Elements are found in different minerals and chemical compounds on earth. Oxygen is dominantly found in silicate minerals which, aside from silica, contain other elements. One way of looking at the chemical form of oxygen is to look at its bonds within the compounds. Is oxygen dominantly bonded to silicon? If so by how much and to what other elements does it most likely bond to?
This answer is basically an expanded version of my comments on this answer here.
If you consider the crust and the mantle, then the most abundant metals are magnesium, silicon, iron, nickel, calcium, aluminium (in that order). Other notable elements would be titanium, chromium, and just about all lithophile elements. The key is to understand that any single oxygen is not exclusively bonded to a single metal. It's bonded to several (different) metals at the same time.
There is an important point to be made here: because we call the minerals "silicates", it's often thought that most of the stuff in there is silicon (and oxygen). In fact, there is actually more magnesium in most of the Earth. Ideal Mg-olivine (forsterite) has double the amount of magnesium than silicon. Enstatite, an orthopyroxene, has the same amount of both elements. It is a common misconception to think of oxygen as being bonded to silicon while the other metals just "hop along". Silica, as SiO2 doesn't actually exist. Not in quartz and not in other silicate minerals. It's not like CO2 that exists as an independent molecule. It's just a convenient, charge-balanced way of representing oxidised silicon that has no existence if there are no oxygens around it. These oxygens are bonded to this silicon cation, but they are also bonded to everything else around it, be it magnesium or iron or whatever. An olivine is not a solid with molecules of SiO2 and MgO, it is an (almost) infinite solid with oxygens bonded to both Si4+ and Mg2+ in a three-dimensional structure. It's just a pile of cations (metals) and anions (oxygen), that happen to conform to combinations of "conventional" oxides because of charge balance.
As mentioned, if you mean Earths crust, then it is bonded mostly to Si and Al (quartz, feldpspars, then other silicates, water, iron oxides etc.). If you mean a whole Earths body, then crust is just a minuscule part compared to mantle - the details are in the previous response. We think that mantle consists of olivine, spinel group etc. Actually recent research suggest much higher water content in the mantle then previously anticipated. But keep in mind that we have no direct evidence of what really is inside the mantle or core.