The Mishna discusses a case where a mother cow gives birth to a child that looks like a donkey. The gemara learns from the pasuk of פטר חמור that the child would not have Kedushas Bechor. It is clear from the gemara that we completely exempt it from any kedusha of bechor, but it isn't clear from the gemara which type of kedusha it would have if not for the pasuk excluding it. Meaning, do we regard it as a cow since genetically it is a cow, just that it looks like a donkey, or do we regard it as a donkey since it looks like a donkey?
It would seem that the answer lies in the end of the Mishna which cites a rule that היוצא מן הטמא טמא והיוצא מן הטהור טהור. The mishna seems to indicate that the genetic makeup of an animal determines the species to which it belongs, therefore an animal born from a kosher animal (even if it looks like a donkey) is kosher. Similarly, an animal born from a non-kosher animal (even if it looks like a cow) is not a kosher animal. If genetics determines the category, we would have to say that a cow that gives birth to a donkey would have been treated like a בהמה טהורה and been brought as a korban on the mizbei'ach, if not for the pasuk excluding it.
However, the implication of the mishna which uses a pasuk of פטר חמור to exclude this case, implies that we really need the pasuk to exclude it from the requirement of פטר חמור, not from kedushas mizbei'ach (although the gemara confirms that it would certainly be excluded from kedushas mizbei'ach as well). The Steipler (4) proves from Tosafos that this is true. Tosafos suggests that for a cow that gives birth to a donkey the pasuk of the mishna (פטר חמור) would be necessary to exclude if from the mitzvah of redemption like a donkey, and the pasuk in the gemara would be necessary to exclude the reverse (a donkey giving birth to a cow) from kedushas mizbei'ach. Although Tosafos concludes that either pasuk would exclude it completely, their logical default for not having a pasuk would be that the animal should be treated as it looks, not as the species of it's mother. Meaning, a cow that gives birth to a donkey would have been treated like a donkey to require redemption despite that it's genetic makeup is that of a cow. The Steipler proves from here that the species of any animal is determined by it's appearance, not by it's genetics. Nevertheless, there is a separate rule that if it looks like a kosher animal but is born from a donkey, it cannot be eaten - היוצא מן הטמא טמא, not because it's a donkey, but because it is a non-kosher cow. Similarly, if it looks like a donkey it is a donkey, yet if born from a cow it can be eaten because היוצא מן הטהור טהור. This approach of categorizing a donkey born from a cow as a "kosher donkey" is significant because for purposes other than eating i.e. eiver min ha'chai it would have status of what it looks like, not what it is genetically.
The gemara 6a seems to strongly support the approach of the steipler that the species is determined by it's appearance rather than it's genetics. According to the rule that having similarities to it's mother can make it have kedusha of bechor, the gemara questions whether a donkey that has similarities to it's mother which is a cow, would have kedusha. The gemara's suggestion of why it is too different from the mother and wouldn't have kedusha is - הא טמאה והא טהורה, הא קדושת הגוף והא קדושת דמים. The gemara describes it as a non-kosher animal and as the value having kedusha rather than inherent kedusha. Clearly, the gemara is understanding that since it looks like a donkey, if it were to have kedusha (because it has similarities to the mother), it would have קדושת דמים like any donkey, not קדושת הגוף like it's mother.