The IUCN/SSC Canid Specialist Group's
African Wild Dog Status Survey and Action Plan
(1997)
Joshua R. Ginsberg, Kathleen A. Alexander, Sarah L. Cleaveland, Scott R. Creel, Nancy M. Creel, Nancy Kock, James R. Malcolm, J. Weldon McNutt, M.G.L. Mills & Robert K. Wayne
Chapter 8 of this Action Plan outlines the information needed to conserve wild dogs more effectively. However, our knowledge has increased dramatically over the past 10 years. This is partly because researchers studying wild dogs in various parts of Africa have made efforts to use similar techniques, making the data they collect directly comparable across ecosystems (e.g. Fuller et al. 1992). In this chapter, we describe some of these established techniques, and also discuss some new techniques which wild dog researchers and managers might find useful. Some of the techniques are only appropriate for use in intensive research programmes, but many do not require that information be collected systematically. Almost all wild dog sightings, even those recorded by observers who have never seen wild dogs before, can be useful, especially if sightings are accompanied by photographs. Wild dog research programmes can, therefore, be organized at local, national and international levels. We would be very grateful if anyone collecting data of any kind on wild dogs would make contact with the Canid Specialist Groupµs Lycaon Working Group, which is chaired by Dr Gus Mills. His address is given at the end of this Appendix.
Some of the most important information to be collected about wild dogs continues to concern their distribution and abundance. The Canid Specialist Group would very much appreciate details of any wild dog sightings anywhere in Africa. Where wild dogs occur, their numbers can be estimated using photographic surveys: these take advantage of the fact that every wild dog has a unique colour pattern on its fur. Thus, photographs taken by researchers, tourists and other observers can be used to obtain a direct measure of wild dog numbers. Furthermore, if such surveys are coordinated over larger areas, they can be used to identify movements of individuals between protected areas: for example, a wild dog from the reintroduced population in Hluhluwe/Umfolozi Park in South Africa was photographed in Itala Game Reserve some 150km away (Maddock 1992).
Since photo-surveys rely on individual recognition of animals, they will produce a minimum estimate of population size. However, given enough photographs, such surveys should be less biased than other survey methods, since repeated sightings of the same animals will be recognized.
Photographs and sightings can be requested from field staff and tourists. People are asked to accompany each set of photographs with details of the date and time of sighting, the locality, and the number of adults and pups (animals approximately half adult size) seen. Results from a large-scale survey in Kruger National Park, South Africa, showed that the same individual dogs were seen together repeatedly, within well-defined areas, and that a similar number of individuals were seen on each occasion. Thus, packs could be identified (Maddock & Mills 1994). Left- and right-side photographs of all individuals from each pack were pasted onto large cards and each was given a reference number. Wherever possible, the researchers followed up sightings by field staff and tourists and took additional photographs - especially useful in matching left- and right-side photographs. This process allowed a direct count of the total number of individually recognizable dogs seen (Maddock & Mills 1994).
In Botswana, J.W.M. (Unpublished data) tried using coloured ear tags to help tourists and field staff to identify individual wild dogs and wild dog packs. However, this proved unhelpful: tourists often failed to see ear tags, and did not report the colour of the tag consistently. Furthermore, the tags caused damage to the ears which did not heal due to licking and further abrasion by other pack members. The tags did not seem to bother the dogs excessively, but they did not prove useful and eventually the researchers removed them.
In Kruger, photographs were collected from tourists by means of a photographic competition run by the Endangered Wildlife Trust. The competition offered worthwhile prizes as incentives to tourists to submit pictures (Maddock & Mills 1994). Other surveys have taken slightly different approaches to attracting tourist photographs: for example, researchers in Selous offer to pay back the costs of sending them copies of photographs (S.R.C. & N.M.C. Unpublished). However, while such reward systems are helpful, they seem not to be essential for the success of surveys. Wild dog researchers carrying out such surveys have found that tourists are usually happy to help, although some tour guides are more reluctant. For this reason, every effort should be made to contact tourists directly. Posters and leaflets requesting wild dog sightings and photographs should be boldly displayed where they will be seen (in lodges and campsites) and should, ideally, be available in several languages. Advertisements in local and international natural history magazines, and through wildlife clubs, may also help to raise awareness of the need for wild dog photographs, and might increase the number of sightings reported from outside protected areas. In all cases, the costs of such surveys - mostly the production of sighting sheets and posters - can be kept down by cooperation among projects. Researchers intending to start such surveys may, therefore, benefit from contacting the C.S.G. to check the current situation of such cooperative efforts.
Sightings by field staff are also crucial to wild dog surveys: rangers represent a body of skilled observers, often travelling in areas which are not frequented by tourists. The efficiency with which such information is collected from field staff can be improved in a number of ways. Game scouts may be given forms to fill in, asking for details of wild dog sightings: motivation may be improved by asking a particular individual in each patrol to take responsibility for recording sightings. A local member of staff - for example, a parks ecologist - should oversee the collection of sightings. In many areas (e.g. Kenya, Tanzania) such techniques are already in operation. However, the Canid Specialist Group would be happy to provide advice and assistance with collating results in areas where such surveys are not yet underway.
As discussed in Chapter4, an important question in wild dog ecology concerns the factors which limit their numbers. Evidence is accumulating that competitors and predators might keep wild dog numbers low (Chapter4). Surveys of spotted hyaenas and lions are, therefore, of interest both in areas where wild dogs occur, and in areas that are being considered as sites for wild dog reintroduction.
A rapid and reliable technique for censusing spotted hyaenas can be achieved by using sound. A six-minute long tape of sounds known to attract spotted hyaenas - the bleating of a wildebeest calf being killed by hyaenas, as well as sounds of hyaenas feeding at a carcass, mobbing lions and involved in territorial fights - is played over two 8W horn speakers pointing in opposite directions. This attracts hyaenas from distances varying from about 2-3.5km, depending on conditions. This tape is played at a succession of calling stations, each situated about 10km apart to prevent double-counting. The tape is played for half an hour at each station, so up to 10 stations can be surveyed on any one night. Copies of the tape and a detailed report of the technique and results obtained can be obtained from Dr Gus Mills, whose address is given at the end of this Appendix.
Lions must be censused using individual identification techniques for accurate estimates. Individuals can be identified by the pattern of whisker spots on the muzzle, so the technique is very similar to the one used to survey wild dogs. We urge that photographs of lions also be collected by people engaged in wild dog censusing efforts.
Censusing lion roars using acoustical playbacks is another way to estimate densities. This method will give only a minimum estimate, but can at least be used to obtain an idea of relative lion densities in different habitats.
Information is accumulating to suggest that wild dogs are not limited by the availability of their prey - although this might not be the case in lower density populations or those inhabiting very arid areas. Testing whether wild dogs are food-limited depends in part upon studying their diet, and, for this reason, we present here a number of techniques currently in use in field studies of wild dogs.
There are several methods by which food acquisition of wild dogs can be measured. The most frequently used is generally termed direct observation. The observer uses a vehicle to follow packs that are foraging, and then records all the kills made, as well as other features of the animals' hunting behaviour. This method provides the least biased data, and must be the first choice. In open, flat habitats like the Serengeti plains, the method is relatively easy to apply. In areas of thicker bush and broken ground, such as the Kruger National Park and Hwange, it becomes more difficult to follow wild dogs and more damaging to vehicles. This can only be accomplished in certain, more open areas, and then only with the aid of radio-tracking apparatus. Furthermore, it is impossible to record details of hunting behaviour as is possible in open areas.
Although wild dogs are predictable in their hunting behaviour in that they are normally crepuscular, they do sometimes hunt at night. Then the problems mentioned above become even more difficult to overcome. Direct observations need to be carefully planned, especially when consumption rate is being measured. The time period for which the dogs are going to be followed should be determined before each observation session, to avoid the temptation to continue observations if the dogs have not killed for a while, or to terminate observations when it seems likely that the dogs will not hunt again for some time.
Where direct observations are impossible, several indirect methods may be used. It is important to realize the biases that these indirect methods introduce. Faecal analysis is a useful indirect method. Dogs often defaecate around daily resting sites, along roads and at dens, so it should be possible to collect a large enough sample when dogs are radio-collared. Because dogs from one pack almost always feed from the same carcass, caution must be exercised in sampling. Only one scat per pack per day should be collected.
Observations of passage rate and number of defaecations per meal for different prey species would help to improve the accuracy of faecal analysis. Such experiments, which could be conducted in captivity, would help wild dog researchers to interpret the results of faecal analysis more accurately.
In areas where field staff and others patrol regularly, opportunistic observations of wild dog kills can provide some information on food habits. The main problem with this technique is that it is biased towards larger prey and is the least accurate method for documenting food acquisition.
Information on wild dogs' feeding behaviour can also be collected from short follows, chance sightings and even photographs, by using 'belly scores' to assess how full the stomach is. Such data give no information about the prey species wild dogs are catching, but can give a measure of food intake. By recording belly scores during continuous follows with good simultaneous data on diet, researchers can calibrate such scores with known categories of food intake.
The number of belly score categories should be kept low (three to five) so that categories remain sufficiently distinct to be reliable. We suggest the following four categories:
Belly scores are visual estimates, and are therefore subjective. Because of this subjectivity, it is important to assess consistency of scores among observers, and to minimize variability of scores. A good method is to prepare a card with drawings (or, better, photographs) of dogs with each belly score. This provides a consistent standard against which belly score estimates may be checked, and should improve inter-observer reliability.
Several researchers have noticed one complication: belly scores may differ systematically between the sexes. These researchers noted that after most kills females had consistently lower scores than did males. It is not yet resolved whether differences in belly scores reflect real differences in food intake or sexual dimorphism in the anatomy of the abdomen. If sex differences in belly scores do not reflect real differences in feeding behaviour, the complication could be resolved in one of two ways: (i) separate scales could be prepared for each sex, or (ii) sex differences in score could be accounted for statistically using a single scale. The second method is preferable because it does not complicate the scoring process, and makes no subjective assessment of which belly sizes represent equal feeding access for males and females. If, however, sex differences in belly sizes prove to reflect real differences in the amount of food eaten by males and females, separate scales for the two sexes would be invalid.
Regurgitation, which occurs at the den, is often easier to record than hunting or feeding. Researchers at the 1992 meeting in Arusha therefore considered the possibility that food limitation of wild dog populations might be quantified from measures of regurgitation rates.
In Serengeti, well over 90% of all regurgitation took place within 12 hours of feeding and any regurgitation therefore provided some measure of food consumption (J.R.M. Unpublished data). In one pack observed in 1985 at a time of food stress, yearlings failed to regurgitate to pups at the den despite their priority at kills. The yearlings stole food regurgitated by the adults to the young pups. This pack eventually abandoned their pups in response to food stress. In this extreme case, regurgitation could have been used to infer food limitation. However, there were many other cues, notably the long absences of the pack from the den, and the poor condition of the pups.
The weight of food available to Serengeti packs correlated with the number of subsequent regurgitations: dogs in packs that had more food available to them regurgitated more often. However, there was a great deal of variation in the rate of regurgitations, both among packs and among different members of the same pack. It appeared that even a small number of adults could easily supply the food needs of pups at least to 10 weeks of age. Even though the packs studied were small, the adults gave only about 10% of the food they ate to the pups at each regurgitation. Dogs on average regurgitated three or four times after a full meal: maybe a third of their intake. That wild dogs can provide more was illustrated by a female that regurgitated 11 times after a meal. The packs in this study did not appear to be food stressed: it was not uncommon to see adults approach the pups as if to regurgitate but leave as no pups begged. Uneaten food often lay around the den. The largest pack, with the highest food availability, gave the smallest proportion of its food to the pups - the pack with least food seemed able to compensate for the lower consumption by increasing the frequency with which each group member regurgitated to the pups.
These results suggest that regurgitation will seldom provide a useful measure of food intake. It is possible that a food-stressed pack, living on comparatively small prey items where a single kill does not provide an immediate glut of food, would regurgitate to pups in proportion to what they ate. Otherwise, the large food items and small food requirements of pups at a den mean that regurgitation rates will rarely provide a measure of the amount that packs are eating.
As discussed in Chapter4, diseases represent an extremely important problem in the conservation and management of wild dog populations. It is very important, then, that wild dogs be screened for infectious diseases whenever the opportunity arises. Screening of live animals usually requires that individuals be immobilized and is only appropriate, therefore, for intensive research projects. Carcasses of dead wild dogs can also be screened for disease, and we describe techniques for collecting such samples in the next section.
Researchers who are capturing wild dogs for any reason should collect blood samples for disease screening. Ideally, collect two blood samples, one into a large (10ml) vacuum tube without anticoagulant, and another into a small tube (2ml) containing EDTA or heparin as an anticoagulant. The larger sample is for serological screening and is the more important one. The smaller sample allows full blood counts to be performed, and also allows screening for blood parasites. Make blood smears from it and fix them immediately in methanol. Keep the rest of the sample cool and submit it to a lab for blood counts within a day if possible.
The larger sample - the one without anticoagulant - must be centrifuged. Pipette off the serum, keep it cool and freeze as soon as possible. The sera should be divided into at least two separate samples, and one sample should always stay in the country of origin with the appropriate government agency. Aliquots of the same sample should be stored in two separate freezers to protect against sample loss as a result of freezer failure. Sera can be stored almost indefinitely at -70·C before testing for antibodies, enzymes etc. Do not forget to maintain (and update) a log of all samples, along with information about the animals from which the samples were taken and the places where the samples are stored.
Screening of samples can often be carried out by local veterinary laboratories. However, if anyone needs assistance in organizing the analysis of samples taken from wild dogs, they should contact Nancy Kock, who is Chairman of the African section of the Wildlife Disease Association. Her address is given at the end of this Appendix. The results of disease screening should be reported to the appropriate government agency in the host country, even if screening is carried out elsewhere.
Quantifying the causes of wild dog mortality forms an important part of assessing their local conservation needs. In particular, examining dead wild dogs for evidence of disease may provide a crucial warning that local disease control is necessary. Even animals that are known to have died from other causes (such as road accidents or predation by lions) may be carrying diseases that threaten other population members. Post-mortem examinations are best carried out by vets or other qualified personnel, but in some cases this is not possible. Since diagnoses often rely upon the collection of tissue samples as soon after death as possible, it may be more useful for inexperienced observers to collect samples immediately, than to wait hours or days for a vet to be available. For this reason, we outline, here, a protocol for post mortem examination of any wild dogs found dead in the field. If at all possible, photographs should be taken throughout the examination - or, better still, the whole process should be recorded on video - this can be very helpful in arriving at a diagnosis. Finally, remember that useful information about wild dog genetics can also be obtained from carcasses - indeed, carcasses which are far too decomposed for disease screening can often yield useful samples for genetic analysis. We describe protocols for collecting genetic samples in the next section.
Keeping in mind that over 75% of the infectious diseases that affect animals may also occur in humans, it becomes obvious that in performing necropsies observers must ensure that they are protected from potential pathogens. Gloves and protective clothing should be worn wherever possible, although this may not be possible in the field. Alternatively, one can guard against infectious diseases by washing up thoroughly afterwards. Do not smoke or eat while carrying out a post mortem examination. Do not cut towards yourself or others, and if accidental cuts do occur, attend to them immediately with appropriate flushing and antiseptic.
Use the correct instruments if they are available - only a few are necessary: strong, sharp knives, a sharpening steel or stone, scissors, forceps, scalpel handles and blades, a hacksaw or rib cutters, and possibly a small hatchet. Perhaps the most important samples to take from wild dog carcasses are brain samples, and for these you will need ordinary drinking straws, about 5mm in diameter.
It is wise to plan ahead for samples that you might submit for bacterial or viral culture. Sterile swabs and transport media are available for bacterial samples, and viral samples can be frozen, preferably in sterile vials. Parasites, along with tissues, can be fixed in 10% buffered formalin. 10% formalin can also be used to store samples of brain tissue. In addition, if possible additional brain samples should be stored in a 50% solution of bidistillated glycerin in phosphate buffered saline mixed with 10-4 thimerosal (also called thiomersal or thiomersalate).
Remember that any tissues you preserve will be interpreted to best advantage if they are not damaged at the time of necropsy, so treat them gently. If you need to palpate something, do not do so until a portion has been safely placed in formalin. Although it is often easier to examine tissues after the blood has been washed off them, take samples first, as water will damage the tissues.
Tissue samples should be 5-10mm thick, and placed in about 10 times their volume of 10% buffered formalin. Hollow organs may be opened and their contents (e.g. faeces) removed before fixation. Once the tissue is fixed, you can drain away most of the formalin, leaving just enough to keep them moist, and submit them for examination by post if there are no local experts who can interpret the results.
It is important to select samples carefully for bacterial culture: remember that by about 24 hours after death invading bacteria may obscure results, making culture for pathogens unwarranted.
It is usually possible to have samples examined by local veterinary laboratories. If this is not possible, Nancy Kock is willing to examine histological samples fixed in formalin. Her address is given at the end of this Appendix.
As discussed in Chapter2, the study of wild dog genetics can yield useful information for their conservation. For this reason, Dr Robert Wayne of the Canid Specialist Group is keen to receive tissue samples from wild dogs for genetic analysis. Samples can be collected from living wild dogs in the course of immobilization by researchers carrying out intensive field studies on wild dog populations. In addition, however, useful information can also be obtained from samples taken from wild dog carcasses found anywhere in Africa - road kills are a good source, and even samples from decomposed carcasses can be useful. Dr Wayne is especially keen to receive samples from West and central Africa, but will welcome any samples that are sent to him. His address is given at the end of this Appendix.
Draw blood samples into vacutainer tubes containing EDTA. You can then follow one of four protocols which are, in order of preference:
Any wild dog carcass can yield useful genetic samples, which are easy to collect. New techniques mean that researchers may be able to extract DNA from almost any tissue that was once living, even materials such as hair, skin and bone, and even if the tissue is several years old and dried or decayed. Please do not throw anything away if it might be important! If you find a wild dog carcass, do please try to collect samples from it. The best tissues are, in order of preference, heart, tongue, skeletal muscle, kidney and liver. Heart and skeletal muscle are the best, but any tissue will do. Collect a sample 1-2 cm across. If at all possible, place the sample in a ziplock bag and freeze it. For liquid nitrogen storage, wrap the samples in foil or place them in cryo-safe freezer vials. These samples can then be shipped packed in dry ice. However, if refrigeration is not available, chop up the sample into 1mm pieces and place it in a container with the preservative solution described above, or 90% EtOH.
Please contact Dr Wayne before sending samples, to avoid problems with importing them into the U.S. Do not hesitate to contact him should you need supplies for collecting genetic samples from wild dogs.
Dr Gus Mills,
Chairman, Lycaon Working Group,
Kruger National Park,
Private Bag X402,
Skukuza,
1350 South Africa.
Dr Nancy Kock,
Associate Professor,
Department of Paraclinical Veterinary Studies,
University of Zimbabwe,
P.O. Box MP 167,
Mount Pleasant,
Harare, Zimbabwe.
Fax: +263 - 4 - 333407 / 335249
Dr Robert K. Wayne,
Department of Biology,
621 Circle Drive South,
University of California at Los Angeles,
Los Angeles, CA 90024, U.S.A.
Tel: ++1 - 213 - 825 - 9110 (work)
++1 - 213 - 825 - 5014 (lab)
++1 - 213 - 470 - 8968 (home)
Fax: ++1 - 213 - 206 - 3987
Return to Wild Dog Action Plan Table of Contents, AWD Species Account, CSG Publications or CSG Home Page.
© 1997 International Union for the Conservation of Nature and Natural Resources.