This is more of a question than an idea.  Are we going to focus on a particular ecosystem service within this theme or is that also something we need to decide?  I guess the one that I immediately thought about when when thinking about linear habitats is biodiversity, but that may just be because that is one of my primary interests.  They of course can also be tied to other services like pollination, natural enemies, water quality, aesthetics, etc.  So, will we be considering all of these, one of these or what in our project? 

I favor exploring how linear habitats can be implemented to increase biodiversity (and whether that contributes to ecosystem services or other goals you may have).  You can approach the aspect of habitat management on many scales, from management of an orchard or other agricultural system, to even deciding which shrubs a homeowner might want to plant in the yard.  I wonder if we can tie together landscape horticulture and agricultural systems.  It seems that species interactions of pollination, predation, etc should be similar, even in these different settings.

How about looking at the relationship between linear habitats and biodiversity at an even larger, regional-level scale? For instance, we could look at factors such as total area in corridors, connectivity, mean corridor length between "natural" habitats (i.e.-forests), and so forth across a range of landscapes and see if we can't find some generalities. This might lend itself well to a meta-analysis...

Another thought, and not necessarily exclusive to the first, might be to ask whether certain types of corridors could be negatively impacting biodiversity by acting as vectors for the spread of exotic invasive species (seems plausible in temperate systems, although I have no idea if such patterns are reasonable in the tropics).