Are all crystal-bearing magmas actually Newtonian?

Crystal bearing magmas have long been known to be non-Newtonian, exhibiting clear shear-thinning features when exposed to shear stresses. However, the micromechanical origin of this shear thinning remains enigmatic and attempts to describe how shear thinning arises in magmas have been equivocal. Here, we demonstrate that in controlled experimental systems, shear thinning is a non-local and scale-dependent artefact of crystal organisation during flow, and is not therefore an intrinsic property of the crystal-bearing magmas sensu stricto. Furthermore, we show that most experimental approaches to crystal-bearing magma rheology sit in a regime in which crystal migration effects will be dominant, explaining why the experimental evidence is that non-Newtonian effects are observed. We use a numerical conduit model for crystal migration physics to demonstrate that in nature, volcanic systems sit in another regime altogether and will not organise crystal cargo on the time- and length-scales of magma ascent. This leads us to tentatively conclude that crystal-bearing magmas on Earth are Newtonian, and that the only non-Newtonian effects of concern relate to bubbles at moderate capillary number, and the melt phase at moderate Weissenberg number. Finally, we note that this goes some way to unify the mismatch in effective viscosity between low-temperature analogue experiments and experiments on natural crystal-bearing melts.