Seasonal Variation in the Positional Behavior of Red Howling Monkeys (Alouatta seniculus)

Transcription

1 PRIMATES, 39(4): , October Seasonal Variation in the Positional Behavior of Red Howling Monkeys (Alouatta seniculus) DIONIS1OS YOULATOS University of Puerto Rico ABSTRACT. Recent studies on the positional behavior of primates reveal that significant seasonal variation occurs in both locomotion and postures that is related to changes in diet and foraging techniques. Howling monkeys (genus Alouatta), which also have a seasonally varied diet, are predicted to have correspondingly varied positional behaviors. Two groups of red howling monkeys were studied in a primary rain forest in French Guiana during the dry and wet seasons. During the dry season, when howler diet is based mainly on leaves, howlers traveled more frequently by quadrupedal walking on large supports, a mode of progression that is probably inexpensive energetically and relatively stable. During feeding, quadrupedal and tripedal stand contributed considerably, a posture probably associated with the equal distribution of leaves within a tree crown. In contrast, during the wet season, when fruit was abundant, howlers fed very frequently by sitting on large supports, probably because fruit consumption required more time for special manipulation. However, most seasonal changes in feeding postures, and in travel and feeding locomotion, were difficult to associate directly with dietary shifts. These behavioral changes may be more highly correlated with slight modifications in microhabitat use (horizontal and vertical daily ranges, similar and alternative arboreal pathways) that are not considered in this paper. Key Words: Alouatta seniculus; Red howling monkey; Locomotion; Postures; Variation; French Guiana. INTRODUCTION Knowledge of the positional behavior of extant primates is important for inferring patterns of positional behavior, habitat use, and foraging strategies of fossil primates. However these inferences require reliable form-function-behavior associations. Most ecological and behavioral parameters show seasonal fluctuations, usually associated with seasonal variation in the environment (such as type, amount, or distribution of resources and potential predators). Since positional behavior is an important means of obtaining access to resources and avoid predators, such variations are likely to cause variation in primate positional behavior (CROMVrON, 1984; DAGOSTO, 1995). A number of regent studies have shown that primates are indeed characterized by personal variability in their positional behavior, mostly associated with changes in diet (CROMPTON, 1984; GEBO & CHAPMAN, 1995; DAGOSTO, 1995). In these studies, the variation observed was highly significant during feeding locomotion, postural behavior, and support size and orientation. These differences are most likely related to the different structural features of the different resources exploited partially during certain seasons, and requiring different foraging techniques (CROMPTON, 1984; DAGOSTO, 1995). Howlers, genus Alouatta, although almost exclusively vegetarian, show significant seasonal variation in their diet, usually tracking resource availability (MILTON, 1980; ESTRADA, 1984; JUILLOT ~ SABATIER, 1993; GUILLOTIN et al., 1994). During the dry season, when fruit availability is reduced, howlers tend to rely almost entirely on leaves, spend most of their daylight

2 450 D. YOULATOS hours resting, and have short daily ranges (MILTON, 1980; ESTRADA, 1984). In the rainy season, when more fruit is available, they travel more and farther, and fruit represents a very important proportion of their diet. Thus, it is reasonable to predict that these different strategies are associated with different patterns of positional behavior. In this paper, I analyze and compare quantitative data on the positional behavior of red howling monkeys (Alouatta seniculus) in French Guiana, collected during two different seasons (wet and dry) in two successive years (1992, 1993). The questions addressed are (1) Is there any seasonal variation in positional behavior, and if so, which behaviors and parameters are most affected; and (2) Are the most seasonably different behaviors and parameters associated with dietary shifts? STUDY SITE, CLIMATE, AND PHENOLOGY Data in both seasons were collected at the Station des Nouragues, a primary lowland wet rain forest site in French Guiana (4~ 52~ The study site is mainly characterized by high forest, where the dominant plant families are Sapotaceae, Lecythidaceae, Chrysobalanaceae, Ceasalpiniaceae, and Burseraceae (SABATIER PREVOST, 1990). The equatorial climate of French Guiana is characterized by a relatively short dry season extending from August to November. The rest of the year is the long wet season, generally interrupted by a short dry period between February and March. Annual rainfall varies from 3,000 to 3,250 mm, and the mean temperature is 26.1 ~ The study site is described in detail by JUILLOT and SABATIER (1993) and ZHANG (1995). In the study site, seasonal fruit availability was strongly correlated with rainfall (ZHANG, 1995). There is a peak in fruit production starting in February and extending until April. Afterwards, fruit abundance declines gradually, reaching a minimum in August and September (JUILLOT & SABATIER, 1993; ZHANG, 1995). These results are further substantiated by data from other sites in French Guiana, where the fruit peak was found to be between February and May, and least available fruit between August and October (GUILLOTIN et al., 1994). SUBJECTS AND DATA COLLECTION I conducted focal animal instantaneous sampling (ALTMANN, 1974) on adult individuals of red howling monkeys (Alouatta seniculus). The same two adult males and two adult females were sampled in both seasons. They all belonged to groups A and B studied by JUILLOT and SABATmR (1993), and were entirely habituated. A focal individual was followed for 15 min, and then I shifted to another individual. If the focal individual was lost from view before the 15-min period ended, I shifted to another individual. Locomotor behavior during traveling (movement from and to sleeping sites and between feeding sites) and feeding (movement within the same or adjacent feeding trees) was recorded at 20-sec intervals. This interval was adequate for recording the following variables: locomotor mode, behavioral context, support size, support inclination, and forest layer where the animal was located. Feeding postural behavior was recorded by time bouts (CANT, 1987). A bout ended when one of the recorded variables changed. The variables recorded were: posture, behavioral context, support size, support inclination, and forest layer. Locomotor modes recorded were: Quadrupedal walk (horizontal quadrupedal progression above and along a single support), Climb (upward or downward vertical quadrupedal progression along a single support), Clamber

3 Seasonal Variation of Locomotion and Postures in Alouatta 451 (upward, horizontal, or downward quadrupedal progression above and across multiple diversely oriented supports), Bridging (cautious gap crossing through deformation of the initial and terminal supports, always involving at least two limbs at a time), Air (gap crossing modes involving an airborne phase such as leaping and dropping), and Suspensory (involved below support activities such as tail swing, hoist, and inverted quadrupedalism). Postural modes recorded were: Sit (crouched posture above one or multiple supports). Quadrupedal stand (standing on three or four limbs above one or multiple supports), Bipedal stand (standing on the two hindlimbs above one or multiple supports), Lie (prone, supine, or lateral lie), Tail-only hang (hanging below a support by only the tail), Tail+Hindlimb hang (hanging below one or multiple supports by the tail and one or two hindlimbs), and Tail+ Fore+ Hindlimb hang (hanging below one or multiple supports by the tail, one forelimb, and one or two hindlimbs). The diameter (d) of the largest support used during an instantaneous observation or a postural bout was visually estimated and recorded according to these size classes: Small (d_<2cm), Medium (2cm<d_ < 10cm), Large (locm<d--<2ocm), and Very large (d>20cm). Support inclination classes were (relative to true horizontal): Horizontal (0 ~ 10~ Moderate (10~176 Steep (45~176 and Vertical (80~176 The forest was divided in three layers: Emergents (trees above 30m height), Main canopy (most of the trees ranging between 15m to 30m), and Understory (trees and shrubs from 50cm to 15m). The first period of data collection was between April 10 and July 2, 1992, during the wet season (total rainfall=974 mm). Sample sizes for this period are Travel locomotion 1,191 instants, Feeding locomotion 372 instants, and Feeding postural behavior 411 min. The second study period was from July 28 to September 3, 1993, during the transitional and dry season (total rainfa11=356 mm). Sample sizes for this period are Travel locomotion 707 instants, Feeding locomotion 283 instants, and Feeding postural behavior 776 rain. Since such data form contingency tables with rows and columns that cannot be ordered in a natural way I used G-tests to statistically compare percentages of behavior, p values of 0.05 and less were considered as significant. RESULTS FOREST USE In the wet season, there were significant differences in forest layer use between travel locomotion and feeding postures (Fig. 1, G=28.24, p<0.001). Comparing feeding postures seasonally, there was significantly greater use of the emergent layer during the dry season (G=13.24, p<0.05). Furthermore, in the dry season, travel occurred more frequently on the emergent layer (G= 14.22, p=0.001). TRAVEL AND FEEDING LOCOMOTION Travel and feeding locomotion differed significantly within both seasons (Table 1), and travel locomotion differed between the two seasons (Table 1). In travel locomotion, Clamber showed the most dramatic increase during the wet season (1992 wet vs 1993 dry, G=42.34, p<0.001), while its increase in feeding locomotion was not significant (1992 wet vs 1993 dry, G= 1.948, p=0.1711). In the dry season and during travel, both Quadrupedal walk and Air showed significant increase (Qw, 1992 wet vs 1993 dry: G=12.61, p<0.05; Air, 1992 wet vs 1993 dry: G= 7.535, p<0.05). During feeding locomotion, Climb was more frequently used in the wet season but the differ-

4 452 D. YOULATOS Fig. 1. Use of forest vertical layers for travel locomotion and feeding postures during the 1992 wet season and the 1993 dry season. Sample sizes: Travel 1992, n=l191; Travel 1993, n=707; Feed 1992, n=411 rain; Feed 1993, n=776 min. Table 1. Seasonal variation in use of locomotor modes in travel and feeding. Travel 1992 wet Travel 1993 dry Feed 1992 wet Feed 1993 dry Quadrupedal walk 34.3 % 42.4 % 21.8 % 29.3 % Climb l Clamber Bridging Air Suspensory Sample sizes n= 1191 n=707 n=372 n=283 Travel 1992 vs Travel 1993: G=46.13, p<0.001; Feed 1992 vs Feed 1993: G=12.08, p<0.05; Travel 1992 vs Feed 1992: G=40.42, p<0.001; Travel 1993 vs Feed 1993: G=69.03, p< ence was not significant (G=3.257, p=0.0989). During the wet season, travel occurred more frequently on small and medium-sized supports, while large ones were frequently used as traveling supports in the dry season (Fig. 2; 1992 wet vs 1993 dry, G= 10.78, p<0.05). No significant difference was found in support size use in feeding locomotion between the two seasons (G , p=0.9194). Support size in both wet and dry seasons also showed significant differences between travel and feeding locomotion (1992 Travel vs Feed, G=10.48, p<0.05; 1993 Travel vs Feed, G=23.02, p<0.001). In both wet and dry seasons feeding locomotion occurred mainly on small supports while small and medium-sized supports were almost equally used in travel locomotion (Fig. 2). No seasonal differences were detected in support inclination use during travel locomotion (Fig. 3; 1992 wet vs 1993 dry, G= 6.752, p ), while during feeding more steep and less moderate supports were used in the dry season (feed locomotion 1992 wet vs 1993 dry: G= 10.01, p<0.05). The most important differences between the two seasons are summarized in Table 3. FEEDING POSTURES The difference in proportion of postural modes between the two seasons was significant (Table 2). Sit and Tail-only hang were much more frequent during the wet season (Sit, 1992 wet

5 Seasonal Variation of Locomotion and Postures in Alouatta % % 60% % 20% - :41.6 5, Is]very large" 123 large Iz]medium 58.7 [--Ismall 0% Fig. 2. Sizes of utilized supports in travel locomotion, feeding locomotion, and feeding postures during the 1992 wet season and the 1993 dry season. Sample sizes: Travel 1992, n=l191; Travel 1993, n=707; Feed locomotion 1992, n=372; Feed locomotion 1993, n=283; Feed postures 1992, n=41l rain; Feed postures 1993, n=776 min. 100% 6.1 :6;5 80% ; % O% 20% A % f---ivertical 37.1 i---isteep I---lmoderate r"lhorizontal 36.1 Fig. 3. Inclination of utilized supports in travel locomotion, feeding locomotion, and feeding postures during the 1992 wet season and the 1993 dry season. Sample sizes: Travel 1992, n= 1191; Travel 1993, n=707; Feed locomotion 1992, n=372; Feed locomotion 1993, n=283; Feed postures 1992, n=411 min; Feed postures 1993, n=776 min. Table 2. Seasonal variation in feeding postural modes during the 1992 wet season and the 1993 dry season. Wet 1992 (%) Dry 1993 (%) Sit Quadrupedal stand Bipedal stand Lie Tail-only hang Tail+H1 hang Tail+F1 +H1 hang Sample sizes n=411 min n=776 rain G=42.49, p< HI: Hindlimb; FI: forelimb.

6 454 D. YOULATOS vs 1993 dry, G=9.996, p<0.05; Tail hang, 1992 wet vs 1993 dry, G=23.25, p<0.001). On the contrary, there was more Stand (quadrupedal and bipedal) and Tail-hindlimb hang in the dry season (Stand, 1992 wet vs 1993 dry, G=7.439, p<0.05; Tail-hindlimb hang, 1992 wet vs 1993 dry, G=9.629, p<0.05). Overall, support size use differed significantly between seasons (G=13.17, p<0.05). More small supports were used during the dry season (Fig. 2; G=7.231, p<0.05), while the percentage of large supports doubled in the wet season (Fig. 2; G=9.473, p<0.05). A significant difference in the inclination of supports was found between the two seasons (Fig. 3; G=53.53, p<0.001). Horizontal supports predominated during feeding postures in the wet season (G=32.48, p<0.001), while moderate ones were used more in the dry season than the wet one (G=38.5, p<0.001). Table 3 summarizes the most important differences in feeding postural behavior. Table 3. Summary of the major differences in positional behavior between the 1992 wet season and the 1993 dry season. Behavior Wet season Dry season Significance Travel locomotion More emergent use * More Clamber ** More Air * More Qw * More small supports * More large supports * Feed locomotion More Qw * More moderate supports * More vertical supports * More steep supports * Feed postures More emergent use * More Sit * More Tail-hang ** More Stand * More Tail+HI hang * More small supports * More large supports * More horizontal supports ** More moderate supports ** Levels of significance: * p<0.05; ** p< DISCUSSION COMPARISON WITH PREVIOUS STUDIES CROMPTON (1984) found that during the summer (heavy rainfall and high temperature), galagos climbed significantly more, using the peripheries of tree crowns more thus using more oblique and small supports. He associated these changes with a radical shift of the galago diet from gummivory to insectivory in the summer (especially flying arthropods found in tree crowns). GEBO and CHAPMAN (1995) also found that red colobus monkeys showed slightly greater climbing in travel locomotion during the wet season than in the dry season. Probably the increase in clambering in red howlers in the wet season indicates a similar pattern. However, given the different definitions of modes in their and my study such a pattern is difficult to assess. Although, GEBO and CHAPMAN (1995) did not associate these changes with shifts in diet or habitat use, this similarity is intriguing given the similar feeding ecologies of howling and red colobus monkeys (STRUHSAKER, 1975; MILTON, 1980).

7 Seasonal Variation of Locomotion and Postures in Alouatta 455 DAGOSTO (1995) found that, in three out of four malagasy lemur species that she studied, leap increased during the dry season. This pattern was also found in red howlers. However, this does not mean that the same patterns result from parallel shifts in diet or in canopy use. According to her data lemurs did not show any radical changes in microhabitat use or diet, contrary to howlers. In contrast, GEBO and CHAPMAN (1995) found that leaping in colobus monkeys increased in the wet season, and that leaping distances were longer. A decrease of quadrupedal walk in feeding during the wet season was observed by CROMPTON (1984), GEBO and CHAPMAN (1995), and in this study. DAGOSTO (1995), however observed the opposite pattern in lemurs. These comparisons show that seasonal variation in the positional behavior of primates does not follow a set pattern. Such a result should be expected for primates, as they exhibit different postcraniai morphologies, have varied diets and foraging patterns, and apparently react differently to seasonal changes in resources. PROBABLE ASSOCIATIONS OF SEASONAL PATI'ERNS WITH DIETARY SHIFTS IN RED HOWLERS Travel Locomotion During the dry season, use of the emergent layer increased during traveling. Since this layer is composed of isolated high tree crowns (OLDEMAN, 1974), a part of the increase in leaping during the dry season is likely to be associated with it. It is difficult to suggest an explanation for greater quadrupedal walk in the dry season. Perhaps it is associated with increased use of larger supports. Slow quadrupedal walk (QW) on large supports seems to be a secure and energetically inexpensive way of moving within the canopy. That would fit with the energy economizing foraging strategy of the howlers during the dry season (MILTON, 1980). However, given that howlers travel less and cover shorter distances in the dry season, an increase in QW (as well as gap crossing modes), where bouts are long, is unclear. The significant increase in clamber in the wet season is also hard to explain, although it is probably associated with the extensive use of small supports and the high percent of main canopy use. GEBO and CHAPMAN (1995) and DAGOSTO (1995) have agreed that travel locomotion varies less seasonally than feeding locomotor and postural behavior. The big significant differences that I observed in howlers certainly would not lead so obviously to such a conclusion. On the other hand, the differences do not appear to conform with expectations deriving from the ecology of the animal. For instance, one problem would be the significant increase of gap crossing modes during the dry season, when howlers are expected to travel less, and cover shorter distances. However, it is true that I have not focused on possible subtle differences in daily horizontal and vertical ranges, similar or alternative arboreal pathways, with available supports showing similar or divergent structural features (see CROMPTON, 1984; GARBER & PREUTZ, 1995; McGRAw, 1996). Feeding Locomotion and Postures Given that seasonal changes in the environment have a direct impact on diet, feeding locomotion and postures are likely to show more seasonal change (CRoMPTON, 1984; DAGOSTO, 1995). During the wet season, red howlers feed on a large percentage of fruit (JUILLOT & SABATIER, 1993). Usually, fruit grows on the terminal branches in tree crown peripheries, while flush

8 456 D. YOULATOS leaves appear to be more evenly distributed in the crown of a given tree. The increase of clamber and climb in feeding locomotion during the wet season is likely to be associated with the location of fruit within the tree crown, implying fine-grained differences in resource dispersion. CROMPTON (1984) clearly showed such differences in feeding locomotion of galagos associated with specific microhabitat utilization during the different seasons. However, galago diet changes are more radical, shifting back and forth between insectivory and gummivory, activities that definitely occur in different parts of a tree crown. In the vegetarian howlers such microhabitat changes seem to be more subtle; in fact differences are probably due to different tree architectural models, depending on the species exploited in each season (see DAGOSTO, 1995). Fruit eating always requires special manipulation (opening of husk, extraction of seeds, aril, etc.) and therefore is much more time consuming than leaf eating (unpubl. data). Fruit eating is more common during the wet season, when howlers used much more sitting and larger and horizontal supports. These features are likely to be interrelated. On the other hand, the more evenly distributed flush and mature leaves can be acquired and processed in all parts of the crown by adopting standing postures, which were more common during the dry season. The enigmatic significant seasonal differences in tail-only hang and tail~hindlimb hang are hard to explain, especially since I had not found any correlation between food type and suspensory posture (although tail-hindlimb hang was more frequently used for leaf acquisition and consumption, YOULATOS, 1994). CONCLUSIONS Red howling monkeys exhibited considerable seasonal variation in many aspects of their behavior. Although previous studies of other primates have shown that feeding locomotion and postures were mostly affected by seasonal variation, red howlers showed some significant differences in traveling locomotion. However, it was difficult to produce any reliable associa'- tions between changes in traveling variables and shifts in diet and foraging patterns. In addition, comparisons with other studies showed that there is no consistent pattern of seasonal variation during travel locomotion. Some associations between feeding techniques and feeding postures were presented above, but unfortunately there are more changes that are problematic to explain than changes that can be explained by dietary shifts. Therefore, I come to the same conclusions as DAGOSTO (1995), that if dietary shifts imply shifts in microhabitat utilization with subtly different structural features, then is probably more likely to obtain more explicit associations between shifts in diet and positional behavior. Seasonal variation appears to play an important and integral role in the positional behavior of extant primates. Differences between the different seasons are sometimes highly significant. Therefore these differences should be considered if one wants to use the positional behaviors of an extant species, as a model for inferring positional behavior in extinct primates. Acknowledgements. I am gratefully indebted to Prof. J.-P. GASC and Dr. R CHARLES-DOMINIQUE for financing fieldwork, and access and help at the study site in French Guiana. Fieldwork was made possible by funds from the "Action Sp6cifique Guyane" of the Mus6um National d'histoire Naturelle, Paris, France, and the Centre National de Recherche Scientifique, France. Many thanks go to BENOIT DE THOISY, for help in the field. Drs. J. G. H. CANT, M. D. ROSE, and D. C. DUNBAR improved significantly earlier versions of the manuscript.

9 Seasonal Variation of Locomotion and Postures in Alouatta 457 REFERENCES ALTMANN, J Observational study of behaviour: sampling methods. Behaviour, 49: CANT, J. G. H Positional behavior of female Bornean orangutans (Pongo pygmaeus). Amer. J. Primatol., 12: CROMPTON, R. H Foraging, habitat structure, and locomotion in two species of Galago. In: Adaptations for Foraging in Nonhuman Primates, RODMAN, P. S.; CANT, J. G. H. (eds.), Columbia Univ. Press, New York, pp DA6OSTO, M Seasonal variation in positional behavior of malagasy lemurs. Int. J. Primatol., 16: ESTRADA, A Resource use by howler monkeys (Alouatta palliata) in the rain forest of Los Tuxtlas, Veracruz, Mexico. Int. J. Primatol., 5: GARBER, P. A.; PREUTZ, J. D Positional behavior in moustached tamarin monkeys: effects of habitat on locomotor variability and locomotor stability. J. Human Evol., 28: GEBO, D. L.; CHAPMAN, C. A Habitat, annual, and seasonal effects on positional behavior in red colobus monkeys. Amer. J. Phys. Anthropol., 96: GUmLOTIN, M.; DtJBOST, G.; SABATmR, D Food choice and food competition among the three major primate species of French Guiana. J. Zool. Lond., 233: JUILLOT, C.; SABATIER, D Diet of the red howler monkey (Alouatta seniculus) in French Guiana. Int. J. Primatol., 14: MCGRAW, W. S Cercopithecid locomotion, support use, and support availability in the Tai forest, Ivory Coast. Amer. J. Phys. Anthropol., 100: MILTON, K The Foraging Strategy of Howler Monkeys. Columbia Univ. Press, New York. OLDEMAN, R. A. A L'architecture de la forfit guyanaise. Mdm. O.R.S.T.O.M., 13: SABATIER, D.; PREVOST, M.-E Variations du peuplement forestier h l'6chelle stationnelle: le cas de la station des Nouragues en Guyane fran~aise. Actes de l'atelier MAB-UNESCO sur l'am6nagement et la Conservation de l'ecosystbme Forestier Humide, pp STRUHSAKER, T. T The Red Colobus Monkey. Chicago Univ. Press, Chicago. YOULATOS, D Mai'trise de l'espace et acc~s aux ressources chez le singe hurleur roux (Alouatta seniculus) de la Guyane Fran~aise: etude morpho-fonctionnelle. Ph.D. Diss., Mus6um National d'histoire Naturelle, Paris, France. ZHANG, S.-Y Activity and ranging patterns in relation to fruit utilization by brown capuchins (Cebus apella) in French Guiana. Int. J. Primatol., 16: Received: December 16, 1996; Accepted: March 23, 1998 Author's Name and Present Address: DIONIS1OS YOULATOS, 35, Agathoupoleos Street, 11252, Athens, Greece.