The setting for my story is a very foggy and windy place. There is a plant that grows on relatively short trees that needs to attract animals to spread its seeds, which are inside its fruit. Because of the wind, smell is not a good way to attract animals as it will be hard to locate the source and it will be quickly diluted. It is often easier to see light than color in fog. So, could the plant have bio luminescent fruit? From my research, it is my impression that the "easiest" way to have bio luminescent fruit would be to have the fruit contain or in some way host a population of bio luminescent bacteria. However, when the fruit gets eaten, the bacteria will be in the animal and/or its excrement and no longer the plant. How can the plant either keep or get back its bio luminescence/bio luminescent bacteria to continue spreading its genes?
Foxfire is produced by a fungus feeding on decaying wood. A similar process could be elicited by some symbiosis between the fruit and fungus, with a coating which feeds the fungus. It produces an eerie pale green light (which I found unsettling when I first unexpected saw it in the middle of a dark night while on a camping trip in the deep woods of West Virgina.)
Well something about the plant attracts the bacteria in the first place. Perhaps they are in the soil?
Also, why wouldn't the plant spread its seeds using the wind?
EDIT: In retrospect, you gave yourself the answer. The bacteria are in the excrement . . . and so are the seeds! They are co-located. It's not a big leap to say they stay that way.
As already pointed out, your set-up has serious problems with internal logic. The most obvious is that if the wind is that "strong", then why are insects (etc.) needed at all? (The wind should be quite suitable for spreading seeds.) The second is your claim that strong winds so dilute any "scent" (i.e. chemical signal) that finding the source would be difficult or impossible. You should ask yourself which is less effective: zero wind or a gale-force wind to do the signalling you want. The answer is "it depends". The stronger the wind, the farther the signal travels, but IF it can be diluted (i.e. is a chemical vapor or aerosol) the more sensitive the detector will have to be. (We can easily detect dust from Africa blown across the Atlantic into S. America. The idea that a tree couldn't develop a signalling mechanism effective in strong winds is, I believe, wrong. (Think of a message in a bottle drifting in the ocean, was the message lost because it was diluted? even after being in contact with millions (billions?) of gallons of water, the message isn't confused (but it does have to be found/detected).) Anyway, to answer your question. The only reason a symbiotic signal would be preferred to a inherent one is that it is more "efficient". You could make a case either way, there is no right nor wrong here, although would bacteria, say, really want to develop a signal for themselves to be eaten? If so, why? (survival of the fittest, of course, but more specifically, so that the bacteria could multiply - implying that they wouldn't be (effectively) digested/killed when eaten.) Using the rule of KISS, (keep it simple, stupid), I'd go with an internal signal - you'd need one anyway to signal the bacteria (or fungi, or algae, or whatever) to light up.
The pollination is carried out by small insects (like ants) which are "infected" by the bacteria. The ants glow in the dark as well and need the plant to protect them, feeding solely off the sap and pollen. The bacteria could infect the stigma of the flower which could in turn infect the newly fertilized fruit.
The ants might get to the plant on the backs of a ant-eating bird when they swarm over the bird to protect the colony. The bird lands on a new tree. The ants fall off and since there are no new ants they start a colony. Might be better without a queen-dependent insect, but that would be the general idea.