If I understand the way recessive genes work, for them to activate they'd need to come from both parents. This works if the non-Caldari parent also has some Caldari blood, or if the trait appears elsewhere at random.
Another possibility, albeit an extremely rare one in our own biology, is Y-linked recessive traits passed from a father to son. The trait sits on the Y chromosome and is passed on by the one parent, in that case. Of course, since this would limit Kresh drinking to males only, something which is observably not the case, this whole post serves no purpose other than to show off that I've been listening in Biology class...
Another more useful possibility for consideration is the idea of incomplete dominance. In that case, someone could be heterozygous dominant (Aa), where the little "a" represents the Kresh-digesting enzyme (or whatever) and yet the recessive gene could still be partially expressed (still capable of limited enzyme production and therefore limited digestion). In practical terms they might not be able to smash a whole bottle, but they also won't die from a little sip. In that possible scenario, the recessive genes can come from just one parent and yet still be expressed.
The only reason the Caldari went through the slow process of acclimation to Kresh is because it was the only green plant that actually survived the worst of the Caldari Prime winters - according to stuff I've read. For that reason, being able to tolerate the particular neurotoxin in Kresh was such a pro-survival trait that Caldari who had the ability seem to have bred out those who didn't.
Fascinating stuff. It reminds me of lactose intolerance, actually, which evolved in Anglo/Euro societies where there was a selective advantage to having the lactase enzyme. Meanwhile, in Africa where cows weren't domesticated, most never evolved it. Maybe a tall glass of cold milk is the Englishmen's version of Kresh.