Fluids, whether slab-derived or not, and even any additional phases, thermodynamically tend to lower the temperatures at which partial melting (and also crystallisation) is commenced. This is a result of depressing of the liquidus and solidus with the eutectic points. In subduction zones, this is typically driven by addition of the H2O- and CO2-rich fluids which are released from the descending hydrated crust at ~100 km (amphibole breakdown to prx). This is the basic of the subduction magmatism. This process do not require any reasonable increasing in the temperature of the mantle wedge. Also, when any mineral phase in the source (e.g., cpx) is consumed, melting may be ceased in response to re-increased melting points. You can find further in some books. Anyway, I will send you some lecture notes.
I entirely agree with the reply given above by Prof. E. Ersoy. It is the addition of H2O and CO2 that helps in lowering the eutectic melting temperature that causes partial melts extremely rich in volatiles that give rise to highly explosive eruptions.
Please find enclose paper. The paper is going to help you for interpretation to the your analysis results using by some trace elements and isotope compositions.
"The addition of small amounts water to peridotite will result in a decrease in its melting temperature. This is largely due to the electrically polarized nature of a water molecule (there is an unequal distribution of electrons around the water molecule). The electrical polarization causes a decrease in cation-anion bond strengths within minerals, and so at very high temperatures the bonds may be broken so that atoms may move freely from one another to form a magma. This process results in partial melting of the mantle rock." (http://mtweb.mtsu.edu/cribb/100magma.html).
Thank you very much all, and especially Dr. Hasan Çelik.