Weird question that brain-wormed me recently.
Hmm? I suspect it will probably be something that relies heavily on the seafood department.
Fun question to ponder.
By taxonomically diverse, do you mean the greatest number of individual species, or the greatest number of different kingdoms included in one dish?
For instance, would a salad with 20 different vegetables be more diverse or less diverse than a meat cooked with a vegetable?
For most kingdoms I’m going to go with leftover pizza. It covers (intentionally) plants, animals, and fungi, and (unintentionally) bacteria.
For greatest number of species, I’m thinking almost any intentionally fermented food. Wildcrafted sourdough, sauerkraut, crock pickles, kombucha. . . They all have many species of bacteria and yeast in varying quantities. I’ve probably eaten thousands of species just this morning, but only a few of those were large enough to be seen.
I would imagine there are some curries or other well-seasoned dishes that could involve a large number of species just of spices and herbs, never mind the main ingredients
if you are counting microbes, you want to choose a wide variety of hosts and substrates. it also limits cooking techniques to those which are less destructive (eg no boiling, no excess salt, watch the pH etc) to maximize the biota.
a lot of fermented food has only a few species in it. they’re safe and edible because we carefully control the setting in order to support the target, non-toxic species and restrict the growth of others.
Strict species diversity would be tricky to accurately judge, especially with spice mixes and microbial species. I dunno – I’d be interested nonetheless.
Cross-kingdom/phylum etc might be more challenging (and impressive) though.
I guess my answer would be: defend your choice!
I begin to wonder if the answer is every toddler’s favourite snack: a fistful of wet dirt!
Wait. . . You think we have to eat the species alive for them to count for the diversity? I was thinking cooked microbes would count, just like cooked meat or vegetables. I think I’ll pass on animals if they have to be alive. 
Restrict, yes, but not completely eliminate. I was considering both the main agents of fermentation and the many other species (including so-called “harmful” bacteria) that exist in harmlessly small numbers in naturally home-fermented foods.
In that case I’m going to argue (vehemently, because this is such an important topic) that their should be a points system. A dish gets 1 point for each species, plus 2 points for each genus, 3 for family, 4 for order, 5 for class, 6 for phylum, 7 for kingdom, and 8 for dominion. Plus a bonus half point for subspecies.
that’s a good point, not quite what I was getting at. my question is, at what point of destruction does a microscopic organism cease to be an organism under biology, and begin to be a mash of stuff under organic chemistry? carbonized bacon wouldn’t count as pig…
I guess I’d say it has to still be recognizable in some way. a few strands of denatured DNA and some cytoplasm don’t cut it
Gumbo.
Just looking at macroscopic organisms I think a bowl of ramen can have a pretty incredible phylogenetic span. Theoretically I think you could get all six times complex multicellular life evolved in there, e.g.
-Red Algae (e.g. nori)
-Brown Algae (e.g. kombu)
-Green Algae (e.g. aosa nori)
-Symbiomycotan Fungi
-Animals
-Land plants
In my dinner, which I just finished:
Vietnamese brown rice noodles with green tea (Poaceae, Theaceae)
Homemade chicken stock (Phasianidae, Rosaceae for apple cider vinegar, Acetobacteraceae)
Chicken (Phasianidae)
Several varieties of chiles (Solanaceae)
Wakame (Alariaceae)
Coconut (Arecaceae)
Leek (Amaryllidaceae)
Kimchi (Brassicaceae, Apiaceae for carrots, probably Lactobacillaceae)
Half a bagel to wipe the pot (Poaceae, Saccharomycetaceae, and sometimes Amaryllidaceae and Papaveraceae but not today).
ETA: FCLO (Gadidae and I don’t know what bacterium) and EVOO (Oleaceae)
The correct answer is the Every Meat Burrito. It is a joke from a cartoon but I knew some friends from Santa Cruz that made what I would call the Biodiversity Burrito. It had so many things.
I think that any meal that is subject to deterioration, and been left out in the heat for a while before consumption, would contain exponentially more diversity of species than a meal eaten when relatively “fresh”. I mean if you want the highest number of species consumed, the range of microbes, bacteria, etc., will be rather extreme if you let a meal sit out in warm conditions for several hours before eating it. This will far exceed any dish that depends on intentionally used ingredients for species diversity.
it will be difficult for any vegan meals to contend because of our penchant for brassica 
Arguably the microbes and bacteria that collect on an uneaten dish sitting out shouldn’t count as part of the meal.
lorb wrote:
“Arguably the microbes and bacteria that collect on an uneaten dish sitting out shouldn’t count as part of the meal.”
Why shouldn’t they count?
1.) The dish was already finished. Anything that is added after preparation has been completed is accidental and not part of it.
2.) They are not part of the dish in a conceptual sense. Like, “a hotdog” as a type of food surely contains some kind of sausage and a bun. An individual hotdog could be made with wasabi sauce and dried ants (for example), but these aren’t usually considered hot-dog ingredients.
3.) Only things that are intended for human consumption (with all their constituents) should count as proper food.
Edit: Human (and other species) hair and dipterans (flies) are known to occasionally appear in soup, should those count for the diversity score of soup?
Yes, absolutely.
Agreed.
Also, it’s important that whatever dish we land on comes from a public salad bar without a sneeze guard, so we can include as many viruses as possible.
I was wondering today how many species wildflower honey would add. The nectar might not really count once processed into honey, but there’s pollen in with natural honey too, and that’s definitely still identifiable to species. When honey from a whole apiary is mixed together during harvest, this could include trace amounts of pollen from most of the flowering plants in the region.
That is exactly the kind of thinking we need to get the overall species count of a meal up into the hundreds, or even into the thousands. Great thought!