The initial question here is whether reserve in the printer’s manuscript is a mistake for preserve. The nearby reading in Alma 18:2 suggests that preserve is correct:
In Alma 17:31, the 1849 LDS edition replaced reserve with preserve, and this reading has been followed by all subsequent LDS editions. But the RLDS text has kept the earlier reserve, despite its obvious difficulty for modern readers. In support of the 1849 emendation, there is one clear example in the text where Oliver Cowdery misread preserve as reserve:
Here Oliver misread preserve as reserve when he copied from 𝓞 into 𝓟. We might conjecture that the same error occurred in Alma 17:31. If so, the 1849 LDS emendation to preserve restored the original reading.
On the other hand, the earliest extant reserve will work. The Oxford English Dictionary lists under definition 7 for the verb reserve the meaning ‘to retain or preserve alive’; this meaning is identified as “now rare”. All but one of the OED examples listed under this meaning date prior to 1650 (in the following, the accidentals have been regularized):
The 1848 citation refers to the traditional Christian view that John the Beloved did not suffer death. In fact, Jameson’s own use of the verb reserve in this sentence is preceded by the same language in a citation from an earlier source, not identified:
Here the passage is given in its original accidentals; see page 139 in volume 1 of Anna Jameson, Sacred and Legendary Art (London: Longman, Brown, Green, and Longmans, 1848). Note, in particular, the original quote marks in Jameson’s citation.
From the OED as well as Literature Online and , we obtain a few more examples in Early Modern English of the specific phrase reserved alive, including the following ones after 1650 (here I maintain the original accidentals):
There is also an example from the early 20th century, which shows that this phrase, although rare, has maintained itself:
But perhaps the most interesting example is from William Whiston’s 18th-century translation of Josephus’s War of the Jews. Within the same passage (book 7, chapter 8, section 7), Whiston used both reserved alive and preserved alive (just like the earliest Book of Mormon text does in Alma 17–18 with its instance of reserve the flocks followed by preserving his flocks):
The original OED editors (working from 1884 to 1928) considered the older meaning of ‘to preserve’ for reserve as “now rare”. In addition, the 1849 LDS change to preserve in Alma 17:31 is a clear sign that in general reserve had lost its meaning of ‘to preserve’ for either Orson Pratt (the editor of the 1849 LDS edition) or the British compositor in Liverpool who set the type for that edition. In general, the historical information, especially the Whiston quote showing variation, argues that the verb reserve with the meaning ‘to preserve’ may very well be intended in Alma 17:31.
There is, however, another possibility for this verb in Alma 17:31. Earlier in verse 29, Ammon silently considers how he might take advantage of the difficult situation resulting from the scattering of the flocks:
This earlier passage strongly suggests that verse 31 originally read as follows:
Under this emendation, both references to getting back the flocks read almost identically. Moreover, the emendation suggests that Ammon’s words to his fellow servants (in verse 31) are directly based on what he had just thought of (in verse 29). And finally, it is quite possible that an original restore could have been misread as the visually similar reserve, either when Oliver Cowdery copied the text from 𝓞 into 𝓟 or earlier when Joseph Smith dictated the text. In fact, we have an unambiguous example in the text where the verb restore was misread by Oliver as he copied from 𝓞 into 𝓟:
This misreading of restored as raised shows that a similar misreading of restore as reserve is quite possible, although in this case the raised was prompted by two previous occurrences of that word in Alma 41:4 (for discussion, see under Alma 41:5).
Ultimately, the virtual identity as well as the logical connection and proximity of the phrases “restoring these flocks unto the king” (verse 29) and “restore the flocks unto the king” (verse 31 as emended) argues that the original verb in verse 31 was most probably restore, although reserve (the earliest reading) and preserve (the 1849 LDS emendation) are also possible. The preposition unto seems to work especially well with restore, just as it does in the following passage in the book of Ether that refers to other possessions being restored to the people of Morianton:
Also note that with the verb preserve, when it is correctly used in Alma 18:2, does not use the preposition unto (“and he had learned of the faithfulness of Ammon in preserving his flocks”). Thus the critical text will accept the conjectural emendation restore in Alma 17:31 since “restore unto X” works the best.
Summary: Based on the nearby reading in Alma 17:29 (“in restoring the flocks unto the king”), Alma 17:31 should probably be emended to read “we will restore the flocks unto the king”; the earliest extant reading, reserve, is also possible (given its archaic meaning ‘preserve’), as is the 1849 LDS emendation preserve (especially in light of the accidental change of preserve to reserve in Alma 37:18).