I originally published this on January 31, 2007.
If you really read this blog "for the articles", especially the chronobiology articles, you are aware that the light-dark cycle is the most powerful environmental cue entraining circadian clocks. But it is not the only one. Clocks can also be entrained by a host of other ("non-photic") cues, e.g., scheduled meals, scheduled exercise, daily dose of melatonin, etc.
Clocks in heterothermic ("cold-blooded") animals can also entrain to temperature cycles. Lizards can entrain to temperature cycles (pdf) in which the difference between nightime low and daytime high temperatures is as small as 2 degrees Celsius. When taken out of a warm-blooded animal, the SCN clock can also be entrained (if you are a regular here, you recognize the name, don't you) by temperature cycles (presumably a nice feedback loop that stabilizes the mammalian rhythms: the clock entrains body temperature cycles and body temperature cycles entrain the clock).
Some rodents can phase-shift (and thus presumably entrain if presented daily) their clocks under the influence of conspecifics odors or pheromones. In an old study (which was not very good, but enough can be concluded from the data), rats held in groups in constant conditions entrained their rhythms to each other (while the quail did not), suggesting some kind of social entrainment, perhaps mediated by smell.
Social animals are supposed to be sensitive to social cues and it is presumed that their clocks can be entrained by social cues as well. It is also widely believed that no other animal's clock is as sensitive to social cues as the human's.
Everyone who's been in this field has heard the anecdotes about the experiments conducted by Jurgen Asschoff and others at Andechs, Germany in the 1950s and 60s, in which human volunteers were kept in constant light conditions for prolonged periods of time in old underground bunkers (I think Asschoff's bunkers are now preserved as monuments to science, just like the Knut Schmidt-Nielsen's camel chamber is preserved over at Duke University with a nice brass plaque). According to the lore of the field (were those things ever published?), social cues like newspapers, or physical appearance of technicians called in to bring in the food (e.g., sleepy look, or the 5-o-clock stubble) were sufficient cues to entrain human subjects.
It is always difficult to directly test the relative importance of different environmental cues. Sure, one can put them in direct competition by having, for instance, a light-dark cycle and a temperature cycle being 180 degrees out of phase and see to which one of those animals actually entrain (such a study in Neurospora was published a few years back). But, how do you know that the intensities are equivalent? What is the equivalent of 1000 lux in degrees Celsius? Ten, twenty, a hundred?
So, perhaps one should look at the ecologicaly relevant levels of intensity of environmental cues. But how does one dissociate two synchronous cues out in nature in order to do the experiment? Well, of course, use humans for this experiment as the society has already made sure some cues get dissociated! And that is exactly what Till Roenneberg, C. Jairaj Kumar and Martha Merrow did in a new paper in Current Biology: The human circadian clock entrains to sun time (Volume 17, Issue 2 , 23 January 2007, Pages R44-R45)
What they did is take advantage of the fact that time zones are very broad - about 15 angle degrees each. This means that the official (social) midnight and the real (geophysical) midnight coincide only in a very narrow strip running smack through the middle of the time zone. Most of Europe is one time zone. If it is officially midnight in Europe, i.e., the clock strikes 12, it is really midnight (as in "Mid-Night") in a place like Munich, but it is already something like an hour later in Bucharest, and still something like an hour to wait for it in Lisbon.
So, in this paper, they looked at actual entrainment patterns of more than 21000 Germans to see if they entrain to the real midnight - suggesting that light cues are stronger, or to official midnight, suggesting that social cues are stronger. They controlled for age, sex, chronotype (owls/larks) and general culture (former East and West Germanies) and what they found was very interesting: in small cities, towns and villages, people entrain to the light-dark cycles and mostly ignore the official time. However, bigger the city, more independent the entrainment was from the real light-dark cycle. The phase was delayed and more in sync with the official time.
It is hard to interpret the findings, really. Do people in big cities entrain to official time due to stronger social cues (the busy big-city life and social scene) or because they are better sheltered from the natural light-dark cycle and, due to all the light pollution and technology, better able to impose on their clocks an artificial light-dark cycle. I am assuming that untangling this question is going to be their next project.
But, one thing this study did was make us take a more skeptical look at all those Andech bunkers anecdotes. Sure, social cues may work in the absence of all other cues, but they are not THAT powerful and do not seem to be able to overcome the effects of natural light cycles in places in which people are able to perceive a natural light cycle. I guess one can view the life in a big city ("black box") as being in a laboratory experiment in which the society acts as an experimenter, imposing the light-dark cycle on people, while the life out in the country is more like a field experiment in which the human subjects are exposed to the natural environmental cues.