As a general academic principle you should not adjust the titles -- or other essential bibliographic data*, e.g. the form and ordering of the authors' names -- of papers that you reference in any way, with the possible exception of necessary typographical concessions for "exotic" characters.
(For instance, in my branch of mathematics is it common to use the cyrillic letter Sha for something named after a Russian mathematician whose name begins with this character. If you simply don't have cyrllic characters available, you have to do something else. But even in this case it would be better to try to take a little trouble to "get cyrillic characters".)
So my answer to your question is no: it is obviously not your fault that different papers you cite refer to teals in different ways. So if there is any untidiness and/or inconsistency, you are simply faithfully replicating it, as you are obliged to do.
This answer comes from someone with precisely zero avian-specific academic knowledge. So it couldn't hurt to get a second opinion from someone in your field.
*: I confess that I sometimes mess with "inessential" bibliographic data. For instance, most journals come with a volume number followed by a number of the issue within that volume, and I usually omit the second number: you simply don't need it to access the paper. I don't specifically endorse this practice and mention it only for "professional honesty", but in my defense I just checked around and I am far from the only one who does this.
Moreover in some fields there are different formats for referencing. In my field (mathematics), on the one hand we are not too picky about the format, and on the other hand we have standard repositories of all papers: MathReviews/MathSciNet and Zentralblatt, from which I assume that most contemporary authors simply copy the bibliographic data (and then the philistines among us delete the number of the journal within the volume). It is a good idea to choose a consistent format throughout all your bibliographic references: e.g. put last names first always or never, and so forth. But still: the title is not yours to monkey with, as far as I know.