Black cottonwood trees

show limited response to Lenora W. Davis*, Hilary Rose Dawson, Jamie Wright, Lucas Silva
historical flood event *Corresponding author lenorad@uoregon.edu

Background Methods Species (black cottonwood; Results

1996 Flood
Like fingerprints or barcodes, tree rings hold information Populus trichocarpa) criteria: Forest
n=7
about the tree (e.g. age, growth) and historical ● Abundant 0.3% increase
from mean
environmental conditions, (e.g. droughts or floods). ● Widespread in site
Sample tree criteria:
Climate change changes flood frequency and intensity with
● Old/thick enough (>10in
more precipitation and less snowpack1. Floods can
2 diameter)
decrease carbon storage with reduced tree growth rates.

Ring Width Index (RWI)

● Not rotten
Common impacts of floods include uprooting trees and North Slough
Data collection: n=8
reduced growth from sediment, reduced water and 5% increase from
1 ● GPS tagging mean
nutrient uptake, producing decreased ring widths .
● DBH (diameter at breast
Reconstructed flood information from tree growth might
height)
predict some issues that climate change presents.
○ DBH range was 17-73 in
The McKenzie River flooded in February 19963 ● Increment borer core
● Record rainfall and warm temps melted snowpack extraction
South Slough
● It is unknown whether it reached the entire study site Figure 2: Author Lenora Davis coring a ○ Age range was 24-78+ n=3
Populus trichocarpa in the North Slough 18% decrease from
zone of the site.
years mean
1. Did the 1996 McKenzie River flood affect tree growth? ○ 22 trees were cored, 4 cores
2. How did proximity from the river during the 1996 flood discarded for age or error
affect tree growth? Tree core preparation
● Sanded and scanned cores at high resolution
Study Site
5
● Annular growth rings counted and measured in ImageJ
Figure 4: RWI time series by zone. Shaded area represents standard deviation. Blue line
marks the 1996 flood. Percent change compares 1996 growth to mean of all other years.

● Land is the 1. The flood did not sig. affect tree growth (P = 0.06), but
indigenous Figure 3: Populus trichocarpa tree core with increased contrast for ring visibility.
growth decreased compared to previous year.
homeland of Tree core analysis 2. Proximity from river did not affect tree growth (distance
the Kalapuya4 ● Detrended widths using ring width index (RWI) in R 6, 7
from river P = 0.81, interaction term P = 0.65).
Future Research
○ Standardizes ‘age trend’ where younger rings are larger
due to a smaller trunk diameter
● Why did the trees experience such limited disturbance?
● Cropped data to 10 years on either side of the 1996 flood
● How does tree age affect flood response?
(1986-2006). ● How did the flood affect other species on the site?
○ Sample depth greatly decreased further back ● How did the 1964 flood impact tree growth?
Source: EWEB, Google, UO

● Tested significance with an ANOVA test ● What can old trees on the site reveal about past climatic events?
Figure 1a: Map of study
site, Springfield, OR with

Acknowledgements & Citations

Lane County Flood
Hazard Zones and tree
sampling points. Made in Thank you to all Soil Plant Atmosphere Research Lab members for their support with this project, especially Mike Farinacci for assistance in the field and sharing knowledge of tree
QGIS. cores. Thank you to Owen Jetton for assistance with R coding and graphics. This project with UO Noyce ESPRIT is funded by NSF DUE 1660724.
1b inset: University of 1
Ballesteros-Cánovas et. al. (2015). A review of flood records from tree rings. 2Ellison et, al. (2017). Trees, forests and water: Cool insights for a hot world. 3Lane Co. (2016). Flood of
Oregon in relation to the ‘96. 4Lewis. (2016) Native Community History of Eugene Area. 5Rueden et al. (2017). ImageJ2. 6R Core Team.(2020). www.R-project.org. 7Bunn et al. (2020). dplR v. 1.7.1.
Lenora Davis, 2020.
field site. Source: EWEB, LCOG, FEMA https://CRAN.R-project.org/package=dplR