magnolialover said: ... But sometimes neither is true.
Exactly. I think the question has to be taken pretty much on a case by case basis, or at least group by group.
In the case of my own cross L. leicthtlinii x L. maculatum (pod parent x pollen parent), because both parents are species, there is little variation in the F1 generation.
size favors leichtlinii
leaves are midway
flower aspect variable
flower size and shape midway
flower color definitely favors maculatum
Other crosses I have are not really mature enough for data to be conclusive.
----------------------------------------------------------------------------------
However, an interesting side note is a daylily cross I did of Hemerocallis 'Siloam Ury Winnifred' x H. citrina.
form favors citrina
size is midway
leaves favor citrina
flower scape midway
flower form favors Ury Winnifred
flower color midway, I guess you would say
----------------------------------------------------------------------------------
As I thought about the original question, the answer became more and more complex. And I found myself delving more into Mendelian genetics and logic, rather than so much actual knowledge of real outcome. I had written this yesterday, and now rereading it today, I don't see it as so relevant. Still, a few of you may find it interesting. For the rest of you not interested in this drivel, skip to the last paragraph.
1) In the case of a species crossed with a hybrid with greater genetic diversity, from the viewpoint of one-to-one correlations of genetics, the species would have the upper hand in progeny characteristics.
In a simplified example, all progeny would have a set of basic (x) characteristics donated by the species. But with matching genetics coming from a hybrid, this same basic set could be a, b, c, d or e, even though it is only (a), for instance, that is expressed in the hybrid parent. Thus the progeny would be:
xa or ax
xb or bx
xc or cx
xd or dx
xe or ex
Obviously, this favors the species characteristics, whether pod or pollen parent, as the set (x) is in every combination. whether this is true in practicality is yet another question.
(2) Using this same simplistic concept, with two hybrids of equal genetic diversity, there would be no favoring of groups.
3) Sometimes simple genetics from a certain clone is overwhelming in regards to the pollen genetics, so much so that even an expert might not be able to tell if a seedling is apomictic or merely a mimicking hybrid. This would favor the pod parent characteristics. Such an example might be with the Claude Shride Martagon lily. I am not sure if these genes with extra "clout" would have the same effect as a pollen donor. My guess is that in some clones the answer is yes, some no.
3) Apomixis has not been studied well in lilies to know what factor(s) trigger it. Their could be and probably is the capability in many species (and hybrids) generally thought never to do it, just as we have been finding out with bulbil production.
4) There is always the common "problem" of dominance that mires simple calculations.
5) Where incompatibility might factor into breeding possibilities, this releases another set of variables. A couple are:
(a) Even partial incompatibility could be a trigger for apomixis.
(b) Even though the embryo itself is a bred hybrid, at least some of the other parts of the seed are not. They come solely from the pod parent genetics. As you probably already know, it is the incompatibility of the hybrid embryo and the endosperm (solely from the pod parent) that often prevents successful natural germination. So this favors hybrid embryos that are compatible with the endosperm, and therefore with more of the pod parent characteristics. (Remember, this is why embryo rescue was "invented".)
6) In polyploid lilies, it is generally known that a pod parent with x-ploidy, might be pollinated with x-ploidy or (x+y)-ploidy pollen, where y is a one (or greater) integer.
Dealing with only one allele in a chromosome, with an example of a triploid (pod parent) x tetraploid that yields triploid progeny: it's obvious that donor combinations from the triploid parent are far less than that of the tetraploid. Similar to (1) above, with the donor sets from the triploid parent being less than that of the tetroploid, characteristic would tend to favor the pod parent triploid.
In addition, all theses factors (and more) might interact with one another.
While the favoring of pod parent characteristics in progeny does seem to happen more often than not, the reasons are not always clear, and there is certainly no one general guiding principle that could explain the phenomenon. Since many factors (known and unknown) are at work, a generalization cannot be made for all different kinds of crosses because we would never know when, if or what factors are at work.
A fairer statement would be:
Don't be surprised if the progeny seems to favor the pod parent's characteristics.
