Long-distance plane travellers afflicted by jet-lag become acutely aware of their internal body clocks. The body’s internal circadian clock normally ticks away for about 24 hours, with fine-tuning every day by ambient light at dawn and dusk. It is widely known that day/night cycles in body functions are regulated by the hormone melatonin, produced in the brain by the tiny pineal gland. Many travellers now swallow melatonin pills to combat jet-lag. But it is far less well known that long-lived animals have a second kind of biological clock, also regulated by melatonin, that governs body functions across the year. Countless animal experiments have shown that changes in day-length (the interval between sunrise and sunset) regulate this circannual clock. In many mammals, especially those living in regions with sharply contrasting winters and summers, mating and birth are tightly linked to specific times in the annual cycle. Humans, of course, have no obvious breeding season, but we are nonetheless subject to annual cycles with more subtle effects. Many have tried to explain these cycles as direct responses to environmental conditions, but we also need to consider another possibility: Long-term evolutionary processes may steer observed seasonal patterns of conceptions and births to match average year-round cycles in the environment.
Exploring human birth seasons
Adolphe Quetelet, a Belgian intellectual who made pioneering contributions to astronomy, mathematics, statistics and sociology, was among the first to identify a seasonal pattern in human births. His influence lives on through his Body Mass Index (BMI), still used — with minor tweaks — as a key indicator in human biology. In a treatise on birth and mortality published in 1869, Quetelet discussed not only birth timing over the day/night cycle (see my September 14, 2015 post When Is the Best Time to Give Birth?) but also the annual pattern. His graph showing data for the Netherlands over the 12-year period 1815-1826 showed a main peak in February/March and a marked dip in July. Noting that the seasonal pattern was more marked in villages than in towns, he attributed this to environmental temperature differences. But information from the Southern Hemisphere convinced him that the sun’s position in the sky governed seasonal patterns.
A rising flow of reports of annual variation in human birth rates eventually followed Quetelet’s ground-breaking account. Three papers by biologist Ursula Cowgill, all published in 1966, represent a major milestone. In one of them, a worldwide review of birth records for many different human populations revealed that seasonal variation is virtually universal. But patterns of peaks and troughs differed between geographical regions, and one crucial finding was a general 6-month shift in seasonal patterns between Northern and Southern Hemispheres. Cowgill concluded that annual birth patterns are controlled primarily by local climatic conditions, but also influenced by cultural factors. She suggested that ambient temperature might affect conception rate and noted that urbanization and industrialization disrupted seasonal patterns to varying degrees. Another of her 1966 papers, drawing mainly on parish baptism records for the English city of York for 1538-1812, illustrated this. Prior to 1752, two annual peaks were evident, a main one in February-April (corresponding to Quetelet’s graph for 19th Century Dutch data) and a subsidiary one in September-November. But over the next 60 years annual variation in births was far less marked.
Several later studies in various European countries showed gradual modification of an original seasonal pattern with a major birth peak in spring, albeit starting at different times. A notable 2007 paper by Ramón Cancho-Candela and colleagues analyzed data for more than 33 million births in Spain over the period 1941-2000. For the first 20 years (1941-1960), a two-peak pattern was clearly evident, with peak birth frequency in April and a smaller rise in September. Thereafter, the pattern slowly changed, with peaks becoming less obvious and eventually disappearing after 1990. So gradual loss of a seasonal birth pattern occurred far more recently in Spain than with Cowgill’s data from York.
But seasonal birth patterns in North America present an unexplained enigma. Many studies report an autumn peak, commonly in September. This matches the pattern often reported for the Southern Hemisphere rather than that typical for northern latitudes. Cowgill suggested that social factors primarily control conception rhythms in North America, whereas environmental factors predominate elsewhere.
Little lemurs with rigid timetables
In 1968 I gained my first experience of primate behaviour under natural conditions in southeastern Madagascar by studying lesser mouse lemurs — relatively primitive primates weighing only two ounces. I managed to piece together an overall picture of the annual cycle: In late September and early October (the southern counterpart to our northern spring) adult females become receptive in close synchrony. Most conceive at once and give birth in late November/early December after a two-month pregnancy. I found year-to-year consistency in timing of the breeding season. When I returned to my Madagascar field site in 1970, breeding followed exactly the same pattern. The mating and birth seasons of lesser mouse lemurs are tightly bounded, each lasting just a few weeks. This, of course, differs starkly from the human pattern of year-round births with peaks and troughs. With 9-month pregnancies and natural breastfeeding for several years, seasonal changes in the environment have fuzzier effects.
What steers lesser mouse lemurs to mate at a specific time of the year? Matings and births both occur during Madagascar’s wet season in October-March. This is also the hottest part of the year, so either rainfall or temperature could directly influence breeding. However, the circannual clock, responding to changing day-length, might also trigger mating internally. This can be directly tested with laboratory experiments. When studying a breeding colony of mouse lemurs at University College London, I used a special light-clock that automatically replicated the annual pattern of day-length variation in Madagascar. This enabled me to induce these tiny primates to breed at a convenient time of the year. At one stage, I even managed to reduce the interval between breeding seasons by compressing the annual day-length cycle into 9 months.
Factors influencing seasonal patterns
In the attempt to identify factors driving human birth patterns, investigators have often sought a direct link to environmental conditions, emphasizing climatic factors such as temperature and rainfall (potentially influencing food availability). One suggestion is that energy balance influences ovulation; another is that high temperatures inhibit sperm. It is also possible that decreased sexual activity during the hottest months yields lower conception rates. Some authors have therefore examined conception as the prime focus, backdating 9 months from recorded births. But annual patterns of variation in temperature and rainfall remain fairly consistent between years in any given region, so seasonal patterns that change over time cannot be convincingly explained as a direct response to climatic factors.
Long-term experiments on human subjects to explore factors underlying seasonal patterns in reproduction are impossible, so explanations rely heavily on circumstantial evidence. In two 1990 papers, Till Roenneberg and Jürgen Aschoff applied a sophisticated statistical analysis to distinguish biological and social factors influencing annual patterns in human conceptions, backdated 9 months from births. They examined a massive worldwide dataset including 3,000 years of monthly birth rates from 166 different regions. Observed patterns clearly depend on latitude: Seasonality becomes more pronounced at higher latitudes and there is a 6-month shift between Northern and Southern Hemispheres. Moreover, Roenneberg and Aschoff were able to demonstrate convincingly for the first time that variation in day-length does influence human reproduction on a global scale. Ambient temperature was found to be important as well, and they also identified daily sunshine duration as a factor influencing annual conception patterns.
Possible origin of human birth seasons
Looking back into our evolutionary past, any biological basis for seasonal birth patterns in modern humans would presumably have become established in Africa. Outside Africa, fossil evidence for our species dates back at most to 120,000 years ago in the Middle East (a relatively short space of time in evolutionary terms), and it is widely accepted that ancestral humans migrated out of Africa. So the patterns we see today are likely to reflect adaptation to environmental conditions in northeastern African. Taking the Afar region of Ethiopia as an example, with relatively limited variation in average daily temperature (21-28 oC) and two separate rainfall peaks in April and August, birth frequency is typically above average between January and September, but progressively declines between October and December. This would correspond to a higher probability of conception between April and September. The various patterns that we see today in the Northern Hemisphere are perhaps derived from such an original distribution through adjustment to local temperature conditions and cultural influences. While increasing use of artificial lighting and heating has gradually modified and even suppressed the original pattern, perhaps it persists like an underground stream, with subtle effects on our biology.
Becker, S. (1991) Seasonal patterns of births and conception throughout the world. Advances in Experimental Medicine & Biology 286:59-72.
Cancho-Candela, R., Andres-de Llano, J.M. & Ardura-Fernandez, J. (2007) Decline and loss of birth seasonality in Spain: analysis of 33 421 731 births over 60 years. Journal of Epidemiology & Community Health 61:713-718.
Chandwani, K.D., Cech, I., Smolensky, M.H., Burau, K. & Hermida, R.C. (2004) Annual pattern of human conception in the state of Texas. Chronobiology International 21:73-93.
Cowgill, U.M. (1966a) The season of birth in man. Man, n.s. 1:232-241.
Cowgill, U.M. (1966b) Season of birth in man: Contemporary situation with special reference to Europe and the Southern Hemisphere. Ecology 47:614-623.
Cowgill, U.M. (1966c) Historical study of the season of birth in the city of York, England. Nature 209:1067-1070.
Dorélien, A.M. (2013) A time to be born: Birth seasonality in sub-Saharan Africa. Population Studies Center Research Report 13-785:1-60.
Quetelet, A. (1869) Physique Sociale ou Essai sur le Développement des Facultés de l’Homme. Paris: J.-B. Baillière et Fils.
Roenneberg, T. & Aschoff, J. (1990a) Annual rhythm of human reproduction. I. Biology, sociobiology or both? Journal of Biological Rhythms 5:195-216.
Roenneberg, T. & Aschoff, J. (1990b) Annual rhythm of human reproduction. II. Environmental correlations. Journal of Biological Rhythms 5:217-239.