I'm investigating occupancy patterns in an amphibians' community in traditional wood-pastures. My sample unit is 'pond' and I must visit each pond at least twice in order to adequately calculate detection probability. It was clear from the beggining that most ponds are seasonal, which means that they dry out at the end of spring/beggining of summer. This would be ok, because I could have enough time to at least have two visits per pond. However, although I am still in the first visit round, I have already noticed how some water bodies have dried out, and fear that some of my sampled ponds in the first round, will be dry when I come back for the second visit. In the literature I've found, authors just skip the data from those ponds. Others just sample the nearest not dried out pond. Is there other way to deal with this? Many thanks in advanced!