All the possibilities that you mention are possible and just need interested researchers and resources. The problem is that no scientific study has ever documented good evidence for animal sensing earthquakes before they happen. So few sophisticated mainstream researchers are trying the studies, although thousands of such studies have been tried in the past, particularly in China.

The bottom line is that the overwhelming odds are that there is no precursory signal for the animal to sense. However, it is very difficult to prove a negative.

If the program said only that Berkland's ideas are controversial, rather than Berkland's ideas have been tested carefully again and again with thorough and scientific projects and found to be baseless, I'd say that part of the program was not a fair assessment, but rather striving for spectacle over truth.