Techniques for the Study of Primate Population Ecology (1981)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

9 Commentary Transect censusing is useful for determining relative abundance when densities for censused areas are to be compared. Such com- parisons are most useful when the censused areas are comparable in terms of forest type and patterns of human disturbance. Transect estimates of density most nearly reflect real density for those primate species that are relatively large (> 1 kg) and con- spicuous in their activities. To date, the best transect estimates of density have been developed by experienced field workers working in a familiar habitat type who are well trained in observing and detecting the primate species indigenous to the area. One must be careful to distinguish between a density estimate and the actual density for a given population in a designated region. Intensive studies in small sample plots are mandated for small primates and those primate species that are cryptic or noc- turnal if the aim is a close approximation to actual density. Table 9-1 reflects current judgments concerning the appropriate census technique for a range of primate species. In light of these con- siderations, the following comments can fairly be made: 197

198 TECHNIQUES IN PRIMATE POPULATION ECOLOGY 1. Transect length must be determined empirically. A transect line should be long enough to cross all habitat types in the census area. In the example used in this report (see p. 52), Struhsaker found that 4 km was large enough to enable an observer to en- counter the primate species present in his forest. 2. In a broad survey it may not be possible to census more than once. When this is the case, the length of the transect line is determined by pacing during the census. For most censuses the transects should be repeated along a predetermined line. 3. The sample size must be determined empirically by calculating 95% confidence limits as data are collected. In the sample data used in this manual, Struhsaker found that 20-30 repeat transects were adequate. Spacing samples over an annual cycle is the only way to determine the effects of seasonal changes in habitat use. 4. Because detection distances vary with many habitat characteristics, especially foliage density, the transect width must be determined empirically in each habitat. Two distances are measured: the observer-to-animal, or sighting, distance and the transect line-to-animal distance, or perpendicular distance. The total strip width may then be calculated on the basis of a fixed or variable strip width for each method. The maximum reliable detection distance is the most commonly used type of variable strip width. 5. The application of line-transect census methods to forest primates is sufficiently recent that there is no best method, other than inspection of graphed or tabulated data. Examples of three useful methods are given. Much of the theory for census methods has been developed on the basis of surveys of animals in a two- dimensional space; it is assumed that the animals do not move in response to the observer before being detected. Forest primates must be censused in a three-dimensional space; they are often detected by sound before they are sighted; and they often move from their original positions before being detected. Three stan- dardized studies are available that compare the accuracy of cen- sus methods for forest primates with density estimates deter- mined from detailed studies. Struhsaker (see Chapter 4, this volume) found that an em- pirical inspection of the frequency distribution of sighting

Commentary 199 distances with a cutoff distance for reliable detectability (50 m in this example) was the most accurate census method for making density estimates for a common species in his study forest. Density estimates based on perpendicular distance estimates were inadequate under Struhsaker's survey conditions because many sightings made over the trail (a zero distance) resulted in an overestimation of the population. In contrast, density estimates based on sighting distances proved to be less adequate than perpendicular distance estimates under Janson and Terborgh's (in press) survey conditions because they overestimated population densities based on detailed studies of their target species. Eisenberg (see p. 65) describes a nonlinear density plot method that requires prior information on home range size and mean group size. This method also provided density estimates com- parable with those estimated for a common species from a de- tailed study. 6. Records of all individuals actually counted provide a known minimum. It is often useful to record a separate estimate of the number of additional animals in the vicinity (range of group size). For density calculations it is best to estimate densities of groups (excluding sightings of single animals that may be separated from groups or may be transients). Detailed counts of group size should be made at a time separate from standardized transect periods when the groups can be followed. The best group counts are obtained when the animals leave or enter sleeping trees or when they cross a restricted or open area during a progression. Total population densities should be calculated by multiplying the number of groups by the average number of individuals counted in several groups. Duplicate counts of the same groups (due to group movements or multiple observers) can be avoided by knowing the travel rate of the species and by identifying groups by individually distinct size/sex compositions or distinct individuals. 7. Most census methods have depended on age and sex classes that are scaled by size to the largest adult. Observers often sub- divided the immature categories beyond the basic age classes of infant, juvenile, subadult male, adult female, and adult male.

200 R ^. w If (N X E ^ i^ 8 — "§ Q -°. . ""- S a .5 2 J {2 III- V •o i 1 25" ^S 3 "5 •o di "S3 Ji « t/l S + tJ ++t 1"8 4 I ^ 3 + 0 a + + + + 1 E w ^ .S c •o O W CL U V tJ ^_. $ C b 2 lS H T M 1 1 | 13 tt "5! "tfl w5 wl </) .S u *° cu 2 H S K 2 S S O O O O O 0 u T3 u -s 00 8l!ltfl II |4l 1 s lll-S 1 111 *] llsl°^ |l^3 Ji >l PU ia CQ £-H t/5

201 o .E i^> g j- § c ^ . y - | u Rathbun, 1980 + + Thorington et al. + + Janson and Terbc » « ^ ^ 3 •« ^ BJ W »« *Q -. IS ^" oo oo *> e ^ E Sign 1 2 | 1 .a °- 8 E e 18~"2-S E O O 18 O £ ^22 -8" 11 + + a rt Si •o £ i « « S .M +j 5 -o *- « n « 6 g i/i [A .s e M w rt O O 3 ~ O 5 MOO V K 5, 1l ebidae Aotus trivirgatus Aotus trivirgatus Cebus apella 1 ? a ., * "8 1 | bill 1 1 1 1 1 ] ! -8 I 1 l^ls 4 | o ™ « o a . !• ^ X -t g § § «J 11 U U Q

202 TECHNIQUES IN PRIMATE POPULATION ECOLOGY These size classes have been roughly correlated with age classes on the basis of information from captive animals. These tech- niques have been satisfactory for general comparisons of the status of a species in different parts of its range. Using recent field data of known individuals, Dittus (see Chapter 5, this volume) showed that aging techniques can be fur- ther refined by taking into account the fact that captive in- dividuals grow and mature faster than wild individuals, and the size classes may be expected to vary between sexes, places, and years, reflecting the nutritional status of those populations. The more precise the age classifications, the more accurate the life tables and the understanding of population dynamics. 8. Transect census methods provide density estimates of a general nature that are useful when comparisons based on an order-of-magnitude difference are useful. The accuracy of cen- susing is improved with standardization and an increased charac- terization of habitat productivity and seasonality. Long-term studies provide increasingly accurate data on den- sities and additional data on such topics as home ranges and overlap between neighboring groups, seasonal changes in habitat use and diet, and records of groups with known age and sex com- positions. 9. So far as possible, habitat should be characterized by a description of the vegetation profile, and the description should include notes on the distribution of water and a gross classifica- tion of drainage and soil type. In more intensive surveys, vegeta- tion should be analyzed to gain an understanding of the relative abundance and dominance of woody plant species. 10. In longer-term studies, seasonal changes in rainfall and temperature should be recorded, and the findings should be aug- mented by phenological studies of the trees because ultimately the carrying capacity of the habitat is related to plant productiv- ity. In long-term studies of primates, some method for individual recognition must be devised that utilizes either natural or ar- tificial markings. Chronological age can be estimated by em- pirically verifying growth patterns of known individuals or by in- directly estimating age by morphological changes or tooth wear. 11. In order to understand carrying capacity for different species, long-term studies concentrated on feeding ecology and

Commentary 203 use of space must be undertaken. There is reason to believe that many primate species are food-limited in their natural habitats. 12. Before controlled cropping can be undertaken, the popula- tion must be analyzed to determine whether it is at equilibrium, expanding, or in decline. The construction of a life table is an essential first step before rational utilization of primate resources can be planned. 13. Emphasis is still on data gathering. It is to be hoped that standardization of methodology will make possible field data that are comparable between investigators and study areas and will yield predictive statements concerning the behavior of natural populations.