Very high GARP traffic after Sophos 550r2 failover

Sophos 550r2 in HA pair. On our downstream (inside) switches, we are noticing very high GARP requests coming from the Sophos' after a Sophos failover (approximately 62,000-84,000 pps). The Sophos serves as the L3 boundary for a /19 RFC1918 l2 network (inside).

Is this a normal amount of GARP traffic? It seems like an excessive amount.