Animals eating at all would be a major improvement--I've only ever seen grazers eat. Carnivores should consume (most of) their kills, herbivores should eat appropriate wild plants and even farmed crops that they encounter, certain birds should dive for fish, etc. I once had a Giant Armadillo that sat on the exact same tile for over 5 years, because it was never coded to seek & destroy termite colonies (speaking of which, termite mounds should sometimes appear as multi-tile structures).
While the wiki signals issues with CARNIVORE token in 0.34 made creatures such as goblins starve, no such issues are signaled for the current version.
OTOH, the wiki notes that, for the current version, CAN_LEARN is needed for any creature belonging of a civilization to need to eat and drink; similarly, for the CARNIVORE tokens, it signals "If the creature goes on rampages in worldgen, it will often devour the people/animals it kills."
So it seems carnivore eating meat is still being under development.
I can not actually figure out: Why is this not a thing yet? Even in some basic form, even if its just carnivores consuming unrotten corpses and scavangers eating rotting ones. Having absolutely no system in place for creatures eating stuff on-map(it happens in worldgen if a LARGE_PREDAOR or MEGABEAST goes on a rampage) is a pretty big oversight, and it is a oversight, as there are no barriers in place preventing implementation. The only reason i see for not implementing it so far: Not wanting to track the hunger level of every creature on the map. Which is something that does not need to be done, the creatures could just every once in a while "get hungry" and (try to) seek out some food.
Besides having beasts help clean up the corpses would be a nice way for the game improve FPS on its own.
Also this:
Animals eating at all would be a major improvement--I've only ever seen grazers eat. Carnivores should consume (most of) their kills, herbivores should eat appropriate wild plants and even farmed crops that they encounter, certain birds should dive for fish, etc. I once had a Giant Armadillo that sat on the exact same tile for over 5 years, because it was never coded to seek & destroy termite colonies (speaking of which, termite mounds should sometimes appear as multi-tile structures).
And also all of this:
I really hope we get an ”ecology pack” sometime, with non-grazers eating, (partially random) food webs, stomach containers, better animal-made structures/hives/nests/dams, improved animal behaviour (monkeys climbing, am I right?), parasites and pests, and of course plant soil requirements and with it a farming rehaul. Especially since there are more or less enough species to support these improved systems should they be implemented.
stomach containers
For anyone not familiar, the details regarding this can be found in the Stomach capacity, metabolism, and digestion thread, for future reference.
Predation might be difficult to modelize: predators might have to be coded so that they might have to estimate the strength of their prey and compare it to theirs to avoid scenarii where a cat would attempt to take an elephant by himself; we would also have to code for the behaviour of pack of predators.
However, scavengers might easy to code, just sent them to the nearest source of corpses, although we should also account for the amount of food: a battlefield at 20 squares might be more enticing than a mere body at 2 squares.
But I agree, it seems the amount of animals has taken efforts at actually making wildlife
alive.