It is generally considered that significant deviations of zircon δ18O values from the mantle value directly or indirectly result from intra-crustal recycling. Intra-crustal recycling is defined as the interaction of magma with supracrustal materials that ultimately inherit their evolved δ18O values from low temperature processes on or near the surface of the Earth (Valley, 2005). I don't know why elevated δ18O signals can only screen out recycling of altered surficial rocks at specific paleo latitudes. I mean elevated δ18O signals can screen out recycling of altered surficial rocks, which is not necessarily at specific paleo latitudes. What do you think about it? Thank you.
You don't seem to reference who said this (or any full reference), and it is difficult to understand what is being said without its context. What paper says "elevated δ18O signals can only screen out recycling of altered surficial rocks at specific paleo latitudes", as you do not mention it in your previous sentence. Is this Valley 2005, and what is the full reference?
The only latitude dependence I can think of would relate to the temperature-dependent variation of δ18O of meteoric water with latitude (and altitude, but that is surficial anyway). This is in turn relected in things like hydrothermal alteration of near-surface rocks by heated meteoric water, where altered rock δ18O compositions show some latitude variation. One would expect a similar trend in regolith alteration by weathering. and rocks undergoing late diagenesis in meteoric-dominated sedimentary sequences in basins. Conceivably one could have surficial altered rocks at certain (probable near-polar?) latitudes giving a δ18O component to magma the same as that of "primitive" δ18O, so that apparently "primitive" had in fact a significant surficial surficual component . However I am surprised that one could actually use this approach directly for specific magmas except in young situations where δ18O of meteoric water likely to be involved could be interpreted, rather than just using it as an argument perhaps for a general broader change with time because of crustal evolution.