2. Background
• Hubbard Brook whole-watershed transpiration
response to wollastonite (CaSiO3) addition
(Green et al. 2013)
3. Background
• Measuring sap flow is one method to estimate
transpiration (Granier, 1987)
• Taking sap flow measurements at MELNHE
wollastonite plots could help clarify
mechanisms underlying HB W1 transpiration
response
• Preliminary sap flow measurements from HB Ca
suggest increased transpiration in wollastonite
treatment (Zahor et al. 2013)
4. Research Question
• Does a wollastonite
addition increase tree
transpiration?
• Expect that it does, possibly
due to increased xylem and
fine root growth
• Alternatively, sap flow
measurements could reflect
the transition to decreased
transpiration observed in
W1 (Green et al. 2013)
5. Methods
• Sap flow measured by the Granier method: reference probe 10 cm
below thermocouple heating probe with constant source of heat,
measures temperature difference (ΔT) (Granier, 1987)
• Measurements collected by data logger every 30 seconds, average
recorded every 15 minutes
• ΔT converted to sapflux (Js, g x m2 x s-1) using BaseLiner software
(Oren and Parashkevov, 2012)
(Max Planck Institute for Biogeochemistry)
6. Methods
• Sap flow to be measured in
wollastonite and control
plots at mature stands in
Bartlett Experimental Forest
(C8), Hubbard Brook
Experimental Forest (Ca)
and Jeffers Brook
• 9 canopy trees measured
per plot: 3 American beech,
3 sugar maple and 3 yellow
birch
• Measurements collected for
~5 days per stand
7. Results
25
20
15
10
5
0
Bartlett C8
171.5 172 172.5 173 173.5 174 174.5 175 175.5 176 176.5 177
Sapflux (Js, g x m2 x s-1)
Ordinal Date and Time
C8 Control AB
C8 Control SM
C8 Control YB
C8 Ca AB
C8 Ca SM
C8 Ca YB
19. Conclusion
• Data suggests increased sap flow in Bartlett C8
and Jeffers Brook, but decreased sap flow in
Hubbard Brook
• Statistical analysis
• Consider additional covariates and research
questions moving forward, why is Hubbard Brook
different?
• Continuous dataset to work with!
• Step toward understanding how CaSiO3 impacts
forest health and hydrology amidst acid rain,
could influence management techniques
20. References & Acknowledgements
• Granier, A. (1987). Evaluation of transpiration in a Douglas-fir stand by means of
sap flow measurements. Tree Physiology 3: 309-320.
• Green, M.B., et al. (2013). Decreased water flowing from a forest amended with
calcium silicate. Proceedings of the National Academy of Sciences
110(15):5999-6003.
• Oren, Parashkevov, & Duke University. (2012). BaseLiner (Version 2.4.2)
http://ch2oecology.env.duke.edu/orenlab/sofware.html
• Michele Pruyn
• Mariann Johnston
• Mark Green
• Ruth Yanai
• Adam Wild, our glorious leader
• Shoestring Crew
Editor's Notes
My project also aimed at figuring out what happened at W1. Justin already introduced this whole-watershed transpiration response to the 1999 wollastonite addition, but I’ll show you this graph again. Transpiration increased significantly from 2000-2002, and then began to decrease again. This was unexpected and puzzling and it’s still not well understood.
Some sap flow work was done last summer by Michele Pruyn from Plymouth State and her grad student Lily in the MELNHE plots, and their data seemed to suggest increased transpiration in wollastonite treatment… but, they had technical difficulties and had trouble getting continuous data
So, my project this summer was to continue the sap flow work and hopefully create a more complete dataset. The question I was hoping to answer was whether a wollastonite addition increases tree transpiration, which we might expect based on the transpiration response at W1, possibly due to increased xylem and fine root growth. BUT!!! If you compare timelines for the Hubbard Brook addition and the MELNHE addition, we are
When there is higher sap flow up through the xylem, it cools the heating probe and the logger measures a lower temperature difference. Shout out to Michelle Pruyn for teaching me everything I know about sap flow. She loaned us most of the equipment we used, her precious data loggers that have names. I got to know Lenny and Ella and Ivan real well, Joey and Kaylee were temperamental and stayed home most of the time.
This give you a general overview of the data from Bartlett C8, the red lines are the calcium silicate trees and the blues the control trees, averaged by species. You can see how sap flow fluctuates over the course of the day, peaking in early afternoon and going back down at night. It varies day to day based on weather conditions, so the third day here was probably a sunnier day than the rest. They really freak out when it rains. And here you can sort of see a trend of higher sap flow in the Calcium silicate trees compared to control
So then I constructed representative diurnal curves for each tree using hourly median sapflux. I used a script in R that picked out the medians per hour for each day of data, and averaged them over all the days of data. This is Bartlett C8 American beech trees, red is Ca, and blue is control. So there seems to be increased sap flow in the AB calcium silicate trees.
Interestingly, for sugar maple, there was less of a trend. I think sugar maples tend to be the odd one out in some of these nutrient addition experiments. Matt Vad noticed a growth response in sugar maples only in W1, so perhaps sugar maples are moving along the transpiration timeline more quickly
Higher sap flow in yellow birch calcium silicate trees
BUT, I observed the opposite at Hubbard Brook old. For beech, sugar maple, and birch, there appears to be lower sap flow in the wollastonite addition trees.
Interestingly, this correlates with Justin’s soil moisture data for Hubbard Brook Old calcium, he noticed slightly wetter soils at HBO, suggesting decreased sap flow and transpiration in the HB calcium silicate plots
At this point, we were noticing a fair bit of variation across individual trees, so we decided to change the methods to hopefully reduce some of that variation and increase statistical power
This is the individual sugar maple data from Jeffers Brook, each line is one tree. You can see there’s still a fair bit of variation, but there do appear to be obvious trends. Also, the bottom red line is showing fluctuations different from the characteristic diurnal fluctuations with highest sap flow in the middle of the day, so that suggests some issue with the sensors or cable so it’s probably not trustworthy
Ideas for variation
Ideas for variation
Trends were fairly consistent across species except for sugar maples in C8 (why). Consider what it is about HB that could cause the transpiration response to be further along the timeline if comparing to W1, soil data, humidity and temperature, since these sites were measured about two weeks apart, also increase sample size for sugar maples and birch at C8 and Hubbard, look into stomata regulation, fine root growth that could possibly be influencing sap flow, a lot of possibilities. Stats to increase confidence in these results.