Christopher,
Epigenetic regulation is definitely a “thing” in plants.
In addition, there was a genetic study of F1 “hybrid” corn where the two inbred strains carried different alleles at a specific locus. In some years, the maternal allele was expressed, but the paternal counterpart was silenced. In other years, only the paternal was expressed. And in still other years, both alleles were expressed.
In interspecific and intergeneric hybrids there are instances of maternal OR paternal OR both expressions in the F1 generation. Here’s a bibliography of some.
http://bulbnrose.x10.mx/Heredity/King/Election-Variegation.html
I must add that “variegation” also occurs in some of these cases, particularly where expression of both parental traits cannot “average out”. For example, the Trifoliate orange crossed with the common orange yields some few offspring that attempt an intermediate condition. Some leaves are simple, some trifoliate, and some have only one side leaflet (stipule).
Similarly, hybrids of raspberry (cap separates from core) and blackberry (cap remains attached) are often unstable in the expression of attachment: some caps slip off the core, some are attached, some are half-atached, etc.
We see this kind of unstable (or variegated) expression in some roses. The Brownells wrote, “An illustration of this is the variety Orange Everglow in which these two types of blooming habit are present and segregated. Certain confirmation lies in the fact that if the once-blooming cane-growth is not removed it may by its vigor smother and prevent the establishment of reblooming wood.”
http://bulbnrose.x10.mx/Roses/breeding/BrownellClimber1944.html
Beaton (1815) gave another example:
“No one seems to like Gloire de Rosamene for a bed; but by a particular management it makes a splendid bedder, indeed the very richest of all the roses. For bedding, this rose should be treated as a biennial, and no more; that is, to put in cuttings of it every year in April (they will root anywhere, if you stick them firm in the ground), and to plant them in the flower-bed next March, or whenever the bed is ready for them in the spring. Then, from the first of June to the end of August, every shoot which looks very strong, and is likely to run away with the sap, as gardeners say, must be stopped when it is six inches long. In this way all the shoots over a whole bed need not differ much in strength, and they will not stop from flowering in July or August, as this rose is apt to do when older plants are used. After the beds have done flowering in December, the plants must be disposed of, for all the gardeners in the country could not make a regular bed of them the second season”
http://bulbnrose.x10.mx/Heredity/Beaton/BeatonBedRoses1851.html
And we should not forget the unstable Rosa foetida bicolor/lutea that changes from one version to the other at long and irregular intervals. I have read of another variant that was even more unstable, switching between lutea and bicolor repeatedly in each flower, giving striped or sectored petals. Any striped rose is an example of this instability.
Some further reading:
http://bulbnrose.x10.mx/Heredity/Serra/serra3.htm#B
http://bulbnrose.x10.mx/Heredity/King/Epigenetics.html
http://www.cell.com/cell/fulltext/S0092-8674(05)00653-7