San José State University
Thayer Watkins
Silicon Valley
& Tornado Alley

Gestation Periods and Animal Scale

The gestation period of humans is 266 days, 8 days short of nine months. Many such biological characteristics have an explanation in terms evolutionary survival value. Even some seemingly cultural characteristics of human society have such a biological explanation. For example, in all known hunting-and-gathering societies it is the men who do the hunting and women the gathering. It was not until a few decades ago when in the U.S. people started doing marathon running that the explanation was discovered. Women marathon runners found they lost their periods and medical examination revealed that under the stress of distance running women stopped ovulating. In hunting-and-gathering societies the hunters cover about thirty miles a day, day after day. In a society in which the women participated in the hunting there would be few babies born and that society would quickly disappear. Thus the only hunting-and-gathering societies that survived would have the men doing the hunting and the women doing the gathering.

It is therefore tempting to try to find a biological explanation for the nine-month gestation period of humans in terms of seasons. What time of the year would be the best time for babies to be born in order to maximize there chances of survival? And what time of the year would be most conducive to human mating? In a temperate zone springtime mating would lead to midwinter births which surely would not be the optimal birth time. If springtime were the optimal time for births then a nine-month gestation period would imply wintertime mating, which is not the season that people associate with urges to merge. It would seem that a one-year gestation period would give the best fit between mating time and optimal birth time.

But any appeal to seasonality for an explanation of the human gestation period is fruitless because humans did not evolve in a temperate zone. There might have been some dry-wet seasonality in the savannahs of Africa but it is uncertain. And the gestation periods of chimpanzees (237 days), gorilla (257 days) and orangutan (260 days) indicate that the approximate nine-month gestation period predates the emergence of human beings as a species.

The gestation periods of other animal reveals that the length of the gestation period is largely a matter of the animal specie's size. Asian elephants have a gestation period of 645 days and African elephants 640 days. Dogs and other canines have a gestation period of about 60, which is also the period for cats. Even within the species of apes and monkeys gestation period seem to be a matter of size. For Rhesus monkeys it is 164 days and baboons 187 days. For small animals such as rabbits the period is about 33 days and for mice about 20 days.

So it takes a longer time for a big animal such an elephant to grow a baby than for a smaller animal such as a mouse. The baby of the bigger animal is of course bigger but there are the bigger resources of the big animal to promote that growth so it is not obvious that bigger animals would have longer gestation periods. But, deespite the bigger resources, it does take a longer time.

It seems to be a matter of scale. If all of the dimensions of an animal are doubled the double-scaled animal would have eight times the volume and hence eight times the weight of the smaller scaled animal. But the cross section area of the umbilical cord through which all of the nutrients for growing the baby would have to pass is only four times as large. So, all other things being equal, it would take twice as long for the nutrients for the eight-times larger baby to pass through the four-times greater capacity umbilical cord. The blood pressure in the scaled up animal could be twice as great but the length of the umbilical cord would be twice as long giving twice the resistance to the flow of the nutrients. Thus the gestation period should be proportional to the scale of the animal. The volume and hence weight of an animal is proportional to the cube of the scale so scale is proportional to the cube root of the weight of the animal. This would apply to animal of different sizes but the same general shape.

According to the above analysis if one plots the logarithm of gestation period versus the logarithm of the weight of the females there should be a linear relationship with a slope equal to 1/3.

The analysis needs to be modified to allow for the fact that smaller animals find it so easy to grow offspring that they grow them in litters rather than one at a time. Also one needs to know how the ratio of the weight of the offspring in proportion to that of the mother varies with animal size. For humans the weight of the baby is about 5 percent of the weight of the mother and for elephants it is about 4 percent. For cats and dogs the litter weight is a higher proportion of the mother's weight. Thus the slope of the statisitical relation between the logarithm of the gestation period and the logarithm of weight would be less than 1/3.

The data for a number of animal species are given in the table below. In tabulating the data if a source gave a range of values for a datum the midpoint of the range is used. The female body weight that would be relevant would be body weight without fat. There are structural differences between bipeds and quadrapeds so it is appropriate to carry out the analysis separately for the two types of animals. First the primates are considered.

Gestation Periods and Female Weights
for Some Primates
Gestation Female
log(Gest) log(wt)
Species (days) (kgs)
human 266 50 2.425 1.699
chimpanzee 227 40 2.356 1.602
gorilla 257 70 2.410 1.845
orangutan 260 40 2.415 1.602
baboon 187 20 2.272 1.301
monkey, Rhesushesus 164 5 2.215 0.699
monkey, Patas 167 5.5 2.223 0.740

This data gives the follow scatter diagram.

Regression analysis gives the best fitting line for the statistical relationship as

log(Gest) = 2.075 + 0.189(log(Wt)

with a coefficient of determination (R2) equal to 0.9. The results are promising. The slope of the line is significantly below the 1/3 value suggested by the analysis but that could be accounted for by the variation in baby weight as a ratio to mother's weight with size.

The data for a number of quadrapeds are shown below. These are the cases in which the values are readily available in an encyclopedia.

Gestation Periods and Female Weights
for Some Quadrapeds
Gestation Female
log(Gest) log(wt)
Species (days) (kgs)
cow 284 730 2.453 2.863
bison 270 600 2.431 2.778
moose 245 550 2.389 2.740
llama 330 113 2.519 2.053
goat 150 15 2.176 1.176
sheep 148 35 2.170 1.544
bear, black 210 295 2.470 2.322
wolf 64 40 1.806 1.602
elephant, Asian 645 4000 2.810 3.602
elephant, African 640 5000 2.806 3.699
lion 108 150 2.176 2.033
leopard 94 50 1.973 1.699
pig, domestic 114 80 2.057 1.903
rabbit 33 1 1.519 0

Regression analysis gives the best fitting line for the statistical relationship as

log(Gest) = 1.527 + 0.333(log(Wt)

with a coefficient of determination (R2) equal to 0.81. The value of 0.333 is sheer coincidence. The data is too uncertain and spotty to say that this confirms the analysis precisely. However the fact that the coefficient is less than 1.0 and in the neighborhood of 1/3 does indicate that the analysis is sound and that the pursuit of more and better data would be worthwhile. The conclusion is that the nine-month gestation period for humans is merely a matter of human size and not the result of some evolutionary optimization. It is all in the umbilical cord. Babies are like blossoms on the stem of the umbilical cord. And just as there is

The force that through
the green fuse
drives the flower
(Dylan Thomas)

there is the force that builds the baby and it is subject to limitations of size and scale of the umbilical cord.

HOME PAGE OF applet-magic
HOME PAGE OF Thayer Watkins