In what ways can an evolutionary trend develop during the history of a genus or a family?

I don't think that most changes happen at the genus or family level - we only have those categories in retrospect and I think you're looking for something predictive. Evolution is typically thought of as happening at the population level (subspecies level) so it may not have an impact at all on a genus or family (and if it did we'd only recognize it thousands of years later).