Since the mean and variance are so similar I think that a poisson distribution would be a good fit.

If you know a lot about what you are sampling from then you should be able to figure out the type of distribution without any data. The data then allows you to decide what the parameters of that distribution are equal to.