The reason for this is pure psychology. As Daniel Gilbert, psychologist at Harvard has pointed out, we over-react to immediate threats and under-react to long-term threats, especially those that occur slowly and over time. Thus, no matter what many B.C.-ers at the polls indicated about their alarm over the environment, in the privacy of the voters booth it was not the alarm that sounded the most loudly.
Instead, the risk raised by B.C. Liberals that being custodial towards the environment would harm the B.C. economy by cutting it off from tangible benefits such as Liquified Natural Gas development turned out to be the one with the greatest psychological purchase in this election. In contrast, the risks of damaged coastlines and global warming fell back out of first place into the category of an abstract and removed risk.
Obviously this kind of risk analysis is not entirely rational, as either a bitumen spill, whose risk some researchers have suggested to be over 90% if Enbridge puts through the Northern Gateway Pipeline (regarding which the B.C. Liberals have been equivocal), or a globe devastated by more than 2 degrees rise in temperature--as is more likely to come about if hydrocarbon development in Canada is not curtailed--will prove much more devastating to numerous British Columbians than a case in which its LNG promises remain unfulfilled.
But what this election shows is that you probably can't expect a voting populace to perform dispassionate and rational risk assessment in an election context. People will not make the hard decision to sacrifice local economic benefits to avert a widely distributed environmental cataclysm which, might frankly, occur anyway. Those kinds of globally responsible choices require leadership and vision, and it will consequently be up to governments that are in power to make them.