This is a follow up to the creature 'griffins' I sort of created here: Anatomically Correct Griffins
The idea of the final griffin was a small tree based species that will leap down on much larger prey and kill it quickly with attack to jugular. Hunting is very dangerous because griffins are frail compared to their prey and because there are land based predators that make any time spent on land dangerous.
The are very selective in their kills. Their 'pack' spreads out over the trees looking for the perfect kill. When good prey is found the spotters will alert the other's and try to flush their chosen prey towards a defined 'kill zone' by positioning viably positioning themselves for a kill. If they prey doesn't respond the spotters will attack it once in the prefect position, if the prey attempts to flee so the spotter can't attack the spotters maneuver it to the kill zone; a location where their best hunters are hidden just right to be able to make the best/safest leap to have the highest chance of a kill at low risk. Other griffins will converge on the prey to keep driving it towards the killzone once called.
A killzone probably can't be a single predefined location, the griffins would be spread out over a large area and likely need to move around the killzone to be a location close to the prey rather then driving the prey long distance to a single chosen killzone.
There are two problem with this. The first is that the prey will eventually develop some habitual understanding of the griffins hunting style and attempt to avoid being herded to a killzone. If they prey can anticipate where the griffins want it it will attempt to avoid that, even if it means putting itself at risk of being attacked by a spotter.
Second, there are other predators, ones that hunt/kill griffins and scavengers who would be happy to drive off the much smaller griffins to take their kill.
Thus the Griffins need a way to communicate over a distance with their tribe, to communicate where prey are found, decide on a kill zone to drive it to, and work together to flush the prey to the right spot. Yet they need to be able to do it without the prey or other predators learning/evolving to recognizing the calls well enough to develop enough of an understanding of their communication to know where the kill is likely to happen.
I'm open to any viable communication option. I'll mention one example I like, but am not committed to, because it places evolutionary pressures to drive the Griffins towards sapience. That's the idea of griffins communicating via vocal calls, but with some evolved obfuscation that keeps predators or prey from developing a habitual understanding of it; ie it's far more complex the a few predefined food calls similar species use.
In this example when griffins are hunting, or spread apart for any reason, they always keep up a constant 'chatter' across the whole tribe, so common and spread across such a large hunting region as to not be useful to predator's or prey, becoming mere background noise. When they find prey they communicate it with this 'chatter', but without raising the volume or localizing the chatter, and with complex enough language/syntax that other species are unable to tell that they have transitioned from griffin equivalent of 'small talk' to hunt coordination.
This idea would work once evolved, but can it evolve? Can proto-griffins that are just developing the tactics of coordinated hunting and driving of prey species, that haven't reached the level of sapience that this tribe hunting technique will drive them towards, be expected to develop not just long distance communication but some obfuscation technique like this? With there being almost as much evolutionary pressure on the prey to learn the griffins language as there is on the griffins how do the griffins evolve understanding of this language without their prey doing the same?
I'm open to any communication system that can't be learned/exploited by other species, not just the above example, but the main question is the same, not only can one exist but how can it realistically evolve?