printed by
www.postersession.com
From 1990From 1990--2006 we observed the timing of woody vegetation development during the growing season. Bud break (50% leaf emergence) and leaf development (75% final size) on two to five permanently tagged individuals of each species are recorded at 3-7 day intervals from April through June. We have also recorded fall phenology since 1991, with the exception of 1992. Weekly observations of percent leaf coloration and percent leaf fall begin in September and continue through leaf fall.
These observations have documented substantial (up to three weeks difference) interannual variation in the timing of leaf emergence (50%) and leaf development (75% final size), but good relative consistency among species and among individuals within species. In general the fall dates have shown less variability than the spring dates. Although no trend has been evident in the timing of leaf emergence and development in the spring over the course of the study, these trees have shown a trend toward later leaf senescence and leaf fall of up to five days over the course of the study.
-The relative timing of observed bud break, leaf development and leaf fall were fairly consistant among individuals within a species and among means for species.
-The timing of spring events among individuals within a species and among means for species exhibited about twice the range of variability observed in the timing of leaf fall.
-Although spring events remain more variable there is greater consistent separation among species means in the timing of leaf fall.
-Although there is a slight trend toward later leaf out, a stronger trend toward later leaf fall has resulted in a very weak trend toward an increase in “leaves on days” or growing season during the course of this study.
From 1990-2006 we observed the timing of woody vegetation development during the growing season. For the first twelve years (1990-2001) we observed bud break (50% leaf emergence) and leaf development (75% final size) on two to five permanently tagged individuals of 33 woody species at 3-7 day intervals from April through June. Bud break is defined as when 50% of the buds have recognizable leaves emerging from them. Leaf development (75%) is defined as when at least 75% of the leaves have reached 75% of their final size. This point is used rather than “fully developed”because the leaves are growing rapidly during this period and it permits better estimation of a date between observations. Flowering and fruit development were also recorded if they occurred during this period.
These observations documented substantial (up to three weeks difference) interannual variation in the timing of leaf emergence (50%) and leaf development (75% final size), but good relative consistency among species and among individuals within species during these twelve years. Therefore, starting in 2002 we maintained the same observation protocol, but reduced the number of species observed to nine, including red maple (Acer rubrum), sugar maple (A. saccharum), striped maple (A. pensylvanicum), yellow birch (Betula alleghaniensis), beech (Fagus grandifolia), white ash (Fraxinus americana), witch hazel (Hamamelis virginiana), white oak (Quercus alba), and red oak (Q. rubra). This subset of important, representative species allows us to continue to characterize leaf development each spring, and document inter-annual variability while reducing the time resources required for the study.
We have also recorded fall phenology since 1991, with the exception of 1992. Weekly observations of percent leaf coloration and percent leaf fall begin in September and continue through leaf fall. In 2002 we reduced the species observed in the fall to fourteen, including striped maple (Acer pensylvanicum), red maple (A. rubrum), sugar maple (A. saccharum), shadbush (Amelanchier laevis), yellow birch (Betula alleghaniensis), black birch (B. lenta), paper birch (B. papyrifera), beech (Fagus grandifolia), white ash (Fraxinus americana), black gum (Nyssa sylvatica), black cherry (Prunus serotina), white oak (Quercus alba), and red oak ((Q. Q. rubrarubra).).Plant phenology is an important ecological signal integrating the atmosphere and
biosphere. It responds to climate fluctuations and, especially in deciduous systems, drives seasonal changes in albedo, water transport, cloud cover and carbon exchange. Despite this importance, long-term phenological data sets are quite rare. The Harvard Forest data set is especially valuable because Harvard Forest is a Long-Term Ecological Research site with extensive historical records and current measurements of both the vegetation and atmosphere as well as carbon exchange.
As part of our Schoolyard LTER program the protocols for this study have been adapted and presented for use by K-12 teachers on their school grounds. Through workshops at Harvard Forest and presentations at other environmental education workshops 20 teachers from 17 schools have begun using these protocols and submitting phenology data. Further information about our Schoolyard phenology program can be found at: http://harvardforest.fas.harvard.edu/museum/phenology.htmlAdditional information on our research on fall phenology is available at:http:// harvardforest.fas.harvard.edu/research/leaves/autumn_leaves.htmlThe Harvard Forest phenology data set can be accessed at: http://harvardforest.fas.harvard.edu/data/p00/hf003/hf003.htmlQuestions about the data set can be addressed to: [email protected]
Figure 1. The study is being conducted along a two kilometer loop (in red above) at the Harvard Forest (42deg 32’ 30”N and 72deg 11’ 00”W) in Petersham in north-central Massachusetts. The loop starts and ends in open areas at the Harvard Forest headquarters and is primarily in upland mixed deciduous-coniferous forest surrounding a moderate sized wetland. The elevation of the loop ranges from 335 to 365 meters and habitats range from closed forest, through forest-swamp margins to field edges. The majority of the area was agricultural land (primarily pasture) in the mid-nineteenth century. Much of the present forest is 65-85 years old following harvesting early in the twentieth century and extensive hurricane damage and salvage in 1938. Climate data is collected at a weather station located near the Harvard Forest headquarters at the southwest end of the loop and at the EMS tower approximately 1 kilometer east of the loop.
SUMMARY
STUDY AREA
METHODS RESULTS (CONT.)
OUTREACH AND DATA ACCESS
INTRODUCTION
Harvard Forest Phenology Study 1990-2006 John F. O’Keefe
Harvard Forest, Harvard University, Petersham, MA 01366
Figure 2. In each photograph (red oak on the left and beech on Figure 2. In each photograph (red oak on the left and beech on the right) approximately twothe right) approximately two--thirds of the buds have thirds of the buds have recognizable leaves emerging, so this would represent just past recognizable leaves emerging, so this would represent just past 50% bud break in each case.50% bud break in each case.
RESULTS
100
120
140
160
180
200
220
240
260
280
300
1990
1992
1994
1996
1998
2000
2002
2004
2006
YEAR
DA
Y O
F YE
AR
BEAL1BB
BEAL2BB
BEAL3BB
BEAL1L75
BEAL2L75
BEAL3L75
BEAL1LF50
BEAL2LF50
BEAL3LF50
100
120
140
160
180
200
220
240
260
280
300
320
1990
1992
1994
1996
1998
2000
2002
2004
2006
YEAR
DA
Y O
F YE
AR
ACRUBB
BEALBB
QURUBB
QUALBB
ACRU75
BEAL75
QURU75
QUAL75
ACRUL50
BEALL50
QURUL50
QUALL50
MEAN BUD BREAK AND 50% LEAF FALL 4 SPECIES (N=15)
110130150170190210230250270290310
1990
1992
1994
1996
1998
2000
2002
2004
2006
YEAR
DA
Y O
F YE
AR
MEANBBMEANLF50Linear (MEANBB)Linear (MEANLF50)
Figure 3. The three individuals of yellow birch show Figure 3. The three individuals of yellow birch show similar patterns in the timing of leaf development similar patterns in the timing of leaf development (BB=50% bud break and L75=75% leaf development) (BB=50% bud break and L75=75% leaf development) and leaf drop (LF50=50% leaf fall) and their timing and leaf drop (LF50=50% leaf fall) and their timing with respect to each other remains relatively similar. with respect to each other remains relatively similar. There is approximately twice as much variability in There is approximately twice as much variability in the spring events as in the fall.the spring events as in the fall.
Figure 4. Just as the individuals within a species Figure 4. Just as the individuals within a species show similar patterns of leaf development and leaf show similar patterns of leaf development and leaf drop, so too do the mean patterns for each species. drop, so too do the mean patterns for each species. Four species are plotted here: red maple, yellow Four species are plotted here: red maple, yellow birch, red oak and white oak. Although the spring birch, red oak and white oak. Although the spring remains more variable there is greater consistent remains more variable there is greater consistent separation among the species in the fall.separation among the species in the fall.
Figure 5. The summed mean bud break Figure 5. The summed mean bud break and leaf drop dates for the four species and leaf drop dates for the four species shown individually in Figure 4 both shown individually in Figure 4 both show a trend toward later occurrence show a trend toward later occurrence over time. However, the trend for leaf over time. However, the trend for leaf fall is greater than for bud break.fall is greater than for bud break.
Figure 6. The net effect of a slightly later leaf out Figure 6. The net effect of a slightly later leaf out and a more pronounced later leaf drop during the and a more pronounced later leaf drop during the course of the study has been a slight trend toward course of the study has been a slight trend toward a longer growing season between 1991 and 2005. a longer growing season between 1991 and 2005.
Figure 7. The mean 50% bud break date Figure 7. The mean 50% bud break date for all individuals continuously observed for all individuals continuously observed in the study shows a slight trend toward in the study shows a slight trend toward later leaf emergence, considerable later leaf emergence, considerable variability and, not surprisingly, does not variability and, not surprisingly, does not appear related to the date of last frost.appear related to the date of last frost.
Figure 8. The mean 50% leaf fall date for Figure 8. The mean 50% leaf fall date for all individuals continuously observed in the all individuals continuously observed in the study shows a trend toward later leaf drop study shows a trend toward later leaf drop of about five days. This trend may be of about five days. This trend may be somewhat related to first frost dates, which somewhat related to first frost dates, which have trended later during the study.have trended later during the study.
RR22=.546=.546
RR2=2=.116.116
rr22=.054=.054
LEAVES ON DAYS - MEAN FOR 4 SPECIES (N=15)
150
155
160
165
170
175
1990
1992
1994
1996
1998
2000
2002
2004
2006
YEAR
NU
MB
ER O
F D
AYS
LEAVES ON DAYS
Linear (LEAVES ONDAYS)
RR22=.054=.054
50% LEAF FALL
R2 = 0.4754
R2 = 0.5177
250
260
270
280
290
300
1991
1993
1995
1997
1999
2001
2003
2005
YEAR
DAY
OF
YEA
R 1st frost50% fall5yr mean fallLinear (50% fall)Linear (1st frost)
50% BUD BREAK
R2 = 0.0429
R2 = 0.0403110
120
130
140
150
160
1990
1992
1994
1996
1998
2000
2002
2004
YEAR
DAY
OF
YEA
R last frost50% BB5 yr mean BBLinear (last frost)Linear (50% BB)