I have been struggling with a problem for some time.

Namely: We have an XML source document that has a structure similar to this below.

<p xml:id="_1563890625404-000000640-004725783">Several states, including nine African countries, supported by EU Member States, wish to list all African elephant populations under Appendix I of the CITES. For the time being, only elephant populations in Central African countries<authorialNote placement="bottom" xml:id="_1563890625446-000000641-272312170"> <p xml:space="preserve">See Appendix I in the site of the Convention.</p></authorialNote> are listed under Appendix I, thus providing them with full protection. The remaining elephant populations (in southern Africa) are currently listed in Appendix II (which affords them strong protection but allows for trade in ivory). This double classification has the effect of encouraging ivory trafficking in Central African countries<authorialNote placement="bottom" xml:id="_1563890625468-000000643-943032708"> <p xml:space="preserve">www.traffic.org/.../p>

< authorialNote> is a footnote. As you can see, it appears twice in the one paragraph.

After "Convert to Translatable Format" and "Copy to Target Languages" paragraph is spitted into segments according to rules from Project Template. In my case is something like this:

As can be seen we have 8 segments. The problem is with segment 19 and 21 which are one sentence. They are split by the content of < authorialNote> which should be ideally at the end of paragraph or at least after the real segment. 

Is there any way to do it without manipulating with source document?