The received wisdom of gardeners and horticulturalists everywhere (I’m generalising, go with it for now) is that it’s pointless to grow fruit trees from seed because they don’t come true to type. (True to type means that the fruit of the daughter plant will be the same, or very nearly the same, as the fruit of the mother plant.)
There are exceptions – mangos and mot citrus produce what are called polyembryonic seeds, which means that there are multiple embryos in each seed, and only one is the result of sexual recombination – the others are asexually produced, and will grow into clones of the mother plant. And of course there are the fruit trees that don’t produce seeds at all anymore – bananas, some citrus, some grapes, and fruit breeders are working on creating reliably seedless apples (yes, this is possible)..
But broadly speaking, we don’t grow fruit trees from seed anymore. Unless we happen to be fruit breeders, in which case – we do exactly that. Which begs the question: why don’t the rest of us (at least the more patient among us) try it?
The heritability of flavour profiles, colour, and storing qualities are actually quite well studied for most major fruit species. There are a few obvious ones; for example the pale aril colour in pomegranates is, as far as we can tell, a simple recessive – with the standard deep red or pink colour being dominant; darker reds and the red-black colour of some fruit results from a complex of other genetic factors affecting levels of anthocyanins, but the white/pale pink arils are easy to breed for. Most traits, though, are complex and have a low % heritability. Apples, for example, have a 10 – 15% heritability for characteristics like acidity, juiciness, crispness, and sweetness, and only about a 30% heritability for astringency (useful in cider apples, but a negative for dessert or cooking apples). Fruit breeders use fairly high tech methods to test for useful characteristics, including looking for molecular markers for traits known to be useful (such as disease and pest resistance traits) but a lot of the mroe useful flavour traits are too complex to effectively test for using molecular markers and genetic analysis. Even fruti breeders are often reduced to crossing two likely parent trees and growing out the seeds to see if they produce good fruit – but fruit breeders aren’t just looking for pleasant tasting fruit, they’re also looking for something different from existing varieties.
The home grower isn’t generally going to be that picky. If your seedling apple tree produces fruit which is really very much like (but not identical to) a gala or fuji apple, that’s a net positive if you don’t need ot market the tree as a brand new distinct and different variety.
And the secret that geneticists and plant breeders don’t tell anyone is that the chances of getting reasonably good fruit from a random cross between two parent trees that produce good fruit is quite high. It might not store well, or ship well, and it might not be unique and distinct from any existing variety; it might not fruit earlier or later than other common varieties, and the tree may well be susceptible to disease (or it may be resistant). But the chances of edible fruit are really quite high. And every seedling tree which is allowed to grow to adulthood and join the breeding population enhances the health of the species (clone trees don’t add much to the species, since they’re all copies of the one individual and don’t add any genetic diversity to the gene-pool – which is one of the reasons that people try to preserve some of the older varieties, to keep that genetic diversity in the gene-pool).
So why don’t we grow more trees from seed? It’s not hard to do. Apple seeds will often germinate in the fruit, and will otherwise germinate quite happily if you put them between some pieces of damp cotton wool on a saucer in a sunny spot. Stone fruit pits should be put in the fridge for 6 – 8 weeks, sealed in a plastic baggie or a container with some damp sand so they don’t dry out and then planted out once the weather warms up. I haven’t had much luck with pomegranates (only 4 germinations out of nearly 100 seeds), but in theory you just plant the seeds directly into soil or potting mix in warm weather and keep them damp but not soggy. Citrus like the damp cotton wool or damp sand in a saucer and a sunny windowsill to germinate, but you can also plant them directly in a pot of soil or compost. In my experience, mangos like to germinate in the compost heap or worm farm (warm, humid, moist conditions), so just throw the pits in there and rescue the plants once they sprout. Loquats and guavas will self-seed if you let the fruit fall (which i don’t recommend, because of fruit fly – but you can plant the seeds in some good compost for the same effect). I don’t know about grapes yet, but I bet they grow easily from seed; most Mediterranean plants do.
Go forth and plant 🙂