I can get a little jealous of people who have research systems in their labs, or do fieldwork nearby. You can just run experiments year-round if you want. A manuscript needs a more data for the revision? Go ahead and knock that experiment out. If you want flexibility when you get to do research, then having research right at home works quite well.
Then, why is it that some of the most successful researchers that I know have research systems that are geographically far away from the university? And the people — at teaching institutions — with the most tractable, easy-to-use systems can have trouble getting stuff done? (I think there’s a whole other set of problems with model systems on small campuses, but that’s a whole other diatribe post.)
Being far away from your research system can be a recipe for success. Among people I’ve known, a marine ecologist might have to drive eight hours to the rocky coast. Some physicists have collaborative projects at big national or international labs on the far coast of the US, Europe, and Japan. Anthropologists have sites in Southeast Asia and Central and South America. Humanities researchers rely on archives that are in libraries in distant cities. Others might study ephemeral events that occur locally, with no control over the timing of the events.
There are also successful people who work locally, too. Regardless, it is very clear that having your research system on the other side of the world doesn’t preclude success, even if you’re based in a small pond. That strikes me as counterintuitive.
In my own circumstance, I think having all of my fieldwork based out of Costa Rica has been a great boon for my productivity. If I was able to do research in the local mountains or desert, I don’t think I’d really get anything done. I’d never compartmentalize the time that it takes to fully focus on the work.
I’ll consider this with a social insect analogy.
Some of the most “advanced” social insect societies (as some call them) have workers that demonstrate temporal polyethism. That is: workers are born as nurses, then are promoted to guard duty or nest maintenance, and then they spend the last phase of their lives doing the most risky task, foraging. It’s well described in a variety of species.
This temporal division of labor makes for higher productivity, as a result of higher efficiency and organization of labor. (This is at least true in large colonies with a lot going on. The jury is still out on species with small colonies.) A big ant colony would be in disarray if all individuals tried to do everything at the same time. And so would I.
If I tried to run a field research program while doing every other part of my job, I doubt I’d be able to get high quality fieldwork done. I’ve figured out, in a clearly suboptimal fashion, how to juggle writing, teaching, analysis, mentoring during the year, service, and all that stuff. I can’t imagine adding “data collection” to that list of things to juggle during the academic year.
(And, of course, my greatest responsibility and source of joy is being a parent. But this isn’t a Daddy blog, even though I wish such a genre existed. Even though I spend my time writing here about research, don’t be mistaken. I’ve already established that parenting and spousal duties are more important than everything else.)
When I finish a field experiment, it’s over. One project might build upon the other, but I work with discrete ending points, and that’s when I pull the flags from our field sites and pack things to go home. I’ve hired people to do things in my absence for bigger projects, but for most work, I don’t have the option of just returning to do more. If an editor or reviewer asks for another sample, you know what? They’re out of luck, and I’m out of luck. They can buy their own plane ticket to Costa Rica to get that additional data point, if they don’t want to publish the paper without it.
My field site.
This finality of data collection helps me to get stuff done. I have no doubt when I need to start analyzing and writing the manuscript. It’s as soon as I leave the country.
I never think to myself, “Here is a little something which is missing from this project to make it complete.” Instead, I tell myself, “I have to package this as a complete project, and accept the fact that there are some missing holes.”
There’s another reason that working far away lets me get more work done. When I go to my field station, I’m in 100% data-collection mode. We’re running experiments full time, and I’m usually working my butt off. And I’m working my students’ butts off. There’s no way I could give so much focus to work like that while I’m at home, because I’d have to get home and cook dinner, and I’d choose to hang out with my kid at times. When I’m in the field, my responsibility to home is an evening video chat date, which is sometimes missed on one side or the other.
There’s also no way that I would be able to get so much dedication and effort from the students in my lab, without taking them to a kind-of-remote rainforest. When you plop people down in a place where there is nothing to do but fieldwork and labwork, and that’s mostly what you get. (If you bring the right people. I’m getting better at that over the years, but there are always flukes. Flukes, you know, are a kind of parasite.)
I’d guess that work happens by students on site about 12 hours per day, in one form or another. You don’t get that kind of consistent work at that level for an extended period at home. (I lament that the internet has gotten faster on station, because those with an internet addiction have a hard time fully dedicating themselves to their work.) So, at the end of a field season, we have a relative ton of data, much more than I’d have than if I tried to work locally or in the lab.
Some lab work does happen during the academic year, mostly dealing with samples that we collected during the summertime. However, we reserve the academic year for writing manuscripts and preparing for the next field season. Data only gets collected in intermittent bursts, and that has been more than enough for my lab. The fact that I can’t collect data except when I fly to Costa Rica forces me to spend my time writing up the results. That gives me a lot of time to write without any other research-related distraction.
If I block away time during the academic year, it’s usually not to do lab work, it’s only to analyze and to write. When I do research while abroad, it’s only to collect data, and not to write. This temporal polyethism is what allows me to get stuff done.
Small Pond Science 
I have been following your blog for a while now, and it has quickly become one of my favorite science blogs. I can relate very strongly to this post, since most of my research has been done in the Amazon floodplains, alternating between 3,000 and 10,000 km from my main workplace throughout my career. I agree with most of your points, especially with being into “100% data collection mode” when you’re in the field. But I still often dream about doing research in a more accessible region, even though a good part of my work is based on remote sensing.
My main reasons are:
1) Cost: a 10-day field mission in the Amazon usually involves hiring a large boat and a couple of fast boats, with their respective pilots and crew, and enough fuel to travel hundreds of kilometers. Then there’s airfare and meals for all researchers involved, plus the outrageous overweight fees to bring all the research equipment. When you multiply this by two (at least) or four (ideally), because you need to account for the stark seasonal differences in your sampling, it becomes all but impossible to do it without having the largest grants (in Brazilian terms).
2) Immersion: I think the ability to experience your study environment on a frequent basis is invaluable. Imagine if you could take a walk through the rainforest every weekend, how many unique/rare events or interactions would you see? (this goes back to the discussion on natural history that sprung on Dynamic Ecology). Or if you could follow the subtleties of the daily and seasonal changes and how your system responds to them, and how they vary between years? Or if you could just be there when an unusual climatic/environmental event happened? How do you think your understanding of the system would change? By visiting our field sites once a year or less, and usually at the same time of the year, we only get snapshots of what is going on, and I constantly think back on “what I’m missing” by not being there.
3) Practicality: This is a corollary from the other two reasons. There are some studies/experiments that you just can’t do without being there constantly (most of the manipulative studies and anything that requires monitoring); and you can also miss the opportunity to be in the field when something unexpected happens (e.g. the extreme droughts and floods in the Amazon in 2010 and 2009, respectively). You need this sort of access if you want to understand how the system responds to disturbance.
In my relatively short time as a scientist, I have worked in multiple systems spanning from humans where we used mice models, to where I am now working with amphibians. All of my current research is concerned with the effects of (mostly) anthropogenic factors on the current amphibian declines. Because I am focused specifically on the tadpole stage, I have discrete field seasons. My research begins in March when my species-of-interest breeds, and ends when my tadpoles morph out. We do not keep amphibs over the winter in the lab and do not have a breeding colony, so we’re limited by season. Since our field station is relatively close (~45 minute drive), we are temporally, not spatially constrained. But it is a similar issue. During those spring/summer months, I am usually in the field. I get work (MS writing, blogging, etc.) accomplished at nights, so I’m not 100% on, 100% off, but my priorities do shift.
The primary difference I recognize with spatially or temporally constrained systems versus “year-round” labs is the “if you can do experiments whenever you want, why are you not more productive” issue. I worked in a pulmonary lab for a year prior to coming to grad school, and in that time I was constantly running experiments. Revisions came back on a paper? Run the experiment. Have a neat idea? Run the experiment. It was really nice to have the flexibility and I know my PI was very happy. However, the lab wasn’t just pushing out paper after paper, even though we were constantly working. To this day, I’m still not sure what the hold up was.
In my current system, I run as many experiments per season as I can (while still having some decent quality control). If one fails, that’s it for that season and I have to wait another year to try again (sounds oddly like the NSF pre-proposal system). As a grad student, that’s not really an option, hence the multiple projects per year. And while I recognize the general shortcomings of this approach, I still prefer it to year-round experimental availability. Each fall I lock myself in my office and write, whether it’s grants, manuscripts, or teaching materials. My productivity always wains in the spring, but it’s expected as my time is allocated to running, not writing about, experiments. So while my system is a little different (full-on vs. full-off, temporal vs. spatial separation), I can relate (at some level at least).
Cool thoughts, guys. The longest I’ve ever lived at my field site continuously is less than six months. I have a feel for how things are down there throughout the year, but during the North American fall, I’ve never been down there after all these years. Clearly, I’m missing out. I have hired technicians who have done year-round sampling for me over multiple years, but that’s not the same and I recognize that I am losing some insights.
The “snapshot effect” in rainforest biology is probably a huge problem. In the temperate zone, people build in seasonality and interannual variation into the experiment, either by design or by constraining interpretations. There are big seasonal variations, and stochastic variation, in tropical rainforests that are often overlooked. It took me years to appreciate exactly how different an experiment in January is from an experiment in June, and also recognize that there’s even more that I’m missing.
But, the bottom line for me is that even if my system was local – or a mesocosm in my own lab – i still wouldn’t be immersed because my mind and body would be elsewhere. In the classroom, on the computer, at conferences. I actually think that, given my own personal situation, I am able to spend more time in the field because it’s on a different continent than if it was my own backyard.
The finality of our (local) field surveys is encouraged by November 15th being the start of deer gun hunting season. November 14th is always our last day of field work for the year. :)
Yes, that. Yikes.
And, I suppose for people who live in places that still freeze over in the winter, lots of things can only be done in a certain seasonal window. Fieldwork in the desert for many taxa would reasonably be constrained to a relatively brief monsoon, too.