Zelkova Repotting Experiment

[0soyoung]

Last spring I bought twelve 4th year Zelkova serrata seedlings that were offered by pwk5017 on another forum. I potted them on 15 April according to the same experiment design as I am using for Eastern redbuds and cork oaks (the other 4 are being grown as back-ups).

tree	repot	rhizo	cuts	shoot pru	defol
1	-	-	-	+	+
2	+	-	-	-	-
3	-	+	-	-	+
4	+	+	-	+	-
5	-	-	+	+	-
6	+	-	+	-	+
7	-	+	+	-	-
8	+	+	+	+	+

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

Fundamentally this is an experiment about repotting in Aug/Sep versus the conventional ‘known good’ time in spring. I decided that I will try to assess root growth, in addition to monitoring the ‘above ground’ growth. I am using 6 inch clear orchid pots as mini rhizotrons to do this. Because it is difficult to discern roots from damp MVP, I have filled these pots with dark colored Turface Pro League Heritage Red. Hence I have a new experiment factor, ‘rhizo’ that is a different soil/pot. These ‘rhizotrons’ are nested within a gallon size black plastic pot. Soon I will start looking at them weeky to measure the root length.

I am using Turface for two reasons. First is that MVP is my chosen planting medium. Second, is that it can be removed without disturbing the root mass by just gently agitating the tree roots in a bucket of water. This likely is also the case with non-compacting mixes (e.g., pumice, scoria).

I chopped each to 10 inches tall and collected the data that I’ve tabulated below prior to potting. I used a caliper to measure the stem diameter and a 1gram resolution scale to weigh each. Again, I selected the trees randomly with a simple intuitive procedure:

• Write the id numbers on a potting stick
• Put the sticks, numbered ends down, in a cup
• Pick a seedling/tree
• Pick a stick
• Perform the actions required for that id.

tree id	base dia (mm)	cuts	wt (gm)	clr pot
1	13.33	no	75	no
2	9.36	no	26	no
3	8.61	no	26	yes
4	8.45	no	14	yes
5	9.14	yes	28	no
6	15.00	no	63	no
7	11.68	yes	41	yes
8	10.01	no	21	yes

The 8 zelkovas on 15 Apr 13, just prior to potting

 

[brewmeister83]

Quick question/clarification...

The variable repot '-' (spring) : is "spring" a set month/timespan like repot '+' (Aug/Sep) or is it plant dependent like when the buds on the trees are swelling or just about to open?

 

[0soyoung]

Quick question/clarification...

The variable repot '-' (spring) : is "spring" a set month/timespan like repot '+' (Aug/Sep) or is it plant dependent like when the buds on the trees are swelling or just about to open?

Good question.

My purposes are simply to contrast the conventional spring potting time with the contentious other time of Aug/Sep (i.e., demonstrate that it can or cannot be done successfully). So implicitly repot '-' (spring) = as buds swell, the time we all know to be 'safe'.

Thanks for highlighting this. I will note the state of the seedlings as well as the date when I repot in the 'spring' of 2014. Good point!

 

[0soyoung]

Stem/Trunk Thickening

When I potted my zelkova 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 while I was making measurements during the season.

I’ve rendered my data on the attached chart by the average thickness measured across all 8 trees 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 red dashed curve of the rate of thickening (the slope of the curve through the data points). This shows that the trunk cambium did not activate until late June. The rate of thickening then increased rapidly until ca. 16 Aug and rapidly declined since. This is especially interesting to me because the behavior is so different from the other species with which I am experimenting (Douglas fir, Lodgepole pine, cork oak, redbud). A bit better understanding of this might be interesting for bonsai development purposes.

A couple of details that I should add are that my measurement accuracy was about 0.05mm (surprisingly good) and that the average thickness increment for my zelkova seedlings this year was about 1.15mm (which amounts to about a 14% increase in thickness, on average, for one growing season).

 

[0soyoung]

The First Growth Season

For this past season, the only 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 nested in black plastic pots). In pictures, I don’t see anything that would make me choose to use Pro League over MVP (or vice versa) – which is to say there is no effect.

Zelkovas in Orchid pots of Turface Pro League

Zekovas in 1 gal pots of Turface MVP


I should point out that I did a little impromptu thread grafting on two of these trees and that this tends to distort impressions about the relative growth of the trees. I did check the length of the longest shoots on each tree – the average lengths with Pro League (21 +- 4.51 inches) are ‘ identical’ to those on trees in MVP (i.e., soil/pot type had no effect on shoot length).

id	longest shoot (in)	average	standard error
1	18
2	6
5	30
6	33	22	12.34
3	24
4	23
7	14
8	21	21	4.51

An interesting aside to also notice is how an axillary shoot became the new leader from one of the looped ‘threads’ – apical dominance in action!

Thread grafts:


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 4 Sep, I washed the Turface out of the roots of #2, #4, #6, and #8 so I could check the weights. These four trees added a roughly 10% weight gain in weight this season – (16+-0.71 grams; the group weighted 15.75+-6.8 grams just before they were potted on 15 Apr). Root growth was generally quite robust; #2 was the only tree to not completely populate the Turface.

Completely populated Turface




... hit the attachment limit!
next post: pics of #2, #4, #6, #8 on 4 Sep versus #2, #4, #6, #8 on 15 Apr

 

[0soyoung]

The First Growth Season - continued

#2, #4, #6, #8 on 4 Sep

#2, #4, #6, #8 on 15 Apr

 

[Tieball]

Coooool

For me, the most interesting thing that comes from this is the red dashed curve of the rate of thickening (the slope of the curve through the data points). This shows that the trunk cambium did not activate until late June. The rate of thickening then increased rapidly until ca. 16 Aug and rapidly declined since.

This was very interesting. Thanks for pulling the data together and sharing. I've wondered about this but didn't know how to approach building a fact base. Mighty fine work you did!

 

[ChrisV]

Don't know if it replies to your subject/experiment but I repot my trees every time of year with success. Except for white pines. It depends on the health of the tree though. Only if they are healthy enough.
Also I provide them protection and control watering etc after.
Never lost a tree due to repotting. So in my vision it doesn't matter as long if the after care is right.

 

[drew33998]

Were the seedlings taken from the same mother tree? The only thing I would ask would be about genetic diversity among a particular species. Some trees may be slightly genetically different than others, giving them slight advantages or disadvantages over the others.

 

[0soyoung]

Were the seedlings taken from the same mother tree? The only thing I would ask would be about genetic diversity among a particular species. Some trees may be slightly genetically different than others, giving them slight advantages or disadvantages over the others.

I have no idea. They are simply a group of seedlings that I bought from pwk5017. I suppose seedling #2 might have come from a different parent plant. In fact, I don't even know if they started life from seed or as struck cuttings.

As interesting as genetic effects might be, investigating them is far beyond the scope of my experiment. My goal is just to identify whether re-potting at another time of year (Aug/Sep) is deleterious or not compared to the conventional 'as buds swell' time in the spring. I continue to hope that others do some repotting similarly and share their results so we all can see if there is any indications of genetically/geographically caused differences.

It again needs to be noted that I have randomly assigned the trees so that individual differences (ought to, regardless of the causes) average out between these two experiment groups (spring vs. Aug/Sep). I am comparing the averages of goups of trees (the '+' group versus the '-' group) and not comparing individual trees. This is the key point of how factorial experiments work. Their downfall is when the units in each group vary too widely to be able to discern an effect. I had no a priori notion of what the magnitude of variation among trees might be, so this is also something to possibly be discovered from the experiment. Genetic variations might make it more difficult to discern an effect, just as any other kind of variation among the trees and their treatment will do (such as they didn't all start out with the same thickness, same height, same weight, etc).

 

[0soyoung]

And Now, After 2 Years ....

I completed this experiment!

I've summarized the seedling weight data in the following table. It shows that it is beneficial to repot in the Aug/Sep. Seedlings repotted in fall gained more weight than those repotted in the spring.

ID	repot	medium	cuts	P0 wt	RP1 wt b4	RP1 wt after	RP2 wt b4	repot wt removed	seasonal wt gain
1	-	-	-	75	149	110		39	74
2	+	-	-	26	33	31	203	53	172
3	-	+	-	26	121	73		48	95
4	+	+	-	14	97	71	291	81	220
5	-	-	+	28	78	56		22	50
6	+	-	+	63	211	109	457	148	348
7	-	+	+	41	152	78		74	111
8	+	+	+	21	omitted	77	367	107	295

• Potting time0 was on 15-Apr-13
• RePot1 was on 2-Sep-13 for fall repotting (+); on 23-Mar-14 for spring repotting (-)
• RePot2 was on 5 & 6-Sep-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

o for spring (-) repotting = (RP1 wt b4) – (P0 wt)
o For fall (+) = (RP2 wt b4) – (RP1 wt after)​

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, per se, looking for whether every tree in one group gained more weight that every tree in the other group. 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 ZelkovaTime.jpg is a histogram of the seasonal weights gained taken from the table above. In the first bin I’ve counted how many seedlings had a seasonal weight gain of 50 grams or less. There was one (#5) and it was repotted in spring (repot ‘-‘ treatment). There were two with weights between 50 and 100 grams (#1 and #3). And etc. I think it is very clear that, as a group, the seedlings repotted in Aug/Sep gained more weight than the group of seedlings repotted in spring.

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. In the attached ZelkovaPotMedium.jpg you can see that the weight gains in these two groups was essentially the same. Likewise, the attached ZelkovaCuts.jpg plot shows that cuts had no (appreciable, if you wish) affect either.

Summing up my conclusions:

1. The claim that zelkovas 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 supported.
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 he trunk adversely affect the seasonal weight gain of zelkova seedlings.​

edit/ammendment:

A last bit I forgot to mention is that the two groups of trees were equivalent in the beginning of the experiment. This is apparent from the histogram of the P0 weights in the ZelkovaP0.jpg attachment. Of course, the random assignment of seedlings to treatments that I did in the beginning should have assured this (and did), but it is always good idea to check.

 

[0soyoung]

You Can Too

You needn't live in Anacortes, WA to repot zelkova serrata in late summer or fall as I have done. Anyone should be able to do similarly as long as the relative humidity is above 50% and root temperatures below 90F (33C) when they repot.

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.

In 2013, I observed no wilting whatsoever after repotting - see ZelFall_13_Trh.jpg. It was the same in 2014 (see ZelFall_14mod_Trh.jpg) until 11 Sep when suddenly leafs began wilting - see ZelFall_14_Trh.jpg. It is because of this that I believe 50% relative humidity is about the limit.

Repotting is possible at times other than 'as buds swell'.