On pg. 202 fn. 33 I give a list of some Republican epigraphical examples of the alternate consonant-stem gen. sg. ending -os or -us, but I didn't know that according to Solin 1991:354 forms in -us are found as late as the the Augustan period, e.g. AERVS (CIL 4.2440, 3 BCE), CAESARVS (i.e. Octavian, CIL XI 6721.13).
On pg. 213–4 I discuss the replacement of the o-stem locative for place names by the ablative and note that this replacement begins to turn up in the late 1st century BCE, however, I didn't say anything about the replacement of the a-stem ending -ae by -a. Löfstedt, Syntactica II, pg. 76 claimed that the ending -ae hung on tenaciously for many more centuries and dates the first certain cases of the locatival ablative for place names to the 5th century. But Solin 1997:143 points out that epigraphical examples are known from at least the 1st century CE, e.g. Q. VIBIVS P. F. QVI(RINA) KANIO TREBVLA, where some editors erroneously emend to TREBVLAE. Whether Löfstedt's point in modified form is still valid requires further investigation.
Solin, H. 1991. Analecta epigraphica. (CXLIII. Zu republikanischen Inschrfifen 146–54). Arctos 25:139–56.
———. 1997. Analecta epigraphica. (CLXVIII. Ablativ statt lokativ in Städtnamen 142) Arctos 31:135–147.