The paper's a bit hard to read but I don't think it suggests a mechanism, or that the correlation between CMEs and forest fires is statistically significant for that matter. If you email me (address in profile) I can send you a copy.

It does seem plausible that you could get fires as a result of electrical sparks off of long conductors (transmission and communication wires, pipelines, fences), such as with the Carrington event (https://en.wikipedia.org/wiki/Geomagnetically_induced_curren...).