Tree Physiology 28, 105–112
© 2008 Heron Publishing—Victoria, Canada
Estimating the onset of cambial activity in Scots pine in northern
Finland by means of the heat-sum approach
JEONG-WOOK SEO,1,2 DIETER ECKSTEIN,1 RISTO JALKANEN,3 SOPHIE RICKEBUSCH4
and UWE SCHMITT5
1
University of Hamburg, Department of Wood Science, Division Wood Biology, Leuschnerstrasse 91, D-21031 Hamburg, Germany
2
Corresponding author ([email protected])
3
Finnish Forest Research Institute, Rovaniemi Research Unit, Box 16, FI-96301 Rovaniemi, Finland
4
Swiss Federal Research Institute WSL, Land Use Dynamics Research Unit, Zuercherstrasse 111, CH-8903 Birmensdorf, Switzerland
5
Federal Research Centre for Forestry and Forest Products, Institute for Wood Biology and Wood Protection, Leuschnerstrasse 91, D-21031
Hamburg, Germany
Received March 14, 2007; accepted May 18, 2007; published online October 15, 2007
Summary We estimated the date of onset (Dateest) of
cambial activity by the pinning method in Scots pine (Pinus
sylvestris L.) trees at Vanttauskoski (Site 1) and Laanila (Site 2)
near the latitudinal limit of Scots pine in northern Finland. In
each year and at each site, observations were made on a different set of five trees. The estimated dates of onset of cambial
activity were compared with the corresponding heat sums, calculated in degree-days according to two models. Within years,
Dateest varied among trees by up to 15 days at Site 1 and up to 13
days at Site 2. Among years, mean Dateest varied by 15.3 days at
Site 1 and 12.0 days at Site 2. The overall mean Dateest differed
between sites by 6 days (June 5 at Site 1 and June 11 at Site 2).
Among all trees in all years, the mean number of degree days
(d.d.) calculated from mean daily temperature above a threshold of 5 °C before Dateest ranged from 68.7 to 135 d.d. at Site 1
and from 37.4 to 154.7 d.d. at Site 2. Among years, the mean
heat sum before Dateest ranged from 94 to 112.5 d.d. at Site 1
and from 61.4 to 136 d.d. at Site 2. Variation among years in
heat sum before Dateest at Site 2 was highly significant, indicating that one or more factors other than, or in addition to, heat
sum determines the onset of cambial activity in Scots pine.
Similar results were obtained when heat sum was computed
from the area between the sine wave generated by daily maximum and minimum temperature and the threshold temperature.
Keywords: boreal forest, degree days, Pinus sylvestris, wood
formation.
been used to predict the onset, intensity, duration and cessation
of phenological activities in both herbaceous plants (e.g.,
George et al. 1988, Gillen and Ewing 1992) and trees (e.g.,
Sarvas 1972, Kramer 1994, Karlsson et al. 2003).
Trees at their upper latitudinal and elevational distribution
limits are regularly exposed to large diurnal and seasonal temperature fluctuations. Meristematic tissues are particularly
vulnerable to frost damage in late spring and early autumn
(Häkkinen et al. 1995), but early expansion of vegetative tissue can be advantageous for maintaining site dominance
(Bailey and Harrington 2006), provided that the risk of damage by spring frost is not too great (Hannerz 1999).
Summer temperature is the principal controller of radial
growth of Scots pine in northern ecotones (Laitakari 1920,
Hustich 1948, Mikola 1950, Sirèn 1961). Due to this strong
and stable climatic signal, boreal pines have been used successfully as “climate archives” for past summer temperatures
(e.g., Briffa et al. 1990, Helama et al. 2002, McCarroll et al.
2003). However, knowledge of intra-annual dynamics of wood
formation and its relation to temperature is poor. The objectives of this study were to determine by direct microscopic observation the time of onset of wood formation of Scots pine in
the Finnish northern boreal zone and examine its relationship
with the accumulated heat sum calculated in d.d. according to
two models. One model, based on that of Sarvas (1972), computes heat sum from daily mean temperature. The other model,
based on that of Allen (1976), computes heat sum from the
area beneath a sine wave, defined by the daily maximum and
minimum temperatures, and the threshold temperature.
Introduction
Material and methods
Ecologists often use heat accumulation above a threshold
temperature, most commonly expressed in degree-days (d.d.),
to predict the effect of temperature on biological processes
(Baskerville and Emin 1969). For example, heat sums have
Study sites and trees
Scots pine (Pinus sylvestris L.) trees were selected at two sites
in the northern boreal zone of Finland: Vanttauskoski (Site 1,
106
SEO, ECKSTEIN, JALKANEN, RICKEBUSCH AND SCHMITT
66°22′ N 26°43′ E, 150 m a.s.l.) and Laanila (Site 2, 68°30′ N
27°30′ E, 220 m a.s.l.). Sites 1 and 2 are located about 300 and
80 km, respectively, south of the northern limit of Scots pine’s
distribution. The Arctic Circle goes through Site 1 (Figure 1).
At both sites, the intra-annual cambial activity of a different
set of five dominant trees was monitored weekly over the
growing seasons each year from 2000 to 2004 (Table 1), except in 2004, when only four trees were monitored at Site 1.
The study trees were felled at the end of the year of observation. Mean age at breast height of the sampled trees was 43
years at Site 1 and 42 years at Site 2.
through the bark into the outer xylem, causing the death of
cambial cells adjacent to the pin and inducing the formation,
further away, of modified cells (Figure 2). These features were
observed by light microscopy in cross sections of the stems of
the felled sample trees. The calendar date of pinning could be
matched to the site of abnormal cell formation, thus making it
possible to count the cells between pinning dates and to estimate the date of onset and termination of cambium activity.
On each date, a pin was inserted, about 1.3 m above ground,
at three points about 2.5 cm apart (Seo et al. 2007, Figure 2).
Cambial activity was monitored at a single stem height to
avoid the complication of vertical differences in the timing of
cambial activity (Savidge and Wareing 1984).
The trees were pinned weekly from the end of May to the
end of August allowing the determination of the dates of onset
and termination of cambial activity with an accuracy of about
1 week. The better than 1-week resolution of the time of onset
of cambial activity was based on two assumptions: (1) that on a
weekly time scale, cambial activity proceeds at a constant rate,
and (2) that the cambium of the tree forming the most tracheids
during the week in which the onset of wood formation was observed had been active for the entire week.
The day when the onset of wood formation was observed
was called Dateobs and the day when wood formation was estimated to have started was called Dateest (Table 2). In the
1-week interval before Dateobs, trees formed n tracheids; n tracheids as a percentage of the total number of tracheids formed
by each tree during a study year is called n%. The value referred to as maximum percentage of tracheids per day (Max %
per day) is the highest n% value of any tree in a given year. The
number of days required by each tree to produce n tracheids
before Dateobs (Days, in Table 2) was then calculated by dividing n% by max % per day. This value was then subtracted from
Dateobs to obtain Dateest.
Determining the onset of wood formation
Meteorological data
The onset of cambial activity was determined by the pinning
method (Wolter 1968). A 1.2-mm-diameter pin was inserted
Mean (Tmean), maximum (Tmax) and minimum (Tmin) daily temperatures were measured 2 m above the soil surface; climate
Figure 1. Location of the study sites (䊉) and meteorological stations
(䉱). Site 1: Vanttauskoski and Rovaniemi; Site 2: Laanila and Ivalo.
Table 1. Study trees sampled at Sites 1 (Vanttauskoski) and 2 (Laanila). Abbreviation: SD = standard deviation.
Year
ID
No. of
trees
Height (SD)
(m)
DBH (SD)
(cm)
Mean age and age range
(years) at breast height
Site 1
2000
2001
2002
2003
2004
V01–V05
V06–V10
V11–V15
V16–V20
V21–V24
5
5
5
5
4
15.5 (0.7)
14.8 (1.0)
14.9 (0.3)
14.6 (0.9)
15.5 (0.5)
20.2 (2.6)
17.0 (1.0)
16.8 (0.8)
16.6 (1.7)
18.0 (1.2)
49.4 (46–54)
41.8 (41–43)
42.6 (40–45)
39.8 (37–43)
42.0 (40–44)
Site 2
2000
2001
2002
2003
2004
L01–L05
L06–L10
L11–L15
L16–L20
L21–L25
5
5
5
5
5
11.1 (1.2)
11.0 (0.9)
11.7 (1.1)
10.9 (0.9)
9.8 (1.0)
14.2 (0.8)
14.4 (1.3)
14.6 (1.1)
15.2 (1.1)
15.0 (0.7)
42.6 (40–49)
43.4 (36–47)
43.2 (38–47)
43.2 (42–44)
37.2 (33–39)
TREE PHYSIOLOGY VOLUME 28, 2008
HEAT-SUM DEVELOPMENT AND THE ONSET OF CAMBIAL ACTIVITY
107
Figure 2. Illustration of the pinning method; (a) a diagramatic representation of the pinning device and a photograph of Tree 21 after pinning on
four dates. On each date the pin was inserted three times in a spiral pattern 1.3 m aboveground; (b) transverse (T) and radial (R) sections showing
the pin inserted through the phloem (Ph) and cambium (Ca) into the xylem (EW, earlywood; LW, latewood).
data were obtained from the Rovaniemi and Ivalo meteorological stations, which are located close to Sites 1 and 2, respectively (Figure 3).
Calculation of degree-day sum (d.d.)
We calculated heat sum (d.d.) in two ways. Model 1, a modification of the method of Sarvas (1972), computes degree-day
sum (d.d.) from daily mean temperature (Figure 4a) as:
d.d. =
365
∑( TDiff ) i
(1)
i=j
 Ti – 5
TDiff = 
0
d.d. hd1
0
for T min ≤ k, Tmax ≤ k

1
(2)
=  2 ( Thd1 – k)
for T min ≥ k, Tmax ≥ k
1
π
 2 π ( Thd1 – k)( 2 – θ) + α cos θ otherwise
[
]
where
k − Thd1
)
α
α = 12 ( Tmax – Tmin)
θ = arcsin(
Thd1 = 12 ( Tmax + Tmin)
for Ti > 5
otherwise
where d.d. is the sum of TDiff, i.e., the sum of the differences between the daily mean temperature (Ti) and the threshold of
+5 °C; j is the day of year (DOY) when the mean daily temperature is greater than or equal to the threshold for at least five
consecutive days (arrowhead in Figure 4a).
Model 2, which is a modification of the model of Allen
(1976), is based on daily maximum and minimum temperatures (Figure 4b) and computes the area between the sine wave
generated by these variables and a threshold value, in this case
+5 °C. The d.d. for the first half-day (d.d.hd1, black area in Figure 4b) is computed as:
and k is the threshold temperature, α is the amplitude of the
sine curve and θ is the point (in radians) at which the sine curve
crosses the threshold k.
This computation is repeated with Tmax and Tmin+1 (minimum
temperature of the following day) to obtain d.d.hd2 (light gray
area in Figure 4b); d.d.d is the sum of the two half-days and
d.d.y is the sum for all the days in year y:
d.d. d = d.d. hd1 + d.d. hd2
d.d. y =
∑ d.d. d
for d = 1 to 365
(3)
In Figure 4b, Case 1 (Equation 2) occurs twice in Day 1, Case
2 occurs in the second half of Day 3 and in Day 4, and Case 3
occurs in Day 2 and the first half of Day 3.
TREE PHYSIOLOGY ONLINE at http://heronpublishing.com
108
SEO, ECKSTEIN, JALKANEN, RICKEBUSCH AND SCHMITT
The long-term mean of the annual heat sum accumulated
between spring and autumn was calculated for the 1961 to
1999 period at each site. The heat sum calculation started in
spring as indicated in Figure 4a and ended in autumn, when the
temperature dropped below the +5 °C threshold for five consecutive days. This long-term mean is used as a site-specific
reference.
Comparison of Dateest and the accumulated heat sum
The non-parametric Kruskal-Wallis rank sum test (Breslow
1970) was performed to assess whether the means of the dates
for the onset of wood formation and of the accumulated d.d.
differed significantly between study years.
Results
Onset of wood formation
Onset of wood formation was determined by observations on
traverse and radial sections of stems pinned throughout the
growing season in the years 2000 through 2004 (Tables 2a and
2b, column Dateest). Within years, Dateest differed among trees
by up to 15 days at Site 1 and up to 12 days at Site 2, reflecting
within-population genotypic variation, microsite differences,
or error in the determination of Dateest. Among years, mean
Dateest differed by up to 15 days at Site 1 and up to 12 days at
Site 2, reflecting both variation among sample tree populations and climatic differences among years. For all trees and
Table 2a. Onset of wood formation in Scots pine and corresponding degree-day sums (d.d.) calculated according to Models 1 and 2 at Site 1
(Vanttauskoski). Abbreviations: ID is the tree identification number; Dateobs is the first date when newly formed wood was observed; n number of
tracheids newly formed before Dateobs; n% is the value of n as a percentage of the total number of tracheids formed in the study year (bolded text
marks the highest n% value in a study year); Max % per day is the maximum percentage of tracheids formed per day before Dateobs, calculated
from the highest n% value among the sample trees divided by 6; Days is the number of days to produce n% tracheids; Dateest is the estimated date
of onset of wood formation; and SD is standard deviation. Dateobs and Dateest are given as day of year (DOY).
Year
2000
ID
V01
V02
V03
V04
V05
Dateobs
(DOY)
158
158
158
158
158
Tracheids
n
3.0
1.5
2.5
1.5
3.0
Days
n%
7.5
5.1
8.3
3.5
8.3
Max %
per day
1.4
1.4
5
4
6
2
6
Mean (± SD)
2001
V06
V07
V08
V09
V10
162
162
162
169
169
2.0
1.0
1.0
1.5
4.0
4.9
3.3
2.3
3.8
7.1
1.2
4
3
2
3
6
Mean (± SD)
2002
V11
V12
V13
V14
V15
161
154
147
147
147
4.0
2.0
1.0
4.0
1.5
9.3
5.5
1.8
4.7
3.1
1.6
6
4
1
3
2
Mean (± SD)
2003
V16
V17
V18
V19
V20
153
153
153
167
167
5.0
3.5
4.5
6.5
2.5
9.4
5.5
9.5
16.5
6.9
2.7
4
2
4
6
3
Mean (± SD)
2004
V21
V22
V23
V24
173
166
166
166
2.5
4.0
3.0
3.0
13.9
12.7
6.4
8.1
2.3
6
5
3
4
Mean (± SD)
TREE PHYSIOLOGY VOLUME 28, 2008
Dateest
(DOY)
Degree-days (d.d.)
Model
1
Model
2
153
154
152
156
152
110.7
117.5
105.5
123.3
105.5
135.6
142.0
130.5
147.5
130.5
153.4 (1.7)
112.5 (7.8)
137.2 (7.4)
158
159
160
166
163
68.7
76.9
85.0
135.0
104.2
112.6
121.4
129.8
183.0
150.9
161.2 (3.3)
94.0 (26.5)
139.5 (28.1)
155
150
146
144
145
171.2
116.9
76.7
70.6
71.5
192.5
141.8
105.1
96.8
100.2
148.0 (4.5)
101.4 (43.5)
127.3 (40.7)
149
151
149
161
164
85.3
91.8
85.3
128.0
132.8
105.8
112.6
105.8
149.9
156.3
154.8 (7.2)
104.6 (23.7)
126.1 (24.9)
167
161
163
162
129.0
95.4
106.5
100.7
156.2
123.9
134.2
128.9
163.3 (2.6)
107.9 (14.8)
135.8 (14.2)
HEAT-SUM DEVELOPMENT AND THE ONSET OF CAMBIAL ACTIVITY
109
Table 2b. Onset of wood formation in Scots pine and corresponding degree-day sums (d.d.) calculated according to Models 1 and 2 at Site 2
(Laanila). Abbreviations: ID is the tree identification number; Dateobs is the first date when newly formed wood was observed; n number of tracheids newly formed before Dateobs; n% is the value of n as a percentage of the total number of tracheids formed in the study year (bolded text
marks the highest n% value in a study year); Max % per day is the maximum percentage of tracheids formed per day prior to Dateobs, calculated
from the highest n% value among the sample trees divided by 6; Days is the number of days to produce n% tracheids; Dateest is the estimated date
of onset of wood formation; and SD is standard deviation. Dateobs and Dateest are given as day of year (DOY).
Year
2000
ID
L01
L02
L03
L04
L05
Dateobs
(DOY)
157
164
164
164
164
Tracheids
Days
n
n%
2.5
3.5
2.5
3.5
6.0
9.3
8.5
5.8
9.2
15.8
Max %
per day
2.6
4
3
2
4
6
Mean (± SD)
2001
L06
L07
L08
L09
L10
169
163
169
163
163
3.0
2.5
7.0
2.5
2.5
7.0
8.6
19.4
7.9
4.9
3.2
2
3
6
2
2
L11
L12
L13
L14
L15
161
161
161
161
161
1.5
3.0
2.5
2.5
5.0
5.1
8.0
9.1
10.2
14.5
2.4
2
3
4
4
6
Mean (± SD)
2003
L16
L17
L18
L19
L20
160
167
167
167
174
2.5
6.0
6.5
3.0
3.5
7.6
15.0
15.3
8.1
11.7
2.6
3
6
6
3
5
Mean (± SD)
2004
153
161
162
160
158
158.8 (3.6)
Mean (± SD)
2002
Dateest
(DOY)
L21
L22
L23
L24
L25
173
173
173
173
173
4.5
7.5
4.0
6.5
4.5
23.7
15.8
7.8
15.5
12.0
4.0
Mean (± SD)
6
4
2
4
3
167
160
163
161
161
Heat sum versus the date of onset of wood formation
The heat unit approach (as defined by Wang 1960) is based on
the assumption that the biological process of wood formation
begins as soon as a certain amount of heat has been accumulated over time. However, there was a strikingly high variability in the d.d. sums calculated according to Model 1 until the
onset of wood formation (Tables 2a and 2b). In absolute values, d.d. before Dateest for individual trees ranged from 68.7 to
135.0 at Site 1 and from 37.4 to 154.7 at Site 2.
Model
1
Model
2
60.6
91.0
96.7
88.7
80.6
71.7
104.5
110.8
101.9
93.4
83.5 (14.1)
96.8
37.4
68.9
51.9
51.9
96.5 (15.2)
110.2
90.0
116.8
102.7
102.7
162.4 (2.8)
61.4 (22.7)
104.5 (10.0)
159
158
157
157
155
154.7
144.7
133.7
133.7
113.4
172.5
162.4
150.7
150.7
132.2
157.2 (1.5)
136.0 (15.4)
153.7 (15.1)
157
161
161
164
169
81.9
87.8
87.8
88.9
101.1
117.7
127.6
127.6
130.8
144.7
162.4 (4.5)
89.5 (7.0)
129.7 (9.7)
167
169
171
169
170
87.4
89.1
100.5
89.1
93.3
113.8
117.0
127.7
117.0
121.1
91.8 (5.3)
119.3 (5.4)
169.2 (1.5)
years, Dateest occurred 6 days earlier at Site 1 (June 5) than at
Site 2 (June 11).
Degree-days (d.d.)
Interannual variation in the relationship between onset of
cambial activity and accumulated heat sum
Heat sum before Dateest varied by 18.5 d.d at Site 1 and 74.6
d.d. at Site 2. The variation among years at Site 2 was highly
significant, indicating that a factor or factors other than, or in
addition to, heat sum determines the onset of cambial activity
in P. sylvestris (Table 3).
Heat sum determined by Model 1 versus Model 2
Model 2 generally yielded higher d.d. values than Model 1, but
the variability of the standard deviation was about the same.
The variation in the standard deviation differed between years;
TREE PHYSIOLOGY ONLINE at http://heronpublishing.com
110
SEO, ECKSTEIN, JALKANEN, RICKEBUSCH AND SCHMITT
Figure 3. Daily mean (bold), maximum (plain), and minimum (dotted) temperature at the Rovaniemi
and Ivalo meteorological stations
from January to June during the
five monitoring years; the shaded
background indicates the area
above the +5 °C threshold.
it was low in 2000 and 2004 at Site 1 and in 2003 and 2004 at
Site 2, but high in 2002 (Site 1) and 2001 (Site 2).
Discussion
In northern Finland, height growth of Scots pine trees begins
between the end of April and the end of May (Salminen and
Jalkanen 2007) and radial growth begins between the end of
May and mid-June (Hustich 1956). The growing season, if defined as the number of days when the daily mean temperature
is above +5 °C, is between 110 and 130 days (Jalkanen 2005),
but if defined in terms of wood formation (height and radial
growth) it is, depending on the year, 77 to 93 days at Site 1 and
58 to 75 days at Site 2. Based on dendrometer readings,
Mielikäinen et al. (1998) estimated that about 100 d.d. are necessary to trigger the onset of tree-ring formation. Our results,
based on the same heat-sum model as in Mielikäinen et al.
(1998), show a 5-year mean of 104 and 93 d.d. for the onset of
wood formation at Sites 1 and 2, respectively. Schmitt et al.
(2004), working from direct microscopic observations of
cambial activity, calculated 85 to 90 d.d. for the onset of wood
formation for the same sites in 1996, which was the coldest
summer in northern Finland during the last 15 years.
In our study, Scots pine close to the northern border of its
distribution (Site 2) started radial growth at breast height between June 1 and 19 depending on the year, corresponding to a
heat sum (Model 1) of between 61 and 136 d.d. (mean of 92
d.d.). About 220 km south (Site 1), the onset of wood formation occurred on average 6 days earlier and with a heat sum
higher by 12 d.d., which suggests that Scots pine has adapted
to a lower heat sum (minimum value: 61 d.d. in 2001) for triggering cambial activity at the northern site than at the southern
site. Wang and Perry (1958) and Johnsson (1974) made a similar observation in birches (Betula verrucosa Ehrh., B. pub-
TREE PHYSIOLOGY VOLUME 28, 2008
HEAT-SUM DEVELOPMENT AND THE ONSET OF CAMBIAL ACTIVITY
111
Table 3. Summary of the Kruskal-Wallis rank sum test at Sites 1
(Vanttauskoski) and 2 (Laanila). Abbreviations: Dateest, estimated
date of the actual onset of wood formation; and df, degrees of freedom.
Dateest
Figure 4. Illustration of the calculation of degree-days (d.d.) according to two models; (a) Model 1 adds the differences between daily
mean temperatures and the +5 °C threshold, starting with the day (arrowhead) when the daily mean temperature is greater than or equal to
the threshold for at least 5 consecutive days for the first time during a
year; (b) Model 2 adds the areas above the +5 °C threshold and below
the sine wave curve drawn through the daily maximum and minimum
temperatures.
escens Ehrh.). Heat sum is possibly not as strictly limiting for
the onset of wood formation further south or it is not the only
limiting factor, for otherwise the trees at Site 1 would have
started growing earlier than was observed. According to Karlsson et al. (2003), leaf emergence in deciduous trees at high latitudes occurs close to the summer solstice. The maximum
growth rate of conifers in cold environments is also achieved
around the time of maximum day length (Rossi et al. 2006). It
may be more favorable, in terms of plant carbon economy, to
start growing as early as possible rather than prolong the
photosynthetic season further into the autumn (Karlsson
1989), to provide enough time to complete cell wall formation
and lignification before winter (Rossi et al. 2006). Timing becomes more important with increasing latitude, because the
growing season shortens toward the north. The growing season is much longer at our southern site (Site 1) than at the
northern site (Site 2). Northern trees therefore need to go
through all the phenological stages with a smaller heat sum, so
that they can complete all growth stages in time to survive the
winter. A similar compromise between maximizing the period
of photosynthetic activity and minimizing the risk of frost
damage is reached in regulating the onset of growth in spring
(Linkosalo et al. 2000), which presumably explains why conditions favorable to spring growth were not fully used by the
Scots pine trees at Site 1.
Acknowledgments
We thank the staff of the Laanila and Kivalo research areas of the
Finnish Forest Research Institute for carrying out the pinning work in
the field. We are grateful to Tarmo Aalto, Pekka Närhi and Reino
Vierelä for cutting the sample trees and delivering the pinned discs.
The study was funded by the EU-project PINE “Predicting Impacts on
Natural Ecotones” (EVK2-CT-2002-00136).
Degree-days
Model 1
Model 2
Site 1
df
Chi-squared
P-value
4
14.2
0.007
4
3.1
0.540
4
2.1
0.720
Site 2
df
Chi-squared
P-value
4
16.2
0.003
4
14.8
0.005
4
20.9
0.000
References
Allen, J.C. 1976. A modified sine wave method for calculating degree
days. Environmental Entomol. 5:388–396.
Bailey, J.D. and C.A. Harrington. 2006. Temperature regulation of
bud-burst phenology within and among years in a young
Douglas-fir (Pseudotsuga menziesii) plantation in western Washington, USA. Tree Physiol. 26:421–430.
Baskerville, G.L. and P. Emin. 1969. Rapid estimation of heat accumulation from maximum and minimum temperatures. Ecology
50:514–517.
Breslow, N. 1970. A generalized Kruskal-Wallis test for comparing K
samples subject to unequal patterns of censorship. Biometrika
57:579–594.
Briffa, K.R., T.S. Bartholin, D. Eckstein, P.D. Jones, W. Karlén, F.H.
Schweingruber and P. Zetterberg. 1990. A 1,400-year tree-ring record of summer temperatures in Fennoscandia. Nature 346:
434–439.
George, M.R., C.A. Raguse, W.J. Clawson, C.B. Wilson, R.L. Willoughby, N.K. McDougald, D.A. Duncan and A.H. Murphy. 1988.
Correlation of degree-days with annual herbage yields and livestock gains. J. Range Manage. 41:193–197.
Gillen, R.L. and A.L. Ewing. 1992. Leaf development of native
bluestem grasses in relation to degree-day accumulation. J. Range
Manage. 45:200–204.
Häkkinen, R., T. Linkosalo and P. Hari. 1995. Methods for combining
phenological time series: application to bud burst in birch (Betula
pendula) in Central Finland for the period 1896–1955. Tree
Physiol. 15:721–726.
Hannerz, M. 1999. Evaluation of temperature models for predicting
bud burst in Norway spruce. Can. J. For. Res. 29:9–19.
Helama, S., M. Lindholm, M. Timonen, J. Meriläinen and M. Eronen.
2002. The supra-long Scots pine tree-ring record for Finnish Lapland: Part 2, interannual to centennial variability in summer temperatures for 7500 years. Holocene 12:681–687.
Hustich, I. 1948. The Scotch pine in northernmost Finland and its dependence on the climate in the last decades. Acta Bot. Fenn.
42:1–75.
Hustich, I. 1956. Notes on the growth of Scotch pine in Utsjoki in
northernmost Finland. Acta Bot. Fenn. 56:1–13.
Jalkanen, R. 2005. Metsäympäristön ja –talouden erityispiirteet.
Metsäympäristö ja luonnonolot. In Metsätaloutta kairoilla— Metsänuudistaminen Pohjois-Suomessa. Eds. M. Hyppönen, V. Hallikainen and R. Jalkanen. Metsälehti Kustannus, Helsinki, pp 10–28.
TREE PHYSIOLOGY ONLINE at http://heronpublishing.com
112
SEO, ECKSTEIN, JALKANEN, RICKEBUSCH AND SCHMITT
Johnsson, H. 1974. Genetic characteristics of Betula verrucosa Ehrh.
and B. pubescens Ehrh. Ann. For. 6:91–133.
Karlsson, P.S. 1989. In situ photosynthetic performance of four coexisting dwarf shrubs in relation to light in a subarctic woodland.
Funct. Ecol. 3:481–487.
Karlsson, P.S., H. Bylund, S. Neuvonen, S. Heino and M. Tjus. 2003.
Climatic response of budburst in the mountain birch at two areas in
northern Fennoscandia and possible responses to global change.
Ecography 26:617–625.
Kramer, K. 1994. Selecting a model to predict the onset of growth of
Fagus sylvatica. J. Appl. Ecol. 31:72–181.
Laitakari, E. 1920. Tutkimuksia sääsuhteiden vaikutuksesta männyn
pituus ja paksuuskasvuun. Acta For. Fenn. 17:1–53. With summary in German.
Linkosalo, T., T.R. Carter, R. Häkkinen and P. Hari. 2000. Predicting
spring phenology and frost damage risk of Betula spp. under climatic warming: a comparison of two models. Tree Physiol. 20:
1175–1182.
McCarroll, D., R. Jalkanen, S. Hicks, M. Tuovinen, M. Gagen, F.
Pawellek, D. Eckstein, U. Schmitt, J. Autio and O. Heikkinen.
2003. Multiproxy dendroclimatology: a pilot study in northern Finland. Holocene 13:831–841.
Mielikäinen, K., P. Nöjd and M. Timonen. 1998. Puun muisti. Kasvun
vaihtelu päivästä vuosituhanteen. Metsäntutkimuslaitoksen tiedonantoja 703, 71 p.
Mikola, P. 1950. Puiden kasvun vaihteluista ja niiden merkityksesta
kasvututkimuksissa. Commun. Inst. For. Fenn. 38:1–131. With
summary in English.
Rossi, S., A. Deslauriers, T. Anfodillo, H. Morin, A. Saracino,
R. Motta, R. Borghetti and M. Borghetti. 2006. Conifers in cold environments synchronize maximum growth rate of tree-ring formation with day length. New Phytol. 170:301–310.
Salminen, H. and R. Jalkanen. 2007. Intra-annual height increment of
Pinus sylvestris at high latitudes in Finland. Tree Physiol. 27:
1347–1353.
Sarvas, R. 1972. Investigations on the annual cycle of development of
forest trees. Active period. Commun. Inst. For. Fenn. 76:1–110.
Savidge, R.A. and P.F. Wareing. 1984. Seasonal cambial activity and
xylem development in Pinuscontorta in relation to endogenous
indol-3-yl-acetic and (S)-abscisic acid levels. Can. J. For. Res.
14:676–682.
Schmitt, U., R. Jalkanen and D. Eckstein. 2004. Cambium dynamics
of Pinus sylvestris and Betula spp. in the northern boreal forest in
Finland. Silva Fenn. 38:167–178.
Seo, J-.W., D. Eckstein and U. Schmitt. 2007. The pinning method:
from pinning to data preparation. Dendrochronologia. In press.
Sirèn, G. 1961. Skogsgränstallen som indicator för klimatfluktuationerna i norra Fennoskandien under historisk tid. Commun. Inst.
For. Fenn. 54:1–66.
Wang, J.Y. 1960. A critique of the heat unit approach to plant response
studies. Ecology 41:785–790.
Wang, C.W. and T.O. Perry. 1958. Ecotypic variation of dormancy,
chilling requirement, and photoperiodic response of Betula species. Proc. 10th Intern. Congr. of Genetics, Montreal, 2:307.
Wolter, E.K. 1968. A new method for marking xylem growth. For.
Sci. 14:102–104.
TREE PHYSIOLOGY VOLUME 28, 2008