Ttdfjt

Putting style above content and reducing the ability to find the paper does not help anybody. My experience with copy editors is that they make and introduce far more errors and problems than they solve.

Anyway, I would discuss this with the editor that accepted your paper and give him/her the scientific reasons. This might help more than discussing the issue with the copy editor.

I have had this issue in the past. As a first step, look through past issues of the journal and see if they consistently apply their style guide. If they don't, provide them a few references to their articles that use your preferred version. If they consistently apply the style, your battle will be harder. Go through the articles in your reference section show them that your usage is preferred. Finally, provide them references to highly cited articles in other journals that show your usage.

Personally, I'd push-back on the copy editor's choice of "ice-berg". If you're unable to convince the copy editor, then as other's have mentioned, speak with, or exchange email with, the editor for your submission, if that's a different person than the copy editor. If you don't get satisfaction from the editor, you can raise the issue with the publication's Editor in Chief. Obviously, your final recourse is to withdraw your paper. Only you can determine how much of an issue this is for you and how far down that road you want to go.

In discussions, I usually find that it's important to have evidence to back up your choice.¹ It would be good to have a sampling of papers in your area showing which version of "iceberg" vs. "ice-berg" is predominately used. I also find that for this sort of discussion, it's often convincing to use information from Google Book's Ngram Viewer (info).

For "iceberg" vs. "ice-berg", Google Book's Ngram Viewer shows that "iceberg" was used 344 times more often than "ice-berg" in 2008, and has been the dramatically predominant form, at least in Google Book's corpus, for more than 200 years:

You can also look "iceberg" up in various dictionaries. All of the ones I checked didn't even give examples of the hyphenated version.

^{1. Something just being your preference is also valid, but that's not the case for this issue.}

Don't worry about it. If your paper is on a topic suitable for the ArXiv, just use your preferred spelling in the ArXiv version, which is more likely to be found by google search anyway. This way people are likely to find your article regardless of which spelling variant of the keyword they use.

You might consider voicing your concern to a member of the editorial board -- probably whoever handled your paper.

If they agree with you, then they will probably contact the journal on your behalf and request that your preferred spelling be allowed to stand. Conversely, if they think that your concerns are unwarranted or unnecessary, then you should probably drop the matter.

I suggest not worrying about it. First, even if a clear majority of the scientists in your field prefer "iceberg", the journal still has to stick to its own style. Second, it's not like people will confuse "iceberg" with "ice-berg".

If it really bothers you, then there's no point arguing with the copyeditor - they don't control the journal's style. You will have to convince the editorial board. Contact the editor who accepted your paper; he/she should know what to do.

I strongly advise allowing the journal their style, since you already put up an argument.

A physics journal change $K$-theory to K theory in one of my papers. I investigated, and found the same publisher did the same thing to a Fields medalist. At that point I figured I had a funny story to tell and let it go.