Population surveys of the Buton macaque, Macaca ochreata brunnescens were carried out on Buton island, south east Sulawesi in 1999 using two different methods of estimation. Simple line transects were carried out and a new method referred to as the track and count method was also used which guarantees a minimum amount of macaques in the area under study. Both methods were carried out around two village areas, Kaweli and Kambowa, in order to compare the results. It was found that the line transect method gave lower estimates than the track and count method in both cases (line transects underestimated known minimum population levels by 43.45% in Kaweli and 67.91% in Kambowa). Finally line transects only were carried out in undisturbed primary rainforest in order to compare the population levels of primary forest with that of the secondary forest around the village areas. Higher population levels were found in the secondary forest than in the primary forest (23.3 macaques/ km² around Kaweli compared to only 12.5 macaques/ km² in the primary forest). Reasons for this include a combination of high food availability from the farmland surrounding the village areas and lack of preventive measures carried out by local farmers. The sparsely distributed food supply in the primary forest is also a contributory factor.

KEYWORDS: Macaca ochreata brunnescens; population density; line transect; track and count method; secondary forest; primary forest.

The genus macaca (from the group Cercopithecine) is very successful and found only in Asia and a small part of Europe. The nineteen species of macaque span from Morocco and Gibraltar in the west to Japan, Taiwan, the Philippines, Sulawesi and Bali in the east. Macaques are medium-sized primates and are relatively generalised in many aspects of their anatomy compared with other members of the sub-family (Fleagle 1988). They are characterised by long snouts, high-crowned molar teeth with very low cusps and long third molars (Rosen 1974). They vary considerably between species in the extent to which they are arboreal or terrestrial, however, all species use both settings to some extent. All macaques are frugivorous, but may consume considerable amounts of leaves, flowers and other plant material as well as animal prey (Fleagle 1988). Macaques live in relatively large, multimale social groups, with troops of some species containing up to or more than fifty individuals. The subject of this investigation is the Buton Macaque, M.ochreata brunnescens. The Buton Macaque is endemic to Southeast Sulawesi and one of its major strongholds is on Buton island. To date there have been no population studies carried out on this species (and relatively few general studies), however, there have been many studies carried out on a closely related species, the Sulawesi Black Crested macaque (Macaca nigra) which inhabits most of north Sulawesi and many of the islands located off the north-east tip. The brunnescens sub-species of Macaca ochreata is found only on Buton Island, however, M.ochreata is found on the adjacent island of Muna and a small part of the south-east tip of mainland Sulawesi. A photograph of M.ochreata brunnescens can be seen on the front cover of this report. The populations of macaque on Buton island are under threat as their habitat is being logged at a rate of up to 10% each year¹ , however, this seems to be the only threat to the macaque species here as they are rarely killed and few are captured for the pet trade. In contrast to the black crested macaque, Macaca nigra, few are eaten on Buton as the majority of the population here are of the Islamic faith and the consumption of monkey is considered ‘haram’ or forbidden.

Originally the main objective of this project was to estimate population densities of M.ochreata brunnescens around village areas on the island and to compare these with estimates taken in an area of Primary forest away from the village influence, however, after recording had started around a village called Kaweli it became evident that the line transect method of estimating population density was giving results much lower than the figures estimated to me by the local farmers of the area. It was for this reason that I altered my investigation. As well as estimating the population densities of the macaque species using the transect method I also constructed another method of my own which involved following troops of macaques about during the day and making accurate counts whilst the macaques were in their sleeping trees at dawn and dusk. This species of monkey, like many other species, move about during the day in order to feed and exercise, however they like to return to a known area and one particular tree or group of trees, depending on the size of the troop, in order to sleep (Fooden 1980). It was observed during my investigation that each troop of M.ochreata brunnescens circulated a large area of land each day (home range² ) in order to return back to the trees, at around 6.00pm, from which they had left.

It is due to this that I will be referring to the main trees in which the macaques were observed at dawn and dusk as sleeping trees. These two different methods are explained in more detail in the materials and methods section and were carried out around two villages on Buton island the first being Kaweli and the second being Kambowa. Map 1 shows the island of Buton and the positions of the village areas that have been studied in this investigation.

Once the two different methods had been carried out and all the data had been collected I was able to make a reasonable comparison between the two methods and deduce whether line transects really do under-estimate the levels of M.ochreata brunnescens present in the areas under study (see results section of this report). Once I had finished the comparison of the two different population estimation techniques I trekked into dense primary forest south of the village of Lawele in a region know as Lasalimu. Here I walked eight simple line transects simply to compare the estimated density of M.ochreata brunnescens in the primary forest with the estimates from the secondary forest around the village areas. I could only use the line transect method in the primary forest due to the size of the area and the absence of local people who knew the whereabouts of the macaque troops making the track and count method virtually impossible.

The main hypothesis for this investigation is that line transects do not give an accurate estimate of the population density of M.ochreata brunnescens around the villages under study and will grossly underestimate population levels. The track and count method gives a known minimum amount of individuals in the area and this will be used to compare the results obtained from the transect method. The high levels of assumption associated with line transects account for the inaccurate population estimates given in this case. The second hypothesis for this study is that I expect the transects carried out in primary forest to give higher population density estimates than the transects carried out in the secondary forest surrounding the villages because primary forest has a much higher carrying capacity due to more natural food supplies, more space for sleeping and breeding and, most importantly, less disturbance caused by human activities.

Two completely different methods of population density estimation were used in this study. One method involved interviews with local people, tracking troops of M.ochreata brunnescens and counting the number of individuals whilst they were in a fixed location (sleeping trees) and the other method used was basic line transects which are used in many scientific disciplines in order to estimate the population size of a variety of different organisms. Both methods were carried out in the two village areas but only the line transects were carried out in the Primary forest in the region of Lasalimu.

Each line transect walked was approximately 1.5 km in length and had an visual width of 100 m, therefore giving a survey area of 0.15 km² for each transect walked. The numbers of M.ochreata brunnescens counted during the transect were cross checked by at least two other observers and trails were walked at about 7.00am and 3.00pm when the macaques were most active. It was noted that troops of M.ochreata brunnescens moved from there sleeping tree(s) at around 6.30am, just after day break and were active until around 10.00am. At this time they tended to slow down due to the excessive heat of the late morning and the afternoon. They started to become more active around 3.00pm when they had a final feeding "rush" until they arrived back to the sleeping tree(s) in time for dusk. The line transect technique that I used was almost identical to that used by Rosenbaum et al in 1996 whilst studying the black crested macaque, Macaca nigra on the north east tip of Sulawesi and the nearby island of Bacan. A straight line was always walked whenever possible (using GPS as a basic guide) and at selected intervals the GPS positions would be recorded so that the line transects walked could later be plotted onto maps. Along each transect the number of M.ochreata brunnescens that were visually identified were recorded. No sound identifications were made as it was very difficult to determine how many individuals of M.ochreata brunnescens were present simply by the distant calls. Walking had to be slow and quiet as not to disturb or antagonise the macaques and great care was taken to ensure that none of the line transects overlapped in any way.

After the line transects were carried out I noticed that the results obtained were considerably lower than the estimates given to me by the local farmers. It was due to this that I constructed the track and count method. This consisted of several key components. The first part to this technique was to interview five of the local farmers around each village and also the head man. I asked them several questions which gave me an idea of the macaque population size in the area. Such questions included how many troops of M.ochreata brunnescens were in the area and how many individuals were present in each troop. The local people gave me their estimates which were invaluable to this investigation as the local farmers come into contact with the different troops of M.ochreata brunnescens at least once everyday. The most important information that I gathered from the local people was the location of the main sleeping tree(s) of each of the different troops. Once I had the information required I set off with a local guide into the secondary forest which surrounded the farmland that bordered the village. We went to the areas where we were told we would find the different troops and tried to identify the key troops of M.ochreata brunnescens. Some troops of M.ochreata brunnescens were located after we had trekked around the range area for some time but most were found at Dusk or dawn when they were in the sleeping trees. All of the information regarding the whereabouts of each sleeping tree was considerably accurate and decreased the length of time that it took to find and count the individual macaques. Once each troop was located the main sleeping trees were marked on GPS¹ . Global positioning system (GPS) is a handheld device which gives you your almost exact location on the globe using satellite systems. This is very accurate with an error of only 5m² . This was a very effective method of recording the sleeping trees of each troop of macaque and made life much easier when it came to revisiting the desired sites. The GPS method enabled me to draw up accurate maps showing the sleeping trees and line transects in relation to the village areas. GPS also assisted me to walk straight lines when carrying out line transects. Each of the sleeping trees for each troop of macaque around the village under study was visited on a number of different days at both dawn and dusk when the macaques were not in motion. We counted the number of M.ochreata brunnescens from each troop that we could see in the trees on several occasions which spanned approximately one week and finally ended up with a maximal count² . I added the maximal counts from each troop together to determine the total density of M.ochreata brunnescens in the area surrounding the village. This result assured a minimum amount of M.ochreata brunnescens around the village under study. Once we had the combined maximal counts for each village area, and the known area of survey was calculated, I could then work out the estimated number of macaques per kilometer squared (population density) around the village area.

In Kaweli five different macaque troops occupying a survey square of 3.4 km² were identified. The total number of individuals of M.ochreata brunnescens that were identified around kaweli was 141. Below is a table summarising the results of the track and count method showing the different counts of the different macaque troops on different observations and also a reference to the troop size as predicted by the local people of Kaweli. The results of the track and count method (table 1) tell us that there is a minimum of 141 macaques in an area of 3.4 km² which gives a population density estimate of macaques around the village of Kaweli of 41.2 macaques/ km² (1.46 troops/ km² ). The above population density estimate differs greatly from the estimates calculated by the line transect method. Table 2 gives the number of M.ochreata brunnescens encountered during the six different line transects carried out in the area surrounding Kaweli. (The accurate plots of the transects walked can be seen on Map 2). The results of the line transects (table 2) give us a population density estimate of M.ochreata brunnescens around the village of Kaweli of 23.3 macaques/ km² . We know that there are a minimum of 41.2 macaques/km² around kaweli calculated by the track and count method. It can be said, therefore, that the line transect method has underestimated the population density in this case by 43.45%.

In Kambowa the exact same procedures were carried out as was for Kaweli. Here I identified six troops in almost double the area than that of Kaweli, however here the troop size of M.ochreata brunnescens was much larger. The results of the track and count method (table 3) indicate that there are a minimum of 231 macaques in an area of 6.8 km² around the main village area of Kambowa. This gives us a population density estimate of macaques around Kambowa of 33.75 macaques/ km² (0.88 troops/ km² ). In contrast to the track and count method eight line transects were completed in the area of study around Kambowa, the length and area being the same as those in Kaweli. The table below (table 4) summarises the observations made while walking the transects and Map 3 shows the positions of the line transects taken from the GPS data. The results of the line transects near Kambowa suggest a population density estimate of 10.83 macaques/ km² . Again we know that there is a minimum of 33.75 macaques/ km² around Kambowa resulting from calculations made by the track and count method, therefore, I can calculate that line transects in this example have underestimated the population density by 67.91%.

To show the different results more clearly Graph 1 shows the difference in the estimation of population densities of M.ochreata brunnescens resulting from the two techniques used in the villages of Kaweli and Kambowa. As you can see from this the line transect method of population density estimation has estimated the levels of M.ochreata brunnescens around the village areas to be much lower than the estimates from my track and count method.

Map 4: Line transects walked in Primary forest.

 

 

 

 

 

 

 

 

 

 

Graph 2: Differences in Macaque density between the two village areas and the Primary

forest resulting from the line transect method.

 

c) Primary forest.

After the two methods had been carried out on the two village areas I set about obtaining the data required to prove the second hypothesis which involves a comparison between the population densities of M.ochreata brunnescens in Primary forest and in secondary forest near village boundaries. To do this I walked eight simple line transects (carried out using the same methods that were used around the village areas) in primary forest near Lasalimu. Table 5 above summarises the data obtained and map 4 shows the positions of the walked transects. A total of fifteen individuals of M.ochreata brunnescens were observed whilst walking all eight line transects in the primary forest. This gives us a population density estimate of 12.5 macaques/ km² for primary forest. If we look at Graph 2 above we can see the difference between the population density estimates of the village areas and the primary forest obtained via the line transect method. By examining the results obtained and graph 2 above it can be seen that there is a higher population density of M.ochreata brunnescens in the secondary forest near Kaweli than in the primary forest near Lasalimu. Kambowa had a slightly lower estimate than the primary forest area, although there is only a difference of 1.67 macaques/ km² .

Statistical analysis of the data collected has proved invalid and so has not been included in this report. Much advice has been sought from different members of both the Animal and Plant Science department and the Geography department of the University of Sheffield and the overall conclusion was that statistical analysis was not appropriate in this report due to several factors. Firstly there are too many differences in variables between the two population density estimation techniques, i.e. the two methods differ in many ways making it impossible to carry out any direct statistical comparison. The end estimates given by the two techniques are clear enough to give an indication of the differences between them. The recording in the track and count method was carried out when the individuals of M.ochreata brunnescens were "clustered", i.e. in the sleeping trees, whereas the recording in the line transect method was carried out at different times of the day whilst the individuals were mobile and distributed quite randomly.

Secondly when comparing the densities from the primary forest and the secondary forest surrounding the villages, there were too few end results present (three in total - two from village areas and one from primary forest) to justify the use of a statistical test. Mean values were not used in this data as it was the maximal count that was most important and needed to calculate population density estimates. Finally time restrictions did not allow me to walk transects more than once which again restricted my analysis of the data. To make the line transects more valuable, walking them a number of times would have been needed which would then enable statistical analysis to be made on the variability of the macaque identifications along each individual transect.

If a future study was to be carried out many more areas would have to be sampled and each area would have to use transects that are walked a number of times each before statistical analysis could be used effectively.

The main two conclusions that can be made from the data collected are that the track and count method gives a more accurate population density estimate in the village areas than does the line transect method (line transects underestimated the results by 43.45% in Kaweli and 55.68% in Kambowa) and secondly that the village of Kaweli supports a higher population density of M.ochreata brunnescens than does areas of Primary rainforest near Lasalimu. Kambowa has only a small difference in population density than that of the primary forest.

This investigation has succeeded to prove my first hypothesis as it was seen that in both village areas my method of population density estimation gave a guaranteed minimum number of individuals of

M.ochreata brunnescens present in the area surveyed and when these figures were compared to the results obtained from the line transect method it was seen that the line transects grossly underestimated the known minimum levels of M.ochreata brunnescens.

Line transects are used in the field by a variety of different scientific disciplines to identify population densities of many different organisms. The most popular type of organism are birds, although the technique has been used by many workers to estimate the levels of many different primates around the globe. Line transects are also used for plant distributions, underwater coral distributions and many insect species. Line transects are one of the most popular methods of population estimation, however, there are many reasons why the line transect method can give inaccurate results (Bibby et al 1992). Firstly line transects assume that all macaques (or indeed any other organism that is under study) are visible in the trees and thus become included in the final count. It was my observation whilst undergoing this investigation that it is not possible to identify all the individuals of M.ochreata brunnescens present in a tree in a short period of time. Quiet sitting and/or following the different troops for periods of up to two hours (sometimes much longer) is needed to correctly identify the true number of individuals present in an unknown troop, for example, at one instance it was observed that a troop of M.ochreata brunnescens consisted of only seven individuals. After ninety minutes of continued observation this figure had risen to eighteen individuals. A line transect does not allow this amount of time to be spent on each troop as they will soon move from the transect area once human presence is detected and they very rarely move in the same direction of the transect path. Another problem associated with line transects is that many individuals of the organism under study may detect human presence before we detect the organism and so move from the transect area out of our visual field and do not become incorporated into the population density estimate, therefore underestimating true population levels. The opposite may also be true, i.e. if an observer walks directly into a stationary troop of individual macaques then an unusually large count is going to be made resulting in an overestimation of the population density given the small survey area of each transect, however judging by the low numbers of macaques identified during the line transects in the primary forest (where the track and count method cannot be used) I don’t think this overestimation has been the case. A smaller assumption is that individuals are counted only once. If a high level of individuals are present along a transect then there is a high risk of double counting especially when disturbed individuals move back into the transect path further along (Bibby et al 1992). This again would result in a large overestimation.

There are also inaccuracies associated with human error. There may be many errors that alter final results which stem from an observers failure to correctly identify all the individuals present and/or double count certain individuals. The error of over-estimation was not encountered during my investigation (if taking the results of the track and count method to be accurate) but the results do suggest an under-estimation in both village areas. The line transect method, however, does have its advantages. Even with the high amount of error associated with them, transect methods are the only type of population density estimation technique that can be carried out in areas such as dense primary forest. The home range of M.ochreata brunnescens in primary forest is much larger than that of M.ochreata brunnescens in the secondary forest near village boundaries. This makes them much more difficult to find, thus making the track and count method impossible and giving line transects an advantage. The track and count method that was used in the village area surveys eliminated many of the assumptions associated with the line transect method as there is little chance of double counting or mistaking a new group with one already counted. The variability in the counts during different observations is due to the arrival of darkness (when counted at dusk) and the movement of the macaque troop into inaccessible areas when my presence was detected (when counted at dawn).

One of the most important elements of the track and count method is the information gathered from the local people of each village area under study. They provided accurate information as to where the different troops could be found at different times of the day and more importantly where the main sleeping trees of the different troops were located. Graph 3 simply shows the estimated population size of each of the troops given by the local farmers in each village compared to the maximal counts made. As can be seen from this graph the information given proved very accurate in both village areas, however it was the information that helped us to located the main sleeping trees used by the different troops that was most crucial. One of the disadvantages of the track and count method is that it is not possible to carry out in areas devoid of human information. Areas where no villages exist, and consequently no local people, identifying the troops of M.ochreata brunnescens can be like trying to find a "needle in a haystack". The main difference between the two methods was time. Line transects usually lasted up to four hours each (dependant on terrain) whereas the track and count method could last up to two days just for one troop of M.ochreata brunnesscens. It can be said that even though the track and count method takes much longer to carry out it gives much more accurate results and a set of results which can be used to calculate the inefficiency of the line transect method. Line transects are, however, relatively easy to carry out, quick and require less physical demand by the person(s) carrying out the investigation.

The second hypothesis for this investigation is that there would be more individuals of M.ochreata brunnescens in Primary forest than around the villages. This has been proven incorrect and can be explained by several factors. The secondary forest near the village boundaries has a rich food supply.

The farms on the forest boundary contain a variety of different food sources including coconuts, sweet potato, bread fruit and cocoa. Photos 1 and 2 below show some of the typical foods grown on the local farms and eaten by M.ochreata brunnescens. The macaques here usually locate their main sleeping trees near the farm/forest border and each day proceed on a ‘tour’ of all the local farms within their home ranges. The owners of each farm observe the macaques trying to take food from the farm at specific times each day.

The high amount of readily available food grown in the farms around village areas compared to the randomly distributed and unreliable food supply present in the primary forest gives a higher carrying capacity for a higher abundance of macaques in the small secondary forest area immediately surrounding the villages. Combined with the high food supply, it was observed on Buton that the populations of M.ochreata brunnescens do not decline as a result of human interference around the village areas. In the areas that I studied very few individuals were killed each year (3-4 macaques each year was an average amount suggested by various local sources) as the local people do not have the resources to eradicate the problem. The local people do not possess any guns to kill off the macaques and simply do not have the time to chase after the macaque troops. Local people do set simple box traps in order to capture the monkeys however this rarely works as the macaques are not enticed by the traps. Another method that was used a couple of years ago in the village of Konde was to put poison in some of the coconuts and sweet potatoes, however the macaques recognised this "alien substance" and did not consume the fruits. Some of this infected fruit later became mixed up with the uninfected fruit and was consumed by local people resulting in many illnesses. The major mechanism for deterring macaque troops from eating local produce is the use of dogs who simply chase the macaques off the land. Alternatively local children are paid money to sit and keep watch and throw stones at the macaques, however in Kambowa this has simply resulted in increased aggression by the macaques. Despite these "primitive" measures of prevention local people in Kaweli estimate that about 50% of all farm produce is lost to troops of M.ochreata brunnescens every year. This high food supply gives rise to the elevated population levels observed in

the secondary forest around village areas. The lower population levels per squared kilometer in the primary forest simply result due to a sparsely distributed food supply, which can be unreliable at various times of the year, that means troops of M.ochreata brunnescens have to adopt a larger home range in order to maximise the food intake. The larger home ranges gives rise to more territorial behaviour which creates even more spread between the macaque troops (O’Brien et al 1997) giving the low densities of macaque per kilometer squared. In contrast less space is needed in the secondary forest around the village areas as the food supply is readily available meaning smaller home ranges for the macaque troops as they do not need to move far in order to find ample food. It may also be the case that the increased rate of logging in the primary forest of Buton has given rise to the lower populations of M.ochreata brunnescens, however, this has not yet been studied and so this theory is only speculation. A report published in 1998 by Rosenbaum et al looked at the populations of the Sulawesi black crested macaque, Macaca nigra, in North Sulawesi. Rosenbaum and his workers found the opposite situation to that on Buton. They found that there were higher densities of M.nigra present in primary undisturbed forest (170.3 macaques/ km² ) than in forest where villages were in close proximity (47.9 macaques/ km² ). (These figures, however, may be under-estimated as only the line transect method of population density estimation was used in this study.) Rosenbaum et al found that the main causes of the observed trend were deforestation carried out by local farmers and very high levels of trapping and killing in the forest near the villages. In the primary forest Macaca nigra use the tall, broad-canopied emergent trees for feeding and sleeping. In North Sulawesi Dracontomelon and Ficus species of tree were favoured by the macaques and occurred at high densities in the primary forest areas. These tree species were drastically reduced around village areas and are cited as one of the main reasons for the low numbers of macaques. If this was the only factor involved then the macaques may still retain high population densities by consuming local produce around the village areas. However in North Sulawesi there is another important factor that lowers the populations of M.nigra around the village areas. The religion of Christianity in North Sulawesi does not pose any restrictions on the consumption of monkey meat, therefore many individuals of M.nigra are captured each year to be used either as food or as an income in the pet trade. These two factors combine to keep population levels around the villages low. The location of my investigation is very different to that of North Sulawesi culturally and this could play a major role in the difference in trends.

The Islamic religion is dominant on Buton, thus forbidding any consumption of macaque meat. Also the people of Buton do not capture or kill many individuals of M.ochreata brunnescens. Both these factors give rise to the elevated densities of M.ochreata brunnescens around the village areas.

It can be concluded, therefore, that it is a combination of the high food levels and little interference by local people that gives rise to the higher population of M.ochreata brunnescens around the village areas. Further study includes looking at many more village areas around the island of Buton to add to the data in this report. When a great deal more villages have been surveyed using exactly the same methodologies as this investigation then an accurate measure of inaccuracy of the line transects could be made which could be used as a correction factor to add to areas such as primary forest where only line transects can be carried out. Many village surveys using both methods would be needed in order to obtain a reliable mean value which would greatly benefit the use of advanced statistical analysis which due to the low number of samples are not valid in this investigation. A study looking at the distribution of the main foods of M.ochreata brunnescens in primary forest in conjunction with population levels would be of great interest and may help explain the large home ranges of macaque troops in primary forest. Also the rate of deforestation needs to be closely looked at on Buton to see how it effects the population levels of M.ochreata brunnescens in terms of space and food supply.

This work was carried out under the supervision of Operation Wallacea. Funding for this project was given partly by Birmingham Local Education Authority and Yardley Educational Foundation. I would like to thank Dr Tim Coles, Steve Oliver, Dedy besar , Dedy kechil, Ade, Dr Jim Berkleman, Cahyo, Pak Freddy, Polly Daniels, Lara Slazenger, Charlie Perfect, Matt Tatum, Ali Bishop, Alex Weir, Dean Jones, all the forest rangers and all the other volunteers and staff of Operation Wallacea 1999. Special thanks and appreciation goes to all the local people of Kaweli, Konde and Kambowa who assisted this work by providing food, accommodation and field assistance during much of the fieldwork.

• Bibby, J. Burgess, T. Hill, P. (1992). Bird Census Techniques. Academic Press.
• Dahl, J.F. (1987). Conservation of Primates in Belize, Central America. Primate Conservation 8.
• Fleegle, J.G. (1988). Primate Adaptation and Evolution. Academic Press.
• Fooden , J. (1969). Taxonomy and evolution of the monkeys of Celebes (Primates: Cercopithecidae). Bibliotheca Primatologica 10.
• Fooden, J. (1980). Classification and distribution of the living macaques. New York: Van Nostrand Reinhold.
• Johns, A.D. Skorupa, J.P. (1983). Responses of Rainforest Primates to Habitat Disturbance. International Journal of Primatology 8.
• Lee, R.J. (1997). The Impact of Hunting and Habitat Disturbance on the Population Dynamics and Behavioural Ecology of the Crested Black Macaque (Macaca nigra). Ph.D. dissertation. University of Oregon.
• MacKinnon, K. (1986). The Conservation Status of Non-human Primates in Indonesia. Springer-Verlag.
• Napier, J.R. Napier, P.H. (1985). The Natural History of the Primates. British Museum.
• National Research Council. (1981). Techniques for Study of Primate Population and Ecology. National Academy Press.
• O’Brien, T.G. Kinnaird, M.F. (1997). Behaviour, diet, and movements of the Sulawesi Black Crested Macaque (Macaca nigra). International Journal of Primatology 18.
• Rickard, A.F. (1985). Primates in Nature. W.H.Freeman and Company.
• Rosen, S.I. (1974). Introduction to the Primates. Prentice-Hall, Inc.
• Rosenbaum, B. (1996). Patterns and Determinants of Abundance for the Sulawesi Crested Black Macaque (Macaca nigra) on Bacan Island, Indonesia, with notes on its Conservation. Ph.D. dissertation. University of Colorado.
• Rosenbaum, B. O’Brien, T. Kinnard, M. Supriatna, J. (1998). Population Densities of Sulawesi Crested Black Macaques (Macaca nigra) on Bacan and Sulawesi, Indonesia: Effects of Habitat Disturbance and Hunting. American Journal of Primatology 44.
• Terborgh, J. (1983). Five New World Primates. Princeton University Press.
• Wilson, W.L. Johns, A.D. (1982). Diversity and Abundance of Selected Animal Species in Undisturbed Forest, Selectively Logged Forest and Plantations in East Kalimantan, Indonesia. Biological Conservation 24.

This appendix simply contains all the GPS data collected throughout the field studies which enabled accurate plotting of the line transects and the main sleeping trees of the different macaque troops.

Troop1: 00.210 56.825

Troop2: 00.565 56.755

Troop3: 01.082 56.798

Troop4: 01.423 56.322

Troop5: 01.798 56.912

Troop6: 01.955 56.387

Philip Lewis Stanier The University of Sheffield 1999