ip subnet zero tells the router that it is OK to use the ip subnet zero.
when a network address is subnetted, the first subnet created after subnetting the network address is called Subnet Zero.
(all the subnet bits of the address when written in binary form are all zeros)
it used to be that the subnet of all zeros and the subnet of all ones were recommended not be used for addressing. (see RFC 950) http://www.ietf.org/rfc/rfc950.txt
"It is useful to preserve and extend the interpretation of these special addresses in subnetted networks. This means the values of all zeros and all ones in the subnet field should not be assigned to actual (physical) subnets." (taken from RFC)
back in the day, alot of systems used the subnetZero address as a broadcast for that subnet domain. nowadays, there are not many systems that use subnet zero for specific functionality and therefore, it is common now to go ahead and use subnet zero.
the "ip subnet zero" command tells the router that using subnet zero of any network address is acceptable.
see this link for more info:
http://www.cisco.com/en/US/tech/tk648/tk361/technologies_tech_note09186a0080093f18.shtml
