensity. experiment in which 2-4 oxpeckers were kept with cattle upon Fly repelling behavior. Frequently observed movements that which ticks were applied found that about 20 percent of larval and animals use to repel or dislodge biting flies are ear twitching, nymph and 100 percent of adult ticks were removed by oxpeckers bead-tossing, leg stamping, muzzle flicking, muscle twitching and (21). Observations of impala, when they were attended by tail switching (73, 118,202, 218). These actions allow herbivores, oxpeckers, revealed that they make postural adjustments to ac- which must spend a great deal of time grazing or browsing, to commodate the tick birds involving lowering their ears to allow the reduce the pain and loss of blood inflictedby flies. Studies in cattle birds to forage inside the pinna or closing their eyes while the have found that when biting fly intensity is high, fly repelling or oxpeckers remove ticks from the eyelids (114). dislodging behavior increases (106,126). Under experimental Movements to accommodate or facilitate the foraging actions conditions with pairs of calves exposed to the same number of of tick birds is evident with some reptiles that are attended by biting flies, calves that most vigorously engaged in fly repelling finches. Two species of Darwin's finches which remove ticks from actions such as tail switching, foot stomping and head tossing had Gal,'ipagus tortoises perform exaggerated hops in rapid succession fewer biting flies on them than more placid contemporaries (280). while facing towards a tortoise's head whereupon the tortoise Lying down. Animals may lie down as a means of avoiding flies. typically extends all four legs and stretches its neck out. The birds Red deer more than doubled their time spent lying down on days then fly to the exposed parts and begin searching for and removing when they were harassed by head flies (74). The effectiveness of 280 HART

lying down was evident from visual inspection for flies and from attractiveness among the horses. the ear flipping frequency which dropped 50 percent after the Perhaps the ultimate in grouping in animals is the bunching animals had been lying down for 15 minutes. Reduction in behavior seen in cattle when fly intensity is severe and in which a exposed body surface, decrease in production of attractants such as group of cows or heifers aggregate head-In'st into a tight circle carbon dioxide and sweat, and the fact that a resting animal does (I 19,267). In an experimental study one group of heifers received not dislodge flies from foliage, were suggested as reasons why fly spray to reduce face fly feeding around the eyes and head while lying down may attract fewer flies (74). another group was untreated. Correspondingly, the untreated Microhabitat seeking. Animals that are pestered by flies and heifers had 4-5 times more flies around their heads than treated mosquitoes sometimes seek a microhabitat where there are fewer animals which had only 3-4 flies per head. The untreated animals insects. By moving to higher altitudes during summer grazing formed bunches much more frequently and spent about 10 times caribou are able to avoid the intense mosquito activity which exists more time in bunches than treated heifers (250). Bunching at lower altitudes (65). Within a given altitude caribou use upper episodes generally occurred when face fly intensity reached 9-12 slopes and ridge tops to further avoid mosquitoes and flies (65). flies per face. The effectiveness of the bunching behavior was Feral horses are observed to seek windy ridges during the time of indicated by the observation that heifers in the most central area of the day when flies are particularly bad even though forage is less the bunch, with the most protected heads, had only 1-2 flies per abundant (145). head compared with heifers on the periphery who had 15-20 flies Evasive behavior. Flies that evoke a specific avoidance response per head. from cervids of North America are the nasal bot flies that deposit A variant of the bunching behavior is seen in the cattle of larvae in the nostrils of their hosts. The hatched larvae migrate up nomadic Fulani herdsmen, which gather head-first around a smoky the nasal cavity and lodge in sinuses where they then mature. One fire made by the herdsmen, practically sticking their heads in the species of nasal bot fly (Cephenemyia apicata) that attack black- fire (172). This is presumably a means of obtaining relief from tailed deer typically stalk the deer by inaudibly hovering near the mosquitoes and biting flies that become pests at night. Since such nose (4). The evasive behaviors used by deer towards these flies fires probably did not play a role in the evolution of the include ducking the head, pushing the nose to the ground, t'we-seeking behavior, the behavior is undoubtedly learned. sneezing, snorting and scratching the lowered nose with a hind The selfish herd model of Hamilton (101), which proposes that food and then cautiously raising the head (4). Flies of a different grouping offers protection from predation, especially for animals species (C. jellisoni) typically attack the deer after landing on in the center of the aggregation (20), seems to apply to the above foliage and perching near a deer, taking advantage of the deer's observations on grouping and bunching where the most centrally curiosity in orienting to flies. The fly then attacks, and deposits placed animals are the most protected from flying parasites. The larvae. When black-tailed deer see these bot flies land near them several fly control behaviors that have been mentioned, including the evasive reactions are turning or lowering their heads and fly repelling movements, forming groups, microhabitat selection slowing walking away (4). and lying down, represent a variety of behavioral patterns some of Gadding behavior. One of the most dramatic forms of evasive which can be used simultaneously and some of which are mutually behavior taken by ungulates is the gadding behavior most fre- exclusive. The particular behavior displayed probably depends quently seen in cattle to the approach of warble or hypodermal upon the insect species and behavioral predispositions of the host flies that lay eggs on the legs and lower abdomen. Despite the fact species. Although some aspects of the behavior are undoubtedly that the flies have no mouth parts for biting (and not even a learned, because of the immediate reinforcement from reduction of digestive tract), cattle act wild, sometimes running frantically in fly bites or harassment, the behaviors presumably have a genetic apparent attempts to evade the flies (249,258). A classic reference basis maintained by selective pressures because the costs to fitness work on parasitic flies of Britain contains a description by Austen of insect feeding or oviposition exceed the costs of the avoidance (9) of cattle running and bellowing for some distance through the behavior. heath to the nearest ponds, streams or canals to escape from the Nest fumigation for ectoparasites and pathogens. Plants con- warble flies. In the water the cattle were presumably unavailable tain a wide variety of chemicals that are repulsive or poisonous to for warble fly oviposition on their legs and lower abdomen. parasites or herbivores and that act as a defense against animals Interestingly, gadding behavior of cattle in response to attacks by that injure or eat the plants. These substances include oxalic acid, nonbiting warble flies is much more extreme than reactions of cyanide, cyanogenic glycosides, alkaloids, toxic lipids, terpe- cattle to biting flies which can cause immediate pain and loss of noids, saponins, flavonoids, tannins, lignins and insect juvenile blood when they bite. hormone analogs (86,244). Although some plants contain sub- Grouping. Social grouping or clustering can be partially related to stances that can kill birds and mammals, other plants contain avoidance of biting and blood sucking flies. It has been noted in chemicals which are toxic for parasites of vertebrates but harmless domesticated reindeer, for example, that when fly harassment to the vertebrate hosts of the parasites. The existence of volatile becomes intense, reindeer groups become much larger, being substances toxic to arthropods and bacteria in plants make nest formed by the coalescence of smaller groups (122). To test the fumigation possible. notion that reindeer grouping actually lowers fly attacks per The fact that long-term occupancy of nests increases the animal, investigators constructed dark colored fly traps the size buildup of nest-borne parasites and bacterial pathogens (57,132) and shape of reindeer and which emitted carbon dioxide to form a led to the nest fumigation hypothesis advanced by Wimberger small simulated herd (122). The traps were placed singly or in a (293) and Clark and Mason (43). Birds of some species go to small group with 4 traps in the center, 8 as a middle ring and 12 considerable lengths to obtain fresh material for nest construction in an outer ring. The single traps caught more flies than any traps (49). Passerine (43) and falconiform (293) species that reuse nest in the group. Within the group traps of the inner ring were attacked sites across breeding seasons are much more likely to use fresh less frequently than traps of the outer ring. plant material in nests than birds of species which infrequently Horses in groups of 8 to 36 have been shown to attract fewer reuse old nest sites. The European starling which reuses nest sites flies per horse than those grazing singly or in groups of 3 (68). over several seasons was the subject of three studies by Clark and When small groups were experimentally merged the difference in Mason (43-45) in examining the nest fumigation hypothesis. In per capita attractiveness to flies between large and small groups their studies in which nest boxes were provided to starlings, nest disappeared, revealing there were no differences in individual construction began in March and April. The birds placed fresh BEHAVIOR AND PARASITES 281

green vegetation into the nest boxes and would weave green plants In the most detailed analysis with regard to type of parasite which into an existing matrix of old dried grass from the previous year. would increase with group size, it was found in bobwhite quail that Green material was continually added until the eggs hatched. Of the number of parasites per animal consistently increased with 66 plant species available to starlings, 9 were selected more group size for helminthic species that utilize only one host species frequently than would have been expected if selection was and have short life cycles (198). Microparasites traditionally random. Commonly selected plants were agrimony, yarrow, rough involve only one host species and have very short generation goldenrod, wild carrot and fleabane (43). times, and would similarly pose a risk for grouping. A correspon- Chemicals contained in the leaves of the preferred plant species dence between group size in mangabey monkeys and intestinal were more effective in inhibiting the growth of common avian protozoa parasites has been noted (82). However, in a relatively pathogens and in retarding hatching of louse eggs than substances closed or isolated group animals are exposed to each other's from nonpreferred plants (43). Field experiments found that viruses and bacteria, and tend to develop immunity to pathogens removing fresh plant material from the matrix of starling nests indigenous to the group. Individuals that remain within a group are resulted in an increase in mites (45). A laboratory study showed less likely to become infected with alien pathogens than are that when nest matrix material from starling nest boxes was placed individuals that move between groups and animals that move in plastic bags with preferred plants (wild carrot or fleabane) larval between groups are likely to carry some pathogens that are emergence of mites from the nest matrix was delayed (45). infectious to residents of established groups. Plants preferred by starlings were found to contain more Given the increased vulnerability of socially grouping animal volatiles than nonpreferred species, and this finding raised the species one would expect to find behavioral patterns related to possibility that starlings may be cuing off of the volatility of plants protecting animals from macroparasites or microparasites. Be- in selecting those to incorporate in nest construction. Physiological cause of the highly structured social groups, primates offer experiments showed that the birds can discriminate plants based on opportunities for examining group dynamics related to possible volatiles (44), and the behavior of starlings as they search for plant parasite avoidance strategy. Primate groups cannot completely materials is consistent with the idea that they select plants based preclude the transfer of individuals between groups because of the upon volatility. Starlings seemed not to select a leaf immediately detrimental effects of inbreeding, and Freeland (81) has suggested upon entering a patch of vegetation, but spent nonforaging time that the introduction of new individuals into the group, which searching within a patch (44). typically takes weeks or months, has a disease avoidance function. The rather conclusive evidence that the use of fresh plant The newcomer is typically kept at the periphery of the group, and material provides some protection from parasites for birds that it is repeatedly threatened or attacked by dominant group members reuse nests brings up the question of similar fumigation practices and often deprived of food (2,266). This peripheralization of by mammals that reuse burrows or dens over several seasons. strange conspecifics is perhaps the main aspect of group behavior Burrow or den fumigation has not been identified in mammals, but dynamics that could be related to the microparasite avoidance a logical species to examine would be the European badger. Field strategy. Freeland (81) suggests that such peripheralization, to- observations of badgers by Neal (209) reveal that badgers select gether with the behavioral and nutritional stress to which the certain plants as bedding depending upon time of year and whether intruder is often subjected, make it likely that if the intruder is the bedding is for adult sleeping quarters or young cub quarters. harboring a latent infection it will becomevisibly sick; the stranger The collection of bedding material is especially pronounced just may then not be allowed in the group at all or may die. Of course, before parturition and Neal in fact conjectures that for successful there are other adaptive explanations for the peripheralization of rearing of cubs the collection of certain plants, especially hay and strange conspecifics such as protection of limited resources and bracken fern, are necessary. Whether these and other selected exclusion of potential breeding rivals by group males. One plants have the appropriate biocidal or fumigant activity remains to adaptive function does not exclude others and at this stage it is not be seen, but bracken fern is known for its wide array of biocidal possible or necessary to select among the alternative explanations. substances including cyanogenic glycosides (51) and the insect Sexual behavior and avoidance of sexually transmitted dis- juvenile hormone analog, alpha ecdysone (93,144). Interestingly, eases. Another area where pathogen avoidance behavior may be temperate zone ferns, including bracken, are relatively unattrac- important is in sexual interactions. Females may not only acquire tive to insect pests (72, 162, 163). respiratory or enteric diseases from males, by virtue of physical contact with them, but may also acquire venereal diseases that Avoidance of Microparasites impair fertility, induce abortion, or cause malformations in their young. Venereal diseases may impair fertility in males as well. Probably because it is difficult to observe and count micropar- Rats and other rodents would appear to be particularly susceptible asites such as viruses and bacteria, avoidance strategies directed to the sexual transmission of genital diseases because they are toward microparasites are not as widely acknowledged as those promiscuous breeders, and they engage in multiple copulatory related to macroparasites. Also the immune system is particularly intromissions prior to ejaculation (248). There is evidence that important in controlling microparasites. Avoidance behavior would postcopulatory genital grooming in rodents acts to protect males be most logically directed towards microparasites to which the from contracting venereal diseases from females and becoming animal has not acquired immunity. A number of studies provide sterile themselves or passing such infections onto females they experimental or field evidence for a variety of behavioral patterns subsequently mate. Extensive genital grooming by male rats related to the control of exposure to microparasites. Some behav- invariably follows copulatory intromissions (61). When collars ioral patterns discussed above such as nest fumigation, or avoid- were placed on male rats to prevent postcopulatory grooming and ance of forage contaminated by feces, serve to reduce an animal's the male rats mated to females in which a genital infection had exposure to microparasites as well as macroparasites. been established with a marker organism (Staphylococcus aureus), Social and territorial behavior as protection from alien patho- they were much more likely to contract an infection with the gens. One of the costs of social grouping is increased transmission marker organism in one or more genital organs than male rats of macroparasites and infectious diseases (1,266). Evidence for wearing no collar or short collars which allowed grooming (112). this statement comes mostly from the study of the prevalence of It was also found that rat saliva was effective in killing two macroparasitic infestations as a function of colony or group size in pathogens implicated in murine genital infections (Mycoplasma prairie dogs (129), cliff swallows (33) and bobwhite quail (198). pulmonis and Pasteurella pneumotropica) (112). Since rat saliva 282 HART

was only slightly antibacterial against the marker organism, the for rats (105,191) and dogs and cats (95, 143,282). The intestinal protective effect of genital grooming in the above experiment was epithelium of neonates is permeable to bacteria for the first 48-72 attributed to the physical cleansing action of the tongue on the hours after birth, and in neonatal dogs E. coli and Streptococcus penis and urethra. The flaring of the penis glans area, forming a canis can be involved in septicemia (95,143). The finding of cup-like structure during postcopulatory genital grooming (111), bactericidal effects of saliva against E. coli and S. canis (113) is probably acts to expose the urethral epithelium to the cleansing supportive of the concept that maternal peripartufient licking of the action of the grooming behavior. The antibacterial effect of saliva mammary and anogenital areas is adaptive in protecting the against specific genital pathogens is added insurance against a newborn from excessive exposure to these potential pathogens. It genital infection. Although the particular antibacterial agents of would appear that the most critical time for a mother to lick the murine saliva have not been determined, likely candidates are anogenital and mammary areas is just prior to parturition and lysozyme, lactoferrin, leukocytes, lactoperoxidase, immunoglo- immediately after birth since that is when the newborn gut is most bins, cationic proteins and substances which interfere with the vulnerable and the newborn have not received any protective ability of bacteria to adhere or attach to soft tissues, all of which colostrum. Interestingly, in rats it is in the last few days before have been found in human saliva (29, 179, 180). The protective parturition that anogenital and mammary area ticking peaks (246). effects of genital grooming may be important in controlling the Rat pups will not attach to nipples that have been experimentally spread of genital diseases in rodents, canids and felids, all of washed, hut attachment can be induced when maternal saliva is which engage in postcopulatory genital grooming (108). applied to the nipples (23). The reluctance of newborn rats to Licking of wounds and promotion of healing. It is not uncom- attach to nipples to which maternal saliva has not been applied mon for animals to be wounded in fights with conspecifics, in may be thought of as a fail-safe mechanism to assure that infants escape from predators, or in confronting prey. The body is do not place their mouths on contaminated body parts. equipped for dealing with bacterial contamination of wounds in a Infant coprophagy and protection from coliform enteritis. In number of ways. The inflammatory process, with the increase in addition to just after birth, young animals are also susceptible to blood flow and effusion of fluids and white blood cells into the intestinal diseases just as they are weaned and start taking solid affected area, represents an immediate response for dealing with food. The introduction of solid food brings with it exposure to possible pathogenic invaders. Before an animal's inflammatory common environmental bacteria, particularly the ubiquitous E. process becomes active in the wounded area, it is common to see coli, some strains of which, as mentioned above, cause colibacil- animals, such as primates, canids, felids and rodents, lick the losis. This disease in weanling rats is often brought on by stress. wounded region. Undoubtedly, wound licking serves a cleansing Adult rats receive protection from E. coli enteritis from deoxy- function for common wound contaminants. As an indication of the cholic acid, a bile acid (191), but weanling rodents do not yet value of the bactericidal effects of saliva in wound hygiene, canine secrete deoxycholic acid. However, a pheromone-behavioral sys- saliva was found to be bactericidal to Escherichia coli and tem provides rat pups with protection from colibacillosis by Streptococcus canis (113). In both wild and domestic canids these leading to their consumption of maternal feces which are espe- would be potential wound contaminants and these bacteria were cially rich in deoxycholic acid (191). Rat pups are attracted to isolated from 10-20 percent of dog wounds (113). Saliva was consume maternal feces by a pheromone carried in the mother's found to not be bactericidal to staphylococcus and this organism feces 14-27 days after the onset of lactation which is when rat was isolated from 46 percent of wounds (113). The resistance of pups start to take solid food. At about day 28, when pheromone staphylococcus to antibacterial effects of saliva is reflected in the production drops and attraction to maternal feces wanes, the rat fact that the most common organism in superficial or deep pup's secretion of deoxycholic acid reaches adult levels (149). cutaneous pyoderma of dogs is staphylococcus (21 I). Several studies have shown that deoxycholic acid can protect The saliva of animals also contains epithelial and nerve growth young mammals, including infant rodents, from coliform enteritis factors (56, 204, 224, 259) which would help promote the closure (19,62). One experiment showed that cold stressed rat pups of wouhds when animals lick their wounds. Evidence that these prevented from consuming maternal feces suffer a much higher growth factors play a role in wound healing comes from findings mortality from enteritis than similarly treated rat pups given that removal of salivary glands in mice retards wound healing deoxycholic acid (192). It is not known whether the young of other (137) and topical application of epithelial growth factor (212) or mammalian species may derive disease protection from consump- nerve growth factor (167) to the wounds of animals, in which the tion of maternal feces. Horse foals discriminate and eat the feces salivary glands have been removed, enhances the closure of of their own mothers in the first 20 weeks after birth, possibly in wounds. response to a pheromone and thus may receive a protective Protection of young animals from disease. Young animals with substance (55). undeveloped immune systems are especially vulnerable to infec- Infanticide and cannibalism as disease protection. Despite the fact tious diseases. For those mothers that invest heavily in the care of that mothers will invest a great deal of their own energy, time and their young, we can expect some disease control behavioral resources in their young, they do, on occasion, kill their young. patterns to be directed towards protecting the young from exposure Infanticide and cannibalism of their young can be viewed as an to disease-producing doses of microparasites. aberrant behavior that is maladaptive. However, there is a disease Licking of mammary region. Licking of the mammary and control explanation of some instances of infanticide that would anogenital areas just prior to parturition is evident in rodents (243), seem to merit some attention despite the lack of experimental cats (252) and dogs (23,239) among other mammals. Newborn evidence. Almost all of the attention directed towards understand- mammals not only have an undeveloped immune system, but they ing infant killing has focused upon that carded out by a conspecific are born with a sterile gut and therefore lack the intestinal bacterial adult that is not the mother of the ones being killed (120). Much flora that is protective against opportunistic pathogens (95). These discussion in the literature over recent years has pointed out that newborn animals are at risk to some of the more virulent strains of this type of infanticide can be viewed as adaptive (120,131). E. coli and other potential pathogens commonly found in feces if Infanticide by the mother of the victims has been viewed as exposed to these microorganisms prior to ingestion of protective adaptive in female hamsters that cannibalize their young to reduce colostrum, as they would be during the birth process or in litter size in accordance with prevailing environmental conditions attaching to nipples prior to suckling. Severe enteritis with high and food availability (59). It is often the smaller weaker pups that mortality, called colibaciUosis, is caused by E. coli and is reported are killed. Other proposed explanations of cannibalism relate the BEHAVIOR AND PARASITES 283

behavior to inexperience of the mother and disturbing environ- endotoxins or other toxic substances. Aversions can apparently be mental stimuli that make the mother nervous (110). produced even to food items associated with the visceral response Other than the above explanation, cannibalism seen in rodents of a heavy dose of parasites as was reported in rats with the and carnivores (110) could function to protect littermates from an subcutaneous injection of nematode larvae (148). infectious enteric or respiratory disease if one member of the litter Herbal medicine, animal style. Throughout human history plants is becoming sick. Although neonatal animals are mostly resistant have been used for health-related purposes with a greater or lesser to opportunistic pathogens by virtue of maternal antibodies passed degree of effectiveness in the practice of herbal medicine. Since through colostrum, the occasional animal may become sick with some plants have been shown to have antibiotic and other biocidal an environmental pathogen if it does not receive sufficient colos- properties (see discussion under the Nest Fumigation section), it is trum or suffers from insufficient genetically acquired resistance. In not surprising to find instances where it appears as though animals this case the neonate could act as a reservoir for multiplication of have acquired the behavior of ingesting some plants as a defense the pathogen in sufficient numbers to constitute a risk to the rest of against, or treatment for, intestinal disturbances. the litter. If a mother, by reacting to the early signs of a sick infant There have been three separate reports by experienced inves- such as inactivity and hypothermia, disposes of the sick infant, the tigators of wild female chimpanzees ingesting plants which were rest of the litter may be saved. For this protective system to work not a part of the regular diet, and in which there is evidence the the mother's cannibalistic behavior would have to be triggered by plants have medicinal properties. In all instances the plants were slight abnormalities so that the sick one was disposed of before it chewed and ingested in a manner different than the way plants are became a reservoir of the pathogen to overwhelm the resistance of normally eaten. The most recent observations were by Huffman the littermates. Of course, by consuming the dead infant rather and Seifu (133) on a female and her 2.5-year-old male offspring than just depositing it outside the nest, the mother is able to add to who were subjects of focal animal observations at the time when her own nutritional reserves. There is no direct experimental the female showed signs of illness. On the two days (afternoon of evidence in support of this view of maternal cannibalism, but day 1, morning of day 2) that the female was observed ill she spent veterinarians, laboratory investigators and breeders seem to have over twice as much time in a day bed and in intermittent rest as her long recognized that it is the sickly or deformed infants that are companions and about half as much time foraging. She appeared most likely to be cannibalized (105,110). A way of testing the to be in pain when defecating and defecated in small amounts. On disease control hypothesis for infanticide would be to examine the the first day of her observed illness the female went to a shrub of behavior of mother rodents or carnivores towards members of a Vernonia amygdalina, peeled off the outer bark and chewed and litter that are experimentally made to appear ill. Opportunistic sucked on the pith for about 20 min. In the afternoon of day 2 of field observations could provide relevant evidence as well. the illness the female appeared to be recovering and foraged and Avoidance behavior related to food intake. Localized disease rested at a rate typical of the healthy chimpanzees. The Vernonia of the intestinal tract, as well as systemic diseases, may be plant eaten is very bitter and is claimed to have medicinal acquired through pathogens or toxic products that are ingested properties (38,283). It is used throughout tropical Africa against along with food. One can point to at least two behavioral parasites and gastrointestinal disorders of people and domesticated mechanisms that have been described in which animals avoid such animals. diseases. These are the cannibalism taboo and conditioned food In another report by Takasaki and Hunt (263), a female aversions both of which are briefly described below. Animals chimpanzee was judged to be sick by virtue of the fact that she left could also avoid or suppress intestinal bacterial disorders by her bed about a half hour later than usual. She then walked about consuming plants with medicinal properties as a type of herbal 100 m and ate leaves from a Lippia species shrub for about 15 min. medicine. This possibility is also discussed. The method of eating the leaves was atypical in that the mouth was Cannibalism taboo. Some carnivores and onmivores, including opened and closed on the leaves several times and the animal rats, show a reluctance to feed on carcasses of dead conspecifics appeared to suck on each leaf for a few seconds. The leaves were even though they may be very hungry and even though conspecific rolled in the mouth and apparently swallowed without being flesh would make an ideal food source since it contains the same chewed. After eating the leaves she sat still for 6 min then moved substances as the eater (80). Hungry rats that refuse to eat another off to a tree, made a bed and remained still for 20 min until a dead conspecific will readily feed on the carcasses of other rats companion came near, whereupon they groomed each other before when the skin is removed revealing that the aversion is to the smell moving off. or taste of the skin or fur (42). The adaptive explanation of the The above 2 reports dealt with observations of single chimpan- cannibalism taboo offered here is that an animal that has not been zees that were apparently ill. Preceding these studies and setting killed in a fight or by a predator, but is found dead, has some the stage of recognition of the practice of herbal medicine was the likelihood of having died of a disease that would be infectious to report of a peculiar type of foraging behavior in champanzees by a conspecific eating the carcass. An apparent contradiction to this Wrangham and Nishida (295). This study dealt with several viewpoint is that adult carnivores, namely lions and hyenas, have individuals in two different groups at different study sites. The been observed to occasionally kill and eat a conspecific (221). plants eaten at these study sites were of Aspilia species. As with However, a conspeciflc that is killed by an animal is not so likely the Lippia plant, Aspilia leaves were taken into the mouth singly, to be ill with an infectious disease as is one that had died a rolled around and swallowed without being chewed. The leaves nontraumatic death. were eaten early in the morning after the chimpanzees had often Conditioned food aversions. If a toxic substance or bacterial walked 20 rain or more to the source of the leaves. No signs of endotoxin consumed with a particular food produces gastrointes- illness were observed in these animals. The feeding rate on Aspilia final illness, the toxin is likely to evoke an aversion to the was very low and could not have provided sufficient caloric value foodstuff with which the toxin was associated (87). In mammals to merit the time taken to eat them given that ripe fruits were such aversions are produced to the taste or smell of the food. readily available the same day. A total of 21 individuals at one Conditioned aversions are most easily produced to novel foods. In study site and 12 at another study site were observed to eat Aspilia the laboratory conditioned food aversions have usually been species leaves in this fashion described. Both Aspilia and Lippia produced by lacing food with lithium chloride or apomorphine are used by local African people for medicinal purposes (11,283) (87). In nature conditioned food aversions would tend to protect and have been found in laboratory or field tests to have biocidal or animals from repeatedly ingesting food items that contain bacterial antibiotic properties (242,263). 284 HART

STRATEGY 2: CONTROLLED EXPOSURE its social group. The practice among several primate species of keeping a stranger at a distance before it is allowed into the group The immunological system is exceedingly complex in animals was discussed above with regard to avoidance of microparasites. and enables them to resist disease-causing microparasites and even This peripheralization behavior also provides a means by which some macroparasitic infections. The specific antibodies devel- the residents may acquire low-level exposure to alien pathogens oped depend upon an exposure to the pathogenic organisms, but permitting the development of immunity before the newcomer is the number of microorganisms needed to produce immunity can be assembled into the group and residents are exposed to large very small. Early in life, before a newborn animal produces its numbers of the newcomer's pathogens. A test of the function of own antibodies, it is protected from pathogens in the environment peripheralization for microparasite control would be to monitor the by maternal antibodies made available to it in mother's milk. The appearance of new intestinal microparasites such as protozoa antibody-rich milk, called colostrum, is secreted only during the among residents with the appearance of strange conspecifics that first few days of lactation. As maternal antibodies gradually are initially peripheralized. disappear, the young animal develops its own self-sustaining antibodies through exposure to pathogenic organisms in doses sufficient to evoke antibody production without causing disease STRATEGY 3: BEHAVIOR OF SICK ANIMALS (39). The antibodies for some microorganisms require antigenic exposure every few months while others require exposure only Despite the employment by animals of rather complex behav- once in a lifetime. An animal may also develop immunity by iors to avoid or control their exposure to macroparasites and becoming sick with an infectious disease and recovering. In the era microparasites and the protection offered by a highly complex of modern human and animal medicine, we think of vaccines as immune system, animals do become ill. Without the advantage of providing the basis for immunity to common diseases, but long antibiotic therapy and medical care many sick wild animals do before the advent of vaccination programs, the immune system of indeed die of disease or succumb to predation or starvation after animals and people was sufficiently activated through natural being weakened by disease or parasites. However, it is clear that means to deal with pathogens in the environment. We might animals can survive acute illnesses and they have a variety of expect, therefore, to find behavioral patterns that play a role in the physiological processes to promote recovery from sickness or controlled exposure of animals, or their young, to small doses of injury. The processes for wound-healing, tendon and bone repair, pathogens common in the environment so as to potentiate the and regeneration of damaged organ tissue are evidence that the development of immunological competence. While experiments or body is organized to repair itself. Before reparative or regenerative studies focused specifically on this strategy have not been re- processes can be of value, animals must sometimes recover from ported, some behavioral patterns may be reinterpreted in the light an infectious disease. This section deals with the response of of a controlled exposure strategy. animals to acute illness. Animals that are acutely sick often behave differently than Exposure of Infants to Antigens healthy animals. They are depressed, lethargic, and show no interest in eating. They fail to exhibit body care. They may stop Most exposure of young animals to immunizing doses of grooming, and develop rough coats. These behavioral patterns are antigens probably occurs in the natural process of daily activities not necessarily maladaptive but would appear to comprise an as animals encounter small doses of viruses and bacteria from adaptive behavioral mode that facilitates recovery from illness. saliva, respiratory droplets, nasal secretions, feces and urine with The evidence for viewing the behavior of sick animals as an which conspecifics contaminate food stuffs, water, and body evolved behavioral strategy to facilitate survival in the face of areas. One way that mother carnivores can expose their infants to infectious disease has been recently reviewed (109) and will only low doses of potential pathogens and parasites is by bringing back be discussed here in sufficient detail to complete the overview of kills, which the young chew or eat in small quantities. Thus, while the five strategies of control of pathogens and parasites. serving to introduce the young to prey food as a prelude to hunting, the mother may be exposing the young to immunizing The Fever Response and Recovery From Illness doses of environmental pathogens which have contaminated the carcass. The behavior of sick animals can be related to potentiation of Another possible example mentioned by Freeland (81) as the fever response. The fever response of sick animals involves a playing a role in controlled exposure to pathogens is seen in some rise in temperature caused by the release from monocytes and primates where group members are allowed or encouraged by a macrophages of endogenous pyrogens which act on the hypothal- mother to handle her infant. This handling by group members will amus to increase body temperature. The principal endogenous expose the infants to a somewhat greater variety of group-specific pyrogen appears to be interleukin-1, and the increase in body microorganisms more rapidly than if the mother does not allow temperature is due to raising the hypothalamic thermal set point. such handling. According to Freeland this is more characteristic of Endogenous pyrogens not only cause increased temperature, but precocial primates, where infants are weaned early and achieve induce lower blood concentration levels of iron and zinc. independence at an early age, than in infants of more altricial Increased body temperature potentiates immunological re- species in which there is a longer period of dependence. In the sponses by enhancing and accelerating lymphocyte proliferative latter instance, Freeland argues, it would be more advantageous if responses to antigens, increasing bacterial killing by neutrophils the young were minimally handled and experienced a more and enhancing antibody synthesis (12,138). There is also a direct gradual introduction to group pathogens since the more rapid suppression on growth of some disease-causing organisms, docu- exposure to pathogens carries some risk of producing disease mented in both in vitro and in vivo studies (6, 135, 154, 157). The rather than just provoking an immune response. reduction in iron plasma concentrations associated with the fever response occurs when free iron is sequestered away in the liver and Exposure of Adults to Antigens spleen. Iron is an essential element for bacteria and they multiply by chelating free iron in the blood (37,285). The reduction in iron, The immune system is dynamic and changes as an animal accompanied by increased temperature, is particularly inhibitive to enters new environments or with the flux of animals in and out of the growth of some bacteria (150,156). BEHAVIOR AND PARASITES 285

Much has been learned about the value of an elevated body interaction of interleukin-1 and cell wall components on sleep temperature in fighting disease (152-154). There is a pronounced control mechanisms (272). In the only study of prolonged admin- decrease in the survival rate of lizards subjected to an infection istration of interleukin-1, continuous infusion over 5 days of the when they are kept in a moderate environment versus a wanner peptide in rats resulted in reduced general locomotor and rearing one that allows the lizards to develop a fever (155). Rabbits in activity during the active phase of their diurnal cycle (219). which a drug is administered to suppress fever are more likely to Animals, including rodents, felids and ungulates, which have die of pasteurellosis than control subjects allowed to develop a been sick for several days often have a scruffy, dirty hair coat. fever (157). This is probably due to a reduction in grooming, since grooming When the thermal set point is elevated animals raise body cleans the fur of dirt and oils. However, the cleaner the pelage the temperature by physiological means such as peripheral vasocon- better its insulating functions, and the more grooming the fewer striction and behaviorally by seeking warm places. When these the ectoparasites, so one might ask why a sick animal would forgo responses are not sufficient, body temperature can be further grooming under the demands of a fever reaction, while letting the increased by shivering. However, raising body temperature is insulating benefit and parasite control of the pelage slip. Grooming metabolically expensive costing an estimated 25 percent increase requires energy for the muscular activity involved. For oral in metabolism for each 2-3 degree C increase in body temperature grooming there is a loss of water through saliva. Thus, for a febrile (66,124). The increase in energy expenditure is a function of the animal in the wild that has restricted access to water, to reduce oral direct effect of increased temperature in accelerating enzymatic grooming will conserve water and energy. It has not been reactions plus the increased metabolism, such as that involved in determined whether the reduction of grooming in sick animals is a shivering, that may be necessary to raise body temperature. It is result of the effects of interleukin-1 or a different endogenous the high cost of the fever response, especially if shivering must be pyrogen, hyperthermia, or the sleepiness and inactivity that activated, plus the advantage of lowering the blood concentration accompanies illness. of free iron, that appears to be related to the behavior associated In a review of the animal and human medical literature it was with sick animals. found that acute illnesses characterized by fever were almost invariably accompanied by anorexia and depression (109). Only one disease entity was found in which fever is noteworthy but Sick Animals Are Anorexic, Depressed, and Lethargic which is not typically characterized by anorexia or depression; this Although the occurrence of anorexia seems paradoxical when disease is African horse sickness, a tropical viral disease highly the body needs energy to produce its fever response, if the sick fatal to equids. With this disease appetite may remain despite a animal does not feel hungry, it has no motivation to move about fever of 2-3°C (88,182). This exception to the "rule" that fever is using energy searching for food and it lessens its chance of raising accompanied by anorexia and depression suggests that the behav- its blood concentration of free iron from consuming food that ioral patterns of anorexia and depression may not always be linked might have iron. By staying in one place it reduces body heat loss to fever in environments such as the tropics where animals can that would otherwise result from body surface exposure. The sometimes easily obtain an increase in body temperature. physiological link between the fever response and the induction of The behavior of being sick is counter-productive if carried on anorexia appears to be through interleukin-1 and administration of for more than a few days. When a wild animal has a fever reaction this peptide markedly reduces food intake (173). Force feeding of accompanied by depression and anorexia, it is at a life or death mice during bacterial infections can reduce survival time and juncture and this extreme mode of behavior may be the only increase mortality (207), while food deprivation two to three days chance it has to survive. Once the disease is overcome, fever before giving a bacterial infection (listeriosis) can increase sur- subsides, depression disappears, and the animal recovers a strong vival rates (294). appetite. Sick animals are commonly characterized as depressed, inac- Since the occurrence of fever is energetically very costly, and tive, lethargic, and disinterested in their surroundings. A common is widespread throughout representatives of the animal kingdom element in these characterizations is increased sleepiness. The involving mammals, birds, reptiles, amphibians, fish and arthro- sleepy, depressed or inactive animal is less likely to move about pods (25, 152-154), Kluger (153,154) has argued that fever must and use up energy. Some animals, such as felids, canids, and have evolved as an adaptive strategy in animals as a way of rodents curl up and insulate themselves to conserve body heat, fighting infectious diseases. The same line of reasoning could be making shivering less necessary. Excessive sleepiness and inac- applied to the behavioral patterns of anorexia and depression that tivity is, therefore, complementary to anorexia in inducing an accompany fever. However, fever, and the accompanying behav- energy conservation mode. The main physiological element un- ioral mode, would not necessarily fit all infections or be successful derlying the appearance of depression in sick animals may be all of the time. In fact, prolonged hyperthermia can be associated enhanced slow-wave sleep. Infection of rabbits with S. aureus with neuronal, hepatic or cardiac tissue damage (75,94). Even so, enhanced slow-wave sleep shortly after injection while treatment Ewaid (94) notes, a high fever would be adaptive if, on average, with an antibiotic reduced the sleep-inducing effects of the bacteria the net benefit from the high fever was greater than the net cost of (207). In further experiments by Toth and Krueger (271) three some tissue damage (94). The value of fever as a rule for host different microbial agents were given to rabbits. Streptococcus defense against pathogens does not mean that in all instances it pyogenes and Candida albicans produced marked increases in will aid the host's resistance to infection. The fever response may slow-wave sleep for 20 hours after inoculation whereas E. coli be invaded by pathogens whose survival or reproduction is produced the enhancement only for 4 hours after inoculation. As in resistant to or even enhanced by fever (25,94). anorexia there is evidence that interleukin-1 is the physiological link between fever and the behavioral change. In rabbits (161,278) STRATEGY 4: HELPING SICK OR INJURED ANIMALS and rats (268) cerebral intraventricular administration of interleu- kin-1 causes prolonged slow-wave sleep. However, muramyl Although animals have the immunological, physiological and peptide chains and lipopolysaccharide components of cell walls behavioral processes to recover from systemic illness or injury, the can also enchance slow wave sleep (159,160), so the enhancement reactions of other conspecific group mates to the sick one must of sleep during microbial infections would appear to be due to an also be considered. For solitary animals their weakened condition 286 HART

may force them to give up a territory to a conspecific that takes members accompanied the injured animal to the nearest termite advantage of the sick one's condition. For ungulates that live in mound, 15 miles away, and remained inside. The next day the loosely organized herds the behavior displayed towards a sick injured animal emerged from the mound hobbling on three legs. conspecific will probably be that of indifference or avoidance. Group members frequently surrounded and groomed the injured However, for animals that typically live in integrated social groups animal. When the day's foraging began it was slow as the animals of related animals or males and females in pair-bonded situations, moved to a new mound. For the next three days instead of moving there can be some advantage to the healthy animals, as well as for from mound to mound, the mongooses returned to the same the sick animals, when the healthy ones assist sick or injured mound each night. It appeared that the injured animal was not conspecifics in recovery. Since animals probably do not differen- capable of catching its own prey and was receiving at least some tiate highly infectious from noninfectious conditions, they can be food from the group members. On one occasion, the injured expected to display the same behavior whether the group mate has mongoose was allowed to take a grasshopper from the alpha an infectious illness, a noninfectious disease or an injury. female and eat it in front of her. By day 6, the injured animal was Primate species, in which females are closely related, are one able to put weight on its damaged leg and at this point the group of the animal types where one would expect to see helping of sick resumed their activity of moving from mound to mound twice a group mates. In the observations of the ill female chimpanzee by day. In both the captive and wild instances the alpha pair was most Huffman and Seifu (133) discussed previously under the topic of active in caring for the sick or injured animals. As noted it is the herbal medicine, other female chimpanzees of the group were alpha pair that stands to gain the most by having group members noted to be aware of the sick one's illness and adjust their travel around to help care for their offspring. pace so that the ill female and her young son could keep up. At Foxes are a species where males and females form pairs or other times the same group of females watched over the son of the small groups of related females. MacDonald (177) has conducted sick female, intervening in fights and following the son when he extensive observations of foxes in the field and has recorded two strayed away from his sleeping mother. The predisposition of instances of provisioning sick foxes by group mates. In one primates to care for sick group members is perhaps also illustrated instance an injured female was provisioned by a sister and in in a report by Cowgill (52) of a small group of captive prosimians another instance a male fox provisioned a sick mate. which apparently saved food for an absent member even when the food supply became limited. STRATEGY 5: SEXUALSELECTION Care of the sick or injured is not restricted to primates and Rasa (230,232) has recorded four instances of care-giving behavior in The four disease control strategies discussed above illustrate dwarf mongooses. Members of mongoose groups are closely how an animal can promote its own health or that of relatives and related; groups are matriarchal and both males and females remain group members. Utilization of these behavioral strategies in with the groups well past sexual maturity. The alpha male and conjunction with a highly complex immunological system can female are the only group members to reproduce and raise young affect an animal's personal and inclusive fitness and thus be successfully (231). Other group members act as babysitters and favored by natural selection. Assuming that these parasite and guards and hence aid in the survival of the alpha pair's young. It pathogen control mechanisms are heritable, females (or males if appears important to maintain a critical group size since there is a appropriate) should choose mates who exhibit signs that they are negative correlation between number of young lost to predators endowed with genes for immunological, physiological and behav- and number of adults in the groups. Almost all attacks by ioral resistance to microparasites and macroparasites. Choosing terrestrial predators are on groups of five or fewer adults (232). mates that appear relatively healthy and free of disease is a form of Mongooses eat mostly insects, therefore, cooperation is not the good genes model for adaptive mate selection (158). necessary for prey capture. One set of observations involved a In addition to choosing mates with genes for resistance to captive group (230). A low-ranking 4-year-old male acquired pathogens and parasites, females can enhance their own fitness chronic nephritis and remained as an invalid in the group until it directly by avoiding contacting or sharing nests with males that are died 38 days later. The sick one was allowed to eat simultaneously carrying ectoparasites or communicable microparasites in the form with the alpha male without the alpha male showing any signs of of respiratory infections, or sexually transmissible diseases (Strat- the aggression normally evoked by such transgression. Members egy 1). M¢ller (188) found, for example, that the number of mites of the group spent much more time than usual resting with the sick in a barn swallow nest is related to the number of mites on the mongoose. The alpha female spent a great deal of time grooming individuals that come to occupy the nest rather than the number of the sick animal, especially when the sick mongoose's self- mites in the nest prior to occupancy. Thus, when females choose grooming was markedly reduced. On two other occasions similar males with fewer ectoparasites they are reducing the likelihood of reactions of healthy members of a group towards sick members acquiring parasites themselves as well as infesting offspring with were observed (230). ectoparasites. The parasite avoidance function of mate selection is In a field study Rasa had an opportunity to observe the behavior favored by Borgia and Collis (28) as an explanation for their of wild mongooses towards an injured member of their group finding the male satin bowerbirds with the least lice tended to (232) and a description of her observations seems instructive from achieve the most matings. Diseases not transferable between the standpoint of providing an example of how field studies can conspecifics but that impair a male's fertility would also represent contribute to the understanding and documentation of helping a threat to a female's fitness. In sage grouse, for example, malaria behavior. Typically mongooses spend the night in a termite can cause decrements in sperm motility or viability (30), making mound, leave early the next day and forage until about noon where it likely that a female may not have all her eggs hatched. At the they then enter a different mound for a siesta. After foraging again present time it will often be difficult to separate the proximate in the afternoon they travel to still another mound for the night. On benefits to an animal of avoiding parasited or diseased mates from one occasion a subordinate male of a group of five adults and two the more ultimate benefit of providing offspring with genes for juveniles was apparently attacked by a snake and the skin over the resistance to parasites. Quite likely both functions will go to- lower abdomen and inner side of the left hind leg was torn off. All gether, and animals will achieve both proximate and ultimate BEHAVIOR AND PARASITES 287

benefits when mate selection involves parasites that are transmis- parasite load than those without territories (139) and are less likely sible directly between the partners or that impair the fertility of a to be preyed upon (140). Sage grouse that appear on leks (and have partner. access to females) are less likely to be infected with malaria than those that fail to appear (141). The effects of parasites in Selection of Males With Good Genes for Parasite Resistance influencing a male's success in competing with other males for territory or females, as well as influencing a male's success in Among the cues that would indicate a mate's state of parasitism attracting females, be it for parasite avoidance or genetic benefits, or disease are general vigor, strength and visual or olfactory may undoubtedly all occur simultaneously since the influences are markers of parasites. Hair coat or plumage condition may indicate not mutually exclusive. degree of grooming behavior and freedom from ectoparasites. When an animal does not groom its haircoat it becomes recogniz- Ornamental Characteristics and Showy Displays ably scruffy. Anemia, stemming from a high parasite load, may be reflected in the color of skin in animals where bare patches are A theoretical approach highlighted by Hamilton and Zuk (104) visible. Intestinal disease may produce visual and olfactory signs in their attention-getting paper on sexual selection, is that the of diarrhea. Respiratory conditions are accompanied by visual and ornamental secondary sexual characteristics or showy displays by auditory signs such as sneezing, coughing and nasal or ocular which females often select males can be indicators of a male's discharge. By-products of metabolism indicating a catabolic state relative freedom from, or resistance to, microparasites and/or characteristic of some diseases are found in urine and may be macroparasites. Thus, in avian species which are characterized by evident from olfactory investigation. Olfactory cues from the bright plumage, or the presence of long ornamental feathers, the genital area, representing chemical by-products of bacterial growth brightness or feather length could reflect the birds' relative or inflammatory processes in the genital organs, could be dis- freedom from parasites and a dull color or shorter feathers would cerned by animals as they engage in precopulatory genital inves- be associated with a heavy parasite burden. The Hamilton-Zuk tigation. Hamilton and Zuk (104) suggest that the signs a female hypothesis addresses the dilemma that while genetic theory pre- might use to choose mates should have much in common with the dicts that any polymorphism for a selected trait will eventually end signs a physician would utilize checking eligibility for life insur- with zero heritability, with no particular male any better for "good ance in which the choosing animal should unclothe the subject, genes" than any other male, mate selection continues to be rather weigh, listen, observe vital capacity, and take blood, urine, and obvious. The concept of sexual selection being based upon fecal samples. selection for genes for disease resistance offers a way out of this Consistent with the above concept are studies on birds and fish dilemma since both the host and parasite are coevolving, ensuring showing that females prefer to mate with males that are the most a continual source of fitness variation in gene types. As hosts free of parasites. The inverse correlation between number of lice adopt defense mechanisms to minimize infection, parasites adopt and number of matings recorded for male satin bowerbirds by new modes of infection to which the hosts must then adapt, and so Borgia and Collis (28) was mentioned above. In another field on. An exact equilibrium is never reached and mate selection study dealing with sage grouse, Johnson and Boyce (141) ob- continues to be adaptive. Mathematical models of host-parasite served that males with noticeable infestations of lice or malaria interactions indicate that there can be, in fact, realistic cycles of secured fewer copulations than those free of the parasites. Simi- coadaptation between host and parasite ~,hich never reach equi- larly, Clayton (46) found that male rock doves infested with lice librium (102,103). were chosen less often by females than clean males. Male red The Hamilton-Zuk hypothesis leads to both intraspecific and jungle fowl that were experimentally infected with intestinal interspeciflc predictions regarding ornamental characteristics or nematodes were chosen by females only about half as much as showy displays. The intraspecific prediction is that when females clean males (297). Guppies that were infested with a gut-dwelling distinguish males on the basis of a greater development of a nematode were chosen by females less often than those without the secondary sexual ornament or showy display the ornamental parasite (147). characteristic will reveal information about the current or past When several males are available, as in large troops of degree of parasitism in the potential male. This intraspecific primates, the opportunity exists for females to select males with prediction has received some support. In sage grouse air sacs are the highest genetic endowment for disease resistance. In fact, there ornaments that are used in courtship displays on the lek. Louse is the interesting possibility mentioned by Freeland (81) that a infestation causes hematomas on the air sacs. In an experimental female can induce a state of behavioral or nutritional stress among study males with artificial hematomas applied to air sacs were a group of males through encouraging competition for her atten- chosen less often by females than control males with clean air sacs tion. This stress might then expose those that are on the verge of (141). In experiments on red jungle fowl roosters infected with an becoming sick or that harbor excessive parasite loads. In solitary intestinal nematode had duller and shorter combs, duller eyes, species females may employ the same disease resistance selection shorter tail feathers, and paler hackle feathers than control unin- strategy by attracting several males to her before she becomes fected roosters (297). Body weight and tarsus length in adults, sexually receptive. however, did not reflect the history of infestation. Mate choice A confounding issue to mate selection for freedom from experiments revealed that when females preferred unparasitized disease (whether for parasite avoidance or genetic benefit to males over parasitized males they appeared to be using differences offspring) is the degree to which parasitism or disease determine in the secondary sexual ornaments (297). how a male fares in competing for territory and/or for females and In the barn swallows long tail feathers are a sexually dimorphic in escaping predation. In satin bowerbirds, for example, males that trait that is used as a cue by females when choosing mates (189). hold bowers are more likely to be free of lice than males that do Mlaller (187) found that tall length reflects the history of mite not hold bowers and only males that hold bowers have access to parasitism in male swallows inasmuch as the increase in tail length females (26). Red grouse that are able to establish territories (and from one year to another was negatively related to experimentally have access to females) are found to have a lower gastrointestinal manipulated mite numbers in nests during the preceding season. 288 HART

Furthermore, that resistance to mites is heritable was shown by teristics or sexual displays, in selecting mates. It does not cross-fostering experiments in which mite loads of swallows at necessarily, however, follow that those species with elaborate or spring arrival the year after the cross fostering were more similar ornamental secondary sexual characteristics are more at risk for to those of their actual parents than foster parents (187). The cost parasitism. of mites to fitness of the swallows was indicated by data which revealed that parasite burdens of nestlings affected adult tarsus The Incest Taboo length and maximum body weight before fledgling. In the study of satin bowerbirds by Borgia and Collis (28) Finally, another aspect of sexual selection that appears to be males with the fewest number of lice tended to achieve the most based at least partially upon disease resistance is that of avoidance matings. The plumage of satin bowerbirds is highly sexually of breeding close relatives. A consequence of inbreeding is a dimorphic and the shiny blue-black plumage refracts sunlight in a reduction in an animal population of genetic variability in loci that manner that creates bright flashes when males flick their wings play a role in disease resistance. The major histocompatibility during courtship displays (27). Although Borgia and Collis (28) locus, which is usually characterized by great genetic diversity, found that levels of lice on individual birds were correlated has received particular attention with regard to the effects of between years, indicating heritability of resistance to lice, and that inbreeding because this locus plays a key role in immune reactions birds that return to bowers between years have fewer lice, to pathogens (151). Pathogens, especially viruses, can rapidly indicating an advantage of being parasite-free, there was no evolve mechanisms of infection that can penetrate a specific major correlation between plumage quality and louse load. The finding is histocompatibility antimicrobial function (215). In one analysis it inconsistent with the specific Hamilton-Zuk hypothesis. Still to be was estimated that viral genes evolve about a million times faster determined, however, is whether the plumage quality would than the same genes in the host (92). This disparity in the reflect involvement of a different type of parasite. evolutionary rate of change in viruses versus their hosts has led to A showy display of guppies consists of a sigmoid movement the proposal that the extreme diversity of the major histocompat- which is only seen in sexual contexts. In one study the rate of the ibility complex seen among animals offers protection against sexual display was suppressed when guppies were infested with a invading viruses by creating a diversity of immunological types gut-dwelling nematode. Females preferred males with fewer or no within the population (296). Thus, if a virus takes hold in one parasites than control guppies, presumably on the basis of sexual individual, it will be blocked in others. The effect of reduced display (147). A subsequent experiment found that sexual display genetic variability from an inbreeding bottleneck has been seen in movements were affected more severely by parasite presence than several species, most notably cheetahs. In cheetahs the genetic nonsexual movements indicating a specificity of the courtship monomorphism has been extended to the major histocompatibility display with regard to parasites (176). complex by a demonstration that reciprocal skin grafts between The interspecific prediction of the Hamilton-Zuk hypothesis is individuals were accepted (216). The implication of this was seen that in species which are particularly susceptible to parasitic when feline infectious peritonitis swept through several captive invasion, and in which some variation in parasite resistance is a cheetah colonies causing 50--60 percent mortality over a three-year heritable component, sexual selection would promote the appear- period (216). Since the same virus in domestic cats has an average ance of ornamental secondary-sexual characteristics such as feather morbidity of 1 percent (seldom exceeding 10 percent), the inves- brightness, or song complexity or variety, by which a female could tigators argued that the virus simply acclimated to the major judge a male's past or present parasite load. Accordingly, those histocompatibility complex of one cheetah and rapidly spread to avian species in which parasites are most prevalent should tend to others which were genetically uniform in their immunological have the brightest feathers or the most elaborate songs. Hamilton defenses. Two species other than cheetahs are mentioned by and Zuk presented data showing a positive correlation between the O'Brien and Evermann (215) as having undergone excessive published incidence of chronic blood parasitic infections and the inbreeding through population bottlenecks and which are unusu- intensity of display of male brightness in over 100 species of North ally susceptible to viral diseases that threaten their existence. American birds. Similarly, the relationship between coloration and These are the black-footed ferret which was threatened by canine parasitism was explored in British and Irish freshwater fish by distemper, and bighorn sheep in which a respiratory syncytial Ward (279) who found a significant correlation between mean virus has recently been found to be endemic. number of parasite genera within a family of fish and mean species Zoos are one place where inbreeding has been unavoidable. sexual dichromatism. An analysis by Read (236) using both The health-related costs of producing inbred young from parent- European and North American birds appeared to confirm the offspring or sibling matings were calculated for 40 mammalian Hamilton and Zuk findings of a relationship between male species (228). Mortality of offspring increased by 33 percent in brightness and parasite prevalence even when phylogenetic asso- such matings, although there were wide variations among species ciations were controlled. However, more recent analysis by Read examined. Harvey and Read (ll6) mention that the 33 percent and Harvey (237), using a new set of judges for brightness, and increase in mortality found by Ralls with inbreeding in zoos is including only species in which at least I0 individuals had been probably an underestimation, not taking into account embryonic sampled for parasites, did not yield a significant correlation. Using deaths and the fact that the young which survived in zoos with quantitative measures of song duration, intersong interval, song medical care would have died in the wild. continuity, song rate, song versatility and song and syllable The advantage of incestuous matings is that an individual is repertoire, Read and Weary (238) found a negative correlation propagating additional copies of its own genes. At the same time, with song continuity and hematozoa parasite prevalence and a however, there is a reduction in diversity of the major histocom- positive correlation with song versatility and the parasites, but the patibility complex and other genetic factors regulating immune positive correlation came about through taxonomic association. responses, thus, the offspring have less of a chance of surviving At the present time it appears as though females of a variety of diseases including those that might have killed the parents. It is species may use visual and perhaps auditory and olfactory markers quite clear that incestuous matings are unusual in wild popula- of health or disease, including elaborate secondary sexual charac- tions, constituting an estimated less than 2 percent of all matings BEHAVIOR AND PARASITES 289

(229). The degree to which inbreeding is avoided in wild animal asitic infestations, the control strategies may play a role in populations as a function of an inherent reluctance to mate with enhancing fitness mostly in marginal situations such as when a close relatives versus an effect secondary to dispersal patterns or female is lactating, the animal is attempting to escape from a intrasexual competition is still a topic of considerable debate. predator, or it is under the stress of other infestations or undernu- Investigators have cited behavioral evidence on both sides of this trition. Sexual selection provides the basis for choosing sexual issue (48, 96, 130, 196,220,227). Inbreeding avoidance probably partners that will maximize genetic endowment for parasite reflects a number of interacting factors, rather than one single resistance in offspring. behavioral predisposition (115). While the host animals are evolving effective disease control strategies, parasites are evolving effective counterstrategies. We may not necessarily see parasites affecting an animal, even though CONCLUSIONS the threat is there if defenses are relaxed. The existence of parasite Unable to isolate themselves from parasites and pathogens in control strategies does not mean we can expect animals to be free the natural environment, animals must live with the ever-present of parasites or not to succumb occasionally to disease or parasit- threat of parasitism and disease. This threat is viewed as respon- ism, any more than predator avoidance strategy means that sible for the natural selection of a number of behavioral adapta- animals will not ever get killed by predators. There are many tions by animals to enable them to survive and reproduce in this influences shaping an animal's behavior including the need to type of environment. The documented examples of such behav- forage, find shelter, avoid predators and acquire suitable mates. ioral adaptations discussed here probably just scratch the surface The threat of parasitism and disease should be considered as one of of the total inventory of behavioral patterns related to living in a the important determinants of behavior. world of microparasites and macroparasites. In wild animals ACKNOWLEDGEMENTS selection pressures are constantly present to shape the various strategies of parasite control. Animals probably do not cognitively The suggestions and/or comments on an earlier version of the manu- script by Drs. S. B. Hrdy, D. B. Hrdy, J. Moore, D. H. Clayton, M. S. or knowingly employ control principles. Rather, behavioral pat- Boyce, M. S. Mooring, J. Theis, D. M. Broom, N. Hillgarth, W. D. terns which contribute to survival and reproductive success are Hamilton, R. M. May, P. H. Harvey, M. Zuk, M. A. Huffman, A. P. part of an inherited repertoire of behavioral predispositions ac- M¢ller, as well as Dr. P. W. Ewald and another reviewer (anonymous), are quired through natural selection and sometimes enhanced or gratefully acknowledged. Preparation of this review was supported in part shaped by learning. In some instances, especially with macropar- by NIH Grant RR05457.

REFERENCES I. Alexander, R. D. The evolution of social behavior. Annu. Rev. 16. Bell, J. F.; Clifford, C. M.; Moore, G. J.; Raymond, G. Effects of Ecol. Syst. 5:325-383; 1974. limb disability on lousiness in mice. HI. Gross aspects of acquired 2. Altmann, S. A.; Altmann, J. Baboon ecology. African field re- resistance. Exp. Parasitol. 18:49--60; 1966. search. Chicago: University of Chicago Press; 1970. 17. Bell, J. F.; Jellison, W. L.; Owen, C. R. Effects of limb disability 3. Amadon, D. Gal~pagos finches grooming marine iguanas. Condor on lousiness in mice. I. Preliminary studies. Exp. Parasitol. 12: 69:311; 1967. 176--183; 1962. 4. Anderson, J. R. The behavior of nose bot flies (Cephenemyia apicata 18. Bennett, G. F. Boophilus microplus (Acarina: Ixodidae): Experimen- and C. jellisoni) when attacking black-tailed deer (Odocoileus tal infestations on cattle restrained from grooming. Exp. Parasitol. hemionus colurabianus) and the resulting reactions of the deer. Can. 26:323-328; 1969. J. Zool. 53:977-992; 1975. 19. Bertok, L. Physico-chemical defense of vertebrate organisms: the 5. Anderson, R. M.; May, R. M. Regulation and stability of host- role of bile acids in defense against bacterial endotoxins. Perspect. parasite population interactions. J. Anim. Ecol. 47:219-247; 1978. Biol. Med. 21:70-76; 1977. 6. Anderson, R. M.; May, R. M. Population biology of infectious 20. Bertram, B. C. R. Living in groups: predators and prey. In: Krebs, J. diseases: Part I. Nature 280:361-367; 1979. R.; Davies, N. B., eds. Behavioral ecology, an evolutionary ap- 7. Anderson, R. M.; May, R. M. Coevolution of hosts and parasites. proach. Sunderland, MA: Sinauer; 1978:64-97. Parasitology 85:411-426; 1982. 21. Bezuidenhout, J. D.; Stutterheim, C. J. A critical evaluation of the 8. Andrews, A. H. Warble fly: The life cycle, distribution, economic role played by the red-billed oxpecker Buphagus erythrorhynchus in losses and control. Vet. Rec. 103:348-353; 1978. the biological control of ticks. Onderstepoort J. Vet. Res. 47:51-75; 9. Austen, E. E. Appendix B, Gad flies. In: Edwards, F. W.; Oldroyd, 1980. H.; Smart, J. British blood sucking flies. London: British Museum; 22. Blass, E. M.; Teicher, M. H. Suckling. Science 210:15-21; 1980. 1939:149-152. 23. Bleicber, N. Behavior of the bitch during parturition. J. Am. Vet. 10. Autenrieth, R. E.; Fichter, E. On the behavior and socialization of Med. Assoc. 140:1076-1082; 1962. pronghorn fawns. Wildl. Monogr. 42:1-111; 1975. 24. Bolles, R. C. Grooming behavior in the rat. J. Comp. Physiol. 11. Ayensu, E. S. Medicinal plants of West Africa. Algonac, MI: Psychol. 53:306-310; 1960. Reference; 1978. 25. Boostein, S. M.; Ewald, P. W. Costs and benefits of behavioral 12. Banet, M.; Fisher, D.; Hartmann, K. U.; Hentel, H.; Hilling, U. The fever in melanoplus sanguinipes infected by Nosema acridophagus. effect of whole body heat exposure and of cooling the hypothalamus Physiol. Zool. 60:586-595; 1987. on antibody titre in the rat. Pflugers Arch. 391:25-27; 1981. 26. Borgia, G. Satin bowerbird parasites: a test of the bright male 13. Barton, R. Grooming site preferences in primates and their func- hypothesis. Behav. Ecol. Sociobiol. 19:355-358; 1986. tional implications. Int. J. Primatol. 6:519-532; 1985. 27. Borgia, G. Sexual selection in bowerbirds. Sci. Am. 254:92-100; 14. Beattie, A. J. The evolutionary ecology of ant-plant mutualisms. 1986. Cambridge: Cambridge Univ. Press; 1985. 28. Borgia, G.; Collis, K. Female choice for parasite-free male satin 15. Bell, J. F.; Clifford, C. M. Effect of limb disability on lousiness in bowerbirds and the evolution of bright male plumage. Behav. Ecol. mice. II. Intersex grooming relationships. Exp. Parasitol. 15:340- Sociobiol. 25:445--454; 1989. 349; 1964. 29. Bowen, W. H. Defense mechanisms in the mouth and their possible 290 HART

role in the prevention of dental caries. A review. J. Oral Pathol. 56. Dagogo-Jack, S.; Atkinson, S.; Kendall-Taylor, P. Homologous 3:266--278; 1974. radioimmunology for epidermal .growth factor in human saliva. J. 30. Boyce, M. S.; The red queen visits sage grouse leks. Am. Zool. Immunoassay 6:125-136; 1985. 30:263-270; 1990. 57. Davis, J. W.; Anderson, K. C.; Karstad, L.; Trainer, D. A., eds. 31. Brooke, M., de L. The effect of allopreening on tick burdens of Infection and parasitic diseases of wild birds. Ames: Iowa State molting eudyptid penguins. Auk 102:893-895; 1985. Univ. Press; 1971. 32. Broom, D. M.; Pain, B. F.; Leaver, J. D. The effects of slurry on the 58. Davis, J. W.; Karstad, L. H.; Trainer, D. O., eds. Infectious acceptability of swards to grazing cattle. J. Agri. Sci. Camb. diseases of wild mammals. Ames: Iowa State Univ. Press; 1981. 85:331-336; 1975. 59. Day, C. S. D.; Galef, B. G. Pup cannibalism: one aspect of maternal 33. Brown, C. R.; Brown, W. B. Ectoparasitism as a cost of coloniality behavior in golden hamsters. J. Comp. Physiol. Psychol. 91: in cliff swallows (Hirundo pyrrhonota). Ecology 67:1206-1218; 1179-1189; 1977. 1986. 60. Day, J. F.; Edman, J. D. Malaria renders mice susceptible to 34. Brown, N. S. The effect of host beak condition on the size of mosquito feeding when gametocytes are most infective. J. Parasitol. Manacanthus stramineus population of domestic chickens. PouR. 69:163-170; 1983. Sci. 51:162-164; 1972. 61. Dewsbury, D. E. A quantitative description of the behavior of rats 35. Brown, N. S. The effect of louse infestation, wet feathers, and during copulation. Behaviour 29:154-178; 1967. relative humidity on the grooming behavior of the domestic chicken. 62. Diaz, J.; Sampson, H.; Kessler, D.; Stamper, C.; Moore, E.; Poult. Sci. 53:1717-1719; 1974. Robisch, E.; Hodson, A. Experimental necrotizing enterocolitis: The 36. Buchenuer, D.; Luft, C.; Grauvogl, A. Investigations on the elimi- possible role of bile salts in its etiology and treatment. Pediatr. Res. native behavior of piglets. Appl. Anita. Ethol. 9:153-164; 1982. 14:595; 1980. 37. Bullen, J. J. The significance of iron in infection. Rev. Infect. Dis. 63. Dinarello, C. A. Interleukin-1. Rev. Infect. Dis. 6:51-95; 1984. 3:1127-1138; 1981. 64. Dobson, A. P. The population biology of parasite-induced changes 38. Burkill, H. M. The useful plants of West Tropical Africa, 2nd ed. in host behavior. Q. Rev. Biol. 63:139-165; 1988. vol. 1. Ken: Royal Botanic Gardens, 1985. 65. Downes, C. M.; Theberge, J. B.; Smith, S. M. The influence of 39. Burnet, M.; White, D. O. Natural history of infectious disease, 4th insects on the distribution, microhabitat choice, and behavior of the ed. Cambridge: Cambridge Univ. Press; 1972. Burwash caribou herd. Can. J. Zool. 64:622-629; 1986. 40. Campbell, J. B.; Woods, W.; Hagen, A. F.; Howe, E. C. Cattle grub 66. Dubois, E. F. Fever and regulation of body temperature. Springfield, insecticide efficacy and effects on weight-gain performance of feeder IL: Charles C. Thomas; 1948. calves in Nebraska. J. Econ. Entomol. 66:429--432; 1973. 67. Dubos, R. J. Man adapting. New Haven, CT: Yale Univ. Press; 41. Carpenter, C. C. The marine iguana of the Gal"ipagos Islands, its 1965. behavior and ecology. Proc. Calif. Acad. Sci. 34:329-376; 1966. 68. Duncan, P.; Vigne, N. The effect of group size in horses on the rate 42. Carr, W. J.; Hirsch, J. T.; Campellone, B. L.; Marasco, E. Some of attacks by blood-sucking flies. Anim. Behav. 27:623-625; 1979. determinants of a natural food aversion in Norway rats. J. Comp. 69. Edman, J. D.; Kale, H. W. Host behavior: Its influence on the Physiol. Psychol. 93:899-906; 1979. feeding success of mosquitoes. Ann. Entomol. Soc. Am. 64: 43. Clark, L.; Mason, J. R. Use of nest material as insecticidal and 513-516; 1971. anti-pathogenic agents by the European starling. Oecologia 67: 70. Edman, J. D.; Weber, L. A.; Schmid, A. A. Effect of host defenses 169-176; 1985. on the feeding pattern of Culex nigripalpus when offered a choice of 44. Clark, L.; Mason, J. R. Olfactory discrimination of plant volatiles by blood sources. J. Parasitol. 60:874-883; 1974. the European starling. Anita. Behav. 35:227-235; 1987. 71. Erhlich, P. R.; Dobkin, D. S.; Wheye, D. The adaptive significance 45. Clark, L.; Mason, J. R. Effect of biologically active plants used as of anting. Auk 103:835; 1986. nest material and the derived benefit to starling nestlings. Oecologia 72. Erhlich, P. R.; Raven, P. H. Butterflies and plants: A study in 77:174-180; 1988. coevolution. Evolution 18:586---608; 1964. 46. Clayton, D. Mate choice in experimentally parasitized rock doves. 73. Espmark, Y. Behavior of reindeer attacked by nostril to grub flies. Am. Zool. 30:251-262; 1990. Zool. Rev. 2:28-37; 1961. 47. Clutton-Brock, T. H.; Albon, S. D. Parental investment in male and 74. Espmark, Y.; Langvatn, R. Lying down as a means of reducing fly female offspring in mammals. In: Kings College Sociobiology harassment in red deer (Cervus elaphus). Behav. Ecol. Sociobiol. Group, eds. Current problems in sociobiology. Cambridge, England: 5:51-54; 1979. Cambridge Univ. Press; 1982:223-247. 75. Ewald, P. W. Evolutionary biology and the treatment of signs and 48. Cockburn, A.; Scott, M. P.; Scotts, D. Inbreeding avoidance and symptoms of infectious disease. J. Theor. Biol. 86:107-176; 1980. male-biased natal dispersal in Antechinus spp. (Marsupialla dasyu- 76. Ewald, P. W. Host-parasite relations, vectors, and the evolution of ridae). Anim. Behav. 33:908-915; 1985. disease severity. Annu. Rev. Ecol. Syst. 14:465--485; 1983. 49. Collias, N.; Collias, E. Nest building and bird behavior. Princeton: 77. Ewald, P. W. Transmission modes and evolution of the parasitism- Princeton Univ. Press; 1984. mutualism continuum. Ann. NY Acad. Sci. 503:295-306; 1987. 50. Constantine, D. G. Bat rabies. Current knowledge and future 78. Feare, C. J. Desertion and abnormal development in a colony of research. In: Nagano, Y.; Davenport, F. M., eds. Rabies. Baltimore: sooty terns Sterna fuscata infested by virus-infected ticks. Ibis University Park Press; 1971:253-262. 118:112-115; 1976. 51. Cooper-Driver, G. A. Chemotaxonomy and phytochemical ecology 79. Feder, H. M. Cleaning symbioses in the marine environment. In: of bracken. Bot. J. Linn. Soc. 73:35-46; 1976. Henry, S. M., ed. Symbiosis, vol. 1. New York: Academic Press; 52. Cowgill, U. M. Co-operative behaviour in Periodictious. In: Martin, 1966:327-380. R. D.; Doyle, G. A.; Walker, A. C., eds. Prosimian biology. 80. Fox, L. R. Cannibalism in natural populations. Annu. Rev. Ecol. London: Duckworth; 1974:261-272. Syst. 6:87-106; 1975. 53. Crofton, H. D. Nematode parasite populations in sheep on lowland 81. Freeland, W. J. Pathogens and the evolution of primate sociality. farms. VI. Sheep behavior and nematode infections. Parasitology Biola'opica 8:12-24; 1976. 48:251-260; 1958. 82. Freeland, W. J. Primate social groups as biological islands. Ecology 54. Crowden, A. E.; Broom, D. M. Effects of the eyefluke, Diplosto- 60:719-728; 1979. mum spathaceura, on the behaviour of dace Leuciscus leuciscus. 83. Freeland, W. J. Mangabey (Cercocebus albigena) movement pat- Anim. Behav. 28:287-294; 1980. terns in relation to food availability and fecal contamination. Ecology 55. Crowell-Davis, S. L.; Candle, A. B. Coprophagy by foals: Recog- 61:1297-1303; 1980. nition of maternal feces. Appl. Anim. Bebav. Sci. 24:267-272; 84. Freeland, W. J. Functional aspects of primate grooming. Ohio J. Sci. 1989. 81:173-177; 1981. BEHAVIOR AND PARASITES 291

85. Freeland, W. J. Parasitism and behavioral dominance among male 112. Hart, B. L.; Korinek, E.; Brennan, P. Postcopulatury genital mice. Science 213:461-462; 1981. grooming in male rats: Prevention of sexually transmitted infections. 86. Futuyma, D. J. Evolutionary interactions among herbivorous insects Physiol. Behav. 41:321-325; 1987. and plants. In: Futuyma, D. J.; Slatkin, M., eds. Coevolution. New ll3. Hart, B. L.; Powell, K. Antibacterial properties of saliva: Role in York: Sinaner; 1983:207-231. maternal periparturient grooming and in licking wounds. Physiol. 87. Garcia, J.; Hankins, W. G.; Rusinak, K. W. Behavioral regulation Behav. 48(3):in press; 1990. of the milieu interne in man and rat. Science 185:824-831; 1974. I14. Hart, L. A.; Hart, B. L. Autugrooming and social grooming in 88. Gibbs, E. P. J. Virus diseases of horses. In: Gibbs, E. P. J., ed. impala. Ann. NY Acad. Sci. 525:399-402; 1988. Virus diseases of food animals, vol. I. New York: Academic; I15. Harvey, P. H.; Rails, K. Do animals avoid incest? Nature 320: 1981:217-230. 575-576; 1986. 89. Gibbs, H. C. Effects of parasites on animal and meat production. In: 116. Harvey, P. H.; Read, A. F. Copulation genetics. When incest is not Gaafar, S. M.; Howard, W. F.; Marsh, R. E., eds. Parasites, pests best. Nature 336:514-515; 1988. and predators. Amsterdam: Elsevier; 1985:7-27. 117. Harvey, T. L.; Brethous, J. R. Effect of horn flies on weight gains 90. Gibney, V. J.; Campbell, J. B.; Boxier, D. J.; Clanton, D. C.; of beef cattle. J. Econ. Entomol. 72:516--518; 1979. Deutscher, G. H. Effects of various infestation levels of cattle lice 118. Harvey, T. L.; Launchbaugh, J. L. Effect of horn flies on behavior (Mallophaga trichodectidae and Anoplura haematopinidae) on feed of cattle. J. Econ. Entomol. 75:25-27; 1982. efficiency and weight gains of beef heifers. J. Econ. Entomol. 119. Hausens, E. J.; Valiela, J. Activity of the face fly in New Jersey. J. 78:1304-1307; 1985. Econ. Entomol. 60:26-28; 1967. 91. Gill, D. E.; Mock, B. A. Ecological and evolutionary dynamics of 120. Hansfater, G.; Hrdy, S. B., eds. Infanticide. Comparative and parasites: The case of Typanosoma diemyctli in the red-spotted newt evolutionary perspectives. New York: Aldine; 1984. Notophthalmus viridesceus. In: Rollinson, D.; Anderson, R. M., 121. Hausfater, G.; Meade, B. J. Alternation of sleeping groves by yellow eds. Ecology and genetics of host-parasite interactions. London: baboons (Papio cynocephalus) as a strategy for parasite avoidance. Academic; 1985:157-183. Primates 23:287-297; 1982. 92. Gojobori, T.; Yokoyama, S. Molecular evolutionary rates of onco- 122. Helle, T.; Aspi, J. Does herd formation reduce insect harassment genes. J. Mol. Evol. 26:148-156; 1987. among reindeer? A field experiment with animal traps. Acta Zool. 93. Goodwin, T. W. Aspects of terpinoid chemistry and biochemistry. Fenn. 175:129-131; 1983. London: Academic; 1971. 123. Helle, T.; Tavainen, L. Effects of insect harassment on weight gain 94. Gore, I.; Isaacson, N. H. The pathology of hyporexia. Am. J. and survival in reindeer calves. Rangifer 4:24-27; 1984. Pathol. 25:1029-1059; 1949. 124. Hensel, H.; Brnck, K.; Raths, P. Homeothermic organisms. In: 95. Greene, C. E. Clinical microbiology and infectious diseases of the Precht, H.; Christopherson, J.; Hensel, H.; Lardaer, W., eds. dog and cat. Philadelphia: Lea and Febiger; 1984. Temperature and life. Berlin: Springer-Verlag; 1973:503-761. 96. Greenwood, P. J. Mating systems, philopatry and dispersal in birds 125. Hewetson, R. W.; Nolan, J. Resistance of cattle tick, Boophilus and mammals. Anita. Behav. 28:1140-1162; 1980. microplus. I, The development of resistance to experimental infes- 97. Gubernick, D. J. Parent and infant attachment in mammals. In: tation. Aust. J. Agri. Res. 19:323-333; 1968. Gubernick, D. J.; Klopfer, P. H., eds. Parental care in mammals. 126. Hillerton, J. E.; Morant, S. V.; Harris, J. A. Control of Muscidae on New York: Plenum Press; 1981:243-305. cattle by flucythrinate ear-tags, the behaviour of these flies on cattle 98. Gunderson, H. Effect of cattle grub treatment on weight gains in beef and the effect on fly-dislodging behaviour. Entomol. Experiment. cattle. J. Econ. Entomol. 38:398-399; 1945. Appl. 41:213-218; 1986. 99. Haddlow, A. J. Field studies on an African monkey Cercopithecus 127. Hoeprich, P. D., ed. Infectious diseases, 2nd ed. New York: Harper ascanius schraidti matshie. Proc. Zool. Soc. Lond. 122:297-394; and Row; 1977:34---45. 1952. 128. Holmes, J. C.; Bethel, W. M. Modification of intermediate host 100. Hamilton, W. D. The genetical evolution of social behaviour. J. behavior by parasites. In: Canning, E. U.; Wright, C. A., eds. Theor. Biol. 7:1-16; 1964. Behavioral aspects of parasite transmission. J. Linn. Soc. Suppl. 101. Hamilton, W. D. Geometry for the selfish herd. J. Theor. Biol. 1:123-149; 1972. 31:295-31 l; 1971. 129. Hoogland, J. L. Aggressive, ectoparasitism, and other possible costs 102. Hamilton, W. D. Sex versus non-sex versus parasite. Oikos 35: of prairie dog (Sciuridae, Cynomys spp.) coloniality. Behaviour 282-290; 1980. 69:1-35; 1979. 103. Hamilton, W. D. Pathogens as causes of genetic diversity in their 130. Hoogland, J. L. Prairie dogs avoid extreme inbreeding. Science host populations. In: Anderson, R. M.; May, R., eds. Population 215:1639-1641; 1982. biology of infectious diseases. New York: Springer-Verlag; 1982: 131. Hrdy, S. B. Infanticide among animals: a review, classification and 269-296. examination of the implications for the reproductive strategies of 104. Hamilton, W. D.; Zuk, M. Heritable true fitness and bright birds: A females. Ethol. Sociobiol. 1:13--40; 1979. role for parasites? Science 218:384-387; 1982. 132. Hubalek, Z. Coincidence of fungal species with birds. Ecology 105. Harkness, J. E.; Wagner, J. E. The biology and medicine of rabbits 59:438--442; 1978. and rodents. Philadelphia: Lea and Febiger; 1989. 133. Huffman, M. A.; Seifu, M. Observations on the illness and con- 106. Harris, J. A.; Hillerton, J. E.; Morant, S. V. Effect on milk sumption of a possibly medicinal plant Vernonia amygdalina (Del.), production of controlling muscid flies, and reducing fly-avoidance by a wild chimpanzee in the Mahale Mountains National Park, behaviour, by use of Fenvalerate ear tags during the dry period. J. Tanzania. Primates 30:51-63; 1989. Dairy Res. 54:165-171; 1987. 134. Hunter, D. M.; Moorhouse, D. E. The effects of Austrosimulium 107. Harrison, C. J. O. Allopreening as agonistic behaviour. Behaviour pestilens on the milk production of dairy cattle. Aust. Vet. J. 24:161-208; 1965. 52:97-99; 1976. 108. Hart, B. L. The behavior of domestic animals. New York: Freeman; 135. Husseini, R. H.; Sweet, C.; Collie, M. H.; Smith, H. Elevation of 1985. nasal viral levels by suppression of fever in ferrets infected with 109. Hart, B. L. Biological bases of the behavior of sick animals. influenza viruses of differing virulence. J. Infect. Dis. 145:520-524; Neurosci. Biobehav. Rev. 12:123-137; 1988. 1982. 110. Hart, B. L.; Hart, L. A. Canine and feline behavioral therapy. 136. Hutchins, M.; Barash, D. P. Grooming in primates: implications for Philadelphia: Lea and Febiger; 1985. its utilitarian function. Primates 17:145-150; 1976. 111. Hart, B. L.; Haugen, C. M. Prevention of genital grooming in 137. Hutson, J. M.; Niall, M.; Evans, D.; Fowler, R. Effect of salivary mating behaviour of male rats (Rattus norvegicus). Anim. Behav. glands on wound contraction in mice. Nature 279:793-795; 1979. 19:230-232; 1971. 138. Janpel, H. D.; Duff, G. W.; Gershon, R. K.; Atkins, E.; Durum, S. 292 HART

K. Fever and immunoregulation. I11. Hyperthermia augments the 164. Lent, P. C. Mother-infant relationships in ungulates. In: Geist, V.; primary in vitro humoral immune response. J. Exp. Med. 157: Walther, F., eds. The behavior of ungulates and its relation to 1229-1238; 1983. management. Int. Union Conserv. Nature Publ., New Ser., 24:1971. 139. Jenkins, D. A.; Watson, A.; Miller, G. R. Population studies on red 165. Leuthold, W. African ungulates: a comparative review of their grouse, Lagopus lagopus scoticus (Lath.), in north-east Scotland. J. ethology and behavioral ecology. Berlin: Springer-Verlag; 1977. Anim. Ecol. 32:317-376; 1963. 166. Lewis, L. F.; Christenson, D. M.; Eddy, G. W. Rearing the 140. Jenkins, D. A.; Watson, A.; Miller, G. R. Predation and red grouse long-nosed cattle louse and cattle biting louse on host animals in populations. J. Appl. Ecol. 1:183-195; 1964. Oregon. J. Econ. Entomol. 60:755-757; 1967. 141. Johnson, L. L.; Boyce, M. S. Female choice of males with low 167. Li, A. K.; Koroly, M. J.; Schattenkerk, M. E.; Malt, R. A.; Young, parasite load in sage grouse, Centrocerus urophasianus. In: Loye, J. M. Nerve growth factor: Acceleration of the rate of wound healing. E.; Zuk, M., eds. Ecology, behavior and evolution of bird-parasite Proc. Natl. Acad. Sci. USA 77:4379-4381; 1980. interactions. Oxford: Oxford Univ. Press; in press. 168. Limbaugh, C. Cleaning symbioses. Sci. Am. 205:42-49; 1961. 142. Johnson, R. T. The pathogenesis of experimental rabies. In: Nagano, 169. Limbaugh, C.; Pederson, H.; Chace, F. A. Shrimps that clean Y.; Davenport, F. J., eds. Rabies. Baltimore: University Park Press; fishes. Bull. Marine Sci. Gulf Caribb. 11:237-257; 1961. 1971:59-75. 170. Little, D. A. The effect of cattle tick infestation on the growth rate 143. Jones, R. L. Special considerations for appropriate antimicrobial of cattle. Aust. Vet. J. 39:6--10; 1963. therapy in neonates. Vet. Clin. N. Am.: Small Anita. Pract. 171. Londt, J. G. H.; Whitehead, G. B. Ecological studies of larval ticks 17:577---602; 1987. in South Africa (Acarina: Ixodidae). Parasitology 65:469-490; 1972. 144. Kaplanis, J. N.; Thompson, M. J.; Robbins, W. E.; Bryce, B. M. 172. Lott, D. F.; Hart, B. L. Applied ethology in a nomadic cattle culture. Insect hormones, alpha ecdysone and 20-hydroxyecdyscone in bracken Appl. Anita. Ethol. 5:309-319; 1979. fern. Science 157:1436-1438; 1967. 173. McCarthy, D. O.; Kluger, M. J.; Vander, A. J. Suppression of food 145. Keiper, R. R.; Berger, J. Refuge-seeking and pest avoidance by feral intake during infection: Is interleukin-1 involved? Am. J. Clin. Nutr. horses in desert and island environments. Appl. Anita. Ethol. 42:1179-1182; 1985. 9:111-120; 1982-1983. 174. McFarland, C. G.; Reeder, W. G. Cleaning symbiosis involving 146. Kelso, L.; Nice, M. N. A Russian contribution to anting and feather Gal,'ipagos tortoises and two species of Darwin's finches. Z. Tierpsy- mites. Wilson Bull. 75:23-26; 1963. chol. 34:464-483; 1974. 147. Kennedy, C. E. J.; Endler, J. A.; Poyton, S. L.; McMinn, H. 175. McKenna, J. J. Biosocial functions of grooming behaviour among Parasite load predicts mate choice in guppies. Behav. Ecol. Socio- the common Indian langur monkey (Presbytis entellus). Am. J. biol. 21:291-295; 1987. Phys. Anthropol. 48:503-510; 1978. 148. Keymer, A.; Crompton, D. W. T.; Sahakian, B. J. Parasite-induced 176. McMinn, H. Effects of the nematode parasite Camallanus cotti on learned taste aversion involving Nippostrongylus in rats. Parasitol- sexual and non-sexual behaviors in the guppy (Poecilia reticulata). ogy 86:455-460; 1983. Am. Zool. 30:245-249; 1990. 149. Kilpatrick, S. J.; Lee, T. M.; Moltz, H. The maternal pheromone of 177. MacDonald, D. W. Running with the fox. London: Urwin Hyman; the rat: Testing some assumptions underlying a hypothesis. Physiol. 1987. Behav. 30:539-543; 1983. 178. MacFarlane, B. A.; Epstein, A. N. Biobehavioral determinants of 150. Kirby, K. A.; Rothenburg, B. A.; Victery, W.; Vander, A. J.; evaporative water loss in the rat. Behav. Neural Biol. 33:101-116; Kluger, M. J. Urinary excretion of zinc and iron following injection 1981. of bacteria in the unanesthetized rabbit. Min. Elect. Metab. 7: 179. Mandel, J. D. The functions of saliva. J. Dent. Res. 66:623-627; 250-256; 1982. 1987. 151. Klein, J. Natural history of the major histocompatibility complex. 180. Mandel, J. D.; Ellison, S. A. Naturally occurring defense mecha- New York: John Wiley and Sons; 1986. nisms in saliva. Microbiol. Abstr. Suppl.:367-379; 1981. 152. Kluger, M. J. The evolution and adaptive value of fever. Am. Sci. 181. Marshall, A. G. The ecology of ectoparasitic insects. London: 66:38-43; 1978. Academic; 1981. 153. Kluger, M. J. Fever. Its biology, evolution and function. Princeton: 182. Maurer, F. D.; McCully, R. M. African horse sickness--with Princeton University Press; 1979. emphasis on pathology. Am. J. Vet. Res. 24:235-266; 1963. 154. Kluger, M. J. Fever in ectotherms: Evolutionary implications. Am. 183. May, R. M.; Anderson, R. M. Epidemiology and genetics in the Zool. 19:295-304; 1979. coevolution of parasites and hosts. Proc. R. Soc. Lond. B219: 155. Kluger, M. J.; Ringler, D. H.; Anver, M. R. Fever and survival. 281-313; 1983. Science 188:166-168; 1975. 184. Michel, J. F. Parasitological significance of bovine grazing behav- 156. Kluger, M. J.; Rothenburg, B. A. Fever and reduced iron: Their iour. Nature 175:1088-1089; 1955. interaction as a host defense response to bacterial infection. Science 185. Miller, F. L. Behavior associated with parturition in black-tailed 203:374-376; 1979. deer. J. Wildl. Mgmt. 29:629--631; 1965. 157. Kluger, M. J.; Vaughn, L. K. Fever and survival in rabbits infected 186. M¢ller, A. P. Advantages and disadvantages of coloniality in the with Pasteurella multocida. J. Physiol. 282:243-251 ; 1978. swallow, Hirundo rustica. Anim. Behav. 35:819-832; 1987. 158. Kodric-Brown, A.; Brown, J. H. Truth in advertising: The kinds of 187. M~ller, A. P. Effects of an haematophagous mite on the barn traits favored by sexual selection. Am. Nat. 124:309-323; 1984. swallow (Hirundo rustica): A test of the Hamilton and Zuk hypoth- 159. Krueger, J. M.; Kubillus, S.; Shoham, S.; Davenne, D. Enhance- esis. Evolution, in press; 1990. ment of slow-wave sleep by endotoxin and lipid A. Am. J. Physiol. 188. M¢ller, A. P. Effects of parasitism by the haematophagous mite 251 :R591-R597; 1986. Ornithonyssus bursa on reproduction in the barn swallow Hirundo 160. Krueger, J. M.; Pappenheimer, J. R.; Ka.rnovsky, M. L. Sleep- rustica. Ecology, in press; 1990. promoting effects of muramyl peptides. Proc. Natl. Acad. Sci. USA 189. M¢ller, A. P. Male tail length and female mate choice in the 79:6102-6106; 1982. monogamous swallow Hirundo rustica. Anim. Behav., in press; 161. Krueger, J. M.; Walter, J.; Dinarello, C. A.; Chedid, L. Induction of 1990. slow-wave sleep by interleukin-1. In: Kluger, M. J.; Oppenheim, J. 190. M¢ller, A. P.; Allander, K.; Dufva, R. Fitness effects of parasites on J.; Powanda, M. C., eds. The physiologic, metabolic and immuno- passerine birds: A review. In: Blondel, J.; Gosler, A.; Lebreton, J. logic actions of interleukin-1. New York: Liss; 1985:161-170. D.; McCleery, R., eds. Demographical, physiological, genetical and 162. Lawton, J. H. The structure of the arthropod community on bracken. behavioral aspects of population biology of passerine birds. Berlin: Bot. J. Linn. Soc. 73:187-216; 1976. Springer-Verlag; 1990:in press. 163. Lawton, J. H.; MacGarvin, M. Interaction between bracken and its 191. Moltz, H. Of rats and infants and necrotizing enterocolitis. Perspect. insect herbivores. Proc. R. Soc. 86:125-131; 1985. Biol. Med. 27:327-335; 1984. BEHAVIOR AND PARASITES 293

192. Moltz, H.; Lee, T. M. The coordinate roles of mother and young in 219. Otterness, I. G.; Seymour, P. A.; Golden, H. W.; Reynolds, J. A.; establishing and maintaining pheromonal symbiosis in the rat. In: Daumy, G. 0. The effects of continuous administration of murine Rosenblum, L. A.; Moltz, H., eds. Symbiosis in parent-offspring interleukin-1 in the rat. Physiol. Behav. 43:797-804; 1988. interactions. New York: Plenum; 1983:45--60. 220. Packer, C. Dispersal and inbreeding avoidance. Anim. Behav. 193. Moore, J. Responses of an avian predator and its isopod prey to an 33:676; 1985. acanthocephalan parasite. Ecology 64:1000-1015; 1983. 221. Packer, C.; Pusey, A. E. Infanticide in carnivores. In: Hausfater, G.; 194. Moore, J. Altered behavioral responses in intermediate hosts--an Hrdy, S. B., eds. Infanticide comparative and evolutionary perspec- acanthocephalan parasite strategy. Am. Nat. 123:572-577; 1984. fives. New York: Aldine; 1984:31-42. 195. Moore, J. Some roles of parasitic helminths in trophic interactions. 222. Pain, B. F.; Leaver, J. D.; Broom, D. M. Effects of cow slurry on A view from North America. Rev. Hist. Nat. 60:159-179; 1987. herbage production intake by cattle and grazing behaviour. J. Br. 196. Moore, J.; Ali, R. Are dispersal and inbreeding avoidance related? Grassland Soc. 29:85-91; 1974. Anita. Behav. 32:94-112; 1984. 223. Pen-ins, C. M. British fits. London: Collins; 1979. 197. Moore, J.; Gotelli, N. J. A phylogenetic perspective on the evolution 224. Petrides, P. E.; Bfhlen, P.; Shivley, J. E. Chemical characterization of altered host behavior: A critical look at the manipulation hypoth- of the two forms of epidermal growth factor in murine saliva. esis. In: Barnard, C. J.; Behnke, J. M., eds. Parasitism and host Biochem. Biophys. Res. Commun. 125:218-228; 1984. behavior. In press. 225. Potter, E. F. Anting in wild birds, its frequency and probable 198. Moore, J.; Simberloff, D.; Freehling, M. Relationships between purpose. Auk 87:692-713; 1970. bobwhite quail social-group size and intestinal helminth parasitism. 226. Pulliam, H. R.; Caraco, T. Living in groups: Is there an optional Am. Nat. 131:22-32; 1988. group size? In: Krebs, J. R.; Davies, N. B., eds. Behavioral ecology, 199. Mooring, M. S. Ontogeny of allogrooming in mule deer (Odocoileus 2nd ed. Sunderland, MA: Sinauer; 1984:122-147. hemionus). J. Mammal. 70:434--437; 1989. 227. Pusey, A. E. Inbreeding avoidance in chimpanzees. Anita. Behav. 200. Morgan, D. W. T.; Baillie, H. D. A field trial to determine the effect 28:543-552; 1980. of fly control using permethrin on milk yields in dairy cattle in the 228. Rails, K.; Ballou, J. D.; Templeton, A. Estimates of lethal equiva- U.K. Vet. Rec. 106:121-123; 1980. lents and cost of inbreeding in mammals. Conserv. Biol. 2:185-193; 201. Moss, W. W.; Camin, J. H. Nest parasitism, productivity, and 1988. clutch size in purple martins. Science 168:1000-1003; 1970. 229. Rails, K.; Harvey, P. H.; Lyles, A. M. Inbreeding in natural 202. Muller, M. J.; Murray, M. D. Blood sucking flies feeding on sheep populations of birds and mammals. In: Soule, M. E., ed. Conserva- in Eastern Australia. Aust. J. 7_,ool. 25:75-85; 1977. tion biology: The science of scarcity and diversity. Sunderland, MA: 203. Murphy, F. A. Rabies pathogenesis. Arch. Virol. 54:279-297; 1977. Sinauer; 1986:35-56. 204. Murphy, P. A.; Watson, A. Y.; Metz, J.; Forssmann, W. G. The 230. Rasa, O. A. E. Invalid care in the dwarf mongoose (Helogale mouse submandibular gland. J. Histochem. Cytochem. 28:890-902; undulata rufula). Z. Tierpsychol. 42:337-342; 1976. 1980. 23 I. Rasa, O. A. E. The effects of crowding on the social relationships 205. Murray, M. D. The ecology of the louse Polyplax Serrate (burro.) on and behavior of the dwarf mongoose (Helogale undulata rufula) Z. the mouse Mus rausculus L. Aust. J. Zool. 9:l-ll; 1961. Tierpsychol. 49:317-329; 1979. 206. Murray, M. D. Effects of host grooming on louse populations. 232. Rasa, O. A. E. A case of invalid care in wild dwarf mongooses. Z. Parasitol. Today 3:276-278; 1987. Tierpsychol. 62:235-240; 1983. 207. Murray, M. J.; Murray, A. B. Anorexia of infection as a mechanism 233. Rau, M. E. Establishment and maintenance of behavioral dominance of host defense. Am. J. Clin. Nutr. 32:593-596; 1979. in male mice infected with Trichinella spiralis. Parasitology 86: 208. Nansen, P. Importance of immunology in relation to parasitism and 319-322; 1983. animal production. In: Gaafar, S. M.; Howard, W. E.; Marsh, R. E., 234. Rau, M. E. Loss of behavioral dominance in male mice infected with eds. Parasites, pests and predators. Amsterdam: Elsevier; 1985: Trichinella spiralis. Parasitology 88:371-373; 1983. 371-386. 235. Ran, M. E.; Caron, F. R. Parasite-induced susceptability of moose to 209. Neal, E. The natural history of badgers. London: Croom Helm; hunting. Can. J. Zool. 57:2466-2468; 1979. 1986. 236. Read, A. F. Comparative evidence supports the Hamilton and Zuk 210. Nelson, W. A.; Bell, $. F.; Clifford, C. M.; Keirans, J. E. hypothesis on parasites and sexual selection. Nature 328:68-70; Interaction of ectoparasites and their hosts. J. Med. Entomol. 1987. 13:389-428; 1977. 237. Read, A. F.; Harvey, P. H. Reassessment of comparative evidence 211. Nesbitt, G. H. Canine and feline dermatology. A systematic ap- for the Hamilton and Zuk theory on the evolution of secondary sexual proach. Philadelphia: Lea and Febiger; 1983. characters. Nature 339:618-620; 1989. 212. Niall, J. M.; Graeme, B. R.; O'Brien, B. M. The effect of epidermal 238. Read, A. F.; Weary, D. M. Sexual selection and the evolution of growth factor on wound healing in mice. J. Surg. Res. 33:164-169; bird song: A test of the Hamilton-Zuk hypothesis. Behav. Ecol. 1982. Sociobiol. 26:47-56; 1990. 213. Norval, R. A. I. Ecology of the tick Amblyomma hebraeum Koch in 239. Rheingold, H. L. Maternal behavior in the dog. In: R.beingold, H. the eastern Cape province of South Africa. I. Distribution and L., ed. Maternal behavior in mammals. New York: Wiley; 1963: seasonal activity. J. Parasitol. 63:734-739; 1977. 169-202. 214. Nutting, W. B. Host parasite relations: Demodicidae. Acarologia 240. Ritter, R. C.; Epstein, A. N. Saliva lost by grooming: A major item 7:301-317; 1965. in the rat's water economy. Behav. Biol. I 1:581-585; 1974. 215. O'Brien, S. J.; Evermann, J. F. Interactive influence of infectious 241. Robertson, A. S.; Patrick, C. D.; Semtner, P. J.; Hair, J. A. The disease and genetic diversity in natural populations. Trends Ecol. ecology and behavior of the lone star tick (Acarina: Isodidae). VI, Evol. 3:254-259; 1988. Response of unfed adults to certain environmental parameters. J. 216. O'Brien, S. J.; Roelke, M. E.; Marker, L.; Newman, A.; Winkler, Med. Entomol. 12:525-529; 1975. C. A.; Meltzer, D.; Colly, L.; Evermann, J. F.; Bush, M.; Wildt, D. 242. Rodriguez, E.; Aregullin, M.; Nishida, T.; Uehara, S.; Wrangham, E. Genetic basis for species vulnerability in the cheetah. Science R.; Abramowski, Z.; Finlayson, A.; Towers, G. H. N. Thiarubrine 227:1428-1434; 1985. A, a bioactive constituent of Aspilia (Asteraceae) consumed by wild 217. Odberg, F. O.; Francis-Smith, K. Studies on the formation of chimpanzees. Experientia 41:419-420; 1985. ungrazed eliminative areas in fields used by horses. Appl. Anim. 243. Rosenblatt, J. S.; Lehrman, D. S. Maternal behavior of the labora- Ethol. 3:27-34; 1977. tory rat. In: Rheingold, H. L., ed. Maternal behavior in mammals. 218. Okumura, T. The relationship of attacking fly abundance to behav- New York: Wiley; 1963:8-57. ioral reponses of grazing cattle. Jpn. J. Appl. Entomol. Zool. 244. Rosenthal, G. A.; Janzen, D. H. Herbivores. Their interaction with 21:119-122; 1977. secondary plant metabolites. New York: Academic Press; 1979. 294 HART

245. Ross, S. Some observations of the lair dwelling behavior of dogs. 272. Toth, L. A. Krueger, J. M. Hematologic effects of infectious Behaviour 2:144-162; 1950. challenge in rabbits. J. Am. Vet. Med. Assoc., in press; 1990. 246. Roth, L.; Rosenblatt, J. S. Changes in self-licking during pregnancy 273. Turner, H. C.; Short, A. J. effects of field infestations of gastro- in the rat. J. Comp. Physiol. Psychol. 63:397-400; 1967. intestinal helminths and of the cattle ticks (Boophilus microplus) on 247. Ruch, T. C. Diseases of laboratory primates. Philadelphia: Saunders; growth of three breeds of cattle. Aust. J. Agri. Res. 23:177-193; 1959. 1972. 248. Sachs, B. D.; Barfield, R. J. Functional analysis of masculine 274. Urquhart, G. M. Application of immunity in the control of parasitic copulatory behavior in the rat. In: Rosenblatt, J. S.; Hind, R. A.; disease. Vet. Parasitol. 6:217-239; 1980. Shaw, E.; Beer, C. G., eds. Advances in the study of behavior, vol. 275. Vogt, R. C. Cleaning/feeding symbiosis between grackles (Quis- VII. New York: Academic Press; 1977:91-156. caius: icteridae) and map turtles (Graptemys emydidae). Auk 96: 249. Schmidt, G. D.; Roberts, L. S. Foundations of parasitology. St. 608---609; 1979. Louis: Times Mirror/Mosby College Pub.; 1985. 276. Waage, J. K. The evolution of insect/vertebrate associations. Br. J. 250. Schmidtmarm, E. T.; Valla, M. E. Face-fly pest intensity, fly- Lima. Soc. 12:187-224; 1979. avoidance behavior (bunching) and grazing time in Holstein heifers. 277. Waage, J. K.; Nondo, J. Host behaviour and mosquito feeding Appl. Anita. Ethol. 8:429-438; 1982. success: an experimental study. Trans. R. Soc. Trop. Med. Hygiene 251. Schmidtmann, E. T.; Valla, M. E.; Chase, L. E. Effect of face flies 76:119-122; 1982. on grazing time and weight gain in dairy heifers. J. Econ. Entomol. 278. Walter, J.; Davenne, D.; Shoham, S.; Dinarello, C. A.; Krueger, J. 74:33-39; 1981. A. Brain temperature changes coupled to sleep states persist during 252. Schneirla, T. C.; Rosenblatt, J. S.; Tobach, E. Maternal behavior of interleukin-1-enhanced sleep. Am. J. Physiol. 250:R96-R103; 1986. the cat. In: Rheingold, H. L., ed. Maternal behavior in mammals. 279. Ward, P. I. Sexual dichromatism and parasitism in British and Irish New York: Wiley; 1963:122-168. fresh water fish. Anim. Behav. 36:1210-1215; 1988. 253. Seebeck, R. M.; Spingell, P. H.; O'Kelly, J. C. Alterations in host 280. Wames, M. L.; Finlayson, L. H. Effect of host behaviour on host metabolism by the specific and anorectic effects of the cattle ticks preference in Stomoxys calcitrans. Med. Vet. Entomol. 1:53-57; (Boophilus microplus). I. Food intake and body weight growth. 1987. Aust. J. Biol. Sci. 24:373-380; 1971. 281. Washburn, S. L.; DeVore, I. The social life of baboons. In: Eisner, 254. Simmons, K. E. L. Anting and the problem of self-stimulation. J. T.; Wilson, E. O., eds. Animal behavior. San Francisco: W. H. Zool. 149:145-162; 1966. Freeman; 1961:317-326. 255. Simmons, K. E. L. Anting. In: Campbell, B.; Lack, E., eds. A 282. Wasteson, Y.; Olsvik, O.; Skanachke, E.; Bopp, C. A.; Fossum, K. dictionary of birds. Vermillion, SD: Buteo Books; 1985:19. Heat-stable-enterotoxin-producing Escherichia coli strains isolated 256. Simon, N. Between the sunlight and the thunder, the wild life of from dogs. J. Clin. Microbiol. 26:2564-2566; 1988. Kenya. London: Collins; 1962. 283. Watt, J. M.; Breyer-Brandwijk, M. G. The medicinal and poisonous 257. Snowball, G. J. The effect of self licking by cattle on infestation of plants of southern and eastern Africa, 2nd ed. Edinburgh: Living- cattle tick Boophilus microplus (Canestrini). Aust. J. Agfi. Res. stone; 1962. 7:227-232; 1956. 284. Webber, L. A.; Edman, J. D. Anti-mosquito behaviour of ciconii- 258. Soulsby, E. J. L. Helminths, arthropods and protozoa of domesti- form birds. Anim. Behav. 20:228-232; 1972. cated animals, 7th ed. Philadelphia: Lea and Febiger; 1982. 285. Weinberg, E. D. Role of iron in host-parasite interactions. J. Infect. 259. Starkey, R. H.; Orth, D. W. Radioimmunoassay of human epidermal Dis. 124:401-410; 1971. growth factor (urogastrone). J. Clin. Endocrinol. Metab. 45:1144- 286. Whatson, T. S. Development of eliminative behavior in piglets. 1153; 1977. Appl. Anim. Behav. Sci. 14:365-377; 1985. 260. Struhsaker, T. T. Social structure among vervet monkeys (Cercop- 287. Whitaker, L. M. A resume of anting, with particular reference to a ithecus aethiops). Behaviour 29:83-121; 1967. captive orchard oriole. Wilson Bull. 69:195-262; 1957. 261. Sutherst, R. W.; Floyd, R. B.; Bourne, A. S.; Daltwitz, M. J. Cattle 288. Wiesbroth, S. H.; Friedman, S.; Powell, M. The parasitic ecology of grazing behavior regulates tick populations. Experientia 42:194-196; the rodent mite Myobia mulculi. I. Grooming factors. Lab. Anim. 1986. Sci. 24:510-516; 1974. 262. Sykes, A. R.; Coop, R. L. Chronic para,~ • ,m and animal efficiency. 289. Wikel, S. K. Immunomodulation of host responses to ectoparasite ARC Res. Rev. 3:41-46; 1977. infestation--an overview. Vet. Parasitol. 14:321-339; 1984. 263. Takasaki, H.; Hunt, K. Further medicinal plant consumption in wild 290. Wilkinson, G. S. Social grooming in the common vampire bat, chimpanzees. Afr. Stud. Monogr. 8:125-128; 1987. Desmodus rotundus. Anita. Behav. 34:1880-1889; 1986. 264. Tashiro, H.; Schwardt, H. H. Biological studies of horse flies in 291. Williams, R. E.; Hair, J. A.; McNew, R. W. Effects of Gulf Coast New York: J. Econ. Entomol. 46:813-822; 1953. ticks on blood composition and weights of drylot Hereford steers. J. 265. Taylor, R. J. Grazing behavior and helmintic disease. Br. J. Anim. Econ. Entomol. 70:229-233; 1977. Behav. 2:61-62; 1954. 292. Williams, R. E.; Hair, J. A.; McNew, R. W. Effects of Gulf Coast 266. Taylor, R. J. Predation. New York: Chapman and Hall; 1984. ticks on blood composition and weight of pastured Hereford steers. 267. Tesky, H. J. On the behavior and ecology of the face fly, Musca J. Parasitol. 64:336-342; 1978. autumnalis. Can. Entomol. 101:561-576; 1969. 293. Wimberger, P. H. The use of green plant material in bird nests to 268. Tobler, I.; Borbely, A. A.; Schwyzer, M.; Fontana, A. Interleukin-1 avoid ectoparasites. Auk 101:615-618; 1984. derived from astrocytes enhances slow-wave activity in sleep EEG of 294. Wing, E. J.; Young, J. B. Acute starvation protects mice against the rat. Eur. J. Pharmacol. 104:191-192; 1984. Listeria monocytogenes. Infect. Immun. 28:771-776; 1980. 269. Toft, C. A. Communities of species with parasitic life-styles. In: 295. Wrangham, R. W.; Nishida, T. Aspilla spp. Leaves: A puzzle in the Diamond, J.; Case, T. J., exis. Community ecology. Cambridge, feeding behavior of wild chimpanzees. Primates 24:276-282; 1983. MA: Harper and Row; 1986:445-463. 296. Zinkernagel, R. M.; Hengartner, H.; Stitz, L. On the role of viruses 270. Toth, L. A.; Krueger, J. M. Alteration of sleep in rabbits by in the evolution of immune responses. Br. Med. Bull. 4h92-97; Staphylococcus aureus infection. Infect. Immunol. 56:1785-1791; 1985. 1988. 297. Zuk, M.; Thornhill, R.; Ligon, J. D. Parasites and mate choice in red 271. Toth, L. A.; Krueger, J. M. Effects of microbial challenge on sleep jungle fowl. Am. Zool. 30:235-244; 1990. in rabbits. FASEB J. 3:2062-2066; 1989.