Sunday 15 May 2011

Foxgloves (15 MAY 2011)

In March 2010 I bought a packet of Foxglove Excelsior seeds for 78pence (0.88E, US $1.28, 3.85MR). At that time, I didn’t know that Excelsior was a man-made hybrid which produces erect flower trumpets around the entire circumference of the stem, and for this reason more sophisticated gardeners turn their noses up at it. At that time, I didn’t own a camera, I didn’t make gardening notes, I had no plans for carrying out experiments with plants.
I have never been disciplined enough to pay attention to instructions on packets. I didn’t sow the foxglove seeds thinly enough, finally I had a tray full of seedlings with a wide variation in size at my disposal. The strongest seedlings were planted outside first, in the most prominent locations of the sunny flower beds and at the forefront of the woodland area at the bottom of my garden. Some of the remaining runt seedlings, exhibiting a willingness to live despite their small size, were planted in empty spaces, as an after-thought. Photos of all ten specimens are listed in Appendix 1.
When I wrote my first post about foxgloves in November 2010, I was just glad that they were still alive. Their foliage alone had already started to feature strongly in my garden. Then in December, I was applying Design of Experiments to an engineering problem at work, when it occurred to me that this technique could be applied to the plants in my garden to understand in one growing season what makes plants tick, knowledge which would otherwise take a gardener years to learn (unless of course they take the trouble to read the instructions on the seed packet).
Simple experiments can be carried out by changing one factor at a time and taking measurements, starting from a base-line control.  Design of Experiments is a systematic way of measuring, in addition, the effects when (or if) factors interact together. Sometimes interactions may demonstrate synergy where the effects of the factors combined are greater than the effects of the individual factors.
The first step of Design of Experiments is to define factors and levels. I chose two factors, upbringing and location. I set  two levels for upbringing, strong seedling versus runt,  three levels for location, sunshine, dappled shade and complete shade. The experimenter needs to be aware of confounding factors which may affect the results. These became apparent as the experiment progressed, unusually hot weather in April caused the plants in the sunshine to wither, aphids attacked some plants and not others. I attempted to deal with these uncontrollable factors by watering the plants in the sunshine just enough to keep them alive and spraying the aphids as they multiplied with a water and household soap solution.
The next step is to choose a parameter to measure which will gauge the effect of the factors. This dogged me from the beginning as I couldn’t figure out how to measure flowering performance. I decided initially to count the number of flower trumpets , but this is easier said than done because the flowers bloom and fade progressively from the base to the tip of the flower spike. I opted instead to measure not only the final heights of the plants, but also the final height to original width ratio to enable the performance of the runts to be compared fairly with the strong seedlings. I accepted that my results would not take into account that some plants produced more flower spikes than others and that I didn’t have a full set of trials in my unplanned study to be able to draw statistically significant conclusions this time.
The graph in Appendix 2 shows that a strong seedling in dappled shade grew the tallest and a strong seedling growing in the sunshine was the shortest. Specimen 6 was the only strong seedling growing in complete shade, it suffered a reduction in width between January and March this year when the rosettes of leaves from last year were superimposed with new growth,  and by May was outgrown by specimen 9, a runt also growing in the shade.
The graphs in Appendix 3 average the results for the different factors and levels to show their effects separately and as interactions. If the results are taken at face value, they suggest that runts whether growing in the shade or dappled shade had the greatest growth rate (comparing final height with original width), plants grown from strong seedlings were very sensitive to levels of sunshine and shade.
My conclusions from these experiments is that if I find myself in the same position this year with runts and strong foxglove seedlings, I’ll plant the strong seedlings in the precious dappled shade locations, the runts in complete shade (because it wont make much difference) and save the sunny locations for other plants. I realise that most gardeners probably knew all of this already !


APPENDIX 1 -  10 FOXGLOVES EXCELSIOR GROWN FROM SEEDS SOWN IN MARCH 2010

FOXGLOVE 1
Grown from a runt seedling. Located in dappled shade. Two extra flower spikes are growing from the base. Infected with aphids.
      


FOXGLOVE 2
Grown from a strong seedling. Located in sunshine. Two extra flower spikes are growing from the base. Leaves shrivelled due to lack of water.
 


FOXGLOVE 3
Grown from a strong seedling. Located in sunshine. Leaves shrivelled due to lack of water.
 


FOXGLOVE 4
Grown from a strong seedling. Located in dappled shade. One extra flower spike is growing from the base, five extra flower spikes are growing from the main stem. Infected with aphids.
 


FOXGLOVE 5
Grown from a strong seedling. Located in dappled shade. Three extra flower spikes are growing from the main stem.
 


FOXGLOVE 6
Grown from a strong seedling. Located in complete shade.
 


FOXGLOVE 7
Grown from a runt seedling. Located in dappled shade. Two extra flower spikes are growing from the base.
 


FOXGLOVE 8
Grown from a strong seedling. Located in dappled shade. One extra flower spike is growing from the stem. Infected with aphids.
 


FOXGLOVE 9
Grown from a runt seedling. Located in complete shade. Infected with aphids.
 


FOXGLOVE 10
Grown from a runt seedling. Located in complete shade.
 


APPENDIX 2 -  MEASUREMENTS OF PLANT WIDTHS & FINAL HEIGHT
Performance of runts versus strong seedlings - grown in sunshine, shade & dappled shade.



APPENDIX 3 -  RESULTS OF A DESIGNED EXPERIMENT TO ANALYSE FOXGLOVE GROWTH
FACTOR1 : UPBRINGING       2 LEVELS : RUNT versus STRONG SEEDLING
FACTOR 2 : LOCATION      3 LEVELS : SUNSHINE , DAPPLED SHADE , SHADE
Plots of means (averages) : final height    &    final height/original width

For more information about Design of Experiments :

©Copyright 2011 b-a-g. All rights reserved.

16 comments:

Anonymous said...

Your experiment with foxglove is a real winner. They are so big and beautiful. I have no luck like that with them, delphinium yes, foxglove no.And they are supposed to seed like crazy, but not here.

Carolyn @ Carolyn's Shade Gardens said...

Wow, I am very impressed. You really can learn a lot from this type of study, but I am way too undisciplined. This foxglove is gorgeous which just goes to show that "sophisticated" gardeners don't know much.

Alistair said...

B-a-g, that is fantastic information that had me smiling from ear to ear.Now I don't know if it applies to Foxgloves, but very often if you choose the strongest of seedlings there is a chance that you have picked all with the same colour, very often the more interesting shades are from those that appear weaker.Sophisticated gardeners turning their nose up at excelsior for having flowers all the way round the stem is a new one on me. I personally have always heard high praise from the best in the land for this improvement which Excelsior gives.

b-a-g said...

Thanks Donna & Carolyn - I hope they do seed like crazy. I'm prepared to dedicate my garden to the homage of foxgloves!

Alistair - I remember you commented once that foxglove excelsior is one of Myra's favourites. Here is Carol Klein's opinion : http://www.telegraph.co.uk/gardening/howtogrow/3300521/How-to-grow-Foxgloves.html

b-a-g said...

Just for reference ... the foxgloves labelled 1 to 10 were also photographed for posts dated 03JAN2011 and 23APR2011. In both of these posts, the photos were listed in this order : 4 , 2 , 9 , 5 , 7 , 8 , 10 , 3 , 1 , 6

Autumn Belle said...

It is interesting to know that you are applying the DOE technique in the garden too. Foxgloves are not common here in my place and the only flower that looks like it is Angelonia.

Donna said...

I love your dedication to experimentation. I love foxgloves of every kind and have them planted all over in hopes that there will be blooms every year as they only bloom every other year. I find the shade is a help but if in a moist sunny location they do alright as well...of course I always push the envelope and plant in conditions not recommended and am pleasantly surprised.

One said...

Wow!!! Did you hear that! I'm impressed. Can you just do something similar with the stock market?

Congrats! Your caption has been selected. I have a link to your blog today. :)

Stacy said...

There's a certain delicious irony in your not considering yourself disciplined enough to read a seed packet while then you construct a months-long experiment (six and counting), keep 10 foxglove seedlings' growth habits straight, and plot the results on graphs, which you describe in 3 appendices. Just sayin'...

The world would be a much less interesting place if sophisticated gardeners had their way.

b-a-g said...

Autumn Belle - I looked up Angelonia, it's pretty with inverted trumpets. I like the dark purple ones.

Donna - Foxglove excelsior is a biennial, but there are foxglove annuals available. I think you need to plant the seeds in autumn.

Thanks One - Yes I heard you all the way from Malaysia! I'm afraid the idea of working in finance has never appealed to me. Thanks for choosing my caption.

Stacy - It does seem ironic, when you put it like that. I was wondering how people would receive this post (1)because it is yet another post about foxgloves and (2)because of the science bit. Your comments are very encouraging.

Laura Bloomsbury said...

Hi b-a-g, vaguely remember how to design experiments but the multifactorial design always foxed me! Following your investigations of the lore of foxgloves has been fascinating. Water is a big factor though i.e. dappled shade does not dry as quickly but the answer still lies in the moisture/nutrient content of the soil. Hence my self-seeders stayed as runts in dryish part-shade. Your spires are stupendous.
Laura

lifeshighway said...

I am very impressed with the whole foxglove experiment. I have to wonder, what was the most enjoyable part of the experiment?

The growing or the charting.

The charting was very impressive.

b-a-g said...

Thanks Laura - I guessed you were a scientist despite your poetic prose. I tried to take out water as a factor by watering the plants if they looked limp. However, now you mention it, I'll have to check ...

Thanks LH - I enjoyed the whole foxglove experience: sowing, growing, measuring, trying to find the meaning of life in the results ... and it's not over ...

Rosie@leavesnbloom said...

Your graphics are very good and appreciate the time you took to show the data in that presentation. I just let mine seed where ever they fancy in the garden and I'm hoping for a good display this year. My self sown seedlings seem to prefer a moist soil in the dappled shade.

I'm a horti historian and I could never write about english bluebells in a Scottish wood that William Wallace used as a hideout!

b-a-g said...

Thanks Rosie. I enjoyed your bluebells post (whatever their nationality!) : http://leavesnbloom.blogspot.com/2011/05/bluebell-woods-perthshire-scotland.html

Alistair said...

Hi b-a-g, I enjoyed Carol Kleins article on Foxgloves. She is the best and most enthusiastic gardening expert/presenter we have. I see her opinion on Excelsior, I don't think it was always quite so.

Related Posts Plugin for WordPress, Blogger...