The extraordinary recent discoveries in human evolution — such as the discovery of a half-Denisovan, half-Neanderthal hybrid — may have obscured a giant black hole in our history: We don’t know which species we came from. Now, a group of scientists has employed statistics and concluded that whoever our ancestral species was, it almost certainly wasn’t the hominin found in South Africa called Australopithecus sediba.
Published in May in Science Advances, the statistical calculations by Andrew Du and Zeresenay Alemseged of the University of Chicago support the opinion of Prof. Yoel Rak of Tel Aviv University and Prof. William Kimbel of Arizona State University, who based their analysis on morphology — that Australopithecus sediba is not our ancestor. For one thing, it lived later than the earliest known member of the Homo genus, as far as we know. And then it went extinct.
If anything, in his opinion: "Australopithecines and Homo are sister species," Rak tells Haaretz. Not our ancestors.
It has become clear that in contrast to past thinking, human evolution was not a neat, linear progression. Our genus split from chimps six to seven million years ago, but it isn’t that since then one Homo species developed, which evolved into the next Homo species, and so on. Like pretty much every other creature out there, hominins underwent some sort of adaptive radiation, with ancestral human species diversifying into multiple forms.
Over the millions of years, continental Africa and, later, Eurasia were thronged with multiple types of hominins, at least some of which we have yet to detect except as faint signals in modern human DNA (genetic “ghosts”). Today, it seems that only one hominin species remains: us! But that is the exception, not the rule.
The upshot has been an enormous amount of arguing as to who was ancestral to whom, which are distinct species and which mere variants, and more.
- Early humans were roasting tubers 120,000 years ago, researchers discover
- New species of hominin found in Philippines changes paradigm of human evolution
- Giant genetic study detects ‘ghost’ species in today’s horses
Multiple archaic human types existed contemporaneously, and they interbred, apparently a lot. (Various archaic chimp species also merrily mated with one another in prehistory, scientists showed in April; the same indiscriminate behavior has been detected in horses. Miscegenation could be the rule, not the exception.)
Case in point: australopithecines. These archaic hominins are roughly midway between the chimpanzee and modern humans, and they existed in more than one type about three to two million years ago. We have no real idea if an australopithecine, any australopithecine, was directly ancestral to us, or if they were a lineage that went extinct. However, there has been a general assumption that some australopithecine was in our ancestry.
Though not the first australopithecine to be found (that was the robust Boisei), gracile Lucy is the most famous of her type: Australopithecus afarensis. Hominin fossils are frustratingly rare, and when any are found they tend to be very partial. In her case, more than 40 percent of her bones were found, back in 1974, in the Afar Triangle in Ethiopia.
Lucy’s remains were highly edifying. By modern-human criteria, Lucy was tiny — only about 1.1 meters (3 foot 6 inches) tall — and lived about 3.2 million years ago. Going by knee structure, australopithecines were capable of walking upright, though the fossil foot of an infant australopithecine found in Ethiopia in 2018 indicated that the wee ones, at least, lurked in the trees, probably to avoid getting eaten by predators.
In 2008, what was apparently a somewhat later sister species to Lucy was discovered by the University of the Witwatersrand’s Prof. Lee Berger in a South African cave: Australopithecus sediba. Two specimens dating to about 2 million years ago were found in a cave called Malapa, over a million years after Lucy fell off a tree and died.
It bears adding that the discovery of a 2-million-year-old hominin fossil doesn’t mean that’s when the species first arose. It just means that’s what we have found of it so far. We have no idea what time range Australopithecus sediba existed in — or what spatial range, for that matter. But could it have been a direct ancestor to ourselves?
Pros of parenthood
Since the sediba’s discovery, the scientific community has been divided about whether it was a distinct species or a mere variant of Australopithecus africanus, a species a million years later than Lucy. Berger things sediba and africanus were different species. Prof. Yoel Rak for one retains the view that sediba was an africanus variant, he told Haaretz.
The scientific community is even more divided as to whether sediba was our mommy. Lacking any solid evidence whatsoever for genetic continuity, interpretation has to be inferential.
The year he discovered them, Berger argued in the Journal of Anthropological Sciences that, barring further evidence in the form of better preserved specimens, sediba could be a contender.
In 2015, further evidence ensued, in the form of an article in Nature pointing out that sediba’s foot was un-humanlike because it lacked a longitudinal arch. People whose longitudinal arches collapse are said to have flat feet. “The two major traits that distinguish the modern human foot from the ape foot are the longitudinal arch and the adducted hallux,” stated the paper. (Also in 2015, Berger and others would pen a new paper, in Scientific Reports, describing fatal injuries in two of the Sediban bodies. For what it’s worth, they’re not ascribing the injuries to premeditated murder, though that behavior has been observed in modern chimpanzees. They think the individuals fell into the pit-like cave and suffered mortal injuries.)
Rak also points out that if he's right and sediba is merely a variant of africanus, well, in any case, "it is known by many as not leading to us – it won't evolve into humans, so sediba can't be our ancestor," he says.
Are we our own grandpaw
The thrust of the case now presented by Du and Alemseged, which supports the non-ancestor theory of Kimble and Rak, is that the known fossils of sediba date to 1.97 million years ago, which is 800,000 years later than the earliest member of the Homo genus. Accordingly, early Homo came before the sedibans.
Awkward, that. Not impossible — as we said, we don’t know the range of sediba in time or space, but add that to the flat feet and dentition discrepancies and sediba as mother isn’t looking likely.
The notion that sediba was ancestral to us was based on theoretical overlap. In other words, an ancient group of sediba could have given rise to the Homo genus, while other groups of sediba continued on their sediban way. Because discovering any fossil human remains is so rare, we just haven’t found the remains of the sedibans that preceded us, goes the theory.
Not that the new paper renders it impossible, but Du and Alemseged used statistical probability models and concluded that observing an ancestor’s fossil horizon that is at least 800,000 years younger than the descendant’s fossil horizon is extremely unlikely (about 0.09 percent).
Du kindly explained to Haaretz what they’re talking about. “A fossil horizon is defined as a time period where at least one fossil from the species of interest is found,” he says. “For example, there are many A. sediba fossils, but they are all from the time period about 2 million years ago, so they all constitute one fossil horizon only.”
What Du and Alemseged did is estimate the probability of obtaining the observed data assuming a particular model. The model they assumed was that sediba is ancestral to Homo — so they calculated the probability of finding an A. sediba fossil horizon that is at least 800,000 years younger than one from earliest Homo, which is what we observe in the fossil record.
Sediba postdates the earliest Homo by 800,000 years in reality, based on the fossil evidence so far. So, the model’s assumption (i.e., A. sediba being ancestral to Homo) is incorrect. Q.E.D.
“Our paper can be considered the first analysis that directly tests the hypothesis of A. sediba being ancestral to Homo using the standard machinery of hypothesis testing,” says Du. And it isn't.
He immediately qualifies that we have exactly one fossil horizon each from A. sediba and earliest Homo, so have no clue what their true temporal ranges were and how much they overlapped.
When he discovered sediba, Prof. Berger suggested they were a new species and thought it might be an intermediate between africanus and Homo sapiens, Rak explains. "Kimble and I concluded that there is no difference between A. sediba and A. africanus; sediba was not substantially different from africanus. This paper says we're right."
So who is our ancestor? "I think the species the best candidate to be ancestral, to both Australopithecine and to Homo sapiens, is a little known creature a million years older than Lucy, discovered by Leakey, Australoithecine anamensis. Amamensis came from the even older Ardipithecus, which lived 5 million years ago. Ardepithecus is very close to the split between the hominids and chimps."
So it is still possible the Malapa fossils are late representatives of a lengthy A. sediba history that predated and then overlapped — but that seems highly unlikely. There is one case (only one) of the earliest-known fossil of a postulated ancestor, Homo erectus, appearing after its proposed descendant — Homo antecessor, Du says. But at the end of the day, if any australopith is our mommy, they think it’s the Lucy branch, Australopithecus afarensis. Rak for one doesn't agree and the story of human evolution, again, takes a giant step forward and one sideways.