I am currently tackling this question myself in my thesis. The challenge with ecological networks is that they can be hard logistically to characterize empirically in such a way that is efficient enough to be usable for management without requiring years of empirical research. We are developing ways to predict the spatial structure of ecological networks (i.e. the spatial distribution of interactions) by combining species distribution models with pairwise species interactions predictions. The end goal is to allow us to explicitly use network level descriptors in a regional cumulative impacts assessment framework in the gulf of St. Lawrence in eastern Canada. I'd be happy to talk more about it and there are more details available in the project describing my thesis and the attached documentation. The poster describing the thesis gives a brief overview of how we are tackling the problem in our context. 

https://www.researchgate.net/project/Evaluating-the-cumulative-impacts-of-human-activities-on-the-structure-and-function-the-trophic-network-of-the-estuary-and-gulf-of-StLawrence

Poster Cumulative impacts on the communities of the estuary and the...

If you plan to quantify ecological networks at a broad scale, try developing spatial assessment using software like FRAGSTAT. You can then use GIS to assess potential land use/ land cover changes to evaluate loss in connectivity and subsequent changes in populations and community structure. Another option is to evaluate metapopulation dynamics using some dynamic models and evaluate impacts of habitat and subsequent biodiversity values. As David mentions, it is a complex process and narrowing your focus to specific questions and hypothesis might help you in this research.