needrain said:Couldn't the larger seeds simply be an effect of fewer seeds in the pods? Seems like a pod only growing four seeds would be likely to have larger seeds than a pod growing thirty seeds.
Yes, there can also be a relationship between the number of seeds and the size of the seeds.
The basic, perhaps most important part of the relationships involved in seed size would be that of the pod parent's genetics. A cultivar used as the pod parent will have a characteristic basic average
One factor that may change that average size of the seeds would be the plant's nutritional status, health, physical size, etc.
Another factor (factors are in no particular order) that may change the average size of the seeds would be the number of seeds in the pod.
Another factor that may change the average size of the seeds might be how many other pods were being developed at the same time on the same scape. Other pods may have stronger or weaker effects on each other depending on the stage of maturity/development of the other pods and their seeds and their exact position on the scape (for example, same branch or different branch, above or below other developing pods, etc.).
Moving up the scale from individual plants and scapes to more general factors, ploidy differences would also have a general effect on the average size of seeds.
All, or some of the factors might affect the average seed sizes of particular cultivars as pod parents and possibly, which factors had a larger influence on the seed sizes in specific pods might differ from pod to pod even on the same plant. Usually I would expect that ploidy, the specific pod cultivar and the number of seeds in a pod would have the most influence on seeds sizes.
The effect of ploidy should be relatively easy to check. One would need to examine the seed sizes of a few converted diploids (to tetraploids) and their original diploids versions growing in the same conditions. One might have a few quibbles since the pollen parents could not be the same, etc. If the diploid and tetraploid versions of each cultivar also had different average numbers of seeds per pod then one might have to weigh the seeds (as a group per pod) to see if the weights of seeds per pod were the same between diploids and tetraploids). Or one might have to compare the seed sizes and their numbers between diploids and tetraploids to see if there was only a simple relationship between seed number and seed size across ploidies. That is, was the relationship the same in both, for example when there were four seeds in a diploid pod they would be the same average size as when there were four seeds in a tetraploid pod and so on for all numbers of seeds.