It started with vSphere 4.1 onwards, and VMFS 3.46 was the first version to support VAAI.
1. Acquire on-disk locks
2. Upgrade an optimistic lock to an exclusive/physical lock.
In vSphere 4.0, VMFS-3 used SCSI reservations for establishing this critical section as there was no VAAI support in that release. In vSphere 4.1, on a VAAI-enabled array, VMFS-3 used ATS only for operations (1) and (2) above, and ONLY when disk lock acquisitions were un-contended. VMFS-3 fell back to using SCSI reservations if there was a mid-air collision when acquiring an on-disk lock using ATS.
Above quote has been taken from VMFS Locking Uncovered - VMware vSphere Blog - VMware Blogs
here meaning of fall back to SCSI reservation is, that on an upgraded VMFS-5 datastore, system will try to make use of ATS, but if contention is high, that ATS might be unsuccessful and system will fall back to SCSI Reservation. this same is true for VMFS 3.46 or higher as well. But if the datastore is a freshly created VMFS-5 then ATS only is going to be used with VAAI aware array systems no matter how high contention is.