INTRODUCTION

It has often been thought that cooperative hunting favors sociality in carnivores because it allows them to kill and eat larger prey (Creel and Creel, 1995). This has been considered true in the past due to observations of groups increasing the size of prey hunted. Wolves (Canis lupus) will hunt moose together while hunting smaller prey such as Dall sheep (Ovis dalli) when alone (Mech, 1970). There are both costs and benefits to belonging to a cooperative hunting group. For example, solitary hunters may gain more food but must consider the cost of predation. It's also likely that group hunting would cost an individual some of the prey it could have eaten for itself. Conversely, by belonging to a group, dilution would decrease the chances of this organism being preyed upon (Packer et al., 1990). However, with social carnivores, one can expect a complex organization for group behavior. These carnivores not only hunt together but often take care of each other's young and protect territories held by the group. These group behaviors can make it difficult to find definite support for hypotheses that predict there is an optimal group size during hunting. Evolutionarily speaking, the optimal group size would be the number that maximizes each individual's fitness. However, each individual's fitness is determined by a number of environmental factors. An individual does not only need enough food, it needs to be able to detect and catch the prey, protect the prey from competitors, and protect itself.

Optimal group hunting size has been determined in past studies by examining hunting success. Many different species have been researched, and it is particularly interesting to observe species that coinhabit an area and utilize the same resources. African wild dogs (Lycaon pictus) are interesting to research because they not only have a declining population, but they compete with lions (Panthera leo) and hyenas (Crocuta crocuta) and use larger hunting groups than most of their competitors. A better understanding of wild dog behavior might help in saving a quickly declining population. The following paper reviews the influences on group size in wild dogs and the possible factors leading to the decline and extinction of the African wild dog.

INFLUENCES ON GROUP SIZE

Optimal group size for African wild dogs is probably not always the same. Birds and mammals have often been thought to have increased hunting success as the number of the group increases (Caraco and Wolf, 1975). Yet, Caraco and Wolf (1975) also assert that social animals have the ability to adjust their group size according to the most important environmental influences. Many factors could cause fluctuations in group size: feeding requirements, prey size, vegetative cover, offspring requirements, and competition are all thought to play a part. Caraco and Wolf (1975) also considered sex ratios of a group when studying lions. They found female lions to be more successful when hunting. Packer et al. (1990) found that the efficiency of a pride of female lions was based solely on maternity groups even though the group size was close to that which would maximize feeding. All of these factors could possibly influence group size in African wild dogs as well.

In Masai Mara National Reserve in Kenya, overall hunting success for a pack of 12-29 wild dogs was 51% (Fuller and Kat, 1993). In the Serengeti National Park in Tanzania, 9-15 wild dogs were observed to have a success rate of 31.6-35.8% when hunting adult gazelles (Fanshawe and Fitzgibbon, 1993). In another study in Tanzania, Creel and Creel (1995) found a 45% success rate in packs consisting of three to 44 wild dogs. Many other studies have shown success rates ranging from 39-85%. In almost all studies, gazelle and impala were chosen more by wild dogs, and produced better success rates, than other prey. This information, however, gives little insight as to an exact number of wild dogs that will maximize their food intake. Fanshawe and Fitzgibbon (1993) as well as others, noted that the entire pack does not always participate in the hunt. They found that the number of hunters depended on the type and age of prey. One to two wild dogs would hunt young gazelle while numerous dogs would hunt adult prey. The only significant difference they found between group sizes was that groups of four had a 73% success rate while single hunters and numerous hunters averaged 23.5%. This does not support the commonly held belief that as group size increases, hunting success increases. Other studies also challenged this belief. Fuller and Kat (1993) found that prey abundance was related to success and pack size was not. Yet, Creel and Creel (1995) continued to support the hypothesis that an increase in group size seems to have an effect on hunting success. They found a 42% success rate in packs of three adults and an increase in success to 67% in packs of 20 adults. They found these larger packs not only killed more, but killed heavier prey and chased the prey shorter distances.

PREY SIZE AND ABUNDANCE

In most studies, prey size and prey abundance were taken into consideration as major influences on optimal group size. Creel and Creel (1995) found Impala, a very common species in Tanzania, were hunted the most, killed the most, provided a 64% success rate, and yielded 20.4 kg/hunt. Of all species hunted, wildebeest provided the greatest mass/hunt: 35.2 kg/hunt. Fanshawe and Fitzgibbon (1993) found gazelle to be the most popular prey. Success rate exceeded 90% when wild dogs hunted young fawns and a rate of 31.6-35.8% when hunting adult gazelles. They also estimated 18kg of meat/gazelle hunted. Again, small abundant ungulates were hunted more in Masai Mara National Reserve in Kenya. Fuller and Kat (1993) calculated 75-100% of young gazelles hunted were captured while adult gazelles yielded a 49-66% success rate. They also found that as prey density increased, pack size increased. However, prey type did not seem to be affected by fluctuations in the pack size. Other research done with lions found that as prey biomass increased, lion group size increased (Caraco and Wolf, 1975). Studies in Tanzania on lions found that during seasons of prey abundance no difference was documented on success of female lions. However, in times of prey scarcity, there did seem to be a shift in group size. Groups of two to four females had the lowest payoff of prey consumption. This data does not seem to clarify the effects prey size and abundance have on pack size. There must be other important environmental factors influencing pack size.

FEEDING REQUIREMENTS

For wild dogs as well as other social carnivores, there is a maximum mass of prey that can be consumed. Creel and Creel (1995) found that as group size increased, the mean mass of prey consumed increased. Packs of three dogs averaged 16kg while packs of 20 averaged 40kg. Fanshawe and Fitzgibbon (1993) estimated that packs of three to four dogs yielded the highest meat/hunt/dog when hunting wildebeest, and solitary hunters yielded the most food when hunting gazelle. This contradicts Creel and Creel (1995), but it is difficult to make sense of this data because in the study by Creel and Creel (1995), the actual food consumption per day was only 2.0-2.5 kg/dog/day. Estimates show that wild dogs have a stomach capacity of 9 kg per day, so anything in excess of this may be irrelevant or could be necessary due to attacks from competitors. It is evident here that more research needs to be done to clarify these relationships. In referring back to prey abundance, it is quite possible that there is no effect on pack size when prey is abundant because there is enough meat for everyone. Maximizing food intake may only be important in times of starvation or decreasing food supply in wild dog habitat.

VEGETATION

Although these packs in some studies seem to kill more than is possible to consume, a pack may need to kill greater amounts due to attacks on the food by other species. The amount of vegetation in the area could play a role in the amount of group members needed to protect a kill as well as chase and capture a kill. Depending on the type of prey being chased, wild dogs may need more members to obtain the kill. For example, with large defensive prey that do not run from attacks, more dogs may be needed to distract or cause confusion so as to break up large defensive circles (pinwheels) that prey such as wildebeest form. In wooded areas, prey do not flee in a straight line, and it may be necessary for one dog to chase the prey while other dogs intercept the prey (Creel and Creel, 1995). However, Fanshawe and Fitzgibbon (1993) found that the amount of vegetative cover did not play a role in the success of these hunts. One must consider vegetation as one way wild dogs could approach their prey without being detected, but in the study by Fanshawe and Fitzgibbon (1993), wild dogs were always detected by gazelles before the chase began. Fuller and Kat (1993) reported only 6% dense woody areas in their study site which provides little information on whether wild dogs use this vegetation to improve their hunting success in any way. So it seems here that vegetation is not being used as a way to sneak up on prey. More research is needed to obtain conclusive evidence of the influences of vegetation. Many wild dog habitats in Africa, however, provide little dense cover, so vegetation may play only a small role in effecting wild dog group size.

COMPETITION

Since it does not seem likely that wild dogs use dense vegetation to hide their kills from other competitors, some other technique must be used to ward off lions and hyenas. Fuller and Kat (1993) found hyenas at 41% of all wild dog kills. Fanshawe and Fitzgibbon (1993) also found hyenas at a large percentage of kills by wild dogs. Groups of more than four hyenas had a significant effect on how long wild dogs would remain at the kill site. One to 18 hyenas were spotted at virtually all kills. They found that hyenas were quite capable of stealing the prey item before the wild dogs were finished consuming it if the prey item was large. Small prey were consumed before hyenas arrived at the kill. Particularly in the Serengeti in Africa, hyenas have been increasing in number in the past few decades. These hyenas may be an important reason for wild dogs to maintain larger packs. If prey supply is low and wild dog packs are small, interspecific competition could cause a decrease in the amount of food these wild dogs can maintain. Similar evidence was found with hyenas scavenging kills by lions (Caraco and Wolf, 1975). They found that if less than four lions remained at a kill, 10% of the meat on average was scavenged by hyenas. Hyenas are not always abundant in wild dog territory, so in some cases it is difficult to assess the impact of these competitors. Occasional competition also came from lions in Tanzania. In all encounters with lions, all of the kills were lost. Creel and Creel (1995) also documented scavenging by wild dogs and found only 10 instances of wild dogs stealing carcasses. Three times, leopards were chased in order to gain kills, one kill was taken from a lion, four kills were obtained from hyenas, and the other two kills were obtained from unidentified sources. The data from Tanzania reflects that interspecific competition may balance out between species since both wild dogs and hyenas scavenge off of each other. Interspecific competition may play a role in group size of wild dogs, but perhaps only in certain areas where scavenging by competitors is significantly high or in areas where there is an abundance of competing species such as lions or hyenas.

REPRODUCTIVE SUCCESS AND PROTECTION OF OFFSPRING

Another factor that has not been fully explored in wild dogs is that social carnivores communally care for their offspring. Some studies of lions provide information on the hypothesis that greater group size increases individual reproductive success (Packer et al., 1990). Because social carnivores tend to have births all at once in a group, protecting and feeding young can be shared by the group members. Larger groups of female lions can prevent infanticide from new males that might try to enter a pride (Packer et al., 1990). However, little evidence provides insight as to whether or not young benefit from greater mother group sizes when foraging. No data was collected on the amount of food consumed by young of large groups of mothers opposed to small groups of mothers. Data from a study on communal rearing of pups in wild dogs provides only the information that there was a positive but nonsignificant correlation between pack numbers and the number of surviving pups (Malcolm and Marten, 1982). There seem to be many factors that play a role in wild dog pup mortality. Further information could be useful in clarifying whether or not reproductive success has an influence on group size in wild dogs.

THE DECLINE OF THE AFRICAN WILD DOGS

Because there are numerous factors effecting wild dog pack sizes, it is understandable that many factors could be effecting the decline of wild dog populations. If all or some of the previously discussed issues restrict or limit wild dog hunting success, wild dog packs could significantly decrease in size. Throughout the research on African wild dogs, their populations have been noted as declining. However, it has been difficult to explore the reasons for this decline because of such a small sample size and the difficulty in controlling possible confounds in the field. Many of the study populations have entirely disappeared, and it is therefore critical to determine what the cause is and if the populations of wild dogs can be aided in recovery.

Several hypothesis have been imposed over the last couple of decades to explain the dramatic decline in the Serengeti National Park and other parts of Tanzania. In the early 1990's, many of the wild dogs were documented as dying from rabies or canine distemper and had been handled by researchers in an effort to vaccinate the populations. Burrows et al. (1994) suggested that the stress introduced on wild dogs by the research handling increased their susceptibility to diseases such as rabies. He asserted that handling may increase production of adrenocorticosteroids that would cause immune suppression and then greater chance of disease. However, other studies believe that alternative hypotheses should be explored to clarify the reasons for the population depletion. Creel (1992) found numerous problems with the techniques used by Burrows (1992) which Creel believes discredits Burrows theory. Another study by Ginsberg et al. (1995) explored density-dependence, deterministic decline and the rapid fluctuations in the number of dogs in the Serengeti in order to clarify the reasons for declining wild dog populations.

An average of 22 individuals existed in the local population from 1976 until the time of extinction in 1992 (Ginsberg et al., 1995). Using data obtained by Burrows et al. (1994), Ginsberg et al. (1995) constructed their theories on extinction by simulating situations on a computer model called VORTEX. After simulating situations in which carrying capacity and density dependence were manipulated, and in which random events and catastrophes were considered, they concluded that the extinction was possible from chance events alone (Ginsberg et al., 1995). However, they also noted that the decline in the carrying capacity of the Serengeti could have been caused by limited food, competition from other carnivores, and factors that influence disease. The wild dogs did not suffer from a lack of food. Thomson's gazelle increased 150% over the period studied and wildebeest increased by 400% (Ginsberg et al., 1995). The competition hypothesis cannot be as easily dismissed because hyenas increased by 150% and lions increased as well, but as previously discussed, the relationship between these carnivores is not clear. It is possible that over a long period of time, the competition could have limited wild dog prey consumption or territory, but this is speculative. Researchers such as Scott Creel are continuing to explore competition as a factor in limiting wild dog populations.

A more common assumption by many researchers is that canine distemper virus was the ultimate cause of the wild dog extinction. Rabies was documented in five dead wild dogs between 1989 and 1991 (Burrows et al., 1994). Data from domestic dogs in the surrounding ecosystem confirmed that 75% of these dogs were positive for the canine distemper virus and that this could have been a reservoir for the disease (Ginsberg et al., 1995). Canine distemper has been documented in a number of endangered or declining species. In North America, black-footed ferrets (Mustela nigripes) in Wyoming nearly went extinct due to the virus, and it is speculated that it has also caused the decline in silver-backed jackals in the Serengeti (May, 1986). The rabies virus has also been observed in Serengeti bat-eared foxes (Macdonald, 1992).

Rabies was also thought to be the cause of death in a pack of wild dogs in Kenya in 1989 (Burrows, 1992). The interesting fact to note here is that a total of 58 dogs from different packs, all died after radio collaring and vaccinations by researchers. After these deaths, the remainder of the packs were vaccinated by air-pressurized darts, and within one year, all of the packs were dead. Many of these dogs had previously been documented as carriers of rabies antibodies, meaning that they must have been exposed to the disease before the vaccinations. Therefore, Burrows (1992) asserts that the stability of these wild dogs was disrupted by researchers who induced stress that then caused the virus to emerge and flourish in the packs. Unfortunately, criticisms of Burrows (1992) theory are made due to his lack of controlled data. Many of his assumptions were based on data that was not measured correctly and data that cannot be analyzed due to the lack of information on how stress chemicals operate in wild dogs. In fact, little information like this exists for any species. For Serengeti dwarf mongooses, one of the only species with known information, handling does not induce dangerous stress levels (Creel, 1992). Creel (1992) further notes that many wild dog populations have disappeared from diseases and were not previously handled. Burrows (1992) had asserted the idea that vaccinations were often not needed because many wild dogs were carriers of the virus antibodies already; however, Macdonald (1992) in response, considers the fact that these dogs may have only been incubating the disease as well as the criticism that several mishaps could have occurred when administering the vaccinations to the wild dogs that died. Whether or not these ideas are true, Macdonald (1992) continues to emphasize that rabies and other fatal canid diseases are a serious risk for wild dogs in an especially small population.

Because wild dogs exist in such small populations, it is possible that extinction risk is higher. General research on the process of extinction considers demographic stochasticity, environmental stochasticity, and catastrophes (Lande, 1993). Lande (1993) considers various statistical models concerning these population issues, and believes that small populations are especially effected by demographic stochasticity, but are also effected by the other factors. Other researchers, however, believe that the most important factor is random catastrophes followed by environmental stochasticity and then demographics. Whether or not there is a specific order of importance, all of these could effect wild dog populations. As carrying capacity decreases, extinction chances increase (Lande, 1993), and factors such as random catastrophes have a potentially greater effect on these diminishing populations. This can be observed in wild dogs to an extent, but more research is needed to make any clear assumptions about the ultimate causes and how these causes fit together.

DISCUSSION

It is clear that more research on wild dogs is needed to develop an overall understanding of wild dog ecology. Many environmental factors play a role in the sizes of wild dog populations. Prey size and prey abundance, feeding requirements, habitat, and competition all seem to have effects on group size. The effects may vary between populations, and future research could provide an hierarchy of the importance of each factor. However, it is more important to understand that there must be benefits outweighing the costs of group living in order for wild dogs to continue their communal living. There is probably a limit or carrying capacity that restricts group size after a certain number has been reached, but this number may differ depending on the environment. The question remains, however, that if living in groups is beneficial overall, and there is, perhaps, safety in numbers, why are wild dogs having such trouble in maintaining their populations?

A definite answer cannot be given. It is known that canine distemper and other viruses have caused a great deal of damage in wild dog populations, but it is important to explore the underlying influences for the decline in numbers. It is possible that stress by humans and competing lions and hyenas could cause greater susceptibility to disease in wild dogs. Another possibility is that a decrease in the environment's ability to provide food has caused decline in the wild dog populations. However, this should probably be dismissed due to large increases in prey populations while wild dogs declined. Although the increase in prey is also at the time when hyena and lion numbers increased, unless the hyenas and lions are consuming more of the increasing prey populations and restricting wild dog consumption, it is unlikely that prey abundance has a significant effect on the declining wild dog populations.

It seems more likely that many factors are influencing the decrease in carrying capacity for wild dogs. Because individual populations are not large, demographic stochasticity could play a role in the decline. The natural numbers of natality and mortality could cause a decrease in wild dogs just by chance. Although this is not the primary cause, it could be linked to other influences on birth and death rates (environmental stochasticity) and random catastrophes such as disease. With all of these factors possibly decreasing the wild dog populations, wild dogs have little chance of survival as their numbers continue to decrease rapidly. It is unfortunate that a clear reason does not exist, but it is evident that if wild dogs are to be saved, direct action must take place in the near future. If further research can better our understanding of this specific declining population, other endangered species could benefit from the understanding of how populations go extinct and how they can be recovered.
 
LITERATURE CITED

Burrows, R. 1992. Rabies in wild dogs. Nature. 359:277.

Burrows, R., Hofer, H. and East, M.L. 1994. Demography, extinction and intervention in a small population: the case of the Serengeti wild dogs. Proc. R. Soc. Lond. B 256:281-292.

Caraco, T. and Wolf, L.L. 1975. Ecological determinants of group sizes of foraging lions. Amer. Natur. 109:343-352.

Creel, S. 1992. Cause of wild dog deaths. Nature. 360:633.

Creel, S. and Creel, N.M. 1995. Communal hunting and pack size in African wild dogs, Lycaon pictus. Anim. Behav. 50:1325-1339.

Fanshawe, J.H. and Fitzgibbon, C.D. 1993. Factors influencing the hunting success of an African wild dog pack. Anim. Behav. 45:479-490.

Fuller, T.K. and Kat, P.W. 1993. Hunting success of African wild dogs in south-western Kenya. J. of Mamm. 74:464- 467.

Ginsberg, J.R., Mace, G.M., and Albon, S. 1995. Local extinction in a small and declining population: wild dogs in the Serengeti. Proc. R. Soc. Lond. B 262:221-228.

Lande, R. 1993. Risks of population extinction from demographic and environmental stochasticity and random catastrophes. Amer. Nat. 142:911-927.

Macdonald, D.W. 1992. Cause of wild dog deaths. Nature. 360:633-634.

Malcolm, J.R. and Marten, K. 1982. Natural selection and the communal rearing of pups in African wild dogs. Beh. Ecol. Sociobiol. 10:1-13.

May, R.M. 1986. The cautionary tale of the black-footed ferret. Nature. 320:13-14.

Mech, L.D. 1970. The Wolf: The Ecology and Behavior of an Endangered Species. Minneapolis: University of Minnesota Press.

Packer, C., Scheel, D., and Pusey, A.E. 1990. Why lions form groups: food is not enough. Amer. Nat. 136:1-19.