Since there isn't any kind of SimYukkuri FAQ (or at least, I didn't find any), I have to figure everything by myself using test runs.
I feel like some kind of a mad mad scientist.
Just now, I studied the behaviour of predatory types Yukkuris.
To this use, I equiped my yukkurium with two beds, one toilet, and one plate of industrial yukkuri food.
I then put one Reimu, one Patchouli and one Remilia inside.
A clear difference immediately appeared between the first two and the Remilia. Reimu and Patchouli did rub-rub with each other, and quickly refreshed, making the Patchouli pregnant. However, they had absolutely no interaction with the Remilia. Whenever the Remilia happened to hover near them, flying on its little wings, they definately couldn't take it easy, and ran away screaming. To quote :
>"Id'sh a Remiliaaaaa !"
>"Mukyuu, a Remilia !!"
However, the Remilia didn't seem saddened by the situation, and kept smiling, taking it easy.
They had different feeding habits. Reimus and Patchoulis need to eat very often. Patchoulis seem to be very distressed by the presence of "mister poo-poo" near the feeding plate. They refuse to approach it, despite the hunger. Remilias, however, don't eat to eat as often, but being predatory yukkuris, they tend to favor another kind of food.
The Remilia snatched tha babies from Patchouli's stalk, and ate them : one baby fed her for a day. Once two of the three babies were eaten, the mother Patchouli spontaneously died, for reasons unknown : she wasn't attacked, nor was she hungry. This reveals the hunter instinct of the Remilia : by feeding on the babies, and not the mother, she preserves her source of food as long as possible, a necessity in the harsh conditions of the mountains where they originate.
Remilia quickly attacked, killed and ate Reimu the following day. What happened next is interesting. Remila was left with the following elements inside her yukkurium : the corpses of two adult and one unborn baby yukkuri, a broken stalk, a full plate of food, and a few portions of yukkuri feces.
The carnivorous nature of the Remilia had let the observers to think that she would favor the dead bodies, but instead, the Remilia went for the industrial food, not seeming distressed at al by this change of alimentation, going "Uu~!" and taking it easy as usual. This led to the hypothesis of Remilias being unable to consume the lifeless, decaying corpses of other yukkuris.
In order to challenge this hypothesis, the bowl of food was taken away. It instantly fell apart when the Remilia ate Reimu's corpse for her next meal, and then followed with the two others. This established that Remilias can indeed act as scavengers, but prefer fresh preys and processed food to dead bean paste.
The final phase of the experiment began when a Flan was introduced in the yukkurium. At this point, no prey or corpse remained : only a broken stalk, several poritons of "mister poo-poo" and Remilia.
It was expected that, due to their status as predators, both Yukkuris would turn on each other, fight to the death, allowing the winner to feed on the body of the loser. Nothing of that sort happened.
The two yukkuris barely interacted : there was a silent respect between them. They didn't rub-rub or try to refresh. They slept separately, and had nothing to do expect flying around and eating. They first ate the stalk : Flan immediatly went for it, but Remilia only touched it when her level of hunger began endangering her life. The stalk lasted them three days.
When Flan ate the last bit, the observers thought they would then begin to fight. They didn't, and turned to coprophagia to sustain themselves, or, to put in yukkuri terms, mister poo-poo became mister food.
At this time, taking it easy was impossible. Their stress level steadily rose, and they always looked sad and defeated, weakened by the fact that they only brought themselves to eat their own excretions when hunger began to threaten their health.
Some hypothesis were made regarding what would happen when the excretions would have been entirely consumed, because they ate a lot more of it than they produced. They could not be tested, because Remilia died of hunger while crawling from mister bed to mister toilet, not to defecate, but to feed.
Only then did Flandre change her alimentary habits : instead of her usual serving of mister poo-poo, she ate the dead Remilia. The corpse lasted her two entire days. It's fascinating how the two pedators couldn't imagine eating each other as long as they were both alive : only death could break this respect between hunters.
The experiment was then terminated.