Here I am in the Kalahari Desert, as far south of the equator as the Sahara is north.  It’s where the equatorial air that turned south finally comes down from on high, thoroughly dried out.  And since the weather systems move across the southern continent from east to west, it makes this mid-continent location a rain shadow as well.  The plentiful water hereabouts is rain that fell elsewhere, and then ran downhill to here.

     This delta is full of low islands, thanks to hard stuff beneath the shifting sands that the flowing water cannot easily cut into.  So when the river comes down out of the mountains of Angola and reaches these sands, it fans out – rather like what happens when a hose is left running in a large shallow sandbox.  This cuts a large land area into irregular parcels, looking from above like a reticulated giraffe’s skin.  (On many such islands, there are even reticulated giraffes nibbling at the tree tops.)  At the moment, the water level is low and so a lot of shallow water is now converted into green fields of delicious grass.

     Islands are always a matter of interest to the evolutionary biologist.  Just as Darwin was the pioneer of the modern theory of evolution, so Alfred Russel Wallace was an equivalent pioneer of island biogeography.  Here we see fragmentation without downsizing, a lovely teaching example.  Adjacent islands merge when the lake level diminishes, and the islands get rearranged when it floods.  Temporarily there may be islands without predators, and others with an oversupply.

     And speaking of grasslands-adapted baboons happily invading woodlands, I got a good dose of it last night.  They dropped out of an overhanging tree onto the roof of my tented cabin in the middle of the night, shaking things like an earthquake every half hour.  At one point three tails could be seen in the moonlight, dangling over the edge of the rain fly.     At least my open-mouthed yawn, when standing at the front door of the tent, was taken as a threat by the baboons, who ran away despite my lack of impressive canine teeth.  They didn’t even yawn back to display their oversized canines.  (I hope no one will argue that human yawning is an adaptation for dispersing baboons.  Some people see adaptations everywhere.  Yawn.)

     I somehow missed the arrival of the three elephants (who tore up the camp’s internal electricity cables a few hours before dawn) – I can’t imagine how I missed them, as I was awake half the night.  Don‘t these animals ever sleep?

Afternoon now, and so quiet that the baboons must be indulging in a siesta.  The game guide says that when baboons are so active at night, it is because they fear leopards.  So I missed seeing a leopard too.

     My cousin (no, not the German cousin again; she’s the Colorado-England-Kenya cousin with whom I discuss African health planning, her field) came over to where I was writing this morning before breakfast.  Just turn around, she said, and notice the elephant on the river bank behind you.  Oops.  But I could hardly miss the other two elephants in between the cabins, as they were busy dismembering a fallen tree while we ate a proper English breakfast and watched them.

     I’ve just seen two examples of how life shapes geography.  The river channels between these islands are often maintained by the hippos when they trudge through any new sandbars.  Hippos thereby contribute to maintaining island isolation for other species, preventing them from wading to the neighboring island at lower water levels.  Furthermore, the islands themselves are built up to somewhat higher elevations by all the termites that glue the sand grains together into harder stuff.  This island is full of tall termite mounds and their subterranean infrastructure.

I almost missed noticing the baobab tree because I thought, from the distance, that it was just another tall fat termite mound wrapped around a nearby tree, like the one I saw after breakfast.  This baobab looks, on closer inspection, like a table decoration, made by standing a potato upright with some toothpick “feet” and then sticking some leafy twigs up near its top, to mimic a broad-brimmed hat.  Baobabs are thought to be very important in hominid gathering strategies, as its leaves, fruit, and seeds can all be used (its pollen even makes a good glue).  And it can provide water.

     Baobabs are another little lesson in surface-to-volume ratios, and how you maximize volume for water storage while minimizing the surface area from which water can evaporate.  Baobabs are drought-resistant trees in a big way.

    Humans (especially marathon runners) sometimes do the opposite, like the high surface-to-volume ratio of the tall skinny Maasai compared to the rest of us.  It is said to be an adaptation for losing heat quickly by evaporation, by maximizing the surface area from which to sweat.  You have to avoid cooking that big brain with the heat from running and from the hot sun, particularly from both at once.  Upright posture itself is a way of minimizing surface area exposed to the hot overhead sun, just head and shoulders taking the full hit rather than a broad back.  You minimize your shadow.  Some anthropologists suggest that upright posture is a savanna adaptation for treeless places where you can’t “shade up,” as most sensible animals do at midday.

     There are so many suggested explanations for upright posture  – the evolutionary biologist J. B. S. Haldane liked to observe that only a human can swim a mile, run 20 miles, and then climb a tree – that it is difficult to say why our ancestors did it and why the other great apes didn’t.  Sustained bipedal running is really more efficient than quadrapedal, especially if you have a heavy head to support (cantilevering it during bouncy locomotion sure takes a lot of neck muscles, compared to balancing it atop the spinal column).  This is an advantage in the long run, which must be distinguished from arguments about how bipedality got started.

     I have a favorite, naturally, but it doesn’t really exclude any of the other candidates for bipedality’s origin.  Most leading features in evolution have a supporting cast.  Indeed it is often like a repertory theater, with the star one night acting as the walk-on butler the next night.  Picking an overall “star” in an evolutionary repertory is often a mere matter of taste, though “which is fastest” is a good criterion when you are trying to figure out how we got here from back there, and so quickly.

As I mentioned in Paris, the chimpanzeelike hips and knees got modified early on, presumably for upright locomotion – but there are nonlocomotion possibilities too, such as upright stance per se.  Seeing over the tall grass was said to be an advantage (until they discovered upright posture came four million years earlier than life on the savanna).  Upright stance is also an advantage if you wade a lot but haven’t yet learned to swim.

     There may be some secondary uses of upright stance, as for picking fruits off trees without having to climb them, but the others seem more likely to have substantial payoffs.  There are also some temporary advantages of upright posture for hunting large animals, as naïve animals tend not to fear them as they do their usual four-legged predators.  This allows the hunter to get closer before they start edging back.  But this advantage doesn’t last, once a herd has been hunted for awhile.

     Most animals who live in the grass and bush have, of course, managed to do it without switching to upright posture, so I tend to favor the wading-and-shallow-diving hypothesis, given the suite of other adaptations that we humans have (subcutaneous fat layer, copious tearing, loss of most body hair, breathing control for underwater, kidneys that are relatively unconcerned with hoarding salt and water, and so forth) that are often seen in the mammals that returned to the sea.  Losing hair for whatever purpose would also tend to promote upright stance because infants would then need to be carried, being unable to find much maternal hair to grasp (we are an exception to the general infant-carrying rule among primates).  Or perhaps the infants lost their ability to cling, forcing carrying.

     This is usually called the “aquatic ape hypothesis” because it involves so many things that are also seen in the land mammals who returned to life in the sea (whales did it 100 million years ago, sea otters only several million years ago) with a salty diet and a need for less body hair.  No one imagines a fully aquatic ancestor, so the name is somewhat misleading, but rather a creature that foraged along shorelines and occasionally swam a little.  One version of this hypothesis about hominids emphasizes islands, as they have a lot of shoreline and, in a drought, all of the remaining resources on the island might have been the fish and shellfish along the shoreline.

     While apes isolated on a chain of islands in the Red Sea would indeed be an excellent setup for doing the aquatic adaptations quickly, it may be that the lakes and rivers of East Africa could have also done the job.  There are lots of fish in shallow lakes that can be herded into modern nets by small boys splashing around.  In the days before nets, one could likely drive them into restricted spaces where someone could heave them ashore.

     Even filling in the fossil gap between the great apes and the australopithecines may not help settle the issue because they will yield mostly bones from the usual sites of preservation (caves, lake margins).  Bones often tell you something about muscle strength and, via the size of vertebral openings for nerves in the chest region, something about how good their breathing control might have been.  But they won’t tell you about fat layers and the extent of hair coverage, nor about how much salt was conserved by their kidneys.

     You might think that, because so many hominid fossils are found at former lake edges, this might be used as evidence for shoreline foraging.  But the pros all know that sites like forests are very unlikely to preserve bones, and that lakeshores are excellent in that regard.  Caves also preserve the occasional skeleton as at Sterkfontein, but no one assumes australopithecines preferred to live there (more often, pits within Sterkfontein became death traps for explorers or those being chased).  And so paleoanthropologists quite understandably treat the water’s edge as simply a great setup for preservation (a flooding lake buries those who die near the water’s edge and moves the shoreline – and the grazing animals who might crush the bones –  back away), and do not also use it as an argument for where they preferred to live.  Still, the evidence is often consistent with shoreline living.

     In forty years, the aquatic aspect has gained few adherents among the pros, even though the savanna hypothesis has recently proven awkward for the australopithecines.  Archaeologists do not buy the aquatic hypothesis, perhaps because there’s nothing in it for them to study yet (a nice trash heap of shells that is 5 million years old might change their minds).  Physical anthropologists don’t like it for similar reasons; their strength is anatomy, and most aspects of the aquatic ape hypothesis are physiological.  When they proclaim “But there is no evidence for that!” about the aquatic hypothesis (and they are quite vehement), what they seem to mean is that there is none of their specialized kind of evidence – except, of course, those hip and knee rearrangements, and they prefer to ascribe other functions to them.  Everybody has a mental checklist for what needs explaining (mine is that chunnel-train list), mostly items within their own technical expertise, and many don’t like to be bothered by things that don’t address their agenda.

     Occasionally in the history of science, facts finally accumulate to the point that the old way of looking at them seems a little awkward compared to another – perhaps a minority view or some new suggestion.  The facts aren’t yet good enough to make the woodland-to-savanna bipedality hypothesis look awkward without a shoreline interlude.  But then the suite of hominid questions that require evolutionary answers doesn’t, for most, yet include the physiological or the neurobiological agenda.

     Now that the evidence for upright posture has reached back to six million years ago, very close to the DNA dating for the common ancestors with chimps, we are faced with a situation where efficient bipedal locomotion (losing those tree-climbing feet to improve running) happens a few million years after upright stance per se.  So maybe infant carrying or something like wading came very early.  Certainly the bipedal apes were sticking close to woodlands and even Homo erectus, though found in more arid environments and adapted to heat stress, probably had the same savanna drawbacks that we moderns do:  our kidneys waste so much (compared to truly arid-adapted animals) that, before canteens were invented, our ancestors had to stay close to drinking water.

We now know a lot about island biogeography, including the fact that evolution seems to operate faster in small populations than on continents with large ones.  Large central populations tend to buffer change, as natural selection for one trait may be diluted or balanced out by selection for another.  Individuals there have a lot of mating choices, and aren’t as likely to mate with someone whose ancestors have been through similar selection regimes.

     But the archaeologists are now starting to emphasize that population density of australopithecine and Homo species may have remained quite low throughout the ice ages, meaning that large central populations may merely be a feature of recent agricultural times.  Low numbers are what you would expect from top predators in the food chain, the same reason why bird-eating birds like peregrine falcons are so few in numbers compared to pigeons, or why it takes large herds of grazing animals to support a few lions.  Or a few hominids.  Maybe it wasn’t until we learned to grow grains and bake bread that human population density could increase significantly.  Still, I’d bet that hunting is what most allowed the hominid range to expand, what with all those naïve herds to tackle.

Happenstance during subdivision may omit typical predators.  For example, in the last warm period when rising sea level converted a peninsula on the coastline of France into the island of Jersey.  The red deer trapped there underwent a considerable dwarfing in stature within only a few thousand years.  That’s probably because their usual predators died out locally – predators that had made large body size a real advantage.  Lacking predators, there is something to be said for maturing early (at a small body size) and having more time to produce more offspring.

     So it’s possible to predict some of the things that might happen if a prehuman population were fragmented into smaller inbreeding subpopulations by an abrupt climate change.  A higher percentage of the total then live on the margins of some habitat (it’s surface-to-volume ratio, once again) and the margin is also where selection pressure is greatest.

     Local extinctions, as when an island population becomes too small to sustain itself, also speed evolution in a way that isn’t immediately obvious –  that’s because no competition is markedly better than some competition.  A local extinction creates an empty niche.  When subsequent pioneers rediscover the unused resources, their descendants go through a series of generations where there is more than enough food.  That means that even the more extreme variations that arise, the ones that in childhood would ordinarily lose out in the competition with the more optimally endowed – such as the survivors of a resident population – can now survive and reproduce for a few generations.

     When the environment again changes, some of those more extreme variants may be able to better cope with the third environment – better, at least, than the narrower range of variants that would reach reproductive age under the regime of a long-contested niche.  So a flipping climate has an ability to get more variants out onto the board in play, as well as providing a recurring stress that culls the less versatile.

     Thinking of the Ice Ages as the “Chattering Ages” with alternating boom-and-bust provides a perspective quite different from adaptationism’s usual focus on efficiency.  Efficiency arguments, as I mentioned, tend to suggest lean mean machines without a scrap of excess baggage.  But the need to discover a new way of making a living within a single generation shows how jack-of-all-trades variants could survive better.  Techniques that were last needed a hundred generations earlier would need to be rediscovered, in order to make use of less-favored food resources.

Randomly-picked small populations are rarely average.  Often they have some odd clustering.  This always seems to surprise people – as, say, when a randomly-selected jury turns out to be all men or all women, not exactly the proportions in the larger population from which they were drawn.

     This usually isn’t a bias in the selection procedure; it’s just how chance sometimes operates when a few are drawn from the many.  This happenstance clustering has some interesting implications for the evolution of social traits, things like language or reciprocal altruism where groups are important.  By chance, some subgroups are strikingly over represented in one trait, woefully lacking in others. Evolution now operates on dozens of subpopulations independently, rather than upon the whole large “average” population – and a subpopulation may thrive relative to others, simply because it chanced to have a disproportionate number of the bearers of some minority trait.

      Culture can pass things along, but a critical mass is sometimes needed to get cultural transmission going; most inventions simply die out.  Others are useful in the long run, but can be easily overburdened in the short run – and that’s the big problem with reciprocal altruism.

In most species that share food more generally than just mothers sharing with their offspring, individuals only share with close relatives.  If they help out someone being attacked, this assistance is also usually limited to close relatives.  That’s kin selection, where you are helping out copies of your own genes by helping the others.

     As human society presently demonstrates, there are great benefits to expanding the circle of beneficiaries to nonrelatives, what is called reciprocal altruism.  But it’s a puzzle:  How could that happen, when everyone loves to freeload?  Cheaters (those who receive without eventually reciprocating) are the norm in animal societies.  Any individual that tended to give away food, or indiscriminately risk life and limb for non-relatives, would be a loser – unless living, by happenstance, in a subpopulation with a lot of other indiscriminate sharers, likely to provide benefits at other times.

     And that’s what the repeated fragmentation of large prehuman populations into many smaller subpopulations could have occasionally created: a group with a critical mass of sharers.  In hard times, when the every-man-for-himself groups were wasting a lot of time and effort at fighting over the remaining food, the groups that shared (and otherwise minimized conflict) might have survived better, successfully raising a next generation when the others were squabbling.  They weren’t competing against each other as in team sports but rather against the downsizing environment, for sheer survival.

     In this manner, natural selection can occasionally operate on groups – and therefore on social traits.  Some things, like language and altruism, only operate between a substantial number of individuals.  If all the subpopulations are lumped together and mixed, as in today’s cosmopolitan societies, it may be hard to initially evolve such traits, simply because there are always enough freeloaders nearby to swamp and sink even a promising startup.  But a history featuring fragmentation, and then amalgamation (and repeating hundreds of times), is capable of accomplishing some things that might otherwise be improbable.  Thinking in terms of the average can seriously mislead you.

     The traditional thinking that dismisses group selection is that, even if a subpopulation happened to have a majority of cooperators, you’d still expect that tendencies to share could be swamped by all the non-reciprocating freeloaders, who would out-reproduce the sharers and slowly sink the altruistic practice.  So the group trait would be leaky, like a car tire going slowly flat.

     If this were the prime consideration, of course, we would also have to conclude that car tires would never work.  Sooner or later, they too all go flat.  We just pump them back up occasionally, and maybe that’s what reciprocal altruism takes.  The bust-then-boom cycle provides both a concentration mechanism (via fragmentation) and a pump (survivors get the eventual re-expansion opportunities).  Such pumping might allow widespread cooperation to become established long enough for other things to be invented that prop up cooperation by combating freeloading.  I sometimes think that the first sentence spoken was “But you owe me!”

If you could interview a chimpanzee about the differences between humans and apes . . . , I think it might say, “You humans are very odd; when you get food, instead of eating it promptly like any sensible ape, you haul it off and share it with others.”

- Glynn Isaac (1937-1985)