International Journal of Primatology

, Volume 28, Issue 4, pp 895–905

Biology of Cheirogaleus major in a Littoral Rain Forest in Southeast Madagascar


    • Department of Animal Ecology and Conservation, Biozentrum GrindelUniversity of Hamburg

DOI: 10.1007/s10764-007-9163-3

Cite this article as:
Lahann, P. Int J Primatol (2007) 28: 895. doi:10.1007/s10764-007-9163-3


Greater dwarf lemurs (Cheirogaleus major) are small nocturnal primates from the rain forests of eastern Madagascar. I investigated a population of Cheirogaleus major in a littoral rain forest of Southeast Madagascar during 2 rainy seasons to supplement the sparse information available for the species. I collected data on morphology, group composition, sleeping behavior, home range, and social organization via mark/recapture, radio telemetry, and focal individual observations. I identified 2 presumed family groups, and my data from radiotracking revealed a monogamous social organization. In each group, I found an adult pair and its presumed offspring sharing home ranges and sleeping sites together. I also observed gregarious behavior of group members during their nocturnal activity. I found no difference in body measurements between sexes, but body mass and tail circumference increased significantly from November to February, indicating a fatting period before hibernating.


Cheirogaleus majorgregariousnesslemurslittoral rainforestMadagascarmonogamousmorphology


Cheirogaleus major live in the rain forests of eastern Madagascar from Taolagnaro in the southeast to Montagne d’Ambre in the north and to the Tsaratanana Massif and the Sambirano region in the west (Harcourt and Thornback 1990; Mittermeier et al.2006; Tattersall 1982). Studies of the species are still rare (Wright and Martin 1995). Anecdotal observations and a few previous studies indicate that Cheirogaleus major spends 3 mo/yr in torpor or hibernation, hiding either in leaf litter at the base of trees or 10–12 m up in large trees (Wright and Martin 1995; Tattersall 1982). They feed on fruits, flowers, and insects (Ganzhorn 1988, 1989; Lahann 2007). The body masses of adults are 350–450 g (Wright and Martin 1995; Mittermeier et al.2006). They gain weight during the rainy season from October to April, and the body mass and tail circumference increase then.

Researchers know little about their social organization, home range size, and social behavior (Wright and Martin 1995), but Cheirogaleus major might behave similarly to its smaller congeneric species, C. medius, which occurs in the western dry forests and the littoral rainforests of the southeast of Madagascar (Hapke et al.2005; Mittermeier et al.2006; Tattersall 1982). Cheirogaleus medius lives in a monogamous mating system in small family groups, consisting of a mated pair and their offspring from previous years (Fietz 1999; Mueller 1999). The family groups remain stable over years and family members share sleeping sites and home ranges (Fietz 1999; Mueller 1999). Nonetheless, individual Cheirogaleus medius are mostly alone at night; therefore, Bearder (1987) characterized them as solitary foragers and Mueller and Thalmann (2000) termed them dispersed. Ganzhorn (Kappeler 1997) reported that 2–5 Cheirogaleus major frequently feed together nocturnally.

In order to supplement the sparse information available on the species, I investigated a population of Cheirogaleus major in the littoral rain forest of Mandena, in southeast Madagascar, which represents the southern border of their distribution. I collected data via mark/recapture, radio telemetric, and focal observation techniques. I present results on morphology, group composition, sleeping behavior, home range, social behavior, and social organization of Cheirogaleus major.


Study Site

I conducted the study during 2 rainy seasons, February–April 2002 and September 2003–April 2004 at the QMM: Station Mandena in Southeast Madagascar. Mandena is a littoral rain forest (24°56′ S, 46°59′ E, 5–20 m above sea level), consisting of different sized forest fragments, 10 km northeast of Taolagnaro. The study area of 25 ha (500 × 500 m) is in a low degraded primary forest fragment M15/M16 ca. 230 ha, with thick understory and a broken canopy at a height of ca.10–15 m. I focused the study in the middle of the forest fragment to decrease possible edge effects. Trails subdivide the site into 50 × 50 m quadrats. Rainfall averages 1680 mm/yr (measured by staff of QMM). The dry season is from May until September, with lower rainfall of 75–149 mm/mo and the rainy season from October to April, with higher rainfall of 152–179 mm/mo. The average monthly temperatures range between 19.8°C (July; dry season) and 26.2°C (January, February; rainy season) (Ramanamanjato and Ganzhorn 2001). Five other sympatric lemur species occur in this littoral rain forest: Microcebus murinus, Cheirogaleus medius, Avahi laniger, Eulemur collaris, and Hapalemur griseus (Creigthon 1990).


I captured Cheirogaleus major in Tomahawk live traps (Model 204, 50.8 cm × 17.8 cm × 17.8 cm, PO Box 323 Tomahawk, WI). I placed 20 traps in the first season and 38 in the second season at 50–m intervals in an area of 400 m × 300 m at a height of 4–12 m. I set the traps for 4 consecutive nights each month during the study period. I baited the traps with ripe banana in the late afternoon and then opened them. I checked them at dawn the next day or during the reproduction time (November and February) in the same night, to avoid separation of lactating females from their infants for a longer period. I trapped group II individuals only at their sleeping tree at a height of 9 m (n = 2) and group I individuals at 5 locations: in the tree next to their sleeping place at a height of 8–9 m (n = 13), at distances of 25 m east, 25 m west, 50 east, and 100 m east from the sleeping place at a height of 8–10 m (each n = 1). I anaesthetized the captured individuals with subcutaneous injections (0.01 ml/100 g of body mass) of 100 mg/ml of Ketavet and then sexed, weighed, and measured them per Schmid and Kappeler (1994). I marked them individually with subcutaneous transponders (Trovan: type ID—100, Fa. Telinjet, D—Roemerberg) and determined the reproductive condition by examining nipples and by abdominal palpation to check for pregnancy per Foerg (1982). I based age determinations on body measurements and capture date and classified individuals as juveniles (<1.5 yr) if they were <300 g (M30). I classified females as subadults (1.5–2.5 yr) if they showed no sign of reproduction such as swollen vulva, pregnancy, or lactating (F31, F33, F34). I classified M33 as subadult in February 2002, because he was a juvenile (<300 g) in December 2000 (data of Andreas Hapke). I classified individuals as adults (>2.5 yr) if they were in estrus, pregnant, or lactating (F32, F30) or if they were already captured in 2000 by Andreas Hapke (pers. com.) as full grown (M32) or if they had larger testes during reproduction time than those of subadults and juveniles (M35). After anesthesia for 10–30 min I kept the subjects in Sherman traps with banana for ≥5 h to enable their recovery from anesthesia. I released them at dusk or at dawn at their respective trapping site location.

Telemetry and Observation

In the first season I equipped 1 adult male (M32), 1 adult female (F30), 1 subadult female (F34), and 1 subadult male (M33) with radiocollars (Fa. Biotrack, UK—Dorset, 2.5–5 g). The subadult female (F34) and the subadult male (M33) lost their collars. I followed M32 and F30 for 1 mo and radiotracked them (Table I) via a TR—4 receiver with a RA—14K antenna (Telonics, Mesa, AZ). In the second season I equipped 2 adult males (M31, M35), 1 juvenile male (M30), 2 adult females (F30, F32), and 1 subadult female (F31) with radiocollars and followed them for 1–2 mo via radiotracking, before their collars were chewed off by other individuals. I recollected all radiocollars by searching on the ground, except the ones chewed off in the sleeping hole. I tracked 1–5 individuals simultaneously between 1800 and 2400 h for a total of 18 nights. I located the position of each radiocollared individual once an hour via 2 bearings taken from 2 different intersections of the marked trail system. Groups and families were not always completely radiocollared.
Table I

Data of studied and radiocollared individuals of Cheirogaleus major in Mandena





Date of first capture

Body mass (g)

Testis-volume (mm3)

Time radiocollared

Gregariousness (%)


HR (ha)






418 (Feb.)


02/01/02–03/15/03 11/01/03–11/22/03









345 (Sept.)











378 (Dec.)











398 (Feb.)











265 (Nov.)

504 (Dec.)










305 (Nov.)

2799 (Nov.)










478 (Feb.)

2082 (Feb.)










295 (Dec.)

266 (Feb.)









409 (Jan.)

2201 (Jan.)





A = Adults; SA = subadults; JUV = juveniles; mo of first measured body mass and testis volume are in parentheses; Gregariousness: the number of encounters divided by the number of total observations; Fix = Number of fixes during radiotracking; HR = home range sizes in hectares.

aData provided by Andreas Hapke.

I triangulated the positions of the individuals later via Ktrail (Kelinski, unpublished computer programme) and analyzed the home ranges via Tracker (Camponotus AB, Solna, Sweden) via the minimum convex polygon technique (Sterling et al.2000) and calculating the monthly home range sizes with ≥30 positions. I located the sleeping sites of the radiocollared Cheirogaleus major on a total of 54 d between 0800 and 1000 h and determined the sleeping group composition via radiotracking and observation, distinguishing between sleeping alone, in a group, or unknown.

I assessed the social behavior via direct observation of individuals during my nocturnal work. At the same time I recorded the number of individuals, their distances from each other, and if possible the name of the individuals and their behavior. I determine the gregariousness by dividing the number of events, when ≥2 family members (n = 88) were together (encounters), by the number of total observation events when I saw ≥1 (n = 181).

Data Analysis

I provide body measurements as mean and standard deviation (SD) and home range sizes as medium and 25/75 Quartiles. I used Mann-Whitney U test to compare body measurements and home range sizes between males and females. I conducted the Kruskal-Wallis test to compare the body masses and tail circumstances between months. I used χ2 analyses to compare the gregarious behavior between months and individuals. Statistical significance is p < 0.05 for all tests. I carried out all statistical tests via SPSS for Windows.


Trapping and Body Measurements

I captured 10 different Cheirogaleus major (5 males and 5 females; Table I). I captured F32, F33, F34, M33, and M35 only once and F31 twice and M30 4 times in the same year. I captured M31 and M32 2 and F30 3 times, but in different years. There is no significant difference between adult males and females in body measurements (Table II), but body masses and tail circumstances change significantly over the rainy season between months and increased from November to February (Table III). I measured the highest body masses in February, with a mean of 414 g (± 46.1 g). In March and April I could not capture individuals during the trapping session, though they were still active.
Table II

Body measurements of males and females of Cheirogaleus major in Mandena


Males n = 5

Females n = 5

Mann-Whitney U test

Mean (mm)

SD (mm)

Mean (mm)

SD (mm)



Head body length







Tail length







Head length







Head width







Ear length







Ear width







Hindfoot length







Testis length




Testis width




ns = not significant with p > 0.05.

Table III

Body masses and tail circumferences, separated in month of Cheirogaleus major in Mandena


November n = 6 (3 f, 3 m)

December n = 6 (3 f, 3 m)

February n = 6 (2 f, 4 m)

Kruskal-Wallis test, d.f. = 2









Body mass (g)









Tail circumferences (mm)









n = number of individuals, m = males, f = females. ***Significant with p < 0.05.

Home Ranges, Sleeping Behavior, and Family Groups

I analyzed the home ranges and sleeping behavior of 4 radiotracked males and 3 radiotracked females (Tables I and IV), but radiotracked only 1 adult female (F30) during both seasons (2001/2002, 2003/2004) (Table I). My analyses of the home range sizes revealed no significant difference between males and females (Mann-Whitney U test, Z = –0.714, p = 0.476), and the median home range size of Cheirogaleus major is 4.4 ha (quartiles: 4.0/4.5 ha; Table I). I identified 2 presumed family groups (Fig. 1): 5 radiocollared individuals belonged to 1 family group (group I) and 2 radiocollared individuals to another neighboring family group (group II). Each group consisted of 1 adult mating pair and their presumed offspring from the previous years. I could not capture and radiocollar some group members, especially the infants, but I saw them during their nocturnal activity and included them in the group compositions (Fig. 1). The home ranges of radiotracked group members overlap from 78% to 94%, whereas home range overlap between the 2 family groups depended on the individual and ranged from 0 to 5.9% (Fig. 2).
Table IV

Characteristics of sleeping sites of radiocollared Cheirogaleus major in Mandena


Percentage of use/individual

Species of tree

Height of tree (m)

Diameter of tree (cm)

Height of tree hole (m)

Group I

Main sleeping site

100% (F31: 81%)

Poupartia chapelieri




Alternative sleeping site of the subadult female F31






Group II

Main sleeping site





Fig. 1

Group composition in Cheirogaleus major 2001–2002 and 2003–2004. Different sizes of the circles indicate the different age classes of the group members ranging from large circles = adults, medium circles = subadults and juveniles; small circles= infants. Unsexed circles are individuals that I saw but did not capture and sex.
Fig. 2

Home range overlap of Cheirogaleus major. M = male, F = female. Two family groups are represented.

I could describe 3 different sleeping sites for Cheirogaleus major (Table IV). Radiotracked family members used only 1–2 sites (tree holes) for sleeping and thus shared a sleeping site for 96.4% of the observation times. The mating pair (100%) and its presumed offspring exclusively used the main sleeping site of group I. They slept in a big tree (diameter at breast height 48 cm) at a height of ca.11 m. One subadult female associated with the group occasionally (19% of observed cases) and used another tree hole for sleeping (8 m high, 47 cm diameter), 124 m from the main sleeping place of group I. Altogether I found only group members sleeping together and observed no sleeping composition with other conspecifics. However, I found individuals of the 2 sympatric living species of the family Cheirogaleidae (Cheirogaleus medius and Microcebus murinus) sleeping in the same trees of group I but in different tree holes.

In 49% of my sightings of Cheirogaleus major during the night, 2–5 group members were together <10 m from one another, normally in the same tree crown. The behavior does not vary across months (χ2 = 3.02, d.f. = 3, p = 0.389). During the mating season from October to December groups of 2–5 individuals occurred in 49% of observations. During the birth season (February, March) individuals were together in 43% of observations. Analyzing only the events when I could identify individuals by their radiocollars, there is a significant difference in gregariousness between the social behavior of females, males, and juveniles (χ2 = 6.78, d.f. = 2, p = 0.034). Being with ≥1 family members occurred 56% of the time among females (F30, F31, F32) and 40% of the time among males (M31, M32, M35). The juvenile male was together with other family members in 67% of observations (Table I).

Hibernation and Reproduction

During my first census walks at the beginning of September, I observed no individual of Cheirogaleus major. At the end of September I captured and observed the first individuals of Cheirogaleus major. I captured 2 pregnant females (F30, F32) in December (12/11/03 and 12/21/03) and saw the first infants traveling away from a nest site at the end of February. Individuals were still active by the end of April, when my observation ended.


My capture rates of Cheirogaleus major were extremely low in Mandena compared to my capture rates of sympatric C. medius and Microcebus murinus there. I captured 10 different individuals of Cheirogaleus major, and recaptured only 4 of them during 11 4-night trapping sessions. I recaptured only the juvenile male (M30) several times, possibly because of either the small population density in the study area (with 2 families presumed) or the difficulty of capturing them. I could only capture individuals by installing the traps high up in the trees (8–10 m) and at the beginning of the rainy season. They seem to be attracted by bananas in the traps because trees with ripe fruit were still rare. The best trees for capturing were the sleeping tree and the tree next to it, which they always passed, when they woke up and started to travel to the food plants. I captured individuals usually in the first hour after sunset (89%). Once captured, they seemed to avoid the traps in the following monthly trapping sessions, independently of which place they were set. Besides the difficulty in capturing subjects, I had problems with the radiocollars. Radiocollared individuals frequently had their collars chewed off by family members. Therefore I assume that radiocollars consisting of plastic strips are not suitable for Cheirogaleus major; metal would be better.

Nevertheless my data on home ranges, sleeping behavior, and observation of radiotracked individuals indicate a family group pattern and therefore a monogamous social organization for Cheirogaleus major. Apparently, an adult male and an adult females form a stable pair and share sleeping sites and home ranges with each other and with their presumed offspring from the previous years. Thus, Cheirogaleus major shows the same social organization in the littoral rain forest of Mandena as its congeneric species C. medius in the dry forests (Fietz 1999; Mueller 1999). However, individuals of Cheirogaleus medius are mostly alone at night. Therefore Bearder (1987) characterized them as solitary foragers and Mueller and Thalmann (2000) described them as dispersed. In contrast, I found the family members of Cheirogaleus major in 49% of all observation events, foraging, traveling and resting together in the crowns of the same trees <10 m from each other during nocturnal activity. Ganzhorn (Kappeler 1997) reported nearly the same gregarious behavior by the species but with lower percentages of associations of ≥2 individuals. He found during a study in the Foret d’Analamazoatra near Andasibe that in 37% of all sightings during a nocturnal census walk 2–5 individuals were feeding together within a range of ≤10 m. Such gregarious behavior with high associations of individuals occurs among many nocturnal primate species, and studies indicate that the individuals show encounters and social interactions in the night more frequently than expected by chance (Charles-Dominiques 1977; Clark 1985; Gursky 2002a, b, 2005; Sterling and Richard 1995).

Gregarious behavior in primates and other mammals may be based on the distribution of (food) resources in space and time, predation pressure, or infanticide avoidance (Gursky 2002a, b; Isbell 1994; Janson 1992; Krebs and Davies 1984; van Schaik and Kappeler 1996). Therefore one explanation for the gregarious behavior of Cheirogaleus major might be the distribution and clumping of food resources. Cheirogaleus major prefers fruits and flowers for feeding, and they also prefer large trees for feeding, resting, and traveling in the littoral rain forest of Mandena (Lahann 2007). Wright and Martin (1995) reported similar observations of large tree preferences from the rain forest of Ranomafana. Large flowering and fruiting trees are rare in Mandena because of the degradation level of the forest, so the food resources are clumped and the proximity of family members in the same trees and their gregariousness might be the consequence of the food distribution pattern.

Infanticide avoidance might be another factor. In Analamazoatra the proportion of gregarious behavior increases during and after the birth season, when infants are in the group up to 46% (December and January), from 16% in the mating season (September–October) (Ganzhorn, in Kappeler 1997). I found no difference between the mating season and the birth season, but I rarely observed infants during the birth season (n = 12). They were always associated with an adult. However, I observed no encounters between family members and unknown or neighboring individuals, which might indicate infanticidal risk. As reproduction in the species is synchronized, and Cheirogaleus major reproduce only once per season, shortened interbirth intervals could not be an explanation for infanticide, because females did not go into estrus earlier.

Predation pressure might play another role in the gregarious behavior of Cheirogaleus major. Predators of lemurs are plentiful in the dry and rain forests of Madagascar (Goodman 2003). Again gregarious behavior should increase during birth seasons because infants are more vulnerable. I have few data on the birth season, but infants always associated with adults when I watched them. During the mating season, juvenile male M30 associated closely with ≥1 adults in 67% of observations, which is significantly higher than the associations of adults with other family members. He was smaller and presumably more inexperienced than the adults. One may interpret being in close associations with an adult as predation avoidance behavior. Accordingly, I conclude that the gregarious behavior might be the consequences of the distribution of food resources and of an avoidance of predation.

The timing and duration of hibernation at the site are only partially understood. Body masses and tail circumferences increased during the rainy season, as is typical for all Cheirogaleidae, which hibernate or go into torpor in the dry season (Fietz and Ganzhorn 1999; Wright and Martin 1995). Though it is clear that Cheirogaleus major does hybernating at this site, it is not clear what the timing of hibernation is at Mandena. I could not observe individuals at the beginning of September, but I saw them emerging at the end of September, which indicates that they sleep until September. After a clear fattening period, they were still active at the end of April in both study seasons. Consequently, it is not clear when they enter hibernation. Studies on the physiological aspects of hibernation are especially necessary for Cheirogaleus major. Wright and Martin (1995) described nest building behavior for Cheirogaleus major at the rain forest site of Ranomafana. However, I did not observe the behavior in Mandena.

Otherwise my findings match the results from previous observations (Wright and Martin 1995; Ganzhorn, in Kappeler 1997). But researchers have rarely studied Cheirogaleus major, as well as the other recently discovered rain forest species, C. crossleyi (Hapke et al.2005). More studies are needed on the social organization, social behavior, and ecology, including more individuals and biogeographic variation of the broadly distributed species, to complement the information on rain forest populations of Cheirogaleidae. The related species Cheirogaleus medius and Microcebus murinus, which are broadly distributed in the western dry forests, show biogeographic differences in ecology, morphology, hibernation and torpor, and social behavior, owing to the different environments in which they live (Lahann et al.2006). I suspect that populations of Cheirogaleus major would also show ecological and behavioral differences between regions. Gregariousness should be studied in other populations of Cheirogaleus major and in C. crossleyi, to determine this is a typical behavior of the species or only a consequence of environmental factors.


I conducted the study under the accord de Collaboration between the Département de Paléonthologie et Anthropologie, the Département Biologie Animale of the Université d’Antananarivo, and the Department of Animal Ecology and Conservation, University of Hamburg. I especially thank Madame Berthe Rakotosamimanana and Madame Gisèle Randria from the University of Antananarivo for their support. I thank the Commission Tripartite and the Ministère pour la Production Animale et des Eaux et Forêts for their permission to work in Madagascar. I thank QIT Madagascar Minerals and their environmental team, headed by Manon Vincelette and Jean-Baptiste Ramanamanjato, for their help and support. I thank Refaly Ernest, Andry Rajaonson, Georg Schwesinger, and Dimitrij for their assistance during field data collection and Andreas Hapke for his ideas and previous experiments in capturing of Cheirogaleidae in Mandena. Finally I thank Jörg Ganzhorn for his support and comments. The DAAD (German Academic Exchange Service) funded the study.

Copyright information

© Springer Science+Business Media, LLC 2007