written in July 2013
My son decided to pick some cherries during last week’s heat wave. Luckily, 5 minutes later, he lost interest and walked away, and my daughter who normally would have been in front of the tree, was at her grandmother’s, learning to bake pineapple-coconut pie. As a result, neither of them was in the way of the 15-foot branch that my wife sent crashing down, and luckily for us, she maintained her balance on the step-ladder.
After sawing it to begin the cleanup, I noticed that the tree rings on the thick branch’s cross section were irregular. Why, I wondered?
In the spring, my wife’s relative, a forestry student, had sent me a set of research papers by Quebec scientists who had found a correlation between temperature and tree ring thickness in the boreal forest, where moisture is rarely the limiting factor. But the cherry tree is not a conifer, and it’s growing in a temperate forest biome. Other hypotheses came to mind.
First I drew a radius through the widest gaps (the rings were not perfect circles) and measured all the rings (figure 1). The branch was at least 11 years old. Then I consulted Environment Canada to obtain precipitation and temperature data in Montreal for that time period. Here’s what I found.
|avg.temp||mm precip||avg.temp||mm precip||avg.temp||mm precip||avg.temp||mm precip|
|avg.temp||mm precip||avg.temp||mm precip|
|aug||sep||AVG temp||total mm||avg july, aug, sep||total july, aug, sep mm|
At first, there seemed to no correlation of any kind. Here’s a summary of the hypotheses I entertained:
|Temperature is the variable responsible||No correlation|
|Total precipitation from April to September is the variable responsible||No correlation|
|A combination of precipitation and temperature plays a defining role.||No correlation|
|If a specific cold spell after flowering affects cherry production, the tree will subsequently have more energy to devote to trunk growth.||No data|
|Erratic fertilization could be the limiting factor||No data|
|Most trunk growth occurs in July, August and Sep after cherry production and temperature during those months will be the controlling factor.||No correlation|
|Most trunk growth occurs in July, August and Sep after cherry production and total precipitation during those months will be the controlling factor.||No correlation|
|Most trunk growth occurs in July, August and Sep after cherry production and a combination of temperature and total precipitation during those months will be the controlling factor.||No correlation|
Then I thought of the possibility that most of the trunk growth occurred in July, and that the controlling variable was moisture. Since we rarely, if ever, water the tree, trunk growth suffers when not enough rain falls. During that month, the days are still close to being their longest, and the tree is no longer devoting energy to producing fruit.
The correlation was stronger, but still disappointing. If, however, out of the eleven years, I ignored two of them, when other factors could have caused anomalies, the correlation coefficient became 0.88.
But, pardon the pun, was I merely cherry-picking the data? Also, was there anything in the research literature to support the idea that the bulk of cherry trunk growth did indeed occur during one single summer month?
|July mm rain||width of ring|
I contacted an Oregon State University horticultural scientist, Todd Einhorn, summarized my adventure and waited for a reply. Surprisingly, he responded on the same Sunday and wrote:
Thanks for your interesting email.
We have documented, as others have, the rapid growth of trunks after harvest. It appears that trunks compete poorly for carbohydrates when other sinks are active; fruit having the greatest sink strength. So, your findings are supported. Regarding the omission of 2 years of data…I imagine you would have to run stats for spurious data. Nine of eleven years isn’t bad!, what do you suppose could have resulted in the deviation? Perhaps low temperature injury (affecting either cambial tissue, or reducing bloom resulting in light crop loads) be a contributing [covariate] factor?
Not only did the broken branch lead to some interesting science, but it yielded over 25 pounds of cherries. Most fruit from an unbroken branch would have remained out of reach but would have been well within the grasp of competing birds. Now that the pineapple-coconut pie has been vacuumed down, we can feast on cherry pie and jam. After that, it would be nice to get data from more trees in the city!