«The Indris of Anjanaharibe-Sud, Northeastern Madagascar Urs Thalmann,1 Thomas Geissmann,1 Arsène Simona,2 and Thomas Mutschler1 Received July 30, ...»
International Journal of Primatology, Vol. 14, No. 3, 1993
The Indris of Anjanaharibe-Sud, Northeastern
Urs Thalmann,1 Thomas Geissmann,1 Arsène Simona,2 and Thomas
Received July 30, 1991; accepted September 10, 1991
During a short field trip to the Special Reserve of Anjanaharibe-Sud in
northeastern Madagascar, data concerning pelage coloration, behavior
(especially vocalization), and ecology of indris were collected.
Anjanaharibe-Sud is the northernmost locality of indri distribution. In comparison to the better-known indris from the southern part of their dis- tribution, the indris in this region show different pelage coloration. Several types of loud vocalizations are analyzed, based on a small sample of tape recordings. Their song structure is more complicated than previously re- ported, containing distinct sequences of duetting. Data on behavior and ecology were collected by interviewing guides and local inhabitants. Some information contrasts with reports on the more southern indri populations.
The conservation status of indris in Anjanaharibe-Sud and the future of the reserve are outlined.
KEY WORDS: indri; Madagascar; Indri indri; vocalization; duet song.
2 Chef de Poste de la RNI 12 Marojejy, B.P. 55, Andapa 205, Madagascar.
357 0164-0291/94/0600-0357$07.00/0 © 1993 Plenum Publishing Corporation 358 Thalmann, Geissmann, Simona, and Mutschler cluding the Masoala peninsula (Petter et al., 1977; Tattersall, 1982). In the northern part of its distribution, the indri is generally considered to be “at best exceedingly rare” (Tattersall, 1982, p. 91). Indris are not known to occur in the Marojejy Reserve situated northwest of the Andapa Basin (Fig.
1). On the other hand, indris were recently reported to be still present in the “Réserve Spéciale” (RS) of Anjanaharibe-Sud, southeast of the Andapa Basin (Nicoll and Langrand, 1989). This appears to be the northern limit of their distribution.
We report preliminary observations on the indris in Anjanaharibe-Sud.
They are of special interest because most previous information on indris was collected in localities in the southern part of their range of distribution, especially in the region of Perinet.
Fig. 1. Location of the RS (Special Reserve) Anjanaharibe-Sud and our camp site (schematic map after Nicoll and Langrand, 1989). Inset map: Madagascar.
Data Collection Three of us (U.T., A.S., and T.M.) visited the southern part of the RS Anjanaharibe-Sud for 6 days (4–9 October 1990) in order to look for indris (camp at about 14°46’S; 49°30’E, altitude 790 m). At that site, we collected data on external appearance (four individuals) and vocalizations, supported by video- and audio-recordings. Additional information was obtained from local guides.
For comparison of pelage coloration, skins of indris from known localities were studied in the British Museum of Natural History in London (n = 8) and the Rijksmuseum van Natuurlijke Historie in Leiden (n = 9).
In addition, photographs of individually distinguishable animals from Pollock (1975b) were analyzed as far as possible (n = 7). All but three individuals [two from Pollock (1975b) and one specimen observed by us in Anjanaharibe-Sud] were probably adult; the others appeared to be subadult. Seven characters of the pelage coloration were evaluated, of which five are considered in this paper (Table I).
For vocal comparison, several tape-recordings of indri vocalizations collected in August 1968 at Perinet were made available by Professor R.
D. Martin. Additional vocalizations contained on several television films were also included in the comparison (Attenborough and Blanchard, 1969;
Attenborough and Parsons, 1980; Herzog and Wothe, 1989; Salisbury et al., 1988). Finally, various gibbon songs tape-recorded by one of us (T.G.) were available for comparison. The song excerpt used for a sonagram was tape-recorded at the Metro Zoo in Miami (U.S.A.) on 31 September 1988.
On some tape-recordings, several indris are vocalizing together. In order to identify the notes uttered by each individual, we relied on characteristics that we were able to recognize on the tape-recordings or on the sonagrams (such as similarity of note shapes, recurrent patterns of note combinations, intensity of the various song contributions, or overlapping contributions), since we were unable to film or to observe the animals directly while they were singing. Although we believe that we correctly assigned most of the notes to the singing individuals, this method is not completely reliable.
Coordinates for Malagasy localities other than Anjanaharibe-Sud mentioned in this article were extracted from the Official Standard Names Gazetteer no. 2 (United States Board on Geographic Names, 1955).
Spearman Rank Correletion Coefficients were computed using the StatView II statistics software.
Sonagraphic Devices All tape-recorded indri vocalizations were digitized on a Macintosh II computer using the SoundRecorder device (Farallon). The sampling rate is defined as “the number of intervals per second used to capture a sound when it is digitized” (Schmidt et al., 1989, p. 222). Unless otherwise stated, all sounds were sampled at a 5-kHz sampling rate, thus removing frequencies above 2.5 kHz (Schmidt et al., 1989). Sonagrams of all indri vocalizations were generated with the program SoundEdit, Version 2.0.1 (Farallon). The sonagrams were then copied to an image-editing program (Image, NIH, version 1.14u), with which background noise and echo effects were removed.
Terms and Definitions A note is a single, continuous sound of any distinct frequency modulation, produced by either an inhaled or an exhaled breadth. A phrase is a larger and looser collection of notes identifying a single vocal activity, whereas sequences refer to vocal and nonvocal behaviors which are more closely related to each other than to preceding and succeeding behaviors and which, when taken together, form a single theme. These definitions were developed by Haimoff (1984) for the study of gibbon vocalizations.
Following the definition of Thorpe (1961, p. 15) a song is “a series of notes, generally of more than one type, uttered in succession and so related to form a recognizable sequence or pattern in time.” A duet is defined here as the joint vocalization of two individuals, coordinated in time and/or in selection of distinct note-types (Wickler, 1974).
Population To judge from the number of loud indri songs we heard, the population density in this region appears to be fairly high. However, it is quite difficult to find and observe the animals. The terrain is very hilly, with steep valleys. Each indri group sings only a few times a day, and even if the observer is very close to the group, the animals may still be well hidden in the canopy at a height of about 20-30 m. We saw a single individual on one day, and a group of 3 animals on the following day, several hundred meters away from the first sighting. According to our guides, the group size is usually two or three individuals; the largest group ever seen by a guide consisted of six animals.
The indris, called “Babakoto” by the local people, are protected in this region by a taboo (fady), and are therefore not hunted by the members of the local Tsimihety tribe. They believe that, in early times, indris gave them medicine against injuries caused by fights against invaders with iron weapons.
Pelage and external characters Our guides had not observed any obvious variability in the fur pattern of the indris in this region. However, they reported a certain degree of sexual dimorphism and an age change in coloration. The following descriptions are based mainly on the 4 individuals we observed, but are in full accordance with reports from our guides. There is a constant pattern of coloration (Figs. 2 and 3a): The indris are black with a white face ring, white fur on the sides of the abdomen and a white pygal triangle on the back, extending caudally to the gluteal region, including the vestigial tail, and the heels are white to yellowish. The eyes are yellow.
Figure 4 shows the coloration pattern and the variation of five easily recognizable characters (Table I) along a geographical gradient from north to south. The figure uses information on 28 indris from known localities.
These include museum specimens, our sightings of indris in AnjanaharibeSud, and photographs in Pollock (1975b).
In the north, indris usually have a light face ring, but no light occipital patch and no laterally extending light collar. The outer sides of the lower arm and the leg are black. In indris from the southern region, the face is usually black or sometimes has pale patches over the eyebrows or on the cheeks, with only one specimen in our sample showing a white face ring (Pollock, 1975b, plate 6.4; Freeman, 1978, p. 28, photograph by Pollock).
In addition, southern indris exhibit pale or white markings of considerable but variable extent which are usually absent in northern indris: There is a patch on the back of the head (occipital patch) and a lateral light collar extending up behind the ears in most southern individuals. The outer sides of the forearm and hindlimb are grayish or whitish (Fig. 3b).
0.5 0.5 Fig. 4. Frequency distribution of selected characters of pelage coloration (each encoded from 0 to 1 on the vertical axis) along a gradient from north to south (horizontal axis). (a) Face ring; (b) occipital patch; (c) lateral collar; (d) outer side of forearm; (e) outer side of hindlimb. The diameter of the circles is proportional to the number of specimens with the same character code. The key to the character code is presented in Table I. Above each graph, the Spearman Rank correlation coefficient (rs) corrected for ties and the error probabilitiy (p) is shown.
0.5 A mixed pattern occurs at the latitude of Mananara (16°23’S, 49°44’E; n = 7): All individuals have a face ring, albeit narrow in most of them. The light occipital patch is present in four individuals (white in three individuals, gray in one of them). Most of the individuals from Mananara have grayish outer forearms (n = 5) and outer hindlimbs (n = 5).
Circadian Distribution of Songs Because they are difficult to spot, the very loud territorial vocalizations (songs) of the indris are frequently the only sign of their presence.
Our main aim was to observe them, and the songs helped us to locate them.
On the mornings of 5 to 7 October, the first morning songs occurred at 0808, 0847, and 0916, respectively. On the first day, we heard the second song at 0815, but we were not sure whether it was produced by the same group or not. A different group was heard at 0848, and 4 min later a third group began to sing. A single individual vocalized at 0903. A particularly intense period of singing occurred from about 1040 to 1100. Standing on a hill, we heard many groups (song cluster). On the second day, we paid special attention to the first group we heard (0847) because it was very close. They vocalized again at 0954, but they uttered “honk” calls instead of songs. The honks were probably a warning signal elicited by our guides, who found the indris at the same time. After 0954 we heard other groups of indris calling from some distance away for several minutes. At about 1100, we again heard songs of several groups, but the group we were observing did not join in, perhaps because of our presence. After 1130 we heard no further calls.
At the time we visited the reserve indri songs were only heard in the morning between 0800 and 1130. Our guides told us that the number of songs depends, to some degree, on the weather; the indris sing more frequently when rain is expected. During the year, the daily frequency distribution of songs appears to change: According to our guides, indris also sing from midnight to morning in April.
Classes of Vocalizations
The three different classes of loud indri calls recorded at Anjanaharibe-Sud are briefly described as follows:
Waa Notes. This class of vocalization appears in both of the two initial stages of songs tape-recorded at Anjanaharibe-Sud. Waa notes are relatively short, each beginning with a steep rise in frequency, which declines slightly towards the end of the note (Fig. 5a, first half of sonagram). Their fundamental frequency range is about 0.5–0.9 kHz.
Three instances of similar notes were also found among the tape-recordings from Perinet, each occurring at the very beginning of a song (Fig.
5b, first half of sonagram). Waa notes were, on all tape-recordings in which they occurred, uttered by two or more animals as a chorus, here termed a waa sequence. Each waa sequence lasted several seconds and occurred only once, at the beginning of a song. Apperantly, no sonagrams of this vocalization have been published previously.
Waa notes were first described by Petter (1962, p. 109) as barks which initiate the song: “aboiements très forts, émis par tous les Indri en même temps, ‘ouai-ouai-ouai-ouai’….” Similar descriptions have also been pro
Fig. 5. Sonagrams showing the transition from the waa sequence to the long note sequence during the initial stages of an indri song. (a) AnjanaharibeSud; (b) Perinet.