Problem solved! Thanks to helpful advice from Alex Schröder, primary creator of the Oddmuse wiki engine that drives the ZhurnalyWiki, the issue described in ZhurnalyWiki Surge Control is fixed, or so I believe. It wasn't evil SpamBots pretending to be Google's web crawler. In brief, it was a glitch in the interaction between the zhurnaly.com web server (Apache) and the ZhurnalyWiki perl script (Oddmuse) which occurred on certain pages that had spaces or other unconventional characters in their names. As the Oddmuse documentation describes it:

[$ScriptName] is determined automatically by the CGI module and used for all the links within the site. You only need to set this if the autodetection is not working. This happens on some systems but the cause remains unknown. In the mean time, just set this option to the correct value in your config file ...

So I've set $ScriptName to the value 'http://zhurnaly.com/cgi-bin/wiki' and now the links to internal wiki pages no longer accumulate clutter and grow limitlessly. Googlebot and other crawlers no longer see an ever-lengthening variety of URLs. Within a few days, I hope, as their caches clear they will stop hitting the zhurnaly.com server with redundant page requests, and my excessive bandwidth usage will cease. Thanks, Alex!

^z - 2009-11-25