The petrography ....more so of textural attributes help you to differentiate the hornblende of an ortho or para amphibolite.

The zoning perhaps reflected by mineralogy can be made out by EPMA. twinning in grains also can be studied. certain trace elements like Ti, Mn, K may vary, typical of igneous hornblende,  I  think

Mineralogy wise the Orthoamphibolites contain primarily of amphibole, albite, with subordinate epidote, zoisite, chlorite, quartz, sphene, and accessory leucoxene, ilmenite and magnetite .

whereas Para-amphibolites generally have the same mineral assemblage as orthoamphibolites, with more biotite, quartz, albite, and based on protolith, more calcite/aragonite and wollastonite.

I think, the  whole rock geochemistry  will identify ortho- from para-amphibolites more clearly.

In some cases, the para-amphibolites from the area also contain microcline, garnet and/or scapolite. By the mineral assemblage, linked with some field criteria, is possible to distinguish between the two types of rock.

But I was wondering if the mineral chemistry of amphibole would change with those kinds of protoliths. At the same metamorphic conditons, etc.

Thank you for your consideration.

There are many papers on geochemical discrimination of this if you Google - minor element ratios based on elements like Cr, Ti, V, Ni etc versus Mg, Fe . Petrography will only help in a few cases, because in many cases they are very similar in gross mineralogy.

I agree with Dr Martin James Hughes: in many cases  only geochemical studies -including mineral chemistry of the amphiboles may   be used to obtain  well constrained parameters for discrimination between ortho and para amphibolites. Specially in high grade amphibolites.