I don't think they measure if plants/bacterium are anesthetized, but in exposing them to things that cause anesthesia in humans, other potentially unrelated effects are noticed.
Many compounds/mechanisms, especially hormones and neurotransmitters, are widely "re-used" across different biologies for completely different things. They're essentially generic semaphores, and the action caused by raising the semaphore can be basically anything. There's a lot of variance in effect even among different instances of human species, often quite unpredictable, contradictory, and profound.
It's sort of like "hey we already have this testosterone thing, it's currently used to call [function A] but we could refactor that to use it to initiate [function B] instead" (testosterone causes growth in many mammals but inhibits growth in lizards, so female lizards are larger than male lizards)
or "hey we already have the genes to make serotonin for gastrointestinal regulation, but it's not used for anything in the brain. The blood-brain-barrier already prevents somatic serotonin from reaching the brain so we could have a completely different function for it in the brain and regulate gut and brain serotonin in isolation of eachother"
or "hey we have this cholesterol thing that we've been using as a signaling hormone ever since we were on version Plant, maybe we could write a factory that modifies the cholesterol we eat and use it to produce new semaphores like estrogen and testosterone to support a more complex messaging system and handle all the new effects rather than overloading the existing semaphore".
Edit: Probably slightly better to think of them as the coefficients for activation functions, but nothing here is meant to be anywhere remotely close to a direct analogy. Taking any of this literally would be a misreading.
Many compounds/mechanisms, especially hormones and neurotransmitters, are widely "re-used" across different biologies for completely different things. They're essentially generic semaphores, and the action caused by raising the semaphore can be basically anything. There's a lot of variance in effect even among different instances of human species, often quite unpredictable, contradictory, and profound.
It's sort of like "hey we already have this testosterone thing, it's currently used to call [function A] but we could refactor that to use it to initiate [function B] instead" (testosterone causes growth in many mammals but inhibits growth in lizards, so female lizards are larger than male lizards)
or "hey we already have the genes to make serotonin for gastrointestinal regulation, but it's not used for anything in the brain. The blood-brain-barrier already prevents somatic serotonin from reaching the brain so we could have a completely different function for it in the brain and regulate gut and brain serotonin in isolation of eachother"
or "hey we have this cholesterol thing that we've been using as a signaling hormone ever since we were on version Plant, maybe we could write a factory that modifies the cholesterol we eat and use it to produce new semaphores like estrogen and testosterone to support a more complex messaging system and handle all the new effects rather than overloading the existing semaphore".
Edit: Probably slightly better to think of them as the coefficients for activation functions, but nothing here is meant to be anywhere remotely close to a direct analogy. Taking any of this literally would be a misreading.