wow, those are some eye-opening stats, esp this table:
basically the opposite of what I have seen, though maybe my perception is off because my career has been spent in energy and finance.
If you follow the link this appears to be a poll of AAAS members, which tends to mean actual scientific researchers with PhD's. If you polled, say, engineers, doctors, geologists (oil/mining, not univ.), etc., with a BS or MS, you'd get very different results. Research scientists tend to be very Democratic, even though many other professionals who use science are Republican.
The reasons for this are complex; partially it has to do with economic status (researchers are more likely to be dependent on government funding and working for non-profits, other science professionals are highly white-collar/private sector/non-union, which is a very GOP combo), and partially it has to do with the complicated way in which science intersects with culture and public policy. There is a certain combination of engineering and technical expertise with disbelief in evolution by natural selection or anthropogenic climate change that is actually quite common among well-educated evangelicals but almost unknown among researchers.