nph18047-sup-0001-SupInfo.pdf

Full text

New Phytologist Supporting Information
Article title: Trait phenology and fire seasonality co-drive seasonal variation in fire effects on
tree crowns
Authors: Nicole N. Bison, Raquel Partelli-Feltrin, Sean T. Michaletz
Article acceptance date: 6 February 2022
The following Supporting Information is available for this article:
Fig. S1 Forced convection data for Acer rubrum.
Table S1 Akaike information criterion (AIC) values for linear, continuous piecewise linear, and
non-continuous piecewise linear regression models.
Table S2 Piecewise linear regression results for seasonal variation in air temperature and fire
intensity.
Table S3 Piecewise linear regression results for phenological variation in bud traits versus
accumulated degree days.
Table S4 Piecewise linear regression results for seasonal variation in predicted bud necrosis
height versus accumulated degree days.
Notes S1 Broadleaf convection correlations.
Notes S2 R code for reproducing analyses and plots used in the study (see separate file).
Dataset S1 Fire intensity data (see separate file).
Dataset S2 Bud trait phenology data (see separate file).
Dataset S3 Convection data for Acer rubrum (see separate file).

Fig. S1 Forced convection data for Acer rubrum. Data were fit using standardized major axis
(SMA) Model II regression (P = 4 x 10-4, r2 = 0.54). Circular and triangular points represent data
collected at the Waterloo Wildlife Research Station in Athens County, Ohio (39.351103, -
82.262976), and the US Forest Service Northern Research Station office in Delaware, Ohio
(40.356559, -83.064635), respectively. Each point represents a bud on a foliated branch from a
different individual tree.

Table S1 Akaike information criterion (AIC) values for linear, continuous piecewise linear, and
non-continuous piecewise linear regression models.
Bolded values represent the model most preferred by AIC. Multiple bolded values in a given
row indicate that models are equally preferred. na, not applicable.
Taxon
Dependent variable
AIC, linear
regression
AIC, continuous
piecewise linear
regression
AIC,
non-continuous
piecewise linear
regression
na
Air temperature
78.49
22.85
15.05
Fireline intensity
244.48
242.68
239.55
P. glauca
Bud necrosis height
48.32
37.74
31.76
Volume
-557.22
-556.61
-611.09
Surface area
-291.48
-288.34
-342.53
Specific heat capacity
260.37
224.63
217.39
Heat transfer coefficient
76.31
78.98
35.10
Density
248.91
250.24
233.36
P. contorta
Bud necrosis
20.34
22.72
12.90
Volume
-575.48
-578.50
-659.23
Surface area
-303.18
-306.39
-370.58
Specific heat capacity
202.90
202.09
200.15
Heat transfer coefficient
112.45
88.59
49.65
Density
239.34
231.19
211.13
P. tremuloides
Bud necrosis
38.42
17.63
14.82
Volume
-721.28
-729.41
-727.74
Surface area
-428.00
-449.72
-447.77
Specific heat capacity
213.67
206.14
199.29
Heat transfer coefficient
28.28
-0.34
-1.18
Density
269.73
234.52
235.52

Table S2 Piecewise linear regression results demonstrating significant seasonal variation in air
temperature and fire intensity with accumulated degree days.
Dependent
variable
Intercept 1
Intercept 2
Slope 1
Slope 2
P value
r2
Air temperature
9.35
25.40
0.01
-0.01
4.05e-11
0.96
Fire intensity
492.14
322.83
-0.80
0.06
0.014
0.42

Table S3 Piecewise linear regression results demonstrating significant phenological variation for
bud traits of Picea glauca, Pinus contorta, and Populus tremuloides with accumulated degree
days.
Taxon
Bud
trait Intercept 1 Intercept 2 Slope 1 Slope 2 P value r2
P. glauca
ρ
1211.79
1062.08
-1.42
-0.14
1.62e-03
0.58
c
3698.95
3996.25
-1.83
-0.91
2.62e-11
0.96
V
-1.21e-06
1.65e-08
6.51e-09
5.98e-11
7.05e-10
0.94
h
85.56
78.57
-0.04
-1.26e-03
1.36e-07
0.88
A
-9.14e-04
-3.75e-06
4.94e-06
6.74e-08
2.15e-09
0.93
P. contorta
ρ
1102.31
1103.16
-1.11
-0.14e-01
1.86e-05
0.77
c
3415.68
3417.58
0.12
-0.17
8.61e-07
0.85
V
-7.27e-07
-3.80e-08
4.79e-09
1.70e-10
1.57e-14
0.99
h
39.36
47.33
-0.03
-9.99e-03
4.81e-11
0.96
A
-6.00e-04
-1.18e-05
4.14e-06
1.16e-07
4.62e-12
0.97
P. tremuloides
ρ
-116.18
1226.12
3.27
-0.07
2.25e-08
0.91
c
3592.64
3340.16
-1.13
-0.34
2.11e-09
0.93
V
6.41e-09
-1.18e-08
-2.41e-12
4.35e-11
7.21e-11
0.96
h
39.75
34.54
-0.03
-7.29e-03
1.36e-17
0.99
A
1.37e-06
-3.92e-05
2.62e-08
8.49e-08
8.63e-13
0.98
ρ, bud density; c, bud specific heat capacity; V, bud volume; h, bud heat transfer coefficient; A,
bud surface area

Table S4 Piecewise linear regression results showing significant seasonal variation in predicted
bud necrosis height with accumulated degree days for Picea glauca, Pinus contorta, and
Populus tremuloides.
Taxon
Intercept 1
Intercept 2
Slope 1
Slope 2
P value
r2
P. glauca
4.91
1.43
-7.55e-03
2.50e-03
5.99e-05
0.73
P. contorta
2.65
2.87
-2.77e-03
-8.71e-04
2.93e-02
0.36
P. tremuloides
4.78
1.73
-7.95e-03
5.18e-04
5.98e-05
0.73

Notes S1 Broadleaf convection correlations.
On 25 September 2008, we collected 19 foliated branches with attached buds from 19 different
Acer rubrum trees at the Waterloo Wildlife Research Station in Athens County, Ohio
(39.351103, -82.262976), and the US Forest Service Northern Research Station office in
Delaware, Ohio (40.356559, -83.064635). Branches were enclosed with moist paper towels in
large plastic bags, and shipped in a cooler with dry ice to Calgary, Alberta for measurement of
convection data. Convection data were measured in a laminar flow wind tunnel on 1 October
2009 following the methods of Michaletz and Johnson (2006). In brief, thermocouples were
inserted into a bud on each branch, branches were frozen in a cooler with dry ice, then frozen
branches were heated using room temperature air (~22.5 °C) across a range of wind velocities.
Forced convection data were characterized using a power law of the form
Nu Re
n
B=
. Here,
the Nusselt number
/Nu hD k=
(dimensionless) and the Reynolds number
Re /UD
ν
=
(dimensionless), where h (W m-2 °C-1) is the heat transfer coefficient, D (m) is the bud diameter,
k (2.37 x 10-2 W m-1 °C-1 ; Vargaftik, 1975) is the thermal conductivity of air, U (m s-1) is the wind
tunnel free steam velocity, and ν (1.5 x 10-5 m2 s-1; Vargaftik, 1975) is the kinematic viscosity of
air. To estimate the normalization constant B (dimensionless) and the scaling exponent n
(dimensionless), the power law was log10-transformed and rearranged to give the linear form
log(Nu) log( ) log(Re)
Bn= +
, where the slope n and y-intercept log(B) were obtained using
standardized major axis (SMA) model II regression (Warton et al. 2006, Warton et al. 2012). Nu
increased significantly with Re (P = 4 x 10-4, r2 = 0.54), with estimated parameters B = 0.0455
and n = 0.613.
References
Michaletz ST, Johnson EA. 2006. Foliage influences forced convection heat transfer in conifer
branches and buds. New Phytologist 170: 87-98.
Vargaftik NB. 1975. Tables of thermophysical properties of liquids and gases, 2nd edn. New
York, NY, USA: Hemisphere Publishing.

Warton DI, Duursma RA, Falster DS, and Taskinen S. 2012. smatr 3 - an R package for
estimation and inference about allometric lines Methods in Ecology and Evolution 3(2):
257-259.
Warton DI, Wright IJ, Falster DS, and Westoby M. 2006. Bivariate line-fitting methods for
allometry. Biological reviews 81(2): 259-291.