Grid sampling should work. You should form grids over the surface of interest and slected grids at random. If there are a total of N grids and you pick n grids to actively search for the snake, then the estimator of the population total would be N*t/n where the t is the number of snakes observed at the n grids. The problem with this approach is that it is very hard to foresee the precision in the estimator unless you have some prior understanding of how many snakes to expect per grid. As a consequence, it is difficult to determine the number of grids (n) to select.

But I cannot offer a better method than grid sampling as capture-recapture is unlikely to work in this case.

Actually, an alternative came to my mind. Do the snake generally dwell in nests or rat holes? Is there a simple way to determine the total number of nests and rat holes? If yes to these two questions, the following idea might work. Select nests/rat holes at random and check the proportion of times a snake is found inside it. Multiplying the estimated proportion with total number of holes/nests would provide another estimator.

I do not know if this procedure would be simpler in the field, but just a thought...

following link may be useful to your study ....

http://pubs.sciepub.com/ajzr/2/3/2/ 

http://www.mes.org.mk/PDFs/3rd%20Congress%20Proceedings/03_Metodija%20Velevski%20&%20Bratislav%20Grubac.pdf