The Tropical Fruit Forum

I read a paper written way back that said the Citrus x  P. trifoliata F2 is less winter hardy than the F1., and that hardiness is less in each generration.
I think I know the explanation. 
In the  F1, all the genes for winter hardiness are there, but only one copy of each.  We all knew that.
In the F2, 25% of those genes are there with 2 copies.  Good.  And 50% of the genes are there with single copies like the F1.  No further loss there from the F1.  But 25% of the genes for winter hardiness are not in a given F2 plant.  And in the F3 25% of the genes  hetrozygous in the F2 are gone, and so one.
It is not just that genes are being lost, the genes have co-evolved and some of the genes remaining won't work well without the genes that are gone. 
If I just worked selecting in the F2, F3, F4, etc., selection won't be effective on those genes that require those genes that are missing.  So selecting only in population 5 won't be effectively selecting for all genes involved in winter hardiness.  Only as some of the genes that have some effect alone are selected will the others start becoming effective and be acted on by selection.
That is why I'll backcross the F1 to P. trifoliata, to have more of the hardiness genes present.  This will allow selection to work on those genes that need other genes to interact with.  A second backcross, Pop 2, may be needed to get the majority of genes available for selections.  That remains to be seen

In the same way,  flavor is the result of many genes interacting.  Population 8 will at first only be able to use cold hardy genes that have significant effect without all those other genes which will be missing,  But it will have all the genes necessary for good flavor, plus some genes for bad flavor which will be discarded.

I've started a citrus breeding program about 3 years ago.  Goal is hardy edible citrus for zone 6.

The plan is to have 9 breeding populations, based mostly on P. trifoliata, oranges, and mandarins.  The populations are pop. 1, pure P. trifolata;  pop 2, 7/8 P. trifoliata 1/8 Citrus species, most orange and mandarin;  pop 3, 2/8 , 6/8 Citrus;  pop 4, 3/8 Citrus, etc, up to pop 9, 8/8 Citrus.
Pop 1 is to be only monoebryonic P. trifoliata x precocious P. trifoliata.  If the F1 isn't monoembryonic, then, it could take a couple more generations to give a monoembruonic, precocious P. trifoliata.  This will be useful in later breeding.
Pop9, at the other extreme, will be to produce good-tasting oranges or mandarins with as much hardiness as possible without poluting it with P. trifoliata.  Others are working on this, so I won't be working on it.
Pop 5 is where my energy has been so far.  Half P. trifoliata, half citrus.  I have about 25 Sanford open-pollinated seedlings, and 2 US 852 seedlings.  I have been promiced some open-pollinated C. ichangensis x P. trifoliata seeds. I have some other citranges which I'll be using for pollen, so my gene base will have some diversity. I am working mostly on this population at first because until these bloom, I can't do selection or crossing to make other populations.  Plans are to select only for precocity, monoembryony, cold hardiness, and flavor.
Flavor selection will only be against P. trifoliata  flavor.  It will still be too sour for most people, but until the P. trifoliata flavor is dealt with, lack of sweetness is a minor problem.
Selection for cold tolerance will be done by keeping cuttings in a freezer for a week and see which survive.  Each population will have its own temperature. which will change slightly in each generation.  I'm doing experiments now to see what temperatures I'll be starting with.
Precocity will be easy to detect, I think, as well as monoembryony.  I have read that one (or more) citrus variety makes monebryonic seeeds via appomixis.  I'll have to watch  out for that by checking that seedlings aren't too uniform.
When population 5 blooms, crossing it with populations 1 and 9 will give populations 3 and 7.
If any of population 3 can survive here in zone 6, then populations 1 and 2 will not be needed.  But I'm not counting on that.
If any of population 7 has good fruit, population 8 and 9 won't be needed.  But I'm not counting on that either.  I have been encouraged by  reading of Dr. Brown's work with cold hardy citrus.  Some of his would be like population 7.  But I haven't tasted any of his citrus, so I don't know how accurate the reports are.
Population 7 would be the place for Troyer citrangequat, which some people enjoy.  I read that its seedlings don't do well, and I am not planning to use it.  Comments about this are welcome.
Crossing populations 7 and 9 give pop. 8.  Some of these should be edible, if not really good.  I'm prepared to select for flavor for a several generations.  And selection for cold tolerance for several generations might give measurably more cold tolerance.
Crossing populations 7 and 9 will give pop 8.  I expect some to taste quite good even in the first generation.  Perhaps I'm too optimistic.  We'll see.  Likewise we'll see if 1/8 P. trifoliata will give any increased cold tolerance.
Crossing populations 5 and 7 will give population 6, 3/8 P. trifoliata and 5/8 citrus.  Selections will be as for pop. 5.
Crossing populations 3 and 5 will give population 4.    Selection will be as for pops 5 and 6.  Selections is most effective generally when genes are around 50% frequency.  So pops 4, 5, and 6 will give greatest improvement.  Pops 7, 8, and 9 will give good fruit sooner though and somewhat increase the range of good citrus.
Crossing pops 1 and 3 will give pop 2.  If pop 3 gives any winter-hardy seedlings, pop 2 will never exist.  And pop1 would be dropped.
I plan to grow out at least 200 plants per population per generation.  More would be better, of course, but space is limited.
Populations will be dropped, or better still, turned over to other people when they have served my purpose, or better still, as someone else sees their potential and wants to take them on.
Monoembryony should quit being an issue in 3 or 4 generations.
Precocity is very important and will always be selected for, but there will be a limit.
Flavor and winter-hardiness will need work longer than I expect to live, and I expect to live a long time.  That is why I'm hoping others with the means will take an interest in this.

I know that cold tolerance alone doesn't give winter hardiness.  A winter-hardy plant must become cold tolerant before the first hard freeze of winter, and maintain cold tolerance in spite of mid-winter warm spells.  But the lower number populations should include winter survivers because they are mostly P. trifoliata, which survives well here.