Page 303 - Earth's Climate Past and Future

CHAPTER 15 • Humans and Preindustrial Climate 279

3
15-3 Testing Climatic Hypotheses with Size of braincase (cm )
Fragmentary Records 0 400 800 1200 1600
0 Homo
Again and again we bump up against the problem of the sapiens
fragmentary nature of the record of hominin remains.
As noted earlier, preservation of fossils in Africa, Arabia, 1 Rapid
and southern Asia is sparse in part because of prevailing evolution?
aridity. And for hominin bones made of easily dissolved Homo erectus
calcium phosphate (Ca PO ), preservation is even worse
3 4
in the acid-rich soils of the rain forests. Because of Rapid
these problems, the total record of human evolution Myr ago 2 evolution?
over 5 million years is based on a few dozen fragments of
skeletons, enough to reveal some of the broad outline of 3 Gradual
human evolution but few of the details. evolution?
We have already seen that sampling records of Australopithecus
orbital-scale climate change even at intervals as close as a Rapid
few thousand years can lead to gross misrepresentations 4 evolution?
of the shape of the actual climate signals (Chapter 7).
This aliasing problem becomes all the more formidable Samples
in records with just one sample every 100,000 years or so.
Aliasing can produce erroneous indications of the 5
time of first acquisition of new physical or technical FIGURE 15-8 Undersampling of a measurable characteristic
skills (such as walking or the use of tools and fire) or of If a slowly evolving characteristic, such as the size of the human
the first or last appearance of a new hominin species. brain, varies widely at all points in time (black double arrows),
With only a few samples, the actual first or last appear- scattered sampling (yellow circles) will permit an interpretation
ances are likely to be missed, and instead we see a much- of either gradual or rapid evolution.
reduced range (Figure 15–7). As a result, the true ranges
of most hominins are probably longer than those shown
in Figure 15–2. Similarly, aliasing also complicates range of natural variation occurs within human (or pre-
attempts to define the relative timing between climate human) populations. For example, your classmates have
changes and the first use of new skills. heads that vary widely in size around the average value
A second undersampling problem pertains to quan- for the entire class. The size of a human head is closely
titative measurements of the evolution of physical traits. tied to the size of the braincase (the part of the skull that
In this case, the fundamental problem is that a broad
cradles the brain), a key trait in the fossil record of
human evolution.
The actual range of braincase sizes present at any
one time in the past is fairly large in comparison with
the evolution of the mean value (Figure 15–8). If the
fossil record provides only one or two well-preserved
* specimens every hundred thousand years or so, a good
chance exists that some of these specimens will not be
Age range representative of the brain size of the entire population
inferred living at that time but will fall above or below the mean.
Samples from True
* samples range Depending on the specific samples collected and ana-
age
scarce
lyzed, an inaccurate picture of the long-term trend
could emerge.
* be possible to obtain a general sense of the direction in
Even if the available fossil record is sparse, it should
which the trend is moving, especially if the net amount
of evolution far exceeds the natural range of variation at
any one time. But the limitations of sparse data make it
FIGURE 15-7 Undersampling of a fossil record If few impossible to define the true rates of change. Figure 15–8
samples of the fossil record of an organism are available, the shows that sparse data on brain size could be interpreted
true first and last appearances of the organism will be poorly equally well as either a slow, gradual trend or as rapid
estimated. bursts of change.