In a lot of ways this is a response to Meg's post yesterday about system envy in which, she and several commenters talk about the grass-is-greener feeling with other biological systems. Indeed, it's easy to go to talks and see what other people are doing and think "Man, that system has everything I would like to have in the one I'm working on!" One thing Meg points out is that, once you dig in however, every biological system has its warts. Her experience has been that delving into a single system has allowed her to pursue things that she might not even know where interesting if she didn't have the depth in that particular topic.
I have no doubt that is true and Meg's work certainly shows it. Lots of people spend careers learning the nuances of a single system and have produced some really elegant work. But let me propose a slightly different trajectory to trainees considering the direction their career might go.
Don't sweat the system, sweat the question.
One can make the argument that knowing the minutia of a system allows you to answer identify testable questions and I won't quibble with that. But, another way to skin the cat is to have a question and try to determine which system will best allow you to answer it. Rather than becoming logistically constrained into asking certain questions that your system of choice can support, this approach allows you to pick the best model for your question.
Best of all, it allows you to demonstrate similarities across independent groups that can tell you fundamental things you can't derive from sticking to a single taxon. As an evolutionary biologist, I am far more fascinated by independent events that result in the same final product because they tell a much broader story about what is important for that trait, once you strip away all the lineage-specific baggage.
This is why I encourage my trainees to take a non-linear career path. If you decide that you want to dedicate your life to ants, fair enough, but a lot of what I have seen as major advances have come from people bringing outside expertise into a novel system and turning things on their head or drawing broadly across life to zero in on core features.
So if you are in the training stages of your career, consider being a system nomad. If you are doing a Ph.D., think about a postdoc in a totally different system. Maybe you like ants, but what can what you learn from them tell you about moles or coral or phytoplankton or snails? How can your work in the field inform a lab-based system, or vice versa? You'll only know if there's a bigger question to tackle if have a broad enough understanding to recognize which systems give you the best opportunity to answer the question, rather than letting your system dictate the questions.