Douglas fir Repotting Experiment

[0soyoung]

I have just potted 16 Douglas fir (Pseudotsuga menziesii var. glauca) seedlings, to start my repotting-plus experiment (pretty much as I described in the “Let’s All Get Some Answers and Not Wait for the Book” thread):

run	repot	rhizo	cuts	shoot pru	defol
1	-	-	-	+	+
2	+	-	-	-	-
3	-	+	-	-	+
4	+	+	-	+	-
5	-	-	+	+	-
6	+	-	+	-	+
7	-	+	+	-	-
8	+	+	+	+	+
9	-	-	-	-	+
10	+	-	-	+	-
11	-	+	-	+	+
12	+	+	-	-	-
13	-	-	+	-	-
14	+	-	+	+	+
15	-	+	+	+	-
16	+	+	+	-	+

o repot ‘+‘ = Aug/Sep
o repot ‘-‘ = spring
o rhizo ‘+‘ = nested clear orchid pot & proLeague
o rhizo ‘-‘ = MVP in black pot
o cuts ‘+’ = cuts at stem base done at repotting
o cuts ‘-‘ = do nothing
o shoot pru ‘+’ = nip shoot tip to just remove terminal leaves
o shoot pru ‘-‘ = do nothing
o defoliate ‘+’ = remove fraction of leaves by cutting through petiole
o defoliate ‘-‘ = do nothing

I am also trying to measure root growth, in addition to monitoring the ‘above ground’ growth. To facilitate measuring root growth, I am using 6 inch clear orchid pots (filled with dark colored Turface proLeague Heritage Red to contrast with growing roots) as a new experiment factor, ‘rhizo’. These ‘rhizotrons’ will be nested within a gallon size black plastic pot except when I periodically remove them to estimate the root growth. The other trees (not in clear orchid pots) are potted in one gallon black plastic pots filled with Turface MVP.

You may notice that originally I planned to have 2 trees per ‘run’ and use just the first 8 runs of the tabl above. This would have meant that any effect of shoot pruning on seasonal weight gain couldn’t be distinguished from the interaction (if any) of repotting time and the soil & pot size of the rhizotrons. If there turned out to be such an effect then I would have to execute the runs 9 -16 to find out whether the effect was due to shoot pruning or the interaction of rhizotrons and repotting time. Since I had 16 trees, I opted to just execute the full 16 run factorial experiment.

I cut off the ‘needles’ from one side of the central leader of those trees getting the defoliation treatment. The idea is to see if back budding preferentially occurs where needles are (ala mugo pines) or where they are not (ala JBP). I left any existing buds untouched.

Lastly, I gave each 1 teaspoon of Osmocote 14-14-14.

id	base dia (mm)	defol	cuts	wt (gm)	clr pot
1	4.55	yes	no	10	no
2	5.12	no	no	15	no
3	4.75	yes	no	12	yes
4	4.2	no	no	13	yes
5	6.3	no	yes	22	no
6	7.48	yes	no	22	no
7	5.28	no	yes	15	yes
8	5.08	yes	no	14	yes
9	6.67	yes	no	31	no
10	5.77	no	no	17	no
11	5.56	yes	no	20	yes
12	5.03	no	no	12	yes
13	4.55	no	yes	9	no
14	4.7	yes	no	13	no
15	6.53	no	yes	21	yes
16	6.3	yes	no	12	yes

 

[fraser67]

I'm curious...why Doug Fir? I assume because they are plentiful up there
Good luck!!

 

[0soyoung]

I'm curious...why Doug Fir? I assume because they are plentiful up there
Good luck!!

That is pretty much it, though the seedlings I am using are the variety that grows throughout the rockies instead of the coastal variety that is indeginous here. The bottom line is that Arborday.org had a deal

 

[joepa82]

The results of your experiment may be a bit more satisfying if you use less variables and more control. That being said it is interesting

 

[garywood]

0, just curious, how are you reconciling the difference in roots and vigor? Without a year for stabilizing, what is the baseline? I applaud your experiments but they could put you in the same place that started your project. Post hoc ergo propter hoc. Cover all the bases.

 

[0soyoung]

0, just curious, how are you reconciling the difference in roots and vigor? Without a year for stabilizing, what is the baseline? I applaud your experiments but they could put you in the same place that started your project. Post hoc ergo propter hoc. Cover all the bases.

I am after the seasonal carbon fixing (weight) as a measure of tree health and/or response to repotting time. This year I potted all in spring. Half of the trees will be repotted Aug/Sep - this will really be the intial Aug/Sep repotting. So it won't be until Aug/Sep 2014 that I have a legitimate season of Aug/Sep repotting. So it will take at least until 2014 to have proper data. The following spring I will have a second season of spring-spring repotting which I suspect will be better because the trees will have had a season of initial 'development' somewhat like the Aug/Sep repot cycle trees. In principle this could be continued year after year - that's part of your answer (if you can decipher my gibberish).

A second part is that trees were randomly assigned to each group. So on the average they are equal in roots and vigor - equal in the sense of a random grouping of seedlings of this age.

Finally, the determination of whether there is a difference or not is based on comparing the grouping averages - the average of 8 trees repotted in spring versus the average of 8 trees repotted in Aug/Sep. Further, because of having 8 trees in each group, I can spot a tree that isn't like the others and exclude it from the analysis of the rest and then look for a reason why that tree is very unlike the others.

Even though I am hoping to prove something, the statistical underpinnings of the experimental design is an hypothesis that none of the factors (repot time, cuts, etc.) matter - every grouping looks like the others. So finding an effect is finding a grouping that isn't like the others (analogous to the odd tree).

... more than you wanted, I imagine.

Bottom line is: averages are how I cope (even with qualitative measures).

That's the theory, as they say. Of course, there is a multitude of ways for this to go awry, but IMHO, the most serious risk of this being a bomb is if many seedlings don't make it this year. I appreciate your concern and welcome anything further.

 

[ghues]

Interesting design and concept…..and I know you are doing it as a fun project so I just had a few thoughts for you on the genetic differences found in seedlings grown at forest seedling nurseries.
You say you got the seedlings from Arbor Day…. so are you sure they are from the same family, grown in the same nursery under the same conditions, was fertilizer added at the nursery to some and not others?
Up here (BC Canada) seedlings are given specific seedlot numbers based on a number of parameters. Some forest seedling nurseries use fertilizer with watering, others use fertilizer prills, some use nothing but H2O (the fert prills can add to growth in the first year).

If the seedlings were grown from wild seed then its very possible the seed came from a number of different trees/parents (dependent on collecting method) so the genetic differences could be large, large enough to distort the results.
However if the seedlings were grown from seed orchard stock then the parents are known and the genetic differences aren’t as great as wild seed, so the results should be more consistent.
The reason I’ve added this is that I’ve been involved in Arbor Day Celebrations and we get different nurseries (usually 3-4) to donate a few hundred seedlings and often get different “seedlots” (a descriptive for one group of seed= same parents/families) of Fdc seedlings from each nursery.
P.S. by the way - For survival, performance (root growth, top growth, branch development), and other seedling data collection field trials we (foresters up here) use a minimum sample size of 100 seedlings, usually (like this spring) we use 1000's.....that way our stats are pretty sound and has way less noise.
Good luck.
Cheers
Graham

 

[amkhalid]

Interesting design and concept…..and I know you are doing it as a fun project so I just had a few thoughts for you on the genetic differences found in seedlings grown at forest seedling nurseries.
You say you got the seedlings from Arbor Day…. so are you sure they are from the same family, grown in the same nursery under the same conditions, was fertilizer added at the nursery to some and not others?
Up here (BC Canada) seedlings are given specific seedlot numbers based on a number of parameters. Some forest seedling nurseries use fertilizer with watering, others use fertilizer prills, some use nothing but H2O (the fert prills can add to growth in the first year).

If the seedlings were grown from wild seed then its very possible the seed came from a number of different trees/parents (dependent on collecting method) so the genetic differences could be large, large enough to distort the results.
However if the seedlings were grown from seed orchard stock then the parents are known and the genetic differences aren’t as great as wild seed, so the results should be more consistent.
The reason I’ve added this is that I’ve been involved in Arbor Day Celebrations and we get different nurseries (usually 3-4) to donate a few hundred seedlings and often get different “seedlots” (a descriptive for one group of seed= same parents/families) of Fdc seedlings from each nursery.
P.S. by the way - For survival, performance (root growth, top growth, branch development), and other seedling data collection field trials we (foresters up here) use a minimum sample size of 100 seedlings, usually (like this spring) we use 1000's.....that way our stats are pretty sound and has way less noise.
Good luck.
Cheers
Graham

We had a similar discussion about the design of his lodgepole experiment

http://bonsainut.com/forums/showthread.php?10604-Lodgepole-Pine-Repot-Experiment&p=135883#post135883

 

[daygan]

Just a note: I do intend to participate in this experiment as well. I started late this year because I was waiting to get a better feel for the local climate and how trees might respond to various weather factors before buying a bunch of seedlings that I might not even be able to keep alive. I do have a slightly better grasp of what I will need to do to care for these, but at this point, still have hesitations about Douglas Firs' ability to survive in a zone 7 (because it's indicated as being hardy for zones 5 and 6). I've found seeds (no seedlings) which I can purchase, and will stratify them in the winter so that they will be growing by next summer. I may not be able to start the repotting experimentation until 2015, but I will be doing it as long as nothing unexpected interferes.

 

[0soyoung]

Progress - initial potting effects

At this point, only one of my 16 Douglas fir seedlings has died. Two have yet to flush as all the rest have.

Given that there is only 1/16 dead, no effects can possibly be perceived – one dead tree can be assigned to some group and not the other by nothing but luck of the draw.

This experiment is continuing according to plan because I have a spare tree to replace the one that died (as well as two more in reserve).

 

[0soyoung]

Second Flushing Already

My Doug firs appear to be starting their second flush even though the first flush shoots are still light green.

 

[0soyoung]

Trunk/stem Thickening

When I potted my seedlings this spring I painted a mark about 10cm (4 inches) above the soil level and used this site for measuring the trunk thickness of each seedling every week or two with an inexpensive digital caliper. Each time I measured the seedlings, I measured each three times so that I could also gage my measurement accuracy and, therefore, whether the individual trees were indeed thickening during the season.

I’ve rendered my data on the attached chart by the average thickness measured across 17 trees (16 primary experiment + 4 spares - 3 died) and have normalized the data to represent the fraction of the seasonal increment (i.e., the normalized average thickness at the end of the season is 1). I’ve also represented the standard error of my measurements by error bars on the attached chart.

For me, the most interesting thing that comes from this is the rate of thickening (the slope of the curve I’ve drawn through the data points). The rate of thickening increased during the season until ca. 9 July and has steadily declined since. This is quite interesting for folks that like to understand how trees work, but the only practical value I see to this for bonsai is to explain why wire is best applied in August or later in the season and removed in the early spring if one doesn’t want to have wire scars on their branches – as though we didn’t know this already.

A couple of relevant details that I should add are that my measurement accuracy was about 0.068mm (surprisingly good, IMHO) and that the model, fit to the data to date, projects that the average thickness increment of my Douglas fir seedlings will be about 3mm (which amounts to about a 70% increase in thickness for one growing season). I will be remeasuring the seedlings about once a month through the winter.

24Oct: Scholarly papers suggest that the maximum growth rate is controlled by daylight hours and that the dormancy period is controlled by (low) temperature.

 

[0soyoung]

The First Season

For this past season, a treatment that was applied (other than to make 4 length-wise cuts at the bases of two spring potted trees, #5, #7) was potting in Turface MVP (in one gallon black plastic pots) versus finer grained Pro League (in clear orchid pots). While there certainly are differences between individual trees, I don’t see anything in the two pictures that would make me choose to use Pro League over MVP (or vice versa) – which is to say there is no effect.

Douglas firs in clear orchid pots of fine grained Turface Pro League

Douglas firs in 1 gallon plastic pots of Turface MVP


Since I’ve learned that I can measure trunk thickening, one might also ask if the trunks thickened more pronouncedly in one group versus the other. With numbers, the principle is the same – are the two groups like we would expect from random chance (no effect) or not? I compared this season’s trunk thickness increment; long story short: there is no effect (yet?).

On 31 Aug, I washed the Turface out of the roots of #2, #4, #6, #8, #10, #12, #14, and #18 (that replaces #16 that died during the season) so I could check the weights. These eight trees more than quadrupled in weight this season – (added 50.75+-19.78 grams; the group weighted 15.67+-3.61 grams just before they were potted on 17 Mar). No wonder these trees can become so huge!

The season’s root growth was respectable. In the ‘before’ and ‘after’ pix below, it is worth noting how #12 (the third tree from the right) barely changed this season. It popped it buds this past spring, but there was very little extension. I note that its roots appear to be unchanged. I’ve expected it to die, but it keeps hanging on. Nevertheless, I have an extra tree mirroring it, just in case.

Douglas firs #2, #4, #6, #8, #10, #12, #14, and #18 on 31 Aug

Douglas firs #2, #4, #6, #8, #10, #12, #14, and #18 on 15 Mar

id	base dia (mm)	repot wt (gm)
2	12.97	50
4	8.12	49
6	10.85	70
8	9.86	38
10	12.47	52
12	5.63	11
14	7.35	39
18 (16)	11.74	50

 

[0soyoung]

Voila

I've summarized the seedling weight data in the following table. It shows that there is no harm done by repotting in the fall instead of the spring. Conversely, based on seasonal weight gain, there is no reason to repot in spring, ‘as the buds swell’ instead of late summer or fall.

Prior to the 2014 season, shoot pruning was done only as part of repotting. While this might be a good practice when trees are in development, it is unacceptably restrictive if one wants to study pruning response for shaping and maintenance. I did shoot pruning, as indicated in the table, on 17 June 2014, carefully collecting and weighing the trimmings (those cells with ‘–‘ were not pruned; I have grayed the data for spring repotted seedlings, because the trimmed weight corresponds to the seasonal weight gain for the season to end in spring 2015). The extent of my shoot pruning can be seen by clicking back and forth between DFprunB4 2014-06-17.jpg and DFprunFTR 2014-06-17.jpg. It seems that the tip of a pruned shoot will inevitably die back to the nearest bud or node (if there is not intervening bud), regardless of the time it was pruned. Spruce, for example, will usually form a terminal bud at a cut tip.

ID	repot	medium	cuts	P0 wt	RP1 wt b4	RP1 wt after	shoot wt pruned *	RP2 wt b4	repot wt removed	seasonal wt gain
1	-	-	-	10	59	40	21.80		19	49
2	+	-	-	15	65	50		112	15	62
3	-	+	-	12	91	66			25	79
4	+	+	-	13	59	49	25.43	164	10	140
5	-	-	+	22	108	80	44.33		28	86
6	+	-	+	22	112	70		226	42	156
7	-	+	+	15	64	52			12	49
8	+	+	+	12	50	31	omitted	79	19	48
9	-	-	-	31	100	70			30	69
10	+	-	-	17	62	52	42.45	137	10	127
11	-	+	-	20	84	72	30.13		12	64
12	+	+	-	12	12	11		21	1	10
13	-	-	+	9	70	42			28	61
14	+	-	+	13	49	39	34.90	139	10	135
15	-	+	+	21	58	51	15.01		7	37
16	+	+	+	16	86	50		135	36	85

• The original #8 died in the course and was replaced by extra #19
• The original #16 also died in the course and was replaced by extra #18
• Potting time0 was on 15-Mar-13
• RePot1 was on 31-Aug-13 for fall repotting (+); on 24-Mar-14 for spring repotting (-)
  • Shoot pruning was done as part of repotting
  • * In-season Shoot pruning was done on 17 June 2014
• RePot2 was on 6 & 7-Oct-14 for fall repotting (+)
• Weight removed in repotting is the difference between RP1 after and RP1 before.
• Seasonal weight gain is the response to the factors repot, medium, and cuts
  • For spring (-) repotting = (RP1 wt b4) – (P0 wt)
  • For fall (+) = (RP2 wt b4) – (RP1 wt after) + shoot wt pruned
• All weights are in grams

I am looking for whether the seedlings repotted at one time, on average, or as a group accumulated more weight than the other seedlings (again, as a group). I am not comparing individual seedlings. Things outside my control can and do make every tree just a little different in some respect. I cope with this fact of life experimentally by looking at the trees as a group and comparing averages.

The attached DF_P0.jpg is a histogram of the original weights of the seedlings, grouped by their random assignment for repotting in spring and for repotting in fall. The first bin is the count of seedlings with and original weight of 5 grams or less, the second bin between 5 and 10 grams, and etc. I think it is clear that both groups were equivalent at the start of this experiment. DFTime.jpg is a histogram of the seasonal weights gained, taken from the table above (each bin is a multiple of 25 grams). There are 4 seedlings that are remarkably heavier than the rest and they are all in the fall-repot group. The remainder, though, appears to still be equivalent between spring and fall. Thus, I cannot conclude that summer/fall repotting is beneficial at this point. Certainly, spring is not any better, as many claim it to be.

I can do likewise, arranging the data by the ‘-‘ and ‘+’ groups to see that potting the seedlings in black plastic pots of Turface MVP as opposed to finer grained Pro League in clear orchid pots had no effect (see the attached DFmedium.jpg). Similarly, one can easily see that cuts had no effect in terms of weight gain either.

Summing up my conclusions:

1. The claim that Douglas firs can only be repotted ‘as buds swell’ in spring has been disproven.
2. A claim that it is better to repot sometime after the summer solstice, such as around August/September, than 'as buds swell' in the spring is NOT refuted.
3. There is no apparent difference in using coarser Turface MVP in black plastic pots than finer grained Pro League in clear orchid pots.
4. As much as I don't care for 'going postal' on my tree trunks, I have no evidence that longitudinal cuts around the base of the trunk adversely affect the seasonal weight gain of Douglas fir seedlings.
5. A pruned shoot will die back to the nearest bud or node, whichever is more distal.

 

[Bonsai Nut]

First... I think this is AWESOME!

Second... you don't have a large enough sample size to draw some of these conclusions.

Don't kick me - but when someone does something as cool as this I start to wonder whether or not they should publish - at which point I start to get into the nitty gritty of the science and the statistics.

 

[0soyoung]

Second... you don't have a large enough sample size to draw some of these conclusions.
...=

- at which point I start to get into the nitty gritty of the science and the statistics.

Thanks.

It certainly is not a 5-sigma conclusion if that is the sort of thing to which you allude. I would not, for example, wager megabucks based solely on this experiment. However, I would operate a bonsai nursery based on my findings (barring the effects of pruning that I don't feel I have investigated adequately yet). [ed. Strictly speaking, I have only shown that there is no effect in repotting time, no effect in pot-style/media, and no effect of cuts administered to the trunk base, using seasonal weight gain as the response - the responses are just a random distribution of accumulated carbon fixation].

I have tried to be clear about exactly what I have done so that anyone could perform a similar experiment and see if they reproduce my results. Not only is this the scientific way, but it was an aim of mine when this all started with the 'Lets All Get Some Answers and Not Wait for the Book' thread. To the best of my knowledge, no one has yet dared to take the dive.

 

[0soyoung]

You Can Too

You needn't live in Anacortes, WA to repot pseudotsuga menziesii var. glauca in late summer or fall as I have done. Anyone should be able to do similarly as long as root temperature is below 90F (33C) when they repot. There is an unknown limitation in relative humidity, but I am confident this limit is below 50% if roots are pruned as I describe below.

The cohesion tension theory is a well-established model of water transport in trees. The root membranes and the ‘series of tubes’ of xylem lumens are represented as a series of resistors. The water potential of the air versus the soil is analogous to the voltage difference in an electrical circuit that is the force that causes the flow of water (current) though the tree. With regard to repotting, we need only be concerned about the water potential of the air and the resistance we create by how badly we savage/ravage the roots.

For this experiment my methodology was to immerse the roots in a bucket of water and gently work out the Turface. I weighed the tree after patting the roots dry with a towel (B4). I then inverted a 1 gal. black plastic pot (identical to the ones used in the experiment) and used the bottom as a template to prune the roots. I combed only enough to draw out snags/snarls so as to have radial roots, taking care to not rip out fine ‘feeder’ roots. The mat was finally trimmed, again using the bottom of the pot template and the reweighed the tree (Ftr) before putting it back into its pot. Cutting off the bottom half to 2/3 of a root saturated pot, combing the perimeter, and analogously pruning the roots should be equivalent.

The water potential of the air is proportional to the Kelvin temperature of times the natural logarithm of the relative humidity. You don’t need to do any arithmetic, I've plotted some lines of constant water potential in the attached charts. The important point to note is that water potential is almost independent of temperature, over the range for which our trees are growing.

Our perceptions that higher temperature means more transpiration stress comes from the fact that on any given day, the relative humidity goes down as the air heats up. I've also plotted the temperature and relative humidity that were recorded every 5 minutes for a period of 7 days, starting with the day I repotted. Notice how this ‘cloud’ reflects the daily trend we all know so well of humidity going down as the air heats up and goes back up when the atmosphere cools down again at night (on any given day, transpiration stress is greatest in the afternoon).

Generally, the warmer it is the faster plants grow, up to a point. For temperate species, growth stops when the temperature climbs above 90F to 100F because metabolic activity is consuming all the food produced by photosynthesis and then some (i.e., there is nothing left over to make more tree). More importantly, however, trees cannot recover from repotting unless the roots can grow, which means they must be at a temperature below about 90F.

I am confident that this can be done in drier climes than Anacortes, however, it is difficult for me to anticipate times when the relative humidity drops below 50%. I do not have a controlled humidity facility to explore where this limitation might be. In other words, I am (or we are) dependent upon someone living in a drier climate (e.g., in CO, WY, MT, ALB) informing me (or us) about their results with repotting after the summer solstice.

Regardless, repotting pseudotsuga menziesii var. glauca is possible at times other than 'as buds swell'.

 

[edprocoat]

Osoyoung you realize that running a Bonsai nursery on these findings would be equal to wagering mega-bucks on the same findings ? Either way good job and I would feel comfortable repotting Doudlas Firs in the fall judging by the effort you put out and your results. Hell I repot junipers in fall as I have successfully done it before 1 time and since have done it several times with the same results.

ed

 

[wireme]

Thanks for the work and sharing, esp with Doug fir, they are one of the more available species for me here so I have an interest in how they work.
I've found the trunk thickening results interesting too. More than once I have read that girth increases in fall after extention growth is done, always seemed strange.

As far as repotting I've noticed roots are slower to move in the springtime than junipers and pines here. I need more time to observe but I'm beginning to believe that buds should be well progressed in opening if repotting in spring. As an example I collected a fir this spring that had fully extended new growth. I didn't want to collect it as I thought it was too late for fir. The problem was the juniper I was digging at the time extended roots under the fir and it was a case of dig the fir as well or endanger the juniper. The fir came out with cut tap roots and that's about it, I was sure it was a goner. It was late in the day and poor decision making to go for the juniper at all, I left a bit depressed and not proud of the days work! Anyways both trees did great through the first season surprisingly.
I know that's only one tree but I have other observations leading me to think a bit later is better than earlier for Doug fir. This still is only relevant to spring repotting but I'm not surprised to see good results from late summer repotting as well.

One conclusion I would dispute is claim #5 in the post a few above. I've been experimenting with pruning on a number of seedlings in the forest behind my house and find that they very often do produce new buds at and below cut sites in similar fashion to spruce.

 

[0soyoung]

One conclusion I would dispute is claim #5 in the post a few above. I've been experimenting with pruning on a number of seedlings in the forest behind my house and find that they very often do produce new buds at and below cut sites in similar fashion to spruce.

When are you pruning to get new terminal buds?

I had expected to see new terminal buds when I pruned in June (after the first flush), but didn't find any.