Running a MANOVA for habitat selection with missing data in R

Question

I am currently preparing data to conduct a second order and third order habitat selection analysis. For the second-order (use/available), I am looking at the mean distance of random locations to each habitat type within the population range, and random points within the home range of each individual. There are a disproportionate amount of habitat types in the population range versus the home range. Therefore, I have a large quantity of NA's within a few habitat types within the "used" home ranges. I have 10 habitat types and 22 individuals. Such as: Avail: Agriculture9.8, 7.5, 8.8, Use: Agric NA, 0.8, NA. I am attempting to model it off a few other reptile papers... but I have not been able to find any detailed explanation on how researchers are able to conduct this MANOVA with missing values. I would rather avoid removing the data as I think it is important to keep those types of habitats available.. but I am unsure of which direction to turn.... Any help would be appreciated