The study’s core interpretation, using the D/P (dicot/grass) phytolith index to infer shifts in olive tree cover and agroecosystem history, is fundamentally compromised by a lack of taxonomic distinction. You explicitly state that no diagnostic phytolith signature for Olea europaea exists, and woody dicot phytoliths (spheroids, blocky, jigsaw) are largely indistinguishable among taxa. Yet you repeatedly interpret peaks in the D/P index as indicative of shrubland, arboreal vegetation,or xerophilous taxa without any mechanism to separate Olea from Pistacia, Quercus, or any other co-occurring woody species.
The Bosco Pisano calibration does not resolve this. It merely shows that a wild olive woodland yields certain D/P values. But because your index cannot distinguish Olea from other dicots, those values could equally represent any mixed shrubland or woodland. In Cozzo del Lampo and Malia, peaks in D/P are therefore uninterpretable as evidence for olive-specific land use—they could reflect natural Pistacia scrub, oak regrowth, or any other non-olive vegetation responding to climate or abandonment. Without a diagnostic Olea phytolith, your central claim of tracing historical olive agroecosystems over millennia is based on a proxy that cannot differentiate the target species from background vegetation.
How do you justify interpreting D/P index peaks as evidence for olive-related land use when your own methods state no taxonomic demarcation for Olea exists, and co-occurring Mediterranean shrubs (e.g., Pistacia, Phillyrea) produce the same phytolith morphotypes? What specifically in your assemblage excludes non-olive woody dicots?
Bosco Pisano is used as a calibration for tree cover density, yet it is a wild olive woodland. Given that grass phytoliths still dominate there (average 9% dicot phytoliths, vs. ~34% grass short cells), how can this serve as a reference for olive cover when the D/P values reflect the overwhelming grass signal, not the olive trees themselves? If olive trees are minor phytolith producers, what exactly is your index calibrating?