, Volume 49, Issue 1, pp 85–88

Pilot survey of avahi population (woolly lemurs) in littoral forest fragments of southeast Madagascar


    • Museo di Storia Naturale e del Territorio, Centro InterdipartimentaleUniversità di Pisa
Short communication

DOI: 10.1007/s10329-007-0061-2

Cite this article as:
Norscia, I. Primates (2008) 49: 85. doi:10.1007/s10329-007-0061-2


This study presents a pilot survey of the avahis (pair-living, nocturnal prosimians) living in threatened littoral forest fragments of southeastern Madagascar. In the period of October–December 2004, I evaluated the density of adult and newborn avahis (carried by the mothers) by counting the individuals encountered during 58 night walks in seven fragments of Sainte Luce and Mandena forests, along one trail/fragment. Along each trail, I used random plots (5 m2) for a preliminary characterization of the vegetation. The density of the population was not correlated with fragment size and number of plant morphospecies, while it was correlated with large tree availability. Possibly due to a low energy diet based on leaves and to specialized and energetically expensive vertical leaping, the loss of large trees by selective logging seems to affect avahi populations more than other variables.


Avahi lanigerAvahi meridionalisDegradationFragmentationLittoral forestMadagascar


Habitat fragmentation, which threatens the survival of lemurs in Malagasy forests (Mittermeier et al. 2006), is only one step of the degradation process, also including habitat loss, edge effects, land use in the surrounding areas, hunting, and harvesting (Tabarelli and Gascon 2005). Population estimates are unavailable for many lemur species, and effective conservation plans cannot be implemented if the effects of habitat reduction on lemurs are not clarified (Mittermeier et al. 2006). Relatively few studies have focused on single species in non-protected fragments of Malagasy littoral forest (cf. Lawes et al. 2000; Ganzhorn et al. 2003).

This study focuses on the populations of nocturnal woolly lemurs (avahi or fotsyfe, Indriid family) of littoral forest fragments of southeastern Madagascar (forest of Mandena and Sainte Luce), recently recognized as Avahi meridionalis (instead of A. laniger; Zaramody et al. 2006). Such populations are worth studying for different reasons. First, they are confined to littoral forests (Bollen and Donati 2006), one of the most threatened ecosystems of Madagascar (with less than 3,000 ha remaining). Secondly, avahis are vertical leapers, and vegetation destruction can create a barrier to migrations from one local population to another, reducing the possibilities that animals persist in metapopulations (Lawes et al. 2000; Thalmann 2003). Third, the species living in monogamous pairs like avahis (Thalmann 2003) may be more prone to “demographic” extinction than the species living in extended groups with promiscuous mating (Dobson and Lyles 1989). Lastly, the removal of trees can lower leaf production and affect the survival of small-bodied folivores like avahis (Thalmann 2003). In Mandena and Sainte Luce, forest clearance is exercised for plantations, charcoal, cooking, and buildings (Vincelette et al. 2003).

This pilot study presents the first survey of avahis covering seven littoral forest fragments of southeastern Madagascar.

Materials and methods

I collected data in the seven forest fragments (southeast Madagascar; Fig. 1) where avahis were reported to live until 2001. Two fragments (S7 and S17) could be reached only via pirogue/boat. QMM (Qit Madagascar Minerals) provided the logistical support.
Fig. 1

Littoral forest fragments (in black) in the Fort-Dauphin region, southeast Madagascar. The seven fragments included in the study are indicated with a letter and a number (M15, M16, M13 in Mandena: 24°57′S, 46°52′E; S9, S8, S17, S7 in Sainte Luce: 24°46′S, 47°11′E; figure modified from Bollen 2003)

In each fragment, I collected vegetation data by choosing eight vegetation plots of 5 m2 using a random procedure along the census trails. Local botanists identified plant morphospecies for trees with a DBH (diameter at breast height) ≥1 cm, thus excluding seedlings and including saplings (DBH 1–4.9 cm) and medium/large size trees (DBH ≥5 cm) (Ganzhorn et al. 1999). The DBH, correlated with crown diameter, is an indicator of leaf availability (Ganzhorn 1995). Via plot vegetation data, I estimated the number of plant morphospecies and of trees with DBH ≥3.2 cm for each fragment (exploited by avahis in Sainte Luce for the 97.5% of the time; Norscia, unpublished data).

In order to estimate the density of adult and newborn avahis, I used the line transect method. The trails (one per fragment, 1 km each) were located 100–200 m from the forest edge to reduce the bias related to trail location. Aided by two field assistants, I performed 58 night walks (ten walks in S9 and eight walks in the other fragments) at a speed of about 1 km/h, from October to December 2004 (weaning season). I collected 161 sightings, from 20:00 to 22:00 h. I evaluated the abundance of adult and newborn avahis (carried by the mothers) by counting the individuals encountered.

I used “Distance 5.0” to calculate the density (D) of individuals per fragment ([D = n/(2Lw); n number of animals encountered; L trail length; and w effective strip width]. Then, I calculated the birth rate as the number of newborns divided by the number of adults. Finally, due to the small sample size (n = 7) I used the exact Spearman test to verify correlations involving patch areas, the number of plant morphospecies, the number of trees with DBH ≥3.2 cm, adult density, and birth rate (STATXACT 6.0, Cytel Studio).


No significant correlation exists between fragment size and the number of trees with DBH ≥3.2 cm (exact Spearman, n = 7, r = 0.64, ns), between fragment size and plant species richness (exact Spearman, n = 7, r = 0.67, ns), or between the number of plant species and number of trees with DBH ≥3.2 cm (exact Spearman, n = 7, r = 0.58, ns). As can be noted in Table 1, I detected no avahi in M13 during the census.
Table 1

Adult density per hectare (AD/ha), birth rate (number of newborns over number of adults) per hectare (BR/ha), patch area in 2004, number of trees with DBH ≥3.2 cm, and number of plant species referring to the seven fragments of the forests of Sainte Luce (24°46′S, 47°11′E) and Mandena (24°57′S, 46°52′E) considered in the study


AD (ha−1)


Area (ha−1)

DBH ≥3.2 cm

Plant species











































Fragment areas J.B. Ramanamanjato, personal communication

As shown in Table 2, a positive correlation was found between adult density (per ha) and number of trees with DBH ≥3.2 cm. Adult density did not, however, correlate either with fragment size or with the number of morphospecies. The birth rate (per ha) did not correlate either with the number of plant morphospecies or the number of trees with DBH ≥3.2 cm. There was a significant and positive correlation between birth rate and fragment size, but it was not significant in the analysis using the Bonferroni technique.
Table 2

Results of the exact Spearman correlation for each couple of variables



Spearman r


Estimated density of adults/ha and fragment size




Estimated density of adults/ha and % of trees with DBH ≥3.2 cm



P < 0.013*

Estimated density of adults/ha and plant species richness %




Birth rate/fragment and fragment size



P < 0.05

Birth rate/fragment and % of trees with DBH ≥3.2 cm




Birth rate/fragment and plant species richness %




Due to repeated correlation involving the same variables, the significance is fixed at P < 0.013 (α/4) following Bonferroni technique (Rice 1989)


Avahis are specialized for folivory and for vertical climbing and leaping (Thalmann 2003). Thus, the correlation between the density of adult avahis and the number of trees with DBH ≥3.2 cm (Table 2) may suggest that (1) large trees are important for movement between vertical supports and (2) low values of DBH, associated to low leaf availability, affect folivores (Ganzhorn 1995). The absence of correlation between avahi density and the number of plant morphospecies may suggest that selective logging, mainly addressing large size trees, has not yet affected the fragments in terms of plant species richness.

However, the lack of correlation between adult density and fragment size may be due to different reasons. First, higher animal densities in smaller habitats can be the outcome of an artificial concentration in refugia. Second, folivores can be less affected by small fragment size than frugivores because they require smaller home ranges (considering leaves are more homogenously distributed in the forest than fruits; Chiarello and Melo 2001; Norscia et al. 2006). Third, avahis may be adapted to a certain degree of fragmentation, considering littoral forests on the coastal plain are naturally patchy (Ganzhorn et al. 2003). Finally, avahis could tolerate forest edges, which can possibly have a positive effect on leaf chemistry (cf. Ganzhorn 1995; Lehman et al. 2006). However, the correlation trend between fragment size and avahis’ birth rate (Table 2) may suggest that adult avahis will be influenced by size reduction in the future.

Overall, avahis’ populations seem to be affected by the number of big size trees and not by the number of plant morphospecies and fragment size. The fact that these three variables did not correlate one to another may suggest that fragmentation (patch size reduction) and other aspects of degradation (deprivation of both big size trees and plant morphospecies) do not proceed in tandem in the littoral forest. Indeed, the logging pressure was lower in the patches falling in the QMM conservation area (M15, M16, and S9; Fig. 1). In M13 avahi’s population had apparently collapsed by 2004 (Table 1; only one individual was seen the same year; S. Rasarimanana, personal communication). However, in 2001 the density (0.6 ± 0.5 ha−1) was comparable to that recorded in M15/M16 (0.5 ± 0.4 ha−1) (Andrianjazalahatra 2002). It is likely that charcoal makers (absent in Sainte Luce) had selectively cut the big size trees by 2004, leaving the fragment outline unchanged. The low densities in M15 and M16 (Table 1) could be related to competition with Hapalemur meridionalis (a folivore absent in Sainte Luce) and predation by Cryptoprocta ferox (the fosa) that reached Mandena (but not Sainte Luce) in 2004. Finally, the low density in S7 may be due to the high pressure of hunting and logging.

Due to their unique combination of features (nocturnality, folivory, and specialized locomotion) avahis can be protected only in situ (Thalmann 2003). The small home range typical of avahis (2–3 ha in Sainte Luce; Norscia and Borgognini-Tarli 2007) may favour the survival in fragments. However, avahis’ mobility in open spaces is limited by vertical locomotion. Consequently, the avahis living in the small patches of littoral forest (<100 ha) may be at risk, considering that a population of 40 adults could not survive more than 40 years in fragments (Ganzhorn et al. 2003). Thus, this first survey of avahis’ populations in seven fragments of littoral forests can represent a basis for conservation and further investigation.


Thanks to the Malagasy Département des Eaux et Forêt, Parque de Tsimbasasa, Université d’Antananarivo, J. Ganzhorn, S. M. Borgognini-Tarli, J.B. Ramanamanjato, M. Vincelette and the whole QMM (Qit Madagascar Minerals) conservation team, S. Randriamanga, G. Donati, E. Palagi, and the assistants G. Sambo and A. Kadoffe.

Copyright information

© Japan Monkey Centre and Springer 2007