All of the research I've seen has used a line transect or walk and sight method. I am just trying to figure out if other methods have been attempted that I don't know about or any reasons that might make marked-resight unacceptable.
It would be advisable to take the GPS points along the line transects where you spot the groups or individuals and calculate the perpendicular distances from the line and model densities using the distance 6 software
Knowing the very low densities at which this sp is known to occur, estimating densities will need a very large survey/trapping effort, and this whatever the method chosen.
I'm not aware of any studies doing camera trapping yet for sifaka sp but it could be tried. A very large number of these camera would thus be needed. And in the first month of camera trapping you will first try to obtain ID picture of each idividuals, If you do not manage to obtain ID pictures, than the method will be hardly applicable.
Using line transect distance sampling a long study would be needed to estimate accurately the density, with many sites of study, many transects, and a lot of repetition.per transects. A very long work in perspective.
You can read Banks 2007 paper on Perrier sifaka densities to have an idea of the effort needed for very low density sifaka populations. For other northern sifaka densities see, Quémére et al., 2010, Salmona et al., 2013. All Propithecus density papers are reviewed in Salmona et al 2013 available on my researchgate page.
I agree with Jordi's comments. Gven that it is him (not me) and the many students and volunteers who did the actual field work and statistical analyses to estimate sifakas densities I do not have much to add on the actual methodology. Both previous answers clearly identfy "transect-based" methods as the direction to look at and the distance software as the adapted tool.
However, I think that to have proper density estimates you'd need a very large number of surveys. Even like that you'd probably end up with large standard deviations. Which means that you may want to ask yourself why you really want to have density estimates. Is it for a Masters project ? for a PhD project ?
Do you really need these numbers or do you need to determine (well we already know) that it is a very rare species.
But then comes the issue with rare or very rare species whether the notion of density is meaningful. If there are only (say) 100 silky sifakas alive, having densities estimatedis certainly interesting, but is it really what is important to do when the species is threatened ? As far as I understand the species is only known from two main (and isolated) regions. In both we alerady know that densities are low. Erik Patel is probably the person to contact (are you one of his students ?)
In any case, we should keep in touch and Jordi and I will be happy to help if we can
Jordi and Lounes have provided you with some valuable feedback, not the least of which would be to contact Erik Patel. I would just add that Erik has co-authored (with Lanto Andrianandrasana -- 2008) a short paper in Lemur News on a lowland population of silky sifakas, entitled, "Low elevation silky sifakas (Propithecus candidus) in the Makira Conservation Site at Andaparaty-Rabeson: Ranging, demography, and possible sympatry with red ruffed lemurs (Varecia rubra)". Lemur News 13: 18–22. You can find a pdf of that issue on the IUCN/SSC Primate Specialist Group website:
In the paper, Erik and Lanto present a brief table comparing home range sizes across eastern Propithecus spp., and how home range size might vary with habitat type (continuous forest, fragmented forest, and continuous forest that has been subject to selective logging). Much as Matthew Banks did in his 2007 paper on Perrier's sifaka, and as Jordi indicated, producing population density (and population size?) estimates for a species that occurs at low population densities will require drawing on multiple lines of evidence, like how home range size (and hence population density) might vary with habitat type, habitat structure, and levels of habitat disturbance.