162 



P.D. JENKINS AND S.M. GOODMAN 



Fig. 6 Exact position of trap in the Vohibasia Forest that captured the 

 holotype specimen of Microgale nasoloi. The trap was placed about 1 .5 

 m off the ground and the trap opening was facing the direction of the 

 canopy and it is most likely that the animal was descending the vine 

 when captured. Note the thick woody understorey of the forest. 

 (Photograph by S. M. Goodman). 



Vohibasia), these climatic shifts involved a mesic period starting 

 before 5000 years Before Present (BP) and an arid period between 

 3500 and 2500 years BP (Burney, 1993). These proposed shifts are 

 mirrored in changes of species representation and habitat types of 

 subfossils excavated from sites in southwestern Madagascar 

 (Goodman & Rakotozafy, 1997) including Ampoza (Goodman, in 

 press). Radiocarbon dates available from Ampoza include an AMS 

 date of 1350 ± 60 BP from a bone of Hypogeomys antimena A. 

 Grandidier, 1869, an endemic large rodent that no longer occurs in 

 the region (Goodman & Rakotondravony, 1996). Further, bone 

 remains of extinct giant tortoises from the site have been dated to 

 1910 ± 120 BP (Mahe & Sordat, 1972) and 2035 ± 35 BP (Burleigh 

 & Arnold, 1986). Although these radiocarbon dates are more recent 

 than Burney's proposed period of aridification, the important point 

 for this discussion is that over the past few millennia there has been 

 significant change in the environment of the Vohibasia and 

 Analavelona region as reflected by the fauna. 



Over the past few years a number of studies have tried to correlate 

 aspects of the speciation of certain Malagasy vertebrates with 



vicariant events derived from information on shifts in vegetational 



Fig. 7 Photograph of the live individual of the holotype of Microgale 

 nasoloi (FMNH 156187). (Photograph by J. Durbin). 



communities during the Quaternary. These paleoecological extrapo- 

 lations are derived almost exclusively from palynological data 

 dating from the Holocene. In many cases several of the hypotheses 

 advanced seem to explain patterns of the distribution of certain taxa, 

 particularly those living in montane zones of the east (Carleton & 

 Goodman, 1996, 1998). A similar argument in the case of Microgale 

 nasoloi may be formulated as follows: during the recent geological 

 past when the region was more mesic, the distinctly more humid 

 forest currently restricted to the upper reaches of the Analavelona 

 Massif would have been more extensive, consequently, M. nasoloi 

 would have had a broader distribution. As the climate became drier 

 and the humid forest retreated towards the summital area of 

 Analavelona, the distribution of this animal also contracted, leaving 

 remnant populations at sites with suitable habitat to support it, such 

 as the Vohibasia Forest. 



For M. nasoloi there appears to be a conflict between aspects of 

 morphological adaptations, namely a species adapted to semi-xeric 

 conditions and the above scenario associated with a more mesic 

 Holocene in southwestern Madagascar. Given these adaptations it is 

 possible that the opposite sequence took place - as more mesic forest 

 dominated the landscape this species was pushed into drier areas of 

 the southwestern Madagascar, and only after becoming more arid 

 was it able to colonize or recolonize this region. On the basis of very 

 limited information it appears that this species is forest-dwelling and 



