You can't do anything unless you know how they are getting in. If Ning had any kind of reasonable logging, you could try and map the originating IP and block it. My understanding is Ning logging is shit. Are there any settings in captcha that can be tweaked to make it more complex ?
considering what appears to be a growing volume of these fake members and posts, maybe we need a "report spam" button (or is there one?). Of course, then there would have to be someone at the other end to do something about the reported spam.
We are currently removing 40 to 50 spammers a day. And for the record, Ning has implemented several "fixes" to keep them out. These work for a while and then the spammers find a way around it.
One option we have is to require approval before someone becomes a member. I have been resisting this because I want to allow nonbelievers to join and get active quickly. We could try it and see what happens.