Does nonnutritive tree gouging in a rainforest-dwelling lemur convey resource ownership as does loud calling in a dry forest-dwelling lemur.код для вставкиСкачать
American Journal of Primatology 72:1062–1072 (2010) RESEARCH ARTICLE Does Nonnutritive Tree Gouging in a Rainforest-Dwelling Lemur Convey Resource Ownership as Does Loud Calling in a Dry Forest-Dwelling Lemur? SOLOFONIRINA RASOLOHARIJAONA1, BLANCHARD RANDRIANAMBININA1, AND MARINE JOLY-RADKO2 1 Faculty of Sciences, University of Mahajanga, Mahajanga, Madagascar 2 Institute of Zoology, University of Veterinary Medicine Hannover, Hannover, Germany Nonhuman primates may defend crucial resources using acoustic or chemical signals. When essential resources are limited, ownership display for a resource may be enhanced. Defending resources may depend on population density and habitat characteristics. Using the Milne Edwards’ sportive lemurs (Lepilemur edwardsi) and weasel sportive lemurs (L. mustelinus) as models, we tested whether two cryptic nocturnal lemur species differing in population density and habitat show differences in their vocal and chemical communication for signaling ownership of resources. L. edwardsi inhabits a western dry deciduous forest in a high-density population, whereas L. mustelinus is found in an eastern rainforest in low density. We followed ten L. edwardsi (six males and four females) and nine L. mustelinus (four males and five females) for 215 hr during the early evening (06:00–10:00 p.m.) and the early morning (02:00–05:00 a.m.) and recorded their behavior using focal animal sampling. We found that both species differed in their vocal and chemical communication. L. edwardsi was highly vocal and displayed loud calling in the mornings and evenings while feeding or in the vicinity of resting places. In contrast, L. mustelinus never vocalized during observations, but displayed tree-gouging behavior that was never observed in L. edwardsi. Tree gouging occurred more often during early evening sessions than early morning sessions. Subjects gouged trees after leaving their sleeping hole and before moving around. We suggest that, in weasel sportive lemurs, non-nutritive tree gouging is used as a scentmarking behavior in order to display ownership of sleeping sites. Altogether, our findings provide first empirical evidence on the evolution of different communication systems in two cryptic nocturnal primate species contrasting in habitat quality and population density. Further investigations are needed to provide more insight into the underlying mechanisms. Am. J. Primatol. 72:1062–1072, 2010. r 2010 Wiley-Liss, Inc. Key words: nonnutritive tree gouging; resource defense; scent marking; vocal communication; nocturnal lemur INTRODUCTION In nonhuman primates, rich repertoires of signals may be used for communication [see Hauser, 1996]. Vocalizations, as acoustic signals, and scent marking, as chemical signals, were described in contexts, such as intergroup communication [e.g. Braune et al., 2005; Irwin et al., 2004; Lazaro-Perea et al., 1999; Méndez-Cárdenas & Zimmermann, 2009; Rasoloharijaona et al., 2006], social and reproductive advertisement [e.g. Di Fiore et al., 2006; Heymann, 2006; Miller et al., 2003], territorial defense and resource usage [e.g. Braune et al., 2005; Lewis, 2006; Mertl-Millhollen, 2006; Palagi & Norscia, 2009; Pochron et al., 2005; Rasoloharijaona et al., 2006]. Nonhuman primates may defend crucial resources, such as feeding areas or safe sleeping sites, by loud calling or scent marking. Territorial r 2010 Wiley-Liss, Inc. defense or display of a resource ownership is costly and should be linked to the defendability of limited resources [Mitani & Rodman, 1979]. If crucial resources are limited, ownership display for a resource may evolve [Horiuchi, 2008]. Defending resources may thus be shaped by population density and habitat quality: Animals living in a poor habitat in a high population density should demonstrate more ownership display behavior when resources are Correspondence to: Marine Joly-Radko, Institute of Zoology, University of Veterinary Medicine Hannover, Buenteweg 17, D-30559 Hannover, Germany. E-mail: firstname.lastname@example.org Received 13 January 2010; revised 16 June 2010; revision accepted 17 June 2010 DOI 10.1002/ajp.20865 Published online 7 July 2010 in Wiley Online Library (wiley onlinelibrary.com). Tree Gouging in Sportive Lemurs / 1063 limited than animals living in a rich habitat in a low population density. Sound and scent may both be used for long-distance communication in mammals [Brown & Macdonald, 1985]. However, Although scent may provide long-term cues, sound provides short-term cues. The use of acoustic or chemical cues may depend on the need to defend resources and on the presence of potential receivers of the signal. Short-term reaction, i.e. use of acoustic cues for signaling ownership, should prevail when the availability of resources is limited and when the presence of potential competitors in the vicinity increases. On the other hand, the use of chemical cues may be useful to leave long-lasting messages for competitors, which may access limited resources at a later time. Nocturnal prosimians are an interesting model to investigate the plasticity in vocal and chemical communication. Although an enhanced visual system and reduced olfactory system are generally considered to be a specific trait in primate evolution, most of the nocturnal prosimians possess a still functional accessory olfaction system in addition to the main olfactory system [see Barton, 2006; Schilling, 1979]. This may reflect an enhanced chemical sensitivity. Although the major role of the main olfactory system is to detect volatile molecules, the accessory olfaction system is mainly involved in the reception of large nonvolatile molecules, e.g. the sampling of fluids, such as urine or saliva [see Evans, 2006]. Besides acoustic communication, which has been thoroughly investigated [see Braune et al., 2005; Méndez-Cárdenas et al., 2008; Rasoloharijaona et al., 2006], information on the function of olfactory communication in wild nocturnal prosimians remains sparse [Braune et al., 2005]. Earlier field investigations on golden brown mouse lemurs demonstrated that olfactory signals, such as urine washing at sleeping sites, could represent an important mechanism to regulate the distribution of different groups in space, whereas acoustic signals control intragroup cohesion and coordination [Braune et al., 2005]. However, to date, olfactory communication remains poorly investigated in other families of nocturnal lemurs. The sportive lemurs (Lepilemuridae) may represent an excellent model to gain more insight into the plasticity of communication modalities of nocturnal prosimians. Twenty-five different species are currently recognized [see Mittermeier et al., 2008]. Comparative data on their socioecology and communication are still rare, because only a few species have been studied so far. The Milne Edwards’ sportive lemur (Lepilemur edwardsi) and the weasel sportive lemur (L. mustelinus) have a limited distribution in Madagascar. Although L. edwardsi inhabits the western dry deciduous forest, L. mustelinus is found in the eastern rainforest [Mittermeier et al., 2006]. Both species are mainly folivorous and show similar feeding habits, but differ in morphology [Rasoloharijaona et al., 2003, 2008], genetics [Andriaholinirina et al., 2006; Craul et al., 2007], and behavioral aspects [Rasoloharijaona, 2001]. The Milne Edwards’ sportive lemur (L. edwardsi) lives in dispersed male–female pairs [Rasoloharijaona et al., 2000, 2003, 2006; Thalmann, 2006; Thalmann & Ganzhorn, 2003]. Individuals usually forage alone at night, but establish long-term pairs and sleep together during the day [Rasoloharijaona et al., 2003]. Pair partners share a home range with suitable sleeping and feeding sites, usually sleeping together in the same sleeping site and using home ranges exclusively [Rasoloharijaona et al., 2003, 2006]. Pairs are highly vocal. They possess an elaborate vocal repertoire consisting of nine structurally different call types [Rasoloharijaona, 2001; Rasoloharijaona et al., 2006]. Vocalizations are shown to carry signatures for sex, individual and pair identity [Méndez-Cárdenas & Zimmermann, 2009; Rasoloharijaona et al., 2006]. Pairs use duetting as a vocal display for signaling territory ownership, and thus limit direct aggressive encounters between neighbors and strangers [Méndez-Cárdenas & Zimmermann, 2009; Rasoloharijaona et al., 2006]. Milne Edwards’ sportive lemurs emit loud calls mostly during feeding events and in the vicinity of their sleeping sites [Rasoloharijaona et al., 2006]. The weasel sportive lemur (L. mustelinus) is a rainforest-dwelling lemur and is distributed in eastern Madagascar [Mittermeier et al., 2008]. Compared with the sportive lemurs of the northwestern and southwestern part of Madagascar, they have not been well studied in the wild. Weasel sportive lemurs were observed to forage solitarily [Ganzhorn, 1988]. Sexes show a sexual dimorphism; although males and females do not differ in their body length, females are heavier than males [Rasoloharijaona et al., 2008]. Both sex use dense vegetation and tree holes for sleeping during the day. Each animal has its own preferred sleeping sites and uses them alone. Only four types of vocalizations were described in L. mustelinus [Rasoloharijaona, 2001]. However, those vocalizations occurred rarely and were only recorded in one male of six focal individuals [Rasoloharijaona, 2001]. Interestingly, we earlier witnessed that L. mustelinus gouge trees [Rasoloharijaona et al., 2008]. Tree gouging has already been described in other strepsirrhines [Patel & Girard-Buttoz, 2008; Petter et al., 1971; Tan & Drake, 2000]. Tree gouging may be used in a nutritive context; fork-marked lemurs and pygmy slow lorises stimulate exudate flow by gouging trees [Petter et al., 1971; Tan & Drake, 2000]. However, recently Patel and Girard-Buttoz  reported the use of tree gouging in a nonnutritive context; wild male Milne-Edwards’ sifakas combine tree gouging with scent marks for overmarking. The function of this behavior in L. mustelinus is so far not clear. Am. J. Primatol. 1064 / Rasoloharijaona et al. The sportive lemur’s density may vary according to the Malagasy study areas. Although Rasoloharijaona  found that the density of L. mustelinus was 1–2 individuals/hectare in the eastern rainforest, the population density of L. edwardsi in the northwestern dry deciduous forest was about five individuals/hectare [Rasoloharijaona et al., 2000]. The home-range size also varied with about 2.3 ha for males [no data available for females; Rasoloharijaona, 2001] in weasel sportive lemurs in Andasibe and about 1 ha for Milne-Edwards’ sportive lemurs in the forest of Ampijoroa [0.9870.4 ha females and 1.0170.25 ha for males; Rasoloharijaona et al., 2006]. The aim of this study was to provide new insight into the resource ownership display behavior, and especially to assess the presence and significance of specific vocal and tree-marking behaviors of two cryptic sportive lemur species. We hypothesize that specific vocal and marking behaviors are shaped by socioecological factors, such as habitat quality and population density. On the one hand, we expect animals living in a poor habitat in a high-density population to display more vocalizations than animals living in a rich habitat in a low-density population. We, therefore, hypothesize that L. edwardsi living in the dry deciduous forest will demonstrate more frequent loud calls to defend crucial resources, such as sleeping and feeding sites, than L. mustelinus, which lives in the rainforest in a low-density population. Moreover, we hypothesized that L. mustelinus use tree gouging as an alternative communication strategy to leave long-lasting cues in the rainforest. We hypothesize that tree gouging is used to signal resource ownership by L. mustelinus, as L. edwardsi use loud calling and we will describe its behavioral context. METHODS Study Sites and Climatic Conditions The study was conducted both in eastern and northwestern Madagascar using comparable methods. In the eastern Malagasy rainforest, we worked at two different study sites close to Andasibe: The Mantadia National Park (S18147.50 , E48126.60 , altitude: 900 m) and the Maromizà private reserve (S181580 , E481280 , altitude: 1,020 m). The direct distance between Mantadia and Maromizà is about 12 km and both sites belong to the district of Andasibe. In the western Malagasy dry deciduous forest, we worked in the Réserve Forestière d’Ampijoroa (S160 190 , E461490 ), located about 115 km southeast of the town Mahajanga in northwest Madagascar. The study was performed at two geographically different study sites, the so-called ‘‘Jardin Botanique A’’ and the so-called ‘‘Jardin Botanique B.’’ In the eastern rainforest, the climate conditions, such as daily minimum and maximum temperatures Am. J. Primatol. and rainfall, were provided by the Service de la Métérologie Nationale [Rasoloharijaona, 2001]. The climate is seasonal with strong temperature fluctuations during the year; the maximum daily temperature was recorded in December (351C), the minimum daily temperature in July (0.21C). Rainfall was highest from December to April with a monthly maximum rainfall of 380 mm, and lowest from May to November with a monthly minimum rainfall of 20 mm. In the western dry deciduous forest in Ampijoroa climate conditions, such as daily minimum and maximum temperatures and rainfall, were measured by the staff of the Durrell Wildlife Conservation Trust. The climate is highly seasonal and characterized by a hot and humid rainy season from November to April, with heavy rains in January and February and a cool dry season from May to October [Rasoloharijaona, 2001]. In 1998, the maximum daily temperature was recorded in November (39.51C) and the minimum in June (10.51C). Rainfall was highest from December to March with a monthly maximum rainfall of 532 mm in February, and no rainfall at all between June and September. Animals and Their Identification We studied weasel sportive lemurs (L. mustelinus) in eastern Madagascar from May to December 1999, from May to July 2005, and in November and December 2005. In 1999, six adult animals (four males and two females) were captured, measured, and radiocollared with TW-3 button cell tags (Biotrack, Dorset, UK; see Table I). In May 2005, three females were captured, measured, and radiocollared (see Table I). We studied sportive lemurs (L. edwardsi) in northwestern Madagascar from May to December in 1998. Ten adult animals (six males and four females; see Table I) were captured, measured, and radiocollared with TW-3 button cell tags (Biotrack; see Table I). The same capture procedures were used at both study areas [see Rasoloharijaona et al., 2003, 2006]. Research conducted in this study adhered to the legal requirements of Madagascar and permission was obtained from the appropriate agencies. Capture procedures, animal handling, and radio-tracking techniques followed standard protocols and were authorized by the appropriate institutional or government bodies. The research complied with the American Society of Primatologists Principles for the Ethical Treatment of Non-Human Primates. Data Collection and Analysis Direct focal observations were carried out on radio-collared individuals by one observer, using focal animal sampling with continuous recording [Altmann, 1974; Martin & Bateson, 1993] for two sessions: 4 hr after the individuals left their sleeping Tree Gouging in Sportive Lemurs / 1065 TABLE I. List of Focal Animals Site Lepilemur edwardsi in Ampijoroa Lepilemur mustelinus in Andasibe Sex Animal number F F F F M M F-01-97 F-01-98 F-05-97 F-07-98 M-08-98 M-09-98 M M M M M-13-98 M-15-98 M-18-98 M-19-98 F F F F F M F-04-99 F-05-99 F-01-05 F-02-05 F-03-05 M-02-99 M M M M-03-99 M-06-99 M-08-99 Total focal time (hours) Total contact time (hours) % contact time June, August October July, September June June June, August, September, November October July, November September September 14.06 6.54 12.09 6.48 6.35 33.20 10.15 4.43 9.38 5.85 5.73 25.94 72.19 67.67 77.61 90.34 90.25 78.13 7.63 12.40 6.39 12.17 5.49 8.3 4.99 11.07 71.96 66.98 78.08 90.97 August, October June July June June August, September, October, November August, November August, September August, September 16.69 4.91 2.99 11.45 9.80 21.8 5.29 2.34 0.5 2.83 2.95 11.7 31.69 47.71 16.68 24.71 30.1 53.67 8.73 16.27 6.75 1.36 8.49 1.36 15.62 52.17 20.09 Months of observations sites (from 06:00 to 10:00 p.m.), and for 3 hr before they returned in the early morning (from 02:00 to 05:00 a.m.). The total focal time (i.e. total duration an animal was followed and observed) and the total contact time (i.e. total duration an animal was seen during the observation) per individual and site are shown in Table I. In the eastern rainforest, it was more difficult to maintain a long-lasting visual contact with the focal animals because the trees are relatively tall and the ground is very uneven and sloping. Lemurs were observed by dimmed light using white headlamps. Behavior and additional information related to spatial and ecological factors were recorded on a dictaphone and subsequently transferred to data sheets. We continuously recorded the behavior and differentiated among the following categories: resting (REST), feeding (FEED), in the sleeping hole (HOLE), scanning (SCAN 5 the animal looked to the left and right side), moving (LOC), selfgrooming (SG), social activity (SOC 5 encounter with one or more conspecifics within 5 m range of the focal animal), vocalizing (VOC 5 loud calling), tree gouging (GOUG 5 the animal gouges the bark of a tree using its tooth comb), and unknown behavior (UNKN). We determined the activity budget for each individual by calculating the percentage of time an animal spent in each behavioral category in relation to the total contact time. We counted all vocalization bouts and all tree-gouging bouts for each individual. A bout was defined as a sequence of vocalizations or tree gouging occurring successively. The bout was stopped as soon as the animal displayed another behavior without vocalization or tree gouging. We then divided the total number of vocalization or treegouging bouts for an individual by the contact time to express it as a rate per hour of contact time. The behavioral rate was calculated for both the early evening (from 06:00 to 12:00 p.m.) and the early morning (from 02:00 to 05:00 a.m.) observation sessions separately in order to compare the two periods. Finally, we assessed the behavioral context in which the vocalization and tree-gouging bouts occurred by analyzing the behavioral sequence 5 min before and 5 min after the occurrence of each bout. We thus, analyzed what the animal was doing just before and after vocalizing or tree gouging. For each 5 min interval range and each individual, we determined and compared the relative duration of each behavior occurring 5 min before and after the vocalization and tree-gouging bouts. Medians and range were calculated for all variables. We analyzed the data using the program Statistica 6.0 (Statsoft Inc., Tulsa, OK) and SPSS 17.0 (SPSS Inc., Chicago, IL) and applying nonparametric exact two-tailed tests. The statistical significance was P 5 0.05 for all tests. To control for multiple testing, the sequential Bonferroni procedure was used [Holm, 1979]. For that purpose, P-values were ordered and the smallest P-value was compared with 0.005 ( 5 0.05/10 when the same question was asked Am. J. Primatol. 1066 / Rasoloharijaona et al. repeatedly for the ten behaviors). If the smallest value was less than 0.005, we rejected the null hypothesis. The procedure was repeated until the smallest P-value could not be rejected. All hypotheses that had not been rejected in the earlier steps were then accepted. RESULTS Activity Budget Both species showed very similar patterns in their activity budget (Fig. 1). Most of the time, animals were seen resting, feeding, in the sleeping hole, and scanning (representing more than 90% of the activity budget for L. edwardsi and more than 80% of the activity budget for L. mustelinus; see Fig. 1). However, both species differed in their feeding duration, loud calling, and tree gouging (same results, Mann–Whitney U test, NL. edwardsi 5 10, NL. mustelinus 5 9, U 5 5, Z 5 3.27, P 5 0.0001; see Fig. 1). L. edwardsi in Ampijoroa spent more time feeding than L. mustelinus in Andasibe (L. edwardsi: median 5 33.81%; range 5 14.78–37.16%; L. mustelinus: median 5 7.16%; range 5 3.75–29.93% of the contact time). L. edwardsi also spent more time making loud calls (Mann–Whitney U test, NL. edwardsi 5 10, NL. mustelinus 5 9, U 5 5, Z 5 3.35, P 5 0.0001; see Fig. 1), whereas L. mustelinus displayed a longer tree-gouging activity (Mann–Whitney U test, NL. edwardsi 5 10, NL. mustelinus 5 9, U 5 5, Z 5 3.64, P 5 0.0001; see Fig. 1). There was no sex difference in activities in both study areas (Mann–Whitney U test, P40.05). Loud Calls and Tree-Gouging Bouts During our focal observations, 124 bouts of loud calls were recorded. They were restricted exclusively to L. edwardsi (Fig. 2A). Both males and females emitted loud calls at a rate of 1.31 calls (range 5 0.4–3.49) per hour. Fifty tree-gouging bouts were recorded. They were exclusively observed in L. mustelinus (Fig. 2B). Both male and female L. mustelinus were seen gouging and there was no sex difference in the gouging rate (Mann–Whitney U Test, Nmales 5 4, Nfemales 5 5, U 5 5, P40.05). The focal animals gouged on the trunks using their tooth comb, making transverse lines about 2 cm in length, Fig. 1. Behavioral activity budgets of L. edwardsi (in white) in the northwestern dry deciduous forest and L. mustelinus (in gray) in the eastern rainforest. Line 5 median, box 5 25–75%, and whisker 5 nonoutlier range. Outliers and extremes are not represented. Abbreviations of the behavioral categories: REST, resting; FEED, feeding; HOLE, in the sleeping hole; SCAN, scanning; LOC, moving; SG, self-grooming; SOC, social activity; VOC, vocalizing; GOUG, tree gouging; UNKN, unknown behavior; see Material and Methods for the definition of each behavior. Mann–Whitney test, Po0.0001. Am. J. Primatol. Tree Gouging in Sportive Lemurs / 1067 Fig. 2. (A) Rate of vocalizations made by both sexes in Lepilemur sp. (B) Rate of tree-gouging bouts made by both sexes in Lepilemur sp. Fig. 3. (A) Tree-gouging marks in the neighborhood of a sleeping hole. (B) Close-up photograph of tree-gouging marks. Photos by Randrianarison ; white arrows indicate some of the visible tree-gouging marks. two to six times in each sequence of gouging (see Fig. 3A and B). We never observed the focal animals eating the wood or licking the bark while gouging or after gouging. Such a gouging behavior was never observed in Ampijoroa. Behavioral Context of Loud Calling Focal individuals emitted loud calls during early evening as well as early morning sessions. There was no difference in the vocalization frequency between both sessions in both sexes combined (Wilcoxon matched-pairs test, N 5 10, Z 5 0.36, P 5 0.72; males only: N 5 6, P 5 0.465; no analysis was possible for the females owing to the sample size being too small; N 5 4). There were no sex differences in the overall vocalization frequency (Mann–Whitney U test, Nmales 5 6, Nfemales 5 4, U 5 10, P40.05). We determined the behavioral context of 124 vocalization bouts. Figure 4A illustrates the behaviors before and after vocalizing. Three behavioral categories prevailed before vocalization: FEED (median 5 32.74%; range 5 0.00–53.35%), SCAN (median 5 30.13%; range 5 0.00–51.98%), and REST (median 5 14.10%; Am. J. Primatol. 1068 / Rasoloharijaona et al. Fig. 4. (A) Behavioral context of loud calling in L. edwardsi. Behaviors which occurred during the 5 min before (in white) and 5 min after (in gray) are expressed in percentage of duration. Line 5 median, box 5 25–75%, and whisker 5 nonoutlier range. Outliers and extremes are not represented. REST, resting; FEED, feeding; HOLE, in the sleeping hole; SCAN, scanning; LOC, moving; SG, self-grooming; SOC, social activity; VOC, vocalizing; GOUG, tree gouging; UNKN, unknown behavior; see Material and Methods for the definition of each behavior. Wilcoxon matched pairs test, 0.05oPo0.07. (B) Behavioral context of a tree-gouging bout in L. mustelinus. Behaviors which occurred during the 5 min before (in white) and 5 min after (in gray) are expressed in percentage of duration. Line 5 median, box 5 25–75%, and whisker 5 nonoutlier range. Outliers and extremes are not represented. REST, resting; FEED, feeding; HOLE, in the sleeping hole; SCAN, scanning; LOC, moving; SG, self-grooming; SOC, social activity; VOC, vocalizing; GOUG, tree gouging; UNKN, unknown behavior; see Material and Methods for the definition of each behavior. Wilcoxon matched pairs test, Po0.01. range 5 0.00–64.46%). After vocalizing (see Fig. 4A), animals were mostly seen feeding (median 5 42.10%; range 5 15.79–69.33%), resting (median 5 28.36%; range 5 11.51–64.58%), and scanning (median 5 16.74%; range 5 0.00–26.04%). After a loud call, L. edwardsi tend to increase their feeding activity (Wilcoxon matched-pairs test, N 5 10, Z 5 0.059, P 5 0.064) and decrease their time spent moving (Wilcoxon matched-pairs test, N 5 10, Z 5 0.059, P 5 0.064) and in the hole (Wilcoxon matched-pairs test, N 5 10, Z 5 0.046, P 5 0.063). However, we could not reject the null hypotheses according to the sequential Bonferroni procedure, and thus these results could have been explained by chance. Behavioral Context of a Tree-Gouging Bout Focal individuals displayed tree gouging during the early evening as well as the early morning sessions. They gouged more at the beginning of their activity (N 5 8; median 5 2.1 tree-gouging bouts per hour; range 5 1.2–2.9) than at the end (N 5 8; median 5 0.0 tree-gouging bout per hour; range 5 0.0–0.7; Wilcoxon matched-pairs test, N 5 8, Z 5 2.52, P 5 0.008). No analysis was possible for either sex owing to the small sample size (Nmales 5 Nfemales 5 4). Am. J. Primatol. We determined the behavioral context of 50 treegouging bouts. Figure 4B illustrates the behaviors and their relative duration before and after gouging. Three categories of behavior prevailed within 5 min before a tree-gouging bout: HOLE (median 5 41.04%; range 5 28.64–93.53%), SCAN (median 5 29.47%; range 5 1.16–68.3%), and REST (median 5 5.53%; range 5 0.00–37.74%). After gouging (see Fig. 4B), animals were mostly seen scanning (median 5 61.87%; range 5 45.98–81.47%), feeding (median 5 11.98%; range 5 0.00–25.46%), and in locomotion (median 5 11.90%; range 5 6.09–27.94%). L. mustelinus spent more time in the hole before than after a tree-gouging bout (Wilcoxon matched pairs test, N 5 9, Z 5 2.66, P 5 0.004; null hypothesis rejected by the sequential Bonferroni procedure). Conversely, the subjects were more in locomotion, scanning (same results, Wilcoxon matched pairs test, N 5 9, Z 5 2.66, P 5 0.004; null hypothesis rejected by the sequential Bonferroni procedure) after than before a tree-gouging bout. DISCUSSION We report here differences in the vocal behavior of two cryptic sportive lemurs differing in habitat and population density. In agreement with our Tree Gouging in Sportive Lemurs / 1069 hypothesis, L. edwardsi, living in a high population density in the dry deciduous forest of Ampijoroa, displayed loud calls, whereas L. mustelinus, living in a low population density, did not display loud calls. Most of the loud calls were produced while feeding or resting, which corroborate earlier results showing that L. edwardsi produce loud calls for signaling ownership of sleeping and feeding sites [Rasoloharijaona et al., 2006]. L. mustelinus did not vocalize but displayed tree-gouging behavior mostly after leaving its sleeping site and moving around. The meaning and function of this behavior is not yet clear. Tree gouging has been reported in primates [see CoimbraFilho & Mittermeier, 1976; Fonseca & Lacher, 1984; Lacher et al., 1981; Lazaro-Perea et al., 1999; Nash, 1986; Rylands, 1985; Smith, 2006]. Two behavioral contexts have been described for gouging a tree: Either a nutritive or a nonnutritive context. Tree gouging for feeding was observed in gum-feeding or insectivorous animals, which used this technique to scratch the bark and access the tree fluid or the insect larvae under the bark in order to eat it [see Nash, 1986]. Nonnutritive tree gouging has been observed in strepsirrhines and New World monkeys for scent marking a site. Wild male Milne-Edwardś sifakas combine tree gouging with chest-anogenital scent marking for overmarking [Patel & GirardButtoz, 2008]. Callithrix may urinate in holes in both dry and fresh wood and perform genital rubbing in their vicinity [Coimbra-Filho & Mittermeier, 1976]. Similarly, marmosets deposit a high number of scent marks at gouging sites where they have previously fed [Lacher et al., 1981]. It has been suggested that these gouging sites function as ‘‘bulletinboards’’ for the intragroup communication of information specific to an individual [e.g. Lazaro-Perea et al., 1999; Rylands, 1985; Smith, 2006]. In this study, Lepilemur never ate the wood or licked the bark while gouging and we, therefore, hypothesize that tree gouging in L. mustelinus was not a feeding behavior but had more a non-nutritive function. Tree gouging leaves long-lasting visible marks which, for instance, may be used either for signaling the presence of an owner of a specific shelter to other individuals or by the owner itself in order to return to a safe shelter. Because the visual acuity of prosimians is low [Pereira, 1995] and sportive lemurs are furthermore nocturnal, an alternative explanation would be that tree gouging in sportive lemurs is not only used for visual marking but for scent marking. As reported for wild Milne-Edwards’ sifakas [Patel & GirardButtoz, 2008], Lepilemur could use a combination of tree gouging and scent marking. Lepilemur have no scent glands, except a scrotal gland in males [Petter et al., 1977; p 289]. Because we did not observe Lepilemur urinating on or close to the gouged sites and males never performed a genital rubbing behavior, we hypothesize that by gouging the tree bark, animals may scent mark by depositing saliva on the substrate. Indeed, saliva can contain hormone molecules which play a role in social behavior in mammals [see Gröschl, 2009]. For instance, mice and pigs use salivary steroids to mark their territory or attract potential mates [see Gröschl, 2009]. In nonhuman primates, the presence of pheromones in saliva has not been properly investigated yet. In fact, relatively little is known about pheromones in primates, but they seem to play a great role in strepsirrhines and platyrrhines [Aujard, 1997; Barrett et al., 1993]. Strepsirrhines and platyrrhines are actually the only primates owning both functional main olfactory and accessory olfactory systems [see Swaney & Keverne, 2009]. We already know that in mouse lemurs the accessory olfactory system is critical for pheromonal communication [Aujard, 1997]. To date, nothing is known regarding potential pheromonal communication in other strepsirrhines. Behavioral Differences Between L. edwardsi and L. mustelinus Our observations showed that tree gouging was observed only in one of the two studied populations, i.e. only in L. mustelinus from Andasibe and not in L. edwardsi from Ampijoroa. Such a difference in the behavioral repertoire may have been shaped by socioecological constraints related to the habitat quality and the population density. Earlier studies showed that in both study sites, sportive lemurs preferred tree holes and that they used only a few sleeping sites [Rasoloharijaona, 2001; Rasoloharijaona et al., 2008]. Since the distribution of lemurs using tree holes is suggested to be associated with the availability of suitable shelters [Lutermann, 2001; Rasoloharijaona et al., 2003], we propose that both study areas have a limited number of safe shelters for sportive lemurs. In both study sites, sportive lemurs should thus defend such an essential resource. However, both study sites differ qualitatively in their food resource availability. Although the rainforest of Andasibe is characterized by dense vegetation and large evergreen trees, Ampijoroa is covered by a dry deciduous forest that is influenced by a higher seasonality. Food resources (mainly leaves, fruits, and seeds for Lepilemur) are thus more accessible in Andasibe than in Ampijoroa, especially during the austral winter during which we conducted the observations. Both study sites differed in sportive lemur’s density. In Ampijoroa, the population density is higher than in Andasibe. Competition for food is thus higher for Lepilemur in Ampijoroa than in Andasibe. We thus suggest that during our study in the dry season Lepilemur in Ampijoroa needed to compete for both feeding and sleeping sites while Lepilemur in Andasibe primarily needed to compete for sleeping sites. Milne-Edwards’ sportive lemurs defend sleeping and feeding sites mainly by loud call displays, which represent a direct Am. J. Primatol. 1070 / Rasoloharijaona et al. and short-time information transmission for neighbors situated less than 200 m apart [Rasoloharijaona et al., 2006]. Weasel sportive lemurs from Andasibe rarely vocalized, but population density is also lower, which limits the probability for a vocalizing animal to be heard by neighbors. In order to optimize social communication, an alternative strategy would be to deposit longer lasting information by using visual or/and scent marking instead of using short-term signals, such as vocalizations. In the case of treegouging animals, they may thus transmit information on not only the ownership of their sleeping site but also potentially on their reproductive status during the mating season. Earlier investigations showed that the vocal activity of L. edwardsi from ten different local populations was not affected by the population density [Rabesandratana, 2006]. However, because the populations were geographically close (all belonging to the Ankarafantsika National Park in northwestern Madagascar), they could still belong to the same metapopulation. Resulting from our findings, we suggest that differences in the vocal and chemical communication between L. edwardsi and L. mustelinus may have actually evolved as a long-term adaptation to the use of short-term signals in high population density areas and to long-lasting signals in low-density population areas. Intra- and interspecific variability in the vocal and chemical communication needs to be investigated further. Possible Costs of the Tree Gouging— Implications for Conservation Scent marking may be costly for animals if predators are able to use the marks to locate a sleeping site. For instance, Franklin et al.  reported on a large population of golden lion tamarins (Leonpithecus rosalia) in Brazil decreasing from approximately 330 to 220 individuals in a time span of 5 years, owing to a dramatic increase in predation at sleeping sites. Natural predators used the scent marking of these cavity-nesting primates in order to localize them. During our observations, gouge marks were not directly present on the sleeping trees, but sportive lemurs were observed to leave the marks on trees surrounding the sleeping sites. Further investigations would be needed to determine which surrounding trees are marked, i.e. whether sportive lemurs use particular tree species and whether there is a marking pattern in relation to the distance from the sleeping tree. Indeed, with the help of these gouge marks, we were able to easily localize a sleeping hole or nest of a sportive lemur in the forest. These observations imply two statements: First, gouge marking could be costly for L. mustelinus if natural diurnal predators could use the marks in the vicinity of the sleeping site. Diurnal predators in Andasibe are the fosa (Cryptoprocta ferox) which can be active both Am. J. Primatol. day and night and other birds of prey [mainly eagles, such as Accipiter sp.; see Goodman et al., 2003]. Second, sportive lemurs are among those species which are frequently poached in Malagasy forests [Garcia & Goodman, 2003; Golden, 2009]. They are usually easy to capture during the day while sleeping in tree holes. Indeed, poachers were reported to have followed the gouge marks in order to track down the sleeping holes of sportive lemurs. Particular conservation and protection efforts in areas where sportive lemurs signal their presence by marking trees are thus essential to shield the local populations from poaching. CONCLUSION Altogether, we demonstrate in this article that two cryptic species of nocturnal lemurs display differences in the use of vocal and chemical communication to signal the use or ownership of resources. According to our hypothesis, L. edwardsi, which lives in a high-density population in the dry deciduous forest, displayed loud calls, whereas L. mustelinus, which lives in a low-density population in the rainforest, did not emit loud calls. Most of the loud calls were produced while feeding or in the vicinity of resting places, which corroborate with our earlier investigations that L. edwardsi produce loud calls for signaling territory ownership. For the first time, here we describe tree-gouging behavior, which occurred outside a feeding context in a nocturnal lemur species. We suggest that tree gouging is used by individuals for signaling sleeping site ownership. Further investigations are needed in order to provide more insights into the mechanisms driving the evolution of social communication. ACKNOWLEDGMENTS We thank the Malagasy authorities and, in particular, the Ministère de l’Environnement et des Forêts and the Madagascar National Parks for their permission to conduct this study in the National Park of Ankarafantsika, Andasibe—Mantadia and the private Maromizà Reserve and for their logistic support. For the help we received in Ankarafantsika, we thank Jaofetra Randrianarivony, Michel Bisoa, and Romain Rajaonirinaharison. For the help given in Andasibe, we thank Alexandre Rasolofonirina (MNP Andasibe), Jaofetra Randrianarivony, Pafo Velontiana, and Hermano Raveloarisoa (NAT Foundation). We are indebted to Rose Marie Randrianarison for the photograph of a marked tree. We thank Elke Zimmermann, Ute Radespiel, and two anonymous reviewers for valuable comments on an earlier version of the article. REFERENCES Altmann J. 1974. Observational study of behavior: sampling methods. Behaviour 49:227–267. Tree Gouging in Sportive Lemurs / 1071 Andriaholinirina N, Fausser JL, Roos C, Zinner D, Thalmann U, Rabarivola C, Ravoarimanana I, Ganzhorn JU, Meier B, Hilgartner R, Walter L, Zaramody A, Langer C, Hahn T, Zimmermann E, Radespiel U, Craul M, Tomiuk J, Tattersall I, Rumpler Y. 2006. Molecular phylogeny and taxonomic revision of the sportive lemurs (Lepilemur, primates). BMC Evolutionary Biology 6: 17(online 1–13). Aujard F. 1997. Effect of vomeronasal organ removal on male socio-sexual responses to female in a prosimian primate (Microcebus murinus). Physiology and Behavior 62: 1003–1008. Barrett J, Abbott D, George L. 1993. Sensory cues and the suppression of reproduction in subordinate female marmoset monkeys, Callithrix jacchus. Journal of Reproduction and Fertility 97:301–310. Barton RA. 2006. Olfactory evolution and behavioral ecology in primates. American Journal of Primatology 68:545–558. Braune P, Schmidt S, Zimmermann E. 2005. Spacing and group coordination in a nocturnal primate, the golden brown mouse lemur (Microcebus ravelobensis): the role of olfactory and acoustic signals. Behavioral Ecology and Sociobiology 58:587–596. Brown RE, Macdonald DW. 1985. Social odours in mammals, Vol. 2. Oxford: Clarendon Press. 340p. Coimbra-Filho AF, Mittermeier RA. 1976. Exudate-eating and tree- gouging in marmosets. Nature 262:630. Craul M, Zimmermann E, Rasoloharijaona S, Randrianambinina B, Radespiel U. 2007. Unexpected species diversity of Malagasy primates (Lepilemur spp.) in the same biogeographical zone: a morphological and molecular approach with the description of two new species. BMC Evolutionary Biology 7: 83(online). Di Fiore A, Link A, Stevenson PR. 2006. Scent marking in two western amazonian populations of woolly monkeys (Lagothrix lagotricha). American Journal of Primatology 68: 637–649. Evans CS. 2006. Accessory chemosignaling mechanisms in primates. American Journal of Primatology 68:525–544. Fonseca GAB, Lacher Jr TE. 1984. Exudate-feeding by Callithrix jacchus penicillata in semideciduous woodland (Cerradao) in Central Brazil. Primates 25:441–450. Franklin SP, Miller KE, Baker AJ, Dietz JM. 2007. Do cavity nesting primates reduce scent marking before retirement to avoid attracting predators to sleeping sites? American Journal of Primatology 69:255–266. Garcia G, Goodman SM. 2003. Hunting of protected animals in the Parc National d’Ankarafantsika, northwestern Madagascar. Oryx 37:115–118. Ganzhorn JU. 1988. Food partitioning among Malagasy primates. Oecologia 75:436–450. Golden CD. 2009. Bushmeat hunting and use in the Makira forest, northeastern Madagascar: a conservation and livelihoods issue. Oryx 43:386–392. Goodman SM, Ganzhorn JU, Rakotondravony D. 2003. Introduction to mammals. In: Goodman SM, Benstead JP, editors. The natural history of Madagascar. Cambridge: Cambridge University Press. p 1159–1186. Gröschl M. 2009. The physiological role of hormones in saliva. BioEssays 31:843–852. Hauser MD. 1996. The evolution of communication. Cambridge, MA: MIT Press. 771p. Heymann EW. 2006. Scent marking strategies of new World primates. American Journal of Primatology 68:650–661. Holm S. 1979. A simple sequentially rejective multiple test procedure. Scandinavian Journal of Statistics 6:65–70. Horiuchi S. 2008. High population density promotes the evolution of ownership. Ecological Research 23:551–556. Irwin MT, Samonds KE, Raharison JL, Wright PC. 2004. Lemur latrines: observations of latrine behavior in wild primates and possible ecological significance. Journal of Mammology 85:420–427. Lacher Jr TE, Da Fonseca GAB, Alves Jr C, Magalhaes-Castro B. 1981. Exudate eating, scent-marking, and territoriality in wild populations of marmosets. Animal Behavior 29:306–307. Lazaro-Perea C, Snowdon CT, Arruda MF. 1999. Scentmarking behavior in wild groups of common marmosets (Callithrix jacchus). Behavioral Ecology and Sociobiology 46:313–324. Lewis RJ. 2006. Scent-marking in Sifaka: no one function explains it all. American Journal of Primatology 68:622–636. Lutermann H. 2001. Weibchenassoziationen und Fortpflanzungsstrategien beim Grauen Mausmaki (Microcebus murinus) in Nordwest-Madagaskar. Dissertation. University of Hannover, Hannover. 171p. Martin P, Bateson P. 1993. Measuring behavior. An introductory guide, 2nd edition. Cambridge: Cambridge University Press. 238p. Méndez-Cárdenas M, Randrianambinina B, Rabesandratana A, Rasoloharijaona S, Zimmermann E. 2008. Geographic variation in loud calls of sportive lemurs (Lepilemur ssp.) and their implications for conservation. American Journal of Primatology 70:828–838. Méndez-Cárdenas MG, Zimmermann E. 2009. Duetting—a mechanism to strengthen pair bonds in a dispersed pairliving primate (Lepilemur edwardsi)? American Journal of Physical Anthropology 139:523–532. Mertl-Millhollen AS. 2006. Scent marking as a resource defense by female Lemur catta. American Journal of Primatology 68:605–621. Miller KE, Laszlo K, Dietz JM. 2003. The role of scent marking in the social communication of wild golden lion tamarins, Leontopithecus rosalia. Animal Behaviour 65:795–803. Mitani JC, Rodman PS. 1979. Territoriality: the relation of ranging pattern and home range size to defendability, with an analysis of territoriality among primate species. Behavioral Ecology and Sociobiology 5:241–251. Mittermeier RA, Konstant WR, Hawkins F, Louis EE, Langrand O, Ratsimbazafy J, Rasoloarison R, Ganzhorn JU, Rajaobelina S, Tattersal I, Meyers DM. 2006. Lemurs of Madagascar. Washington, DC: Conservation International. 520p. Mittermeier RA, Ganzhorn JU, Konstant WR, Glander K, Tattersall I, Groves CP, Rylands AB, Hapke A, Ratsimbazafy J, Mayor MI, Louis EE, Rumpler Y, Schwitzer C, Rasoloarison RM. 2008. Lemur diversity in Madagascar. International Journal of Primatology 29:1607–1656. Nash LT. 1986. Dietary, behavioral, and morphological aspects of gummivory in primates. American Journal of Physical Anthropology 29:113–137. Palagi E, Norscia I. 2009. Multimodal signaling in wild Lemur catta: economic design and territorial function of urine marking. American Journal of Physical Anthropology 139: 182–192. Patel ER, Girard-Buttoz C. 2008. Non-nutritive tree gouging in wild Milne-Edwards’ sifakas (Propithecus edwardsi): description and potential communicative functions. Primate Eye 96:283. Pereira ME. 1995. Development and social dominance among group-living primates. American Journal of Primatology 37:143–175. Petter JJ, Schilling A, Pariente G. 1971. Observations écoéthologiques sur deux lémuriens malgaches nocturnes : Phaner furcifer et Microcebus coquereli. Terre Vie 118:287–327. Petter JJ, Albignac R, Rumpler Y. 1977. Mammifères lémuriens (primates prosimiens). Faune de Madagascar No. 44. Paris: ORSTOM-CNRS. 513p. Pochron ST, Morelli TL, Scirbona J, Wright PC. 2005. Sex differences in scent marking in Propithecus edwardsi of Ranomafana National Park, Madagascar. American Journal of Primatology 66:97–110. Rabesandratana AZ. 2006. Variations microgéographiques et bioacoustiques de Lepilemur edwardsi (Geoffroy, 1858) dans Am. J. Primatol. 1072 / Rasoloharijaona et al. le Parc National Ankarafantsika (région Nord-Ouest de Madagascar). Dissertation. University of Antananarivo Madagascar. 104p. Rasoloharijaona S. 2001. Contribution à l’étude comparative de la communication vocale et de la vie sociale de deux formes de Lepilemur (Geoffroy, 1858) (Lepilemuridae, Petter et al., 1977) provenant de la forêt sèche de l’Ouest et de la forêt humide de l’Est de Madagascar. Dissertation. University of Antananarivo Madagascar. 150p. Rasoloharijaona S, Rakotosamimanana B, Zimmermann E. 2000. Infanticide by a male Milne Edwards’sportive lemur (Lepilemur edwardsi) in Ampijoroa NW Madagascar. International Journal of Primatology 21:41–45. Rasoloharijaona S, Rakotosamimanana B, Randrianambinina B, Zimmermann E. 2003. Pair specific usage of sleeping sites and their implications for social organization in a nocturnal malagasy primate, the Milne Edwards’ sportive lemur (Lepilemur edwardsi). American Journal of Physical Anthropology 122:251–258. Rasoloharijaona S, Randrianambinina B, Braune P, Zimmermann E. 2006. Loud calling, spacing, and cohesiveness in a nocturnal primate, the Milne Edwards’sportive lemur (Lepilemur edwardsi). American Journal of Physical Anthropology 129:591–600. Am. J. Primatol. Rasoloharijaona S, Randrianambinina B, Zimmermann E. 2008. Sleeping site ecology in a rainforest dwelling nocturnal lemur (Lepilemur mustelinus): implications for sociality and conservation. American Journal of Primatology 70:247–253. Rylands AB. 1985. Tree-gouging and scent-marking by marmosets. Animal Behaviour 33:1365–1367. Schilling A. 1979. The study of prosimian behavior. In: Doyle GA, Martin RD, editors. Olfactory communication in primates. New York: Academic Press. p 461–542. Smith T. 2006. Individual olfactory signatures in common marmosets (Callithrix jacchus). American Journal of Primatology 68:585–604. Swaney WT, Keverne EB. 2009. The evolution of pheromonal communication. Behavioural Brain Research 200:239–247. Tan CL, Drake JH. 2000. Evidence of tree gouging and exudate feeding in pygmy slow lorises (Nycticebus pygmaeus). Folia Primatologica 72:37–39. Thalmann U. 2006. Behavioral and ecological adaptations in two small folivorous lemurs with different social organization: Avahi and Lepilemur. In: Gould L, Sauther ML, editors. Lemurs: ecology and adaptation. New York: Springer. p 327–352. Thalmann U, Ganzhorn JU. 2003. Lepilemur, sportive lemur. In: Goodman SM, Benstead JP, editors. The natural history of Madagascar. Chicago: University Chicago Press. p 1336–1340.