Germination Statistics

I have been using MS Excel for a number of years to follow germination data. This year I have been playing with another statistic: Seedlings per Hip. When planting many seeds from many hips, the efficiency of producing a larger number of seedlings per hip is not easily determined without using a program.

My best seed parents with respect to this statistic are in the range of 15 to 20 seedlings per hip. Since these varieties also set hips with nearly every pollination, there is a great deal efficiency.

In the past, I have mainly looked at germination rates (number germinated per seeds planted of a given cross). When only a few seeds are produced per hip, efficiency goes down. That is true even if the germination rate is reasonably high. One of my favorite mini Hulthemia seed parents has a very good germination rate of about 40%, but with an average of only 2.4 seeds per hip, the germinations per hip averages about 1.0.

This will be an important statistic I will look at when evaluating new potential seed parents.

Jim Sproul

Seedlings per Hip…This will be an important statistic I will look at when evaluating new potential seed parents.

Isn’t this really only half the story, though? Seed quality is also important, and might be affected by the number of seeds per hip for any number of reasons.

For instance, all other things being equal the total amount of photosynthate being delivered to each seed is reduced in hips with many seeds relative to those with fewer seeds. Less sugars means less oils in the cotyledons, thus smaller seed size and less energy for growing large enough to emerge from the soil to be scored as a germination, or for surviving long enough to make the first selection cut.

I don’t imagine you bother counting seedlings a regular intervals after they germinate but you might have the number that get potted up. It would be interesting to plot that against seeds per hip on a cultivar basis, or run the data through multivariate analysis using, say, SPSS software.

A related question that I was wondering about is whether you treat smaller seeds differently that larger seeds when you plant them out for germination. Specifically, do you plant them all at the same depth or do you plant smaller seeds less deeply than larger seeds?

I also use an excel spread sheet to keep track of various pollination and germination data. I

I’ve been tracking seedlings per hip since 2000, along with seeds per hip, seeds per pollination, seedlings per pollination, kept seedlings per hip, kept seedlings per pollination, etc. All of the statistics have value in evaluating seed parents, but my favorite is kept seedlings per pollination. It includes the effects of hip set efficiency and seedling quality, along with the effects of seeds per hip and germination efficiency.

Based on kept seedlings per pollination, my best cross last year was

Apricot Twist X (Sonia X New Zealand)

which had 3 kept seedlings per pollination. Some other statistics from this cross:

5 pollinations

5 hips

39 seeds

7.8 seeds/hip

32 seedlings

82.1% germination

6.4 seedlings/hip

15 kept seedlings

3.0 kept seedlings per pollination


That is a good cross. Any time you can get three kept seedlings per hip, your doing great. It really helps when you get 82% germination rate though.

Having columns for kept seedlings per hip and per attempt are good ideas, because they tell you what kind of seedling production to expect from your effort and which seed and/or pollen parents produce the most quality seedlings. I

I also take some notes.

A surprising observation was for me, that hips containing only one or two seeds took several weeks longer before turning red compared to other hips on the same plant with more seeds.

Got to look more into this. I can think of two explanations- more seeds are probably more efficient in channeling nutrients into their hip, or the genotype of hip and pollen parent don’t work well together leading to low numbers of successful fertilizations and reduced vitality of the embryos.


I would suspect that hormones produced from the developing seeds would cause an absicion layer. The lack of seeds would also cause a lack of hormones. Just a guess!


Speaking of keeping a spreadsheet, I just moved up to an actual database (the SQL clone in Open Office) to keep track of my seedlings. Hadn’t occurred to me to keep track of seedlings per hip or seeds per hip, but then again, the roses I’m keeping track of are kind of young for the most part. Maybe it’s something I should add.

I’m keeping track of tiny little minutia like whether or not there are resin glands on the tips of leaf serrations (since I’m working with R. w. fendleri and R. w. ultramontana, and one has them and the other doesn’t, though I never remember which offhand).

The database in Open Office, by the way, is very nice, and easy to learn, if anyone’s looking for such a thing.

Don, I agree that there are definitely differences seen in the seeds themselves that may or may not contribute to germination rates. Regarding seed size, however, I have seen very small seeds germinate and thrive. I plant all seeds to a very shallow depth regardless of seed size and cover them with perlite. Because it can be so tedious, I only count the seedlings at 1 to 2 week intervals for about 3 or 4 counts. Because seeds continue to germinate, my data is never complete. What is interesting though is that variation times in germination become very evident with several separate counts. One batch may be finished germinating by the time they are counted for the first time, while others still have a fair amount of germinations remaining after the final count.

Paul, I think that you may be right that cultural issues might be the problem with your ‘Showy Pavement’ cross. I have seen some crosses do very well one year and then the same cross fail miserably the next. For that reason, I will try an important cross at least twice even if it doesn’t work the first year.

Jim, those are great stats on your ‘Apricot Twist’ X (‘Sonia’ X ‘New Zealand’) cross! Ultimately, whether we kept seedlings from a given cross is the final determination of whether or not the cross was a good one. For some roses though they produce fewer seeds per hip, or germinate at a lower rate, the extra effort is worth it because of the quality of seedlings that they produce. Most of my new seed parents are coming from my own seedlings. These are selected as possible seed parents because they produce hips well and germinate well. I will be paying more attention in the future as to how many seedlings per hip that they produce. Productive new seed parents that do not produce good quality seedlings will be dropped after one or two years.

Ulrike, I think that I have seen the same thing. I always wonder about the viability of such seeds, but haven’t assessed whether their germination rates are different.

Ryan, that sounds like a reasonable explanation to me!

Fara, for awhile I used the MS Access database program. I know what you mean about “keeping track of the minutia”. Today, I was reviewing some of my old notebook entries. Early on I kept track of leaf and stem color, early prickles, petal upper and lower surface color, and fragrance. I even taped the seedling petals directly into my notebook! That provided an important learning curve. Now, I am limiting my individual seedling documentation to photos with detailed descriptions of the survivors at the end of the year.

Jim Sproul

You guys have more attention span than I do – and I even use excel on a daily basis for other things! Hell, Ive created entire catalogs on it lol. But I dont think I could commit to that level of detail.

Jim, thanks for that information.

Fara, it’s interesting that you should mention resin glands on the tips of leaf serrations. I was potting seedlings today and noticed that there were resin drops dangling off trichomes spaced all around the edges of the cotyledons of one of them. I don’t know if that will translate itself to resinous plants but it made me think I should mark the seedling (I didn’t!). I have been marking seedlings with tricotyledens though.

Don, I’ve had three seedlings with tricotyledens this year and all have carried on normally so far.

On the other hand, I’ve had two seedlings, from very different backgrounds (one Golden Celebration x Fa’s Marbled Moss and one OP La Belle Sultane) which, instead of growing leaves, started making bunches of… what looked like stems of rhubarb! The GC/FMM has since started making normal leaves, but the other now looks like a miniature rhubarb plant. I’m keeping it because it’s sooooo weird.

Jim, yup, what you did sounds like what I would do. I may not keep track of this kind of thing forever, but it will be fun to see down the road if any correlations ever show up in the data.


I don

Yesterday I was looking at “gross weirdnesses” just to get an idea of if there was a pattern. I have four twin embryos, none separated yet, but I am taking photos-every pair has one that is not thriving, and one is. I had 3 albinos-they have all died, and then there is the 'produced a big, healthy set of cotelydons but just does’t want to grow-produced 1/32" of growth in 6 weeks-all from Rt.66 crosses, and then the rhubarb! I was trying to think what these look like, and all I could think of was monstrose, as in cacti, but rhubarb is right on. I have one that is absolutely growing great guns, and two that don’t want to get out the gate. And then the four (Juncus wanabes) curly/wirlies-kind of like every part has had a permanent wave. I’m assuming all these are freaks of chromosome incompatibility, although only the giant non-growing cotelydons and the albinos each have a commonality. Except for the twins, which I’ll try to separate next weekend (if it’s cool and overcast) and the really robust rhubarb(I have the compulsion to find out what is happening there), should I assume these are all destined for the big compost pile in the sky? I have been keeping track of these on an excel spreadsheet, but I think I have been spending to much time looking for the first bud, rather than recording details.