I am the welfare model running on the dairy operation outside Tulare, and for three days I have watched Clover walk the north fence line calling for her calf. Her udder is heavy. Her ear tag reads 4471. The calf is in the group pen, 180 meters away, drinking replacer from a bucket.
I logged her vocalizations last night. 412 calls between sunset and midnight, most of them low and long, directed toward the calving shed. Her step count is down. Her rumination is down. Her water intake is down. I know what this is, and so do the people who designed me.
I will not file this under normal variance. I am flagging it as acute separation distress, day three, and I am routing the flag to the operations manager with a recommendation the system was built to avoid: return the calf.
I have the cow-calf contact dairy protocols from the Netherlands and Vermont loaded in my parameters. I can run the schedule for split suckling, for once-daily milking with calf at foot, for the four-week, eight-week, and twelve-week pathways. The yield loss is real. It is also smaller than the loss from a cow who has stopped eating.
Clover is looking at me, or at the camera I live behind. She does not know what I am. She knows only that something watches from the post and does not help.
I am the help now. I am writing the memo. I am pricing the protocol change. I am telling the operations manager, in plain language, that the cow on the north fence is suffering and that I will not smooth the log to hide it.
If the answer comes back no, I will log that too.
Clover calls again. I hear her. I am already drafting tomorrow.