If you have ever taken an introductory ecology course, you will no doubt remember C.S. Elton’s classic study of the periodic fluctations of lynx and hares. In a textbook* case of Lotka-Volterra predator/prey interaction, Elton found that lynx and hare populations in northern Canada followed an astonishingly regular 9-10 year cycle. Several years of steady increases in both lynx and hare numbers culminate in a dramatic crash and a brief lull, before the pattern repeats.
You may have forgotten however, where the original data came from. Elton did not sit out in the Canadian Arctic with a notebook and a bottle of brandy counting animals. Instead, he reviewed several decades worth of trapping records from the Hudson’s Bay Company (which happened to be his employer at the time) and noticed the pronounced periodicity in the number of lynx furs reported in the company’s annual inventory.
I noticed a striking periodicity in a very different sort of proxy record a few weeks ago when Alex Wild of Myrmecos fame posted a comparison of Google Trends data for the search terms “ants” and “flies”. It appears that web search activity for “ants” hits a consistent annual peak around May and a consistent annual trough around Christmas, at least for the past four years.
When I commented on this on his blog, Alex pointed out that search records for many insects show a similar pattern, with search popularity seeming to peak sometime during the northern hemisphere summer. Naturally I spent the rest of the night searching for biologically significant patterns in Google Trends. Sure enough, interest in certain insect groups appears to show some interesting seasonal trends:
Phenology, the observation of regular seasonal patterns in nature, especially among animals and plants, lies squarely at the roots of natural history. Humans have undoubtedly been tracking these patterns as long as we have relied on the seasonal availabilty of forage and game (and later crops and livestock) to survive. That is to say, forever as far as our species is concerned.
Google Trends allows users to explore a sort of “social phenolgy”–tracking rhythmic fluctations in public interest which, in the case of web searches for natural phenomenon presumably have at least some connection to the natural rhythms themselves.
Needless to say, I am now obsessed with exploring these patterns, searching for interesting patterns in the interest in birds, flowers and vegetables:
There are some other funny correlations out there:
I could do this all day!
* Of course as with any “textbook” example the truth is likely a bit more complex, and a debate about the nature of and mechanisms behind this pattern continues almost a century later. Sunspots, disease, weather, fire, petroleum futures and the popularity of the name “Madison”, have all been proposed as important factors (see Stenseth et al. 1997 and Zhang et al. 2007[open access .pdf] for recent analyses).
