A welcome for Dave and some questions...

Jennifer, this sounds like a toughie. Much I don't know about your situation, in terms of the "bacterial infection" which is most likely made worse or induced by pollution, etc. . . .

Are you sure the oaks are goners? How are their acorns, have you tried them? They may be worth trying to save . . . Just sayin'.

But, assuming they are goners, in successional terms you have a disturbance happening. What legacies will it leave? That is, what else is there that looks like it will survive the infection? ID and make lists of what is already there and will be left standing and likely take over after the oaks die. Do some niche analysis (AKA needs and yields analysis) of those species, see what is useful or functional, and what is less so. That will be a basis for some of your decisions. You'll also need to spell out your own design goals--which you have already, a little ("I'd rather let nature do it her way"--even though there is no separation and we are nature/nature has become us humans, and the probability is that human effects are supporting or causing the infection . . . sorry, just need to point these things out--I'd say the principle of Shifting the Burden to the Intervenor means that we have to step in to help heal the damage we have helped cause . . . we have an urgent role to play to heal the systems we are messing up--anyway . . .). What does "an effective forest garden" mean to you? Its different for everyone, what that means. Define that please, and that will help you get clear on what questions to ask of and about the system with which you are interacting, which will teach you where to go with your interactions. The goals guide the site analysis and assessment, and the site A & A discovers the design.

I talk in EFG about "patch dynamics" and gap succession (vol 1 chap 6). These may be useful constructs for your situation, unless the dieback is more wholesale than gaps here and there. Gap, clearing, or glade? What size are the holes being left in the forest matrix? This determines much about the successional dynamics that will follow. That and the legacies of remaining trees and shrubs, what foresters call "advance regeneration". You could plant patches of fruits and nuts into gaps or clearings the woods, get these patches self-maintaining, then move to the next patch and establish that. You could more simply guide the succession by further removals of what is left behind after the dieback, encouraging some species or individuals and discouraging others. You could do both in different areas, depending on accessibility. You could also use fire (common in the pine barrens in days of yore, at least, and a primary land management tool all over the western hemisphere before Columbus) or clearing to further disturb patches so you have an easier time establishing the kind of food forest you want, while you get a yield from the clearing work (lots of labor though!). Given what you say, this latter is likely not to fit your goals and or values, but it is worth considering for some areas, probably. Your strategies will likely vary patch by patch.

I also should warn you about safety issues when you have a lot of standing dead trees--beware, your health and life are at stake, and those of your loved ones! And the life and health of any trees you may plant into the area before the dead or dying trees come down, and any nearby buildings or infrastructure. Please do factor that into your considerations. Areas where risk is highest are candidates for clearing work and intensive design and replanting . . .

Sandy acid soils: makes me think of chestnuts, chinkapins, hazels, persimmons, pine nuts, blueberries, raspberries and other rubus species; sweet fern, northern bayberry or sweet gale for n-fixation, perhaps some of the floodplain species if you have any wet areas/seasonally flooded zones (pecans, pawpaws, walnuts, hackberry, etc), black locust is probably a must for n-fixation and for coppice or timber production, mulberries may do well if its not too dry. Coppicing may also be a good strategy in general for the oaks; who knows cutting the oaks may actually rejuvenate them, it does in many contexts. You may find the infected trees resprouting anyway . . . Depending on what this infection is, there are plenty of oaks you could plant that would have high value acorns that I would seriously consider if I were in your situation. Oikos Tree Crops (Michigan) has a nice suite of these to look at, but only if you are clear the infection will not spread to them (or to any other plants in the same family, for that matter--many of the species I suggest or that might be good for your site may be subject to the same disease because of family relationships--check that out!).

Take care, and good luck!

d