The BET surface area of fumed silica decreases when treated hydrothermally. However, the BET surface area of doped alumina on silica after hydrothermal treatment remains relatively close to its initial value.
I'm not sure I understand your question, but fumed silica easily aggregates (and forms strings - hence its use as a rheological modifier in toothpaste, for example) reducing the surface area (and the higher the temperature then the greater the aggregation). If the alumina is held on the silica then it's sterically bound and cannot aggregate. See slides 9 and 10 in:
Is the role of the alumina to sterically hinder the silica to prevent aggregation or does it alter the functionality of the silica, eg acidity or to influence the surface properties?
When we say it's bound onto the silica does that mean it hasn't interrupted the structure just the surface? Are there cases where alumina can be introduced into the structure via doping e.g. replace Si?
The surface is always different from the bulk in oxygenated systems - for example in air. Doping is a different case than applying a coating or having a mixture of components. There is a difference in size between the Si4+ and the Al3+ ions and there is thus lattice strain in the system when foreign ions are introduced. This is out of the scope of my experience, so I cannot comment knowledgeably on this topic. All I can say is that the properties of the coating reflect that material (and the support that material too) whereas with doping there'll be a 'fusion' of properties between the 2 different materials if there is homogeneity (i.e. no segregation or separation) and a single phase results.