The only way any externally influenced phenotypic change can be passed on is if the phenotype change is due to a genotypic change in the gametes. I know of no way knowledge of a predator would cause a gene change in a germ cell. If you have any links to this phenomena I would be interested.
"The only way any externally influenced phenotypic change can be passed on is if the phenotype change is due to a genotypic change in the gametes."
I don't know of any experiments off the top of my head, but my main point is that there is no reason to rule it out. We _know_ about heritable changes caused by the environment, and we _know_ about organismal changes caused by sensing predators. What about either process prevents the germline change from being genetic or even heritably epigenetic? Such things would explain the ability of species like the orchid to quickly make complex adaptations with the local insects.