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.
NEOTROPICAL PRIMATES: ASPECTS OF HABITAT USAGE, POPULATION DENSITY, AND REGIONAL DISTRIBUTION IN LA MACARENA, COLOMBIA Lewis L. Klein and Dorothy J. Klein INTRODUCTION While searching between July and November 1967 for a suitable area in which to make a year-long intensive study of spider monkey social organization, we col- lected limited amounts of information on the distribu- tion of primate taxa in selected parts of La Macarena National Park, Colombia, and a few adjoining areas. These distributional data and our intensive observa- tions of feeding behavior and habitat usage made at a single location form the basis for several hypotheses about regional primate distribution and density. Much of the initial surveying and our subsequent study occurred within the boundaries of La Macarena National Park, which in 1968 was still a national reserve (Figure 1). The park is situated approximately 3Â° north latitude between 73Â° and 74Â° west longitude and com- prised about 11,000 km2 (4,250 sq miles). It consists of three types of terrain: an isolated mountain range, foothills, and the floodplains of the three major rivers and several smaller tributaries. The mountain range, with maximum elevations of about 3,000 m (12,800 ft) runs roughly parallel to the Cordillera Oriental. It has a predominantly north-south axis about 125 km (78 miles) long, and an east-west width of about 25 km (16 miles). The foothills on the eastern slope of the sierra are relatively wide and merge gradually into flood- plains at elevations of 250-350 m (820-1,150 ft) above sea level to the south. In contrast, most of the western slope of the mountains is precipitous. Most of the mountainous and foothill regions of the park is covered by forest presumed to be inhabited by nonhuman primates. However, we were unable to investigate these areas except for several very brief excursions into the foothills, as discussed below. The remaining two-thirds of the park's area consists of floodplains, primarily of three riversâthe Guejar, Ariari, and Guayaberoâwhich in 1968 formed the major natural boundaries of the forest reserve to the north, south, and east. The study site and most of the areas surveyed were located on or near these floodplains, which were inundated either yearly or at least once every several years. The forests found at these inundated locations were not floristically uniform, despite their extensive continuity, taxonomic heterogeneity, and freedom from slash and burn agricultural practices. For exam- ple, in our study site of 780 ha (3 sq miles) on the north bank of the Guayabero, using gross criteria of pres- ence and abundance of several easily identifiable trees, we were able to identify eight different types of forest communities. The existence of these small, spatially distinct floristic communities is supported by the work of some tropical foresters (Sawyer and Lindsey, 1971; Rosayro, 1958). In our region the identification of such communities was frequently used by natives in deciding where to plant crops. The patchwork of forest communities probably re- sults from varying durations of inundation, varying amounts of silt deposition or soil removal, changes in drainage patterns ranging from blockage of small sea- sonal streams to major changes in the riverbed, several types of serai succession, and chemical composition and porosity of the soil. 70
ASPECTS OF HABITAT USAGE, POPULATION DENSITY, AND REGIONAL DISTRIBUTION 71 FIGURE 1 Map of La Macarena region showing lo- cation of study site and areas of preliminary survey. Approx. Scale Legend 60 miles boundaries of fo&st resene. as of Study site. ' .. slies The most interesting aspect of these separable forest communities was the way in which they appeared to differentially affect the seasonal and daily movements and feeding of several of the primate taxa. Our data on these matters is most adequate for Ateles belzebuth (approximately 650 hours of observation). Contrasting aspects of habitat usage will be outlined with compara- tive, but considerably less, material on sympatric populations of Cebus apella, Saimiri sciureus, Alouatta seniculus, and nearby populations of Lago- thrix lagotricha. We feel that further delineation of some of these taxonomically correlated patterns of habitat usage may explain some of the larger-scale anomalies of taxonomic diversity and population den- sities noted in several of the areas explored outside the specific study site.
72 KLEIN and KLEIN FOREST ASSOCIATIONS AND HABITAT USAGE Eight forest types were identified at the study site: (1) heterogenous forest on creek and levee banks, (2) Chrysophyllum mixed forest, (3) "typical" study site mixed forest, (4) Brosimum swamp, (5) lakeside, shrub-tree association, (6) Cecropia stands, (7) clear tree swamp, and (8) aerial root swamp. The first three of the eight types of forest were considerably more diverse than the others with respect to tree species present. About 40-60 percent of the study site area was covered by these three formations. In these associations crown continuity between adjoin- ing trees was well developed, generally occurring at several levels from 15-31 m (50-100 ft) from the ground. Except for very scattered and occasional emergents growing to heights of 40-43 m (130-140 ft) (usually Ficus sp. or Ceiba sp.), the tallest trees were about 34 m (110 ft) high. Four types of tall palm trees were common but scatteredâtwo were associated with one type of diverse associations and two were found in the other two diverse forest types. In general, the understory of these three associations was never thick enough to seriously impede walking and was composed predominantly of saplings. Diverse forests of these three types were inundated in 1968 for periods ranging between 1 week and 2 months. The three associations were distinguished from one another on the basis of absence or presence of specific varieties of palms and several large conspicuous trees, frequently used by Ateles. e.g., Chrysophyllum. The most botanically heterogenous forest com- munities were found on the highest points within the study site, which were probably the remnants of natural levees. In 1968 most of these habitat types were inundated for 2 weeks or less, and in many years they may not have been flooded at all. The relatively greater diversity of trees in this community, when compared to the other two types of taxonomically di- verse forests, is probably the result of a combination of the slightly higher elevation and more rapid drainage. The heterogenous forest on creek and levee banks comprised less than 5 percent of the study site area, yet included those areas in which contacts wilh Ateles occurred for a longer period of time throughout the year (from February through September) than in any other comparably sized forest sector. Levee forests, however, were used most intensively when certain trees, commonly found only there, were producing ripe fruit, particularly a species offfyeronomia. About 50 percent of contact time with Ateles during the February ripening of Hyeronomia occurred in these communities. Although this botanical assemblage was important for spider monkeys throughout the year, marked and radical fluctuations in usage occurred. For example, although we spent much time looking for spider monkeys along the higher creek beds in October and November 1968, when most of the rest of the forest was flooded and inaccessible, we logged less than 10 percent of total observation time with Ateles in this habitat. About one-third of our contacts with capuchin and squirrel monkeys in October and November occurred in forest communities on creek and levee banks. The infrequent utilization of the heterogenous forests by A. belzebuth in October and November was correlated with an enormous increase in the proportion of daylight time they were observed feeding, resting, and moving among four discontinuous botanical com- munities composed of clustered concentrations of an unidentified species of Brosimum, locally known as "guaymero." Brosimum trees, which reached 46 m (150 ft) in height, were among the tallest in the region. They grew in association with several species oflnga, with less frequently occurring specimens ofCalophyl- lum sp. and with an occasional specimen of an enor- mous spreading banyan-type fig, locally called "chivecha." At the edges of these communities, there were commonly clusters of a small palm (Bactris sp.), which is less than 12 m (40 ft) in height. This was the only palm directly associated with clusters of Brosimum at the study site. Collectively, the four major Brosimum communities comprised between 18-32 percent of the study site. These areas were inundated for about 4 months in 1968, but drained rapidly toward their peripheries as the river fell. The marked shifts in area usage and arboreal travel routes coincided with a small but incomplete subsi- dence of water during the months of September and October and the ripening of Brosimum (late September and October) and Calophyllum (October). These fruits comprised about 70 percent of Ateles diet at that time (Klein, 1972). Approximately 60 percent of our obser- vations of spider monkeys during this period occurred in Brosimum swamp communities. This compares with less than 10 percent of the observations during the preceding month and less than 5 percent of the obser- vations from February, March, and April. Even when not bearing ripe fruit, the larger Brosimum, Calophyl- lum, and Ficus trees were occasionally used as resting and sleeping sites. Although a comparable quantitative estimate for the use of Brosimum swamp formations by sympatric primate taxa is impossible to provide, our contacts with other species in these areas during September and October were markedly less frequent than with Ateles. Several other types of botanical formations occurred on the terrain, which was inundated for lengthy
ASPECTS OF HABITAT USAGE, POPULATION DENSITY, AND REGIONAL DISTRIBUTION 73 periods of time in 1967 and 1968. Some of these, in contrast toBrosimum swamp, were rarely if ever used by Ateles. Along some of the borders of the oxbow lake El Tigre and its connecting channel with the river, there were several pure stands of Cecropia, and ad- joining areas were covered by relatively short, willow-shaped trees. These areas comprised only a small part (about 2-4 percent) of the study site and were inundated from 5 to 8 months during 1967 and 1968. Although Ateles belzebuth was neither seen nor followed to these relatively homogenous Cecropia stands, Alouatta seniculus and Saimiri sciureus were. Moreover, both A. seniculus and Lagothrix lagotricha were seen frequently in similar areas on the south bank of the Guayabero opposite the study site and Ateles belzebuth were not. The summed observation time for these three primate taxa was less than 20 percent of the approximate 650 hours of observations logged for Ateles. "Clear tree swamp," another habitat type on the seasonally flooded plains, comprised an estimated 12-24 percent of the study site area. These areas were characterized by being under water for long periods of time (approximately 6 months) and by becoming deeper towards their centers, forming a catch basin for stagnant water. Maximum tree height appeared to be 24-31 m (80-100 ft), but the crowns of individual trees were often massive in diameter. Palm trees were absent except at the boundaries, where stands of small, spiny-trunked Bactris sometimes occurred. Al- though the penetration of sunlight to the forest floor was relatively high in these areas, understories of either saplings or herbaceous vegetation were usually absent or sparse. However, small patches of grasses occasionally occurred. To the observer, these forests appeared to be neither regenerating nor in the process of being replaced by another type. It is suspected that this open type of swamp forest may have developed on terrain where relatively recent changes in drainage patterns adversely affected trees that were already present. We noted little fruit on these trees in the months we were able to view them. What little was seen appeared to be the fruit of epiphytes or of an occasional isolated tree growing on small elevated hillocks. Consequently, while spider monkeys were frequently followed into and through these open swamps, and were frequently observed using trees on the border for resting, they were rarely observed to feed there. In contrast, we saw Saimiri and Cebus in these communities frequently foraging for insects, and on one occasion Cebus was observed eating what were probably frogs' eggs, which were suspended from low branches overhanging the standing water. The last of the identified forest communities, aerial root swamp, was defined primarily on the basis of a single structural featureâthe presence of aerial roots. Communities of this type, which were inundated for periods as long as 5 months, comprised about 3-9 percent of the study site area and were characterized by large numbers of short, 15 m (50 ft) or less, stands of trees with extensive aerial-pneumatophoric roots and lianas. As with Cecropia stands, spider monkeys were neither encountered nor followed to these areas within the boundaries of the study site, although Saimiri and Cebus were. PRIMATE DISTRIBUTION DATA Our preliminary explorations took us to several loca- tions within or just outside the park's boundaries. These included: (1) an excursion by canoe and foot between September 24 and October 4, 1967, from the mouth of Cano Cabra upstream to its junction with Cano Centralâ2 days were spent in a foothill area above the floodpiam: (2) 4 days within forests border- ing the Losada River; (3) 7 days between Barranco Colorado on the Ariari River and the headwater areas of the Cafre River; (4) 5 days on the north bank of the Guejar River between Cano Azul and Cano Chivecha; and (5) a total of 9 days at several sites on the south and north banks of the Guejar River downstream from the settlement of Pinalitos. Throughout the study period, we also made occasional short trips into forests on the south bank of the Guayabero River between and bordering its tributaries La Flauta and La Tigrera. Many of the arboreal communities occurring at the study site also appeared to occur at most of the sites explored briefly. However, in some areas certain communities, e.g., Brosimum swamp, appeared to be either absent, curtailed in size, or infrequent; and at several locations trees and vegetational structuring not seen at the study site were noted. In some instances these floristic differences appeared to be associated with either the presence of primate taxa in addition to the four diurnal species at the study site or, in one case, a possible replacement. The apparent replace- ment was noted in two areasâthe upstream noninun- dated sections of forests bordering Cano Cabra (2 days of observation) and an area near the headwaters of Cano Cafre (1 day of observation). In both these areas Callicebus moloch, an animal not present at the study site, were observed. These observations along with several other indications in the literature (Mason, 1971), suggest a degree of competitive displacement between Callicebus and Saimiri, The fact that those forests in which we observed Callicebus and not Saimiri were composed of trees no taller than 22 m (70 ft) may have had some bearing on finding Callicebus
74 KLEIN and KLEIN but not Saimiri, Extensive and continuous tracts of poorly developed forests did not occur at the study site. Similarly, Callicebus torquatus were observed in the vicinity of the rapids, Angostura I (Salto), and reported to be present elsewhere on the south bank of the Guayabero where they occurred on hillside terrain. There, the trees were shorter than the general average for the region and were similar to those in which C. moloch were seen on the north bank foothills. An interesting complication is added by the reported pres- ence of C. moloch in certain areas of the south bank. However, they were limited to extensively inundated sections of forest, which sometimes supported well- defined stands of trees substantially lower in height than those on adjoining, better-drained terrain. The presence of one species of Callicebus on the south bank of the Guayabero and the probable pres- ence of a second was indicative that the south bank forests everywhere supported a greater variety of primate taxa than did the north bank areas. Lagothrix lagotricha, C. torquatus, and probably C. moloch occurred on the south bank, in addition to the four diurnal primate taxa occurring at the study site. Woolly monkeys were seen on the south bank from Angostura II as far upstream as the forested environs of La Macarena village and frequently and regularly in the forest across the river from our campsite. Yet they were neither seen nor heard on the north bank between the two Angosturas. We were also told by native observers that they occasionally encountered addi- tional primates in upstream areas of the south bank tributaries La Tigrera and La Flauta. Their descrip- tions best fit Cacajao melanocephalus and Cebus albifrons. Unfortunately, we were unable to confirm these reports. Neither of these species was ever observed at the study site. Thus, on the south bank there may have been as many as 10 primate species; five in addition to the four diurnal and one nocturnal taxa resident at the north bank study site. A simple explanation for the greater diversity on the south bank would be that the river acts as a barrier to dispersal. This, however, can only be a partial expla- nation. First, the width of the river bed is only 137 m (150 yd) at the study site. Second, floating vegetation in masses that would easily support the weight of monkeys was frequently seen during certain seasons moving both down and across the river. Thirdly, between La Macarena village and Angostura II. the Guayabero River meanders extensively, resulting in occasional oxbow lake formation and shifts of large sectors of bankside forest, 259-518 ha (1-2 sq miles) from one side of the river to the other (see Figure 2). Areas this large could probably support small groups of primates for periods sufficient for the changeover to become completed without necessitating actual rafting or swimming by individual animals. FIGURE 2 Aerial photo of study site forest, Laguna El Tigre, and south bank forest between Carios La Flauta and La Tigrera. Source: Insti- tute Geograflco "Agustin Co- dazzi," Bogota, Colombia, taken in 1961.
ASPECTS OF HABITAT USAGE, POPULATION DENSITY, AND REGIONAL DISTRIBUTION 75 The greater variety of primate taxa on the south bank of the river appears to be related to the greater heterogeneity of forests on the south bank. For exam- ple, along and between Cano Flauta and Cano Tigrera, there were hillsides and large sectors of undulating and level terrains that were high enough to escape flooding in most, if not all, years. These areas supported well-developed, tall, continuous, and taxonomically diverse forests. These forests existed in addition to the major types of arboreal forests identified at the study site, some of which were described above. An analysis of the distribution of primate genera in the area of La Macarena should probably minimize the importance of rivers as dispersal barriers and maximize the impor- tance of botanical diversification and ecologically competitive interactions between differing taxa. POPULATION DENSITY AND GROUP SIZES Adequate estimates of the study site population den- sity of Ateles belzebuth were obtained. The validity of the density estimates for primates other than A. belzebuth, however, include a large magnitude of probable error. They are not based upon fully defined annual home ranges, nor positive reidentification of the same troops or individuals encountered on sepa- rate occasions. However, they represent estimates of population density in a virtually undisturbed forested floodplain habitat and, as such, may be of some value to future students or to researchers working with populations in other locations. The group size esti- mates for all taxa (Table 1) should be fairly reliable. With one exception, specified below, they are based entirely on minimal contact periods of 15 minutes duration with the group counted. Saimiri sciureus. The population density of squirrel monkeys was estimated to be 50-80 animals per square mile (259 ha). The most satisfactory counts of fairly cohesive bisexual groups resulted in figures of 25-35 independently locomoting animals. Between April and June group size usually increased by the addition of 5-10 infants. On two occasions temporarily isolated male squirrel monkeys were observed following sub- groups of spider monkeys. Between three and six separate groups of S. sciureus ranged over the approx- imately 780-ha (3 sq miles) study site. Cebus apella. The population density of C. apella was estimated to be 15-25 animals per square mile (259 ha). Counts of animals ranged from a single instance of an individual to groups composed of 12 independently locomoting capuchins. Median group size was approx- imately six animals. The number of groups at the study site was estimated to be between four and six. Alouatta seniculus. The estimated population den- sity forA, seniculus at the study site is low relative to the existing data for howlers from other areas (Chiv- ers, 1969; Neville, 1972). There appeared to be be- tween 30 and 75 animals per square mile (259 ha). Aside from several observations of either isolated males or pairs of isolated males, groups ranged in size between three and six independently locomoting individuals, with a median of 4.3. If infants are included, i.e., animals carried by females at most tree crossings, the range of group sizes was from three to eight, with a median of almost five. At least 75 percent of the bisexual groups contained only one fully adult male. There were at least 6 groups of howlers at the study site, but probably no more than 15. Shifts in area usage appeared to be at least as marked as those of the spider monkeys, but were not concordant with A. belzebuth throughout the year. Including isolated males, it was estimated that population structure consisted of 28 percent adult males, 40 percent adult females, 22 percent juveniles, and 10 percent infants. The compa- rable figures for A. palliata on Barro Colorado Island in 1967 (Chivers, 1969) were 22 percent adult males, 41 percent adult females, 20 percent juveniles, and 17 percent infants; for groups of A. seniculus on a ranch in Guarico, Venezuela, 29 percent adult and subadult males, 33 percent adult and subadult females, 21 TABLE 1 Estimated Study Site Population Densities, Group Size and Number of Groups Present Estimated Number of Estimated Population Range of Groups Utilizing 780 ha Density per Square Mile Group Size" (3 sq miles) Study Site 5. sciureus 50-80 25-35 3-6 C. apella 15-25 6-12 4-6 A. seniculus 30-75 3-6 6-15 A. belzebuthb 30-40 17-22 3 " Independently locomoting animals. " See text for methods of determining total group size from subgroup data.
76 KLEIN and KLEIN percent juveniles, and 18 percent infants (Neville, 1972). Ateles belzebuth, Although our estimates of popula- tion density, group size, and composition are much firmer for this species than those just given for Alouatta, Saimiri, and Cebus, they were, of necessity, based on different techniques of censusing. Simply counting observed groups of spider monkeys was of little value in determining either population density or total group size. Groupings of spider monkeys, in comparison to other sympatric primate taxa, were markedly labile and flexible. Counts ranged from iso- lates of both sexes and mothers with infants (15 percent of 498 counts: 3 percent adult male, 10 percent adult female, and 2 percent adult female with infant) to groups composed of 22 independently locomoting in- dividuals (see Table 2). Counts of four or more inde- pendently locomoting animals also revealed no adult males, one adult male, or two or more adult males (see Table 3). Although median number of independently locomoting spider monkeys per count over the year was 3.5, it varied monthly from a high of 5.6 in May to a low of 1.5 in December (see Table 4). This quantitatively complex and variable picture can be used to determine population density, group size, and group composition only if the reliable identifica- tion and reidentification of most individuals is feasible. Fortunately, A. belzebuth has a considerable amount of intrapopulational variability in facial and clitoral pigmentation (see Figure 3). In time, despite the labil- ity and flexibility of grouping tendencies and relative absence of social tendencies resulting in fixed limits to interindividual spatial dispersals, almost all A. bel- zebuth at the study site could be assigned to one of three different and mutually exclusive social groups whose members were rarely assembled in a single location (Klein, 1972). The larger of the two best- TABLE 2 The Percentage of Subgroups Con- sisting of One to Eight or More Independently Locomoting A. belzebuth Subgroup Size Frequencies Total Subgroups (%) 1 75 15 2 105 21 3 72 14 4 78 16 5 36 7 6 36 7 7 20 4 8-22 76 15 TOTAL 498 99 TABLE 3 Composition of Subgroups of Four or More Independently Locomoting A. belzebuth Range of Frequency (%) Subgroup Sizes Bisexual subgroups Bisexual with two or more adult males 76 (33) Bisexual with one adult male 76 (33) SUBTOTAL 152 (66) Single sexed groups Entirely male 2(1) Entirely female with no dependent young 13 (6) Entirely female with dependent infants or juveniles 13 (6) Entirely female with and without dependent young 49(21) SUBTOTAL 77 (34) TOTAL: 229(100) 4-22 4-22 4-10 4-11 4 4-5 4-8 4-11 known groups contained 22 independently locomoting animals and 5 infants in February 1968; the smaller, 17 independently locomoting animals and 3 infants in September 1968. Five fully adult males were members of the larger group; 3 fully adult males were members of the smaller. The ratios of adult males to adult females were 1:2.4 and 1:3.5, respectively. Annual home range was estimated to be on the order of 1-1.5 sq miles (259-388 ha), with overlap roughly 20-30 percent. The study site population density of A. bel- zebuth, including the third group, whose members were only rarely encountered, was consequently esti- mated to be 30-40 independently locomoting animals per square mile (259 ha). TOTAL PRIMATE DENSITY AND COMPARISONS Total primate density at the study site, excluding Aotus, was probably 125-250 individuals per square mile (259 ha). Although no comparable estimates can be made for the other areas visited, there were some indications that spider monkey density on the south bank of the Guayabero was considerably lower than on the north bank study site. They were entirely over- looked on the south bank by resident Indians. We saw A. belzebuth on the south bank only once, and heard them on only a few other occasions in areas in which Lagothrix lagotricha and Alouatta seniculus were quite frequently seen and heard. This strengthens our
ASPECTS OF HABITAT USAGE, POPULATION DENSITY, AND REGIONAL DISTRIBUTION 77 TABLE 4 Intermonth Variations in Subgroup Size of All A. belzebuth Observed at Study Site Number of Encounters with Isolated Encounters with Subgroups of Eight Months Median Mode Range Encounters Individuals (9 ;) or Larger (%) Jan. 2.9 3 -17 64 17 2 Feb. 4.0 2 -18 103 5 16 Mar. 4.3 4 -10 48 4 12 Apr. 3.7 2 -II 82 12 15 May 5.6 3 -20 43 7 40 June-Aug, 2.7 1 -6 19 32 0 Sept. 2.2 2 -11 47 28 6 Oct. 5.2 1 -22 44 21 43 Nov. 2.2 2 -11 36 25 3 Dec. 1.5 1 -4 12 5O 0 TOTAL 3.5 2 -22 498 15 15 earlier suggestion that botanical diversity, ecological competition, and forest productivity must be incorpo- rated into explanations accounting for geographical distribution, taxonomic diversity, and population den- sities of single taxa. NEED FOR CONSERVATION The effects of human activities on nonhuman primate populations in La Macarena may be of interest. We made no specific investigations of this subject, but it FIGURE 3 A. belzebuth pausing subgroup. Note variations in facial pigmentation.
78 KLEIN and KLEIN was a matter of concern when looking for a study site. Several areas were rejected because of forest felling and/or evidence of heavy hunting pressures. The areas most affected were those abutting on the boundary rivers, particularly the higher banks less likely to be inundated. Virtually all suitable locations along the Ariari and Guejar rivers had small homesteads or clearings. In comparison, settlement along the Guayabero was relatively light, presumably because of the difficulties involved with transporting cash crops past the two rapids; nevertheless, extensive areas had been felled below Angostura I and within 50 km of the village of La Macarena (Grimwood, 1968). However, the effects of human habitation on wild primates in the La Macarena area were not uniform. Many areas had been colonized only briefly, and the economic and agricultural practices appeared to have differential effects upon the resident taxa. Hunting, for example, appeared to seriously affect Ateles and Lagothrix. They are a favorite food, since they are large and relatively easy to shoot. For many settlers, particularly in their first few years of coloniza- tion, they provided a major source of protein. Unfor- tunately, from the conservation point of view, Ateles is a slow-breeding animal and Lagothrix may be as well. Spider monkey females, for example, do not appear to give birth before their fifth year, and thereafter only once every 2 years. One recent colonist reported killing 22 spider monkeys in a 2-month period. He claimed it was necessary or he would have starved, since he had lost most of his crops during the 1967 flooding. On the other hand, the elimination of animals such as jaguars, ocelots, and some of the larger snakes, as a consequence of the trade value of their skins at that time, probably had a minimal positive effect on the larger primates such as Lagothrix and Ateles and may have had a greater positive effect on primates such as Saimiri and Cebus, which were not usually shot for food. Systematic trapping for primates, as far as we could determine, was not practiced in the park. Captives seen in rancherias and the markets of Villavicencio, which originated from La Macarena or nearby areas, were usually obtained as a result of shooting females with young. Agricultural activities in the area may also have had both positive and negative effects. Forest was felled selectively and generally in small plots; this was done in part to facilitate claiming maximum amounts of land and in part to allow the planting of a diversity of crops. House sites and sites on which fruit trees and yucca were to be planted were usually located on the highest banks and areas. As noted above, many of these terrains supported the most heterogenous portions of the forestâareas intensively used by most of the primates, particularly Ateles. For crops such as bananas and corn, sites with a slight inundation were selected. This again resulted in the destruction of forest types that, at the study site, were highly favored by spider monkeys. Less-productive forest areas, e.g. clear tree swamps, were not favored areas for cultiva- tion. Our general impression was that the degree and nature of clearing that occurred in most park areas probably had lesser effects upon taxa such as Saimiri and Cebus, who were more likely to be distrib- uted over different types of forest than were Ateles. Differences in locomotion and feeding habits, in com- bination with the practice of felling discontinuous patches of forest, were also likely to interfere more seriously with the travel routes of spider monkeys than with those of squirrel and capuchin monkeys. In the La Macarena region, C. apella was the only primate able to turn agricultural activities to im- mediately positive advantage by raiding crops, particu- larly maize. ACKNOWLEDGMENTS The field research reported in this paper was financed primarily through predoctoral fellowship I Fl MH-31, 722-01. and a supplementary research grant, 1 RO4 MH 13608-01. provided by the Department of Health, Education, and Welfare, Public Health Service, National Institutes of Health. REFERENCES Chivers, D. J. 1969. On the daily behaviour and spacing of howling monkey groups. Folia Primatol. 10:48-102. Grimwood, I. R. 1968. Reports and recommendations on the Sierra Nevada de Santa Marta National Park, the Isla Salamanca Na- tional Park, the Tairona National Park, the La Macarena National Reserve. British Ministry of Overseas Development. Klein, L. L. 1972. The ecology and social organization of the spider monkey, Ateles belzebuth. Unpublished Ph.D. Thesis. Univ. of California, Berkeley. Mason, W. A. 1971. Field and laboratory studies of social organiza- tion in Saimiri and Callicebus. Pages 107-137m L. A. Rosenblum. ed. Primate behavior 2. Academic Press, New York. Neville, M. K. 1972. The population structure of red howler monkeys (Alouatta seniculus) in Trinidad and Venezuela. Folia Primatol. 17:56-86. Rosayro. R. A. de. 1958. Recent advances in the application of aerial photography to the study of tropical vegetation. In Proceedings of the Symposium on Humid Tropics Vegetation. Publication of the Unesco Science Cooperation Office for South East Asia. Sawyer, J. O., and A. A. Lindsey. 1971. Vegetation of the life zones in Costa Rica. Indiana Academy of Science, Monograph No. 2. 214 pp.