.

. ‘ 1
"f. ::
'V" . «
_ 4

D ‘ 7
. _ -- .
..

—..‘
o

‘,-I

_ _. ; _
3‘3: ~ I
- ‘4!- ~ c-. _—

1‘ 55"”:

11111
‘1

‘ ..
7
n
t ‘ ‘V 4
. .
v A 'I‘
. 'M ‘
. I. f,
- —- I.

.1 1 I I
s .61..
. 1 L' tttt fifififl"

 

” 1

‘-

  
 
 

. 4. .
wam‘é'f‘
3v ‘3

:1!"
54.31.13: ‘1'

. .
.
. ‘5'

-
a;

 

 

1
111

1:11: "'.1:1::2_:1: “1:15:
:1 11! :1|
1 1'1 :11:1 IN:

. . .1-
u.11-1‘111... ‘
11:1: J :111111“: 111:1.|

11111:: :1:

1‘ . 11:11.
1111111111 111 211111111 11,151

12.1:(zn

1.11'1111: 111111 '1

1111‘

11:1 11:11:“:

:111111.111111111111111111111211

 

THFeis

n

y

 

                                                                                                                           

1293 01691 4552

This is to certify that the

thesis entitled

Plant Responses to Experimental Warming of a Dry Heath
Tundra at Barrow, Alaska

presented by

Lisa Jeanne Walker

has been accepted towards fulfillment
of the requirements for

Masters degree in Botany and Plant Pathology

49M MW

ajo lprofessor

 

 

Dan: ‘3; .ES.“—}

0-7639 MS U is an Affirmative Action/Equal Opportunity Institution

 

¢ #V ‘f vfi~ ,_- .-

W"“ k- WW_ 4

~‘c - _H

 

LIBRARY

Michigan State
Unlverslty

 

 

 

PLACE IN RETURN BOX
to remove this checkout from your record.
TO AVOID FINES return on or before date due.

 

DATE DUE MTE DUE 0A1! DUE

 

 

 

 

 

, ItZAY‘021399

 

\t,

{4 tr'

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

me Wm“

 

 

 

 

 

PLAT

PLANT RESPONSES TO EXPERIMENTAL WARMING OF A DRY HEATH
TUNDRA AT BARROW, ALASKA

By

Lisa Jeanne Walker

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTER OF SCIENCE

Department of Botany and Plant Pathology

1997

PLA

effects
Models
strong}:
and gro
This rer
t0 exper
warminJ
date of 0
groh'mg
increaSec

growth a]

ABSTRACT

PLANT RESPONSES TO EXPERIMENTAL WARMING OF A DRY HEATH
TUNDRA AT BARROW, ALASKA

By

Lisa Jeanne Walker

The International Tundra Experiment (ITEX) was created to examine
effects of increased temperature, as predicted by current Global Climate
Models, on vegetation in the Arctic. This region is predicted to be the most
strongly effected by temperature change. In ITEX the seasonal development,
and growth patterns of plant species are examined throughout the Arctic.
This report is focused on responses of dry heath vegetation at Barrow, Alaska
to experimental warming. Small fiberglass chambers are used to induce
warming over the tundra. Phenophases were examined according to Julian
date of occurrence, number of days since snow melt, and accumulated
grong degree days. Measurements were made to determine the effects of
increased temperature on the total height of reproductive and vegetative
growth and in 1996 stature was monitored to determine differences in growth
rates. Plant responses to warming were not consistently significant during
the years of this study, or between species, showing that plants respond

individualistically.

ACKNOWLEDGMENTS

I would like to thank Dr. Patrick Webber for introducing me to ITEX
and allowing me the opportunity to work on this exciting project. The ITEX
community and meetings are an excellent opportunity to learn about other
countries and the work that is being done around the circumpolar Arctic. I
would also like to thank Bob Hollister for his help with the abiotic data, and
Lisa Koch and Ian Ramjohn for assistance in the field. I am greatly indebted
to Dr. Christian Bay, who initiated the Barrow ITEX site in 1994, and taught
me about the Arctic and its plants during the 1995 field season. Partial
funding for this research came from the National Science Foundation grant
to Dr. Kaye R. Everett at The Ohio State University, and Dr. Fritz E. Nelson at
the State University of New York at Albany. 1 would like to thank my
committee members, Dr. Stuart Gage, and Dr. Frank Telewski, as well as Dr.
Kelly McConnaughay, my family, and my friends for all their help and

support.

iii

 

USIOFT}
USIOFF]

CHMWER
”EX-THE
lmm
Hyp
Spec

TABLE OF CONTENTS

LIST OF TABLES
LIST OF FIGURES

CHAPTER 1

IT EX - THE INTERNATIONAL TUNDRA EXPERIMENT
Introduction
Hypotheses
Species

CHAPTER 2
EXPERIMENTAL WARMING IN OPEN TOP CHAMBERS
Control and Experimental plot descriptions
Chamber effectiveness
Community composition
Summary

CHAPTER 3
PLANT RESPONSES TO EXPERIMENTAL WARMING
Experimental design summary
Effects on phenophases
Effects on growth and stature
Final discussion

APPENDIX A
APPENDIX B
LITERATURE CITED

iv

 

 

Table 1.
Table 2.
Table 3.
Table 4.
Table 5.
Table 6,
Table 7,

Table 8.
Table 9_

Table 10.

Table 11_
Table 12‘
Table 13.

Table 1.
Table 2.
Table 3.
Table 4.
Table 5.
Table 6.
Table 7.

Table 8.
Table 9.

Table 10.

Table 11.
Table 12.
Table 13.

LIST OF TABLES

Characteristics of the dry heath ridge at Barrow, Alaska. 7

List of all vascular species found within plots on the dry heath ridge. 8

All phenological stages measured for each of the species. 9
A comparison of ITEX and NOAA temperature readings. 22
Average soil temperature at different depths 25
Average percentage cover of common vascular plants 28

Number of responses to warming as a determination of trends
in plant responses. 35

Analysis of Variance result tables - Julian Date of Occurrence 73
Analysis of Variance result tables - Days since snow free 79

Analysis of Variance result tables - Growing Degree Day

accumulation 84
Scheffe Post Hoc test results - Julian Date of Occurrence 89
Scheffe Post Hoc test results - Days since snow free 92

Scheffe Post Hoc test results - Growing Degree Day
accumulation 95

Figure 1.
Figure 2.
Figure 3.
Figure 4
Figure 5.
Figure 6.
Figure 7.
Figure 8.

Figure 9.

Figure 10.

Figure 11.

Figure 12.

Figure 13.

Figure 14.

Figure 15.

Figure 16.
Figure 17.

LIST OF FIGURES

IT EX map showing the locations of established sites.

Open top chambers (OTCs) at Barrow, Alaska.

Location map of OTC and control plots along the beach ridge.
Light distribution within an OTC.

Typical temperature and relative humidity records for 1995.
Spatial variation of temperature within a chamber .
Accumulated growing degree days in OT Cs and control plots.

OF C and control plot degree day accumulations during the
growing seasons of 1994, 1995, and 1996.

Average thaw depth for 1995 and 1996.

The Julian date of occurrence of phenophases in three graminoid
species.

Number of snow free days prior to the onset of phenophases for
three graminoid species.

Growing Degree Day accumulations prior to the occurrence of
phenophases of the three graminoid species.

Effects of OTC warming on Julian Date of occurrence of forb
phenophases.

Number of days snow free prior to the occurrence of forb
phenophases.

Number of accumulated Growing Degree Days prior to the
occurrence of forb phenophases.

Julian Date of occurrence of Salix rotundifolia phenophases.

Number of days the plot has been snow free before a phenophase
occurred.

vi

16
17
19
20
21

26

36

37

38

41

42
45

Figure 18.

Figure 19.
Figure 20.

Figure 21.

Figure 22.

Figure 23.

Figure 24.
Fr gure 25.
Figure 26.
Figure 27.

Accumulated Growing Degree Days prior to Salix rotundifolia
phenophases.

Julian date of occurence of phenophases for Cassiope tetragona.

The number of days that the plot was snow free before Cassiope
reached a particular phenophase.

The average amount of growing degree days for Cassiope to
undergo a phenophase.

Growth responses of Luzula arctica, Luzula confirm, and
Arctagrostis Iatifolia growth to increased warming.

Growth responses of Papaver and Saxifraga to increased
warming.

Annual growth increment of Cassiope tetragona.

Effects of warming on Arctagrostis latrfolia stature during 1996.

Effects of warming on Luzula arctica stature during 1996.

Effects of warming on Luzula confitsa. stature during 1996.

vii

47
49

51

6 1
62
65

67

Chapter 1

ITEX - THE INTERNATIONAL TUNDRA EXPERIMENT

Introduction

As the amount of greenhouse gases continue to increase in the
atmosphere, the likelihood that there will be a significant change in the
climate within our lifetimes also increases (MacCracken, 1995; Cohen, 1990).
Global Climate Change Models (GCMs) are the usual basis for these
predictions. A change such as this is expected to take place in the form of
global warming, being most pronounced at polar latitudes (Maxwell, 1992).
There is evidence of warming in some parts of the Arctic, especially the
Western Arctic including Alaska (Chapman and Walsh, 1993). Arctic
ecosystems are likely to be the most affected by future global warming
(Maxwell and Barrie, 1989), which makes arctic tundra the ideal location to
study possible effects of warming on vegetation. The arctic regions are also
most likely to be the location of the largest impact on climate from
anthropogenic pollution, as demonstrated by the trend of an earlier snowmelt
(since 1945) at Barrow, Alaska, suggesting a longer growing season (Foster,

1989). Arctic plants have adapted to living in conditions that are limited by

1

2

low temperature,‘short growing season, low light intensity, and low nutrient
availability (Chapin, 1987), and arctic ecosystems are characterized as having a
low amount of annual net primary productivity (Haag, 1974). Temperature is
seen as being the most important limiting factor to plant growth in tundra
vegetation (Bliss, 1962), which suggests that arctic vegetation is likely to show
a response to warming. Most tundra plants fall into a height range of 6-8 cm,
often forming a dense mat layer, with most of the aboveground growth of the
plants being in the part of the microenvironment of the tundra that has a
warmer temperature, which is important since relatively small changes in
temperature become highly significant to plant metabolic processes (Bliss,
1962). The small stature of many arctic plants also means that the canopy
temperature is more closely related to soil surface temperature, than to the
ambient air temperature, which should be more responsive to climate change

than air temperature (Foster, 1989). Therefore, tundra plants should show a
response to the warming of (1 - 4°C) predicted by most GCMs.

Many climate models have been introduced as a tool in predicting
possible future climate changes. Currently three dimensional models that
couple a general circulation model of the atmosphere to that of ocean
patterns, are used to predict possible future scenarios of climate change
(MacCracken et al., 1991; Manabe 1997). There are two general types of GCMs
(General Circulation Models, or Global Climate Change Models) used to study
the effects of the increasing amount of greenhouse gases found in the

atmosphere. The first type of GCM has the amount of gas concentration

 

(usue
reach
with ‘
differi
future
repres
genera
greenh
tem per
the clirr
Sensitiv
address
also beir
Chinges,
of the lar-
Th

examine l

3

(usually C02) doubled and the model is run until a new equilibrium is
reached. In the other type of GCM, the gas concentration is increased slowly
with time (the rate at which the gas concentration is increased changes with
different models, usually based on past concentration increases, or predicted
future changes) and includes oceanic behavior, which must be accurately
represented. Although results differ from model to model (using these two
general types of models), most agree that with the continuing increase in
greenhouse gas concentrations there will be a global average surface air
temperature warming in the next century assuming that other forces affecting
the climate do not counteract this effect (MacCracken et a1. 1991). The
sensitivity of these global scale models is being tested on regional levels, to
address environmentally important issues (Grotch, 1991). The models are
also being coupled with biological models that simulate surface (vegetation)
changes, so that exchange processes will be included within the final output
of the larger GCMs (Fennessy and Xue, 1997).

The International Tundra Experiment (ITEX) is a project designed to
examine the effects of experimental warming at the plant canopy level (Henry
and Molau, 1997). ITEX began at a meeting at the Kellogg Biological Station of
Michigan State University on 2 - 5 December, 1990 (Molau and Molgaard,
1996). ITEX examines the responses of vascular plants to experimental
warming at 26 sites in 11 different countries. Climate and geophysical
features are monitored. Developmental stages, as well as quantitative growth

data are collected each growing season in control and experimental plots from

4

each of the sites. By comparing these individual plant responses at each site
to those at different sites, it is anticipated that a better ecological
understanding of how warming effects the tundra, or high latitude systems
will result. As the effects of increasing the ambient growing temperature on
the reproductive cycles of arctic plants are relatively unknown (Moore, 1995), .
the ITEX community is examining these responses on a circumpolar level, by
comparing phenological and growth data from similar species at each site.
Tundra plants appear to have sufficient genetic variability and plasticity to
confer some resistance to climate change (McGraw and Fetcher, 1992).
Temperature influence on plant growth is seen as one of the most important
feedback mechanisms at the global level, which also makes it an important
part of present and future modeling (van Minnen et al., 1995). One of the
major constraints in developing higher resolution regional models is the lack
of information on a smaller scale (Cohen, 1990). In the future, ITEX response
data could possibly fill this need. The ITEX project does not examine all
aspects of global warming or global change, but instead focuses on the effects
of growing season temperature on plant performance. The study described
here is from one of two sites at Barrow, a dry heath ridge. The other site is a
wet sedge meadow.

At each ITEX site the standard basic experiment consists of a series of
small fiber-glass chambers that trap energy and cause plant canopy warming.
The construction and effectiveness of the chambers will be discussed later.

Each of these sites has 20 or more replicates of chambers and control plots. A

5

large number of replicates is needed to monitor this extremely variable
system. The standard basic experiment requires that air temperature, plant
phenology and growth be monitored. It is also desirable for other variables
such as relative humidity, incoming solar radiation, and soil temperature to
be measured at each of the ITEX sites. Of the species that are monitored, each
site is expected to target at least one of the species or genus that is on the ITEX
circumpolar list (Molau and Molgaard, 1996) which ranks the important
species to monitor at each site. This will contribute to the circumarctic goal of
ITEX to understand plant responses to warming. Figure 1 depicts the
locations of the established ITEX sites around the circumarctic, including both
arctic and alpine sites.

Barrow, Alaska is one of three existing sites in the United States (the
others are at Niwot Ridge, Colorado, and Toolik Lake, Alaska; see Figure 1 for
locations). Another site near Barrow has been partially established for future

monitoring at Atqasuk, Alaska (not shown in Figure 1). Barrow is located at
71 °19’N, 156°37'W, in the northernmost portion of the United States, on the

North Slope of Alaska (Table 1). Barrow is within the Arctic Coastal Zone,
which is characterized by cool summers, and relatively warm winters, due to
the buffering effect of the Arctic ocean, and also has a low amount of
precipitation of which more than 50% falls as snow (Zhang et al., 1996). Both
sites at Barrow (dry heath and wet sedge) are an important part of ITEX, as
they represent two very different aspects of Arctic seacoast tundra. Table 1

lists the conditions and components of the dry heath site at Barrow.

Circumpolar Arctic ITEX Map

 

(from Marion et al., 1993)

Figure l. ITEX map showing the locations of established sites.

Table 1. Characteristics of the dry heath ridge, at Barrow, Alaska.

(°C)
season
w
msm

0.5 °W

ne ts, ,
and els
c

Rhacomitrium inosum

Thamnolia vennicularis

 

Hypotheses
The underlying hypotheses of this project are as follows: 1) the dry

heath tundra vegetation will show an acceleration of phenophases as a
response to artificially induced warming inside the chambers; 2) the dry
heath plants will exhibit an increase in stature in response to warming;

3) the dry heath plants will exhibit an increase in growth rate in response to
elevated ambient temperature, and; 4) the species on the dry heath tundra
will respond in an individualistic manner to the increase in temperature
within the chambers. Although tundra plants are seen as having a large
amount of genetic variability and plasticity, the specific species should
respond on a microhabitat level to an increase in temperature, even if the

ecosystem as a whole is somewhat resistant to change.

Table 2. List of all vascular species found within plots on the dry heath ridge.

 

Vascular species
Family Genus and Species
Graminae Alopecurus alpinus Sm ssp.
al pi nus
Graminae Arctagrostis latifolia (R.Br.)
Griseb var. latifolia *
Cyperaceae Carex stuns Wahlenb. ssp.
stuns (Drej.) Hult.
Ericaceae Cassiope tetragona (1..) D.
Don ssp. tetragona
Cruciferae Draba lactea Adams
Cruciferae Draba micropetela Hook.
Juncaceae luncus biglumis L.
Iuncaceae Luzula arctica Blytt
Juncaceae Luzula confusa Lindeb.
Polygonaceae Oxyria digyna (L.) Hill
Papaveraceae Papa ver hultenii Knaben
Papaveraceae Papaver lapponicum
Scrophulariaceae Pedicularis kanei Durand
ssp. kaneii
Graminae Poa arctica R. Br. ssp. arctica
Rosaceae Potentilla hyparctica Malte
Ranunculaceae Ranunculus nivalis L.
Salicaeae Salix rotundifolia Trautv.
Saxifragaceae Saxifraga caespitosa L.
Saxifragaceae Saxifraga cernua L.
Saxifragaceae Saxifraga foliolosa R. Br. var.
fol iolosa
Saxifragaceae Saxifraga flagellaris
Saxifragaceae Saxifraga nivalis L.
Saxifragaceae Saxifraga punctata L. ssp.
nelsoniana (D.Don) Hult.
Compositae Senecio atropurpureus
(Ledeb.) Fedtsch. ssp. frigidus
(Richards) Hult.
Caryophyllaceae Stellaria laeta Richards.
Ericaceae Vaccinium vitis-idaea L. ssp.

minus (Lodd.) Hult.

* Bolded species are those analyzed in this thesis
(Nomenclature according to Hulten, 1968)

9

Table 3. All Phenological stages measured for each of the species.

Arcta rostis olia *

Salix rotundifolia

V

Pl: Emergence of first green leaf
P2: Inflorescence visible

P3: First flower bud visible

P4: First flower open/ visible
P5: First stigma visible

P6: Elongation of peduncel

P7: First flower withering

P8: Stigma withering

P9: In fruit

P10: Inflorescence expanding

Q1: Length of flowering shoot

Q2: Length of longest leaf

Q3: Total number of flowers

Q4: Total number of fruits

Q5: Fruit/flower ratio

Q6: Total number of flowering catkins

Barrow Heath

P1; P2; P10; P11

P1; P5; P13; P14; P15;

Q2; Q5; Q6; Q7:

 

P11: Inflorescence open

P12: Corolla drop

P13: Onset of seed dispersal
P14: First pollen shed

P15: All pollen shed

P16: First yellowing of leaves
P17: Emergence of stem

P18: First anther visible

P19: Anther withering

P 20: First petal drop

Q7: Number of flowers in each catkin
(3: Total number of mature catkins
Q9: Number of capsules in each catkin
Q10: Weight of largest leaf

Q11: Mature catkin/flowering ratio
Q12: Annual growth increment

* Bolded species are those analyzed in this thesis

10
Species:

Twenty-six species are found within the dry heath plots (Table 2) for
which phenological stages were recorded (Table 3). The phenological stages
listed are the standard visible vegetative and reproductive stages for each of
the individual species that could be measured. From this data set it was
determined that the following species had sufficient numbers to be examined:
Arctagrostis latifolia, Cassiope tetragona, Luzula arctica,‘ Luzula confusa,
Papaver hultenii, Salix rotundifolia, and Saxifraga punctata.. The following is
a discussion of the characteristics and distribution of these seven tundra
species of focus and a listing of the species specific measurements. All seven

species are among the ten most common species within the site.

Arctagrostis latifolia

Family: Gramineae. It is a tall, purple grass, often found in wet
meadows, along rivers and on tundra and has a wide, circumpolar
distribution. In Barrow, it occurs most frequently on dry sites and on high .
center polygons, 'well drained banks, and former beach ridges. The
phenophases that were measured include: first green leaf emerged, first
inflorescence visible, first inflorescence expanding, and in 1996 first glume
open(P1, P2, P10, P11). The height of the first individual to emerge in each
plot was monitored throughout the growing season by measuring from the
base of the plant to the tip of the longest leaf. At the end of the season the

height of the three largest individuals was measured in the same way, as well

 

Cassir

evergr
or in t
rare, 0
abund.
first bu
dr0p (I
Of the g
Lulula
f
tundra,
as lWell a
Barrow .-
a Commc

measufer

11

as measuring the three largest reproductive shoots, from the base of the plant

to the end of the inflorescence.

Cassiope tetragona

Family: Ericaceae. This is an arctic bell heather species, that is an
evergreen woody dwarf shrub. It is found on dry heaths and rocks on tundra,
or in the mountains throughout the circumpolar arctic. At Barrow it is quite
rare, of only small stature, and restricted to a few beach ridges. It is more
abundant and robust to the south of Barrow. Phenophases recorded were:
first buds visible, first elongation of buds, first flower open, and first corolla
drop (P3, P4, P6, P12). The annual growth increment was measured at the end
of the growing season.
Luzula arctica

Family: Juncaceae. This is a short rush with flat leaves, found on
tundra, mountain tundra, and moist slopes in central and northern Alaska,
as well as around the circumpolar Arctic. Luzula arctica is common at
Barrow and occurs in several habitats. It is most abundant on dry sites, but is
a common component of most tundras. The phenophases that were
measured are: first green leaf emerged, first inflorescence visible, and first
inflorescence open (P1, P2, P5, P11). The height of the first individual to
emerge in each plot was monitored throughout the growing season by
measuring from the base of the plant to the tip of the longest leaf. At the end

of the season the height of the three largest individuals and the three largest

12

reproductive shoots were measured from the base of the plant to the end of

the tallest leaf or inflorescence.

Luzula confusa

Family: Iuncaceae. This is a rush with narrow leaves, slightly larger
than L. arctica. It is found on dry heaths in mountains and tundra
throughout Alaska, and the circumpolar arctic. Like L. arctica it is common
at Barrow in moist to dry habitats. It is most abundant on the driest sites. The
phenophases that were measured are: first green leaf to emerge, first
inflorescence visible, and first inflorescence open (P1, P2, P5, P11). The height
of the first individual to emerge in each plot was monitored throughout the
growing season by measuring from the base of the plant to the tip of the
longest leaf. At the end of the season the height of the three largest
individuals was measured in the same way, as well as measuring the three
largest reproductive shoots, from the base of the plant to the end of the

inflorescence.

Papaver hultenii

Family: Papaveraceae. This is an arctic poppy that has yellow flowers
and silver gray leaves. It is found on sandy and gravely soil and is restricted
to the northernmost part of Alaska. It is closely related to the Papaver
rad icatum complex which is measured at several other ITEX sites.

Phenophases measured were: first green leaf to emerge, first bud to emerge,

13

first peduncle to elongate, first flower to open, and first flower to wither (P1,
P3, P4, P6, P7). Throughout the growing season the length of the longest
peduncle was monitored from the base of the plant to the tip of the bud, or
flower. At the end of the growing season the length of the three longest

peduncles were measured.

Saxifraga punctata subspecies Nelsoniana

Family: Saxifragaceae. This robust saxifrage has small white flowers,
dark green leaves and is found in alpine meadows, tundra hummocks, and
along creeks throughout Alaska and parts of Siberia. At Barrow it is a wide
ranging plant which reaches its greatest abundance on dry sites. Phenophases
measured were: first green leaf to emerge, first bud to emerge, first peduncle
to elongate, first flower to open, and first flower to wither (P1, P3, P4, P6, P7).
Throughout the growing season the length of the longest peduncle was
monitored from the base of the plant to the tip of the bud, or flower. At the
end of the growing season the height of the three longest peduncles were

measured.

Salix rotundifolia

Family: Salicaceae. This dwarf, prostrate shrub willow, with thin small
annual shoots and roundish leaves, has separate male and female catkins,
and is found on arctic and alpine tundra, as well as on rocky places

throughout southern and northern Alaska, and parts of Siberia. Phenophases

14

measured were: Emergence of first green leaf, first stigma visible, first pollen
shed, all pollen shed, first seed dispersal, and first color change (P1, P5, P13,
P14, P15, P16). At the end of the season the length of the longest leaf was

recorded.

Chapter 2

EXPERIMENTAL WARMING IN OPEN TOP CHAMBERS

Control and Experimental plot descriptions
The standard basic ITEX project uses small open top chambers (OTCs)
to induce warming on target tundra plant communities at the plant canopy

level. Chambers at Barrow, Alaska are 1.5 m2 hexagonal structures, made
from Sun-Lite HP"I (Solar Components Corp., Manchester, NH) fiberglass.

These sheets are 1mm thick and have the following optical properties: high
solar transmittance in the visible wavelengths (86%), and a low transmittance
in the infra-red range (<5%) (Molau and Molgaard, 1996). Figure 2 depicts the
sloping sides of the OTCs.

Experimental plots are all permanently marked with a numbered
identification stake, and by small metal stakes placed at each of the chamber
corners, so that exact locations of the chambers can be determined each year as
the snow melt occurs. Monitoring of OTCs and control plots begins after

snow melt (end of May / beginning of June) and continues until mid-August.

15

 

Figure 2. Open top chamber at Barrow, Alaska.

Control plots are 1 m2 in size with permanent stakes at each corner.
These stakes are placed deep into the soil so that they will remain in the same
location from year to year. Control plots are not manipulated in any way, and
are only used as a base-line comparison to responses observed in OTC
experimental plots.

The dry heath site was established in 1994 by Dr. Christian Bay. Areas
of the beach ridge with important ITEX target species and a uniform species
composition were located. Experimental and control plots were randomly
placed within these areas. Figure 3 shows the locations of each control and
OTC plot along the beach ridge. These areas were co-dominated by Cassiope

tetragona and Salix rotundifolia.

17

2I 3. 1

Dry Heath Site Plot Map

; 2010

O-OTC
1 gl-CUItrd
l____.___J

 

 

24. (Adapted from Bay, 1994)

Figure 3. Location map of OTC and control plots along the beach ridge.

18

Chamber effectiveness:

ITEX hexagonal chambers are designed to warm the plant canopy. Each
side of the hexagonal chamber is at an incline of 60°. This causes the chamber
to act like a greenhouse and trap heat, and also makes the chambers more
favorable to incoming radiation, since the optimal transmittance is at a 90° to I

the surface of the fiberglass (ITEX Manual, 1996). Chamber performance has
been field tested at other locations (Marion et al., 1993). Although the validity
of using greenhouse chambers as a means of examining possible responses to
warming, has been criticized on the basis of possible complex and poorly
understood modifications of climate (Kennedy, 1995), the open top chambers
used in this project have been intensely examined (Marion etal., 1993,
Marion et al., 1997) and have been determined to raise the ambient
temperature in a manner consistent with the predicted global warming.
Although light levels within the chambers are slightly altered, no etiolation
was readily visible, and the relative humidity within the chambers tracks the
temperature as it rises and cools, much as in a natural environment (Figures
4 and 5).

Hobo" and Stowaway “‘ dataloggers and thermistors (Onset Computer

Corp., MA) were used in both OTCs and control plots to record temperature,
and relative humidity. Both of these dataloggers employ small thermistors at

the end of a short cord, which allows the thermistor to be placed at a different

19

location than that of the datalogger. The dataloggers are computer activated
to read for a programmed time, which dictates the frequency of the readings

which are then stored in the dataloggers memory for downloading at a later

 

Light distribution
(Percentage of full daylight)

 

 

 

 

 

 

 

 

Figure 4. Light distribution within an OTC.

date. Figure 6 shows the horizontal and vertical variations of temperature
measurements within and outside of the chamber. The highest temperature
occurs at 16 cm above the soil level. This essentially coincides with the height
at which temperatures are measured in both the OTCs and the control plots.

Relative humidity sensors are also made by Onset Computer Corp., MA, and

20

 

.33 .o. 2:39. b.2835 9528 use 83223.2 .33.; .m 959m

 

 

 

 

 

 

o , ermaredutel

 

.5me00].
0._.Ol

 

 

 

 

22.8 ”so >820

o_.

 

 

.80 .6 9:?
a m m e 9 e
a m m u a e

 

 

 

 

 

 

 

 

 

 

3:0 :60 .833.

.00..

 

7U

 

<- 8628!
J or at
.. 00:»

 

 

 

 

>

 

 

 

 

D
f.

m:<-m :60 >55

KunwnH amass

 

 

21

 

Temperature distribution

2.5 as...
2.6 L...
2.5 as...
2.5 .....
2.8 m.
3.0 u...
3.0 .u.
3.2 ..._. .

 

 

 

 

 

 

 

 

Figure 6. Spatial variation of temperature within a chamber

work in a similar fashion, although the sensor is contained within the
datalogger. Temperature thermistors, and relative humidity sensors rest
inside Gill six-plate containers at 15 cm above the ground, so that they are
screened from both solar and ground radiation. The Gill six-plates also allow
ventilation around the sensors. The sensors, for both relative humidity and
temperature, were set to monitor continuously (every 12 - 16 minutes)
throughout the growing season (approximately 1 June - 20 August) each year.
Table 4 compares ITEX temperature readings at the level of the
vegetation canopy to that of a NOAA (United States National Oceanic and

Atmospheric Administration) meteorological screen, located to the North of

22

the site. This chart illustrates the difference between standardized NOAA
data and ITEX air temperature data, as well as emphasizing the need for the
continuous measurements made at the plant canopy level in both the
chambers and the controls. The importance of obtaining temperature data at
the plant canopy level is demonstrated in Table 4, as the NOAA temperature
data is extremely different than the ITEX data. The average increase in

ambient air temperature within the chambers throughout each growing
season was on the order of 1.5 and 17°C for each of the three years. This is

consistent with the predictions of the GCMs (Chapman and Walsh, 1993).

Table 4. A comparison of IT EX and NOAA temperature readings.

 

 

 

 

 

 

 

 

Location Height Number of Temperature
measurements °C
N OAA 2m 1 / hour 4.9
shelter screen
A Control - 15 cm at least 0.6
Shelter Gill 6 plate every 16 min.
A OTC - 15 cm at least 1.5
Control Gill 6 plate every 16 min

 

 

Growing degree days were calculated in the following manner: from
the continuous seasonal measurements the above 0°C temperatures were

averaged for each day. These daily averages were summed consecutively to
obtain the accumulated growing degree days. This yields an index of how

much energy was accumulated in both the control plots and OTCs for each

.363 .9589 can «UPC 5 when moans—u 9:38» 3.35:5qu N 0.5mm"—

 

>mu 5.3
—

OVN 0“ g 8
b b p
4 d J

 

 

 

 

a .
5‘30 anfiao Burners)

 

 

 

 

 

33:00 I ll
. 0.511

0588953 835...: Ill. ‘11 who . 8.

osuuaeou 502 Ill ..
.- 8m
238883 635.82
a a a .4 3
>3 5:2. @mm P .93 5.3..

 

 

 

 

I
\‘\
\
\
1
I

I
l
I

§
5‘30 861630 Bugmwg

‘.

l
l
I

 

 

 

 

v-

di-

 

 

 

 

 

 

monitored growing season (Figure 7). In Figure 7, the center line in each set
of three represents the mean, with upper line being the maximum, and the
lower line the minimum. In all three years the course of the lines show that
the OTC degree days are separate from the degree days in the control plots.
This indicates that the chambers accumulated more energy than the control

plots. Figure 8 represents the total amount of degree days accumulated in the

24

OTCs and control plots for each of the three growing seasons.

 

 

Degree days

1994

 

1995
Years

 

1996

 

Figure 8. OTC and control plot degree day accumulations during

the growing seasons of 1994, 1995, and 1996

The total degree days accumulated for 1994, and 1996 were approximately the

same, while 1995 was an comparatively cold summer, as seen by the fewer

 

25

number of degree days accumulated. Each summer the OTCs accumulate
more degree days than the control plots (Figure 7).

Belowground temperatures were also measured for a portion of the
summer in 1996. Thermistors were inserted into grooves on a wooden
dowel, at 1cm, 5cm, 10cm, 15cm, and 30cm below the soil surface. The dowel
was inserted into the ground, beneath a chamber and beneath a control plot.
This provides a profile of soil temperature for both OTCs and control plots.
Measurements were made from 12 July, 1996 and until 18 August, 1996. The
data (Table 5) demonstrate little difference between average soil temperatures
between chambers and control plots. Using an ANOVA method of analyses
(With a Box-Cox transformation of square root, see chapter three for a more
detailed description of analyses) this difference was determined to not be

significant. The absence of significant soil warming in the OTCs may be

Table 5. Average soil temperature (°C) at different depths (cm)

 

 

 

 

 

 

Depth of OTC Control
thermistor plots plots
1 6.56 5.98
5 4.79 4.97
10 3.65 3.49
15 2.24 2.35
30 0.63 0.36

 

 

 

 

 

because of the large heat sink properties of the surrounding tundra. Larger

chambers might create such an effect although Hollister (personal

 

communica
chambers, a
soils of the c
the large sm
observation

(Figure 9). l

' Debthktn)-

26

communication) recorded a warming in a wet meadow site with identical
chambers, at Barrow. Therefore, it is likely that the fine silts, sand, and gravel
soils of the dry heath site do not store heat, but instead it may be conducted to
the large surrounding heat sink. Active later measurements confirm the
observation that the OTCs do not warm the soil underneath the chambers

(Figure 9). The active layer (the zone of the soil that melts each growing

 

 

 

 

 

 

    

 

 

 

“ID
—0— CTL
...... ..... OTC /A
an ab , :17" "
A 1 996//// /1
LE, so + / A /1'{,
z I/ [/7] 1 9 9 S
H r‘/
80 -_ 0 :i-H
2m ._ I" fl
. .:
i" .1-
0 1 1 1 1
1415 166 1d) {05 d 240

 

Julian Day

 

 

Figure 9. Average thaw depth for 1995 and 1996.

season and in which the roots and microflora are active) was measured

every day for the first two weeks after snow melt. The active layer was

27

measured by forcing a small metal rod (1cm diameter) into the soil untilthe
rod reached the permafrost. The thaw depths measured throughout the
season show that the thawing of the active layer is not significantly deeper in
the OTCs. The fact that the overall thawing was greater in a warmer year
(1996) compared to a colder year (1995) supports the idea that the OTCs are not
large enough to overcome the massive heat sink of the ground. Active layer

development was greatest in 1996 and thaw began earlier.

Community composition

In the summer of 1995 the community composition of each of the plots
was determined using the point-frame analysis method (ITEX Manual, 1996).
This method utilizes a 75cm by 75cm frame that has a grid with cross-wires at
each 7.5cm interval. This creates a 100 point grid, which at each point, the
name of the plants and the height at which they were sighted was recorded,
down to the ground level. This allows for percentage cover and frequency of
each species to be calculated which permits a comparison of the control and
experimental plots. This data also serves as a baseline measurement for
future studies of community change.

Table 6 contains the average percentage cover, and frequency for the
ten most common vascular plants for each of the plots. A percentage
similarity test indicated that although there are differences between the
percentage cover of species between the OTCs and the control plots, that this

difference (86% similar) is within the limits (80% similarity) acceptable to

 

phyt
have
Cont
l. 3?}
of thi

heath

Table

 

28

heath ridge at Barrow consists of 24 OTCs and 24 control plots.

Table 6. Average percentage cover of common vascular plants.

phytosociologists (PJ. Webber, personal communication). The OTC plots
have 8.95% cover of moss, 24.5% cover of lichens, and 1.67% bare ground.
Control plots have 10.50 % cover of mosses, 21.86% cover of lichens, and
1.03% bare ground. The ITEX experiment is designed to minimize the effects

of this natural variation by having a large number of replicates. The dry

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Frequency Frequency
96 cover i 5.5, 96 cover + S.E. of of
OTC OTC Control Control occurrence occurrence
OTC Control
Cassiope 23.1 1 .9 l 6.0 l .1 100 100
tetragona
Sang: _ 18.4 1.7 20.6 1.5 100 100
rotundlfolra
LUZU’a 3.4 0.7 3.6 0.5 91 100
confusa
Stella”? 2.4 0.5 2.4 0.4 1 00 1 00
Iaeta ‘
Arctagrostis 2.0 1.0 2.1 0.7 62 62
Iatifolia
Potentilla 2.2 0.4 6.0 4.1 96 1 00
Imparctica
Po? 1.2 0.2 1.0 0.2 92 83
arctlca
Luqua 1.0 0.2 0.8 0.2 75 71
arctlca
Carex 1.1 0.7 0.2 0.1 1 2 1 2
_§29s
Saxifraga 0.9 0.2 0.7 0.1 83 92
. gunctata

 

Summary:

 

29

The Dry Heath ridge ITEX site was established at Barrow, Alaska in the
summer of 1994. 24 OTCs and 24 control plots were randomly assigned across
this heath tundra. Plots are permanently marked with stakes and corner

markers. The chambers are constructed of a light weight fiberglass , Sun-Lite
HP" , in a hexagonal shape. Point-frame analysis was performed on all plots

to determine the percentage cover and frequency of species. Cassiope
tetragona and Salix rotundifolia are the co-dominant species. The plant
composition within the chambers and controls were determined to be the
same (86% similarity) with some natural variation.

Chambers were shown to be effective in increasing the air temperature
during the growing season by 1.5 - 1.7°C. Degree day accumulation in the

OTCs was separate and more evident in the OTCs than in the controls
throughout the growing season. Total accumulation was larger in the OTCs.
Chambers were determined to not have an effect on the soil temperature at
different depths, and the active layer depth in both OTCs and controls was
similar throughout the season. Active layer development was greatest in
1996 and thaw began earlier, possibly because 1996 was warmer than 1995, and
snow melt occurred earlier. Each year had a greater growing degree day
accumulation in the chambers than in control plots, with a similar pattern of

accumulation.

Chapter 3

PLANT RESPONSES TO EXPERIMENTAL WARMING

Experimental design summary

The ITEX site on the dry heath at Barrow, consists of 24 OTCs and 24
control plots, in which plant developmental stages and growth are monitored
visually throughout each growing season. Due to a large amount of
variability inherent in tundra ecosystems, a large number of replicates are
needed. Each plot is treated as a separate replicate, with the earliest

occurrence of phenophases for each of the species recorded as a data point.
Chambers increase the ambient air temperature 1.5 - 1.7 °C (see discussion in

chapter 2). Effects of this induced warming are analyzed using an ANOVA,
with a conservative significance level of 0.01, to try and compensate for the

large amount of variability within the data (see pp. 32-33, this chapter)

Effects on phenophases
Phenology is the study of the seasonal timing of plant development.
Each phenophase is a specific stage in this development. Phenophases for

species found both in OTCs and control plots were monitored daily.

30

Phenology
were moni
and in 1991'
length for 1
daily by ex
species had
the first pla
each plot be
method of 1
handbook, 1
Plot, so that
the 6Xperjm
lALN’OVA).
Phenc

OCQHTQRCE C

developmeh

been SHOW f;

the Jillian d a

31

Phenology was monitored for all species contained within plots. In 1994 plots
were monitored from 15 June - 18 August, in 1995 from 12 June - 23 August,
and in 1996 from 28 May - 14 August. Average monitored growing season
length for 1994, 1995, and 1996 was 72 days. Phenophases were monitored
daily by examining each plot visually to determine if any of the contained
species had reached the next phenophase. The Julian date was recorded for
the first plant within a plot to undergo each of the monitored stages, with
each plot being treated as a separate replicate throughout the project. This
method of visually monitoring each plot is also described in the IT EX
handbook, which in addition states that each OTC plot has a parallel control
plot, so that pseudo-replication is avoided (ITEX Manual, 1996). The design of
the experiment allows for examination of the data by Analysis of Variation
(AN OVA).

Phenophase data were examined in three ways: first, the Julian date of
occurrence of a phenophase determines the actual date of occurrence of a
developmental stage of the plant; second, the number of days that the plot has
been snow free before a developmental stage was reached; and third, each of
the Julian dates of occurrence of phenophases were replaced with the average
number of growing degree days that had accumulated up to the time that the
phenophase occurred. These three methods of analyses allow a comparison
of calendar date, days since snow melt, and cumulative temperature to
determine significant differences between development in OTCs to that in the

control plots.

32

Data Desk 5.0.1 software (Data description Inc., 1995) was used to
analyze all data. Data was examined for normality and equality of variance,
and was transformed using a dynamic method. It was determined that
because there was not an a priori reason for transforming the data, the Box-
Cox method was necessary to transform the data. This method utilizes a log-
likelihood function to determine the best transformation, which was -0.5 for
all Julian date of occurrence of phenophases; 0.0 for all growing degree day
accumulations for each of the phenophases; and 0.0, or -0.5 for days since

snow free occurrence of phenophases and growth measurements, (which
allowed for ~ normal distribution, and ~ variances). A three-way ANOVA

was used to determine significance between category variables of species, year,
and treatment for all response variables, except for phenophases undergone
by Sal ix and Cassiope that did not have.corresponding phenophases in other
species, for which a two way AN OVA was performed. All interactions were
examined as well, with Post-Hoc Sheffe tests as a conservative method of
determining all possible contrasts, and partitioning of the variance (Sokal and
Rohlf, 1995). Julian date of occurrence data was analyzed by separating woody
from herbaceous data, so that the assumptions of normality and equality of
variances could be met for the two separate ANOVAs that were performed.
The distributions and variances of the number of days since snow free, and
the accumulated growing degree days allowed for all species to be analyzed

together. Responses were considered to be significantly different in the OTCs

and control pl
than 0.01, so t

For all 1
treatment (OT
These main efl
the main effec
lhe main effec
gmtn'ng degre
Significantly d
P1“ WP? (eith
AP’Pendix A ('
Van'ables testf

P051 ‘ Hoe Scl

33

and control plots, either positive or negative responses, if the p value was less
than 0.01, so that a 99% confidence level was established.

For all phenophases measured the main effects of species, year, and
treatment (OTC or control) were all significant for Julian date of occurrence.
These main effects were also significant for days since snow free, except for
the main effect of treatment (OTC or control) for elongation of peduncles.
The main effects of species, and year were all significant for the number of
growing degree days accumulated, while the main effect of plot type was not
significantly different. These main effects indicate responses across species,
plot type (either OTC or control), and year, regardless of the other effects.
Appendix A (Tables 8-10) contains all ANOVA tables for all dependent
variables tested. Appendix B (Tables 11-13) contains all significance levels of
Post - Hoc Scheffe tests.

Emergence of firetgneenleaf, WW and
W (first stigmas visible, or glume open) were monitored for
the graminoid species of Arctagrostis latifolia, Luzula confusa, and Luzula
arctica. The phenophases MW and first
mm were not significant at the p < 0.01 level for the species of
Arctagrostis latifolia, Luzula arctica, and Luzula confusa for all three years,
regardless of whether Julian date of response, or days since snow free of the
response, were being examined. The Julian date of Wu
was significant (p < 0.01) for Luzula confusa in 1996 (p < 0.001), as well as the

days since snow free (Box-Cox transformation 0.0) in 1996 (p < 0.001). Table 7

records the total

trend of the spet

show the trends

occur earlier in t
12 represents tht
before Arctag ms
the next develo;
energy alccumu].
the .1310th acct
SIigrlificantly Iarg
OTCS {0r Luzu 1a
Figure 12 demo:
Phen0pha595 00
the effeqs of inc
phenophaSe to o
requirement to c
more grorn‘ng d1
that SOme factor

needs to be met
1

34

records the total number of responses for each of the species, indicating the
trend of the species in response to experimental warming. Figures 10 and 11
show the trends for the occurrence of phenophases in both sets of data to
occur earlier in OTCs, although these differences were not significant. Figure
12 represents the average amount of degree days that were accumulated
before Arctagrostis latifolia, Luzula arctica, and Luzula confusa underwent
the next developmental stage. There is a general trend that the amount of
energy accumulated in the OTCs is approximately the same as, or larger than
the amount accumulated in the corresponding control plots. In 1996, a
significantly larger amount of growing degree days were accumulated in the
OTCs for Luzula arctica to have Warm: (p < 0.01).
Figure 12 demonstrates that although Figures 10 and 11 show trends of
phenophases occurring earlier within the chambers, it is not always due to
the effects of increased temperature. Although the trend is for the
phenophase to occur earlier in the OTCs, there is an additional energy
requirement to do this, demonstrated by the trend for OTCs to accumulate
more growing degree days than the control plots. This trend also suggests
that some factor or condition that is not, or less, temperature dependent

needs to be met to reach the next phenophase.

35

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

221: o _ o a new
3:30 : o a 02.. >63 unan 880.3% emaubuum
3:30 _ _ _ 80n— 3:3
5:30 a E...» o w c a new
3:30 0 Av c Q 00.... >63. 0:00 3656.360 36m
3:0.M0 m 2 o 9 03n— 3:3.
03m: o c m m 000
3:30 N o _ _ _ 02.. >65. nuan— 3333 3.650%
3:30 _ _ o N. 0.00 3:3.
06:. .6 .060 .v o o m ODD
3:30 .6 6.60 v G o m 00.: >63 mwao sagas 5:05
3:30 col—3:00 v n _ m 030 3:3.
I 03:. m n o o 000
3:30 6 .030 N o a m 00... >65 «N30 863 5:05
3:30 .6 .060 m o c o 030 3:3. _
3:30 N _ .n b 0010
3:30 o m c Z 00... 30:0 0.30 0.66.33 03.5.0.0
3:30 o w o 2 03n— 3:3.
6.60 m o o N 000
3:30 3 360 v o o m 00.... >65. mxao 56035»: £633.03.
3:30 6 3:3 m o o v 030 3:3
:8de
30:030.. mom—5:000 + 30:03.00 3 mom—6:00.33 3020.3
23: 6.60 .6 a 232.6»: 6 u 33— 6 a 3:30 .6 n .6 0&2. 360nm

 

309530.. “:39 E 06:00. 6 ccza£§0~0v a 00 ”5:303 2 80:05:00 6 33:52 N 030.:

 

330% 205803 00.3. .8. 00:03:80 000.3233: :o 0.20 5::— .o~ 0.5m:—

 

 

36

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

58 03.0.3 30.505 :03 03.03 $0.596
88083:; 02.“. 85083:... 3...... 02¢ 00 85905 8:88ch 00.: 85083:... 0a.... 08: .0 8.3905
-002 32 32 82 32 32 08. 32 3.: ov- 108. 32 $2 82 30. 32 08. 32 3.... o:
09.. — — - cm—
.I r
m: m.
m... m...
u 0
cap 8—
p p
B e
(A K
OON 8N
CNN CNN
NmeCOU N~3N31~ $0.305“ N~=N=d
se 3.... 35%
0:5 8:883 00... 0 o 8 3 .
80. .0 08... “aw—532... 000mg”. 82 3%.: poovn a.
o: 3.550 U
- UFO I
oo—
.I
m.
m.
u
8p
0.
B
K
8N
ONN
3.38:3 mzmogmmut<

 

 

86me 20582» ~85 .8. 88.39.23 3 «8:0 2: 2 3E 93v 3.: 30:... .o awn—=32 .: 95E"—

 

37

 

  
  

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

88 .32, 35% 58 is, 38.5
8508ch am...“ 85085:... a»... 3 850.25 85088:... 85088:... at: 3 8593.6
jog. 32 32 82 n8. 3.: 82 “-8. 32 o 132 32 32 82 32 32 82 “.2: o
_ m 1m
9 S S
l. o— m.
m a
ow S S
w m
M 8
Om .3 4..
w w
ow cm
39.39 «3qu 3.58m «ENE
onus .8. 5.5
53 0620 85088:... an: an! 3 85925 ‘
8.: 82 32 v8. 82 32 32 o podvn .c
m gan—
o. M US I
8 5
gm
S
9m
on .1
s a
£8.52 mtmoafiogc‘

 

 

38

.86on 29:82» 3.2.: 5 88:90:29 .0 8:92:80 2: 2 Sta acorn—2:33 .30 camoo MEBEU :2 959m

 

58 «32> .8. :85 .o 88 23>
8:88.85 Em 85098:... 5...... 889880 5....

umm— mmm— vamp uam— mmm— vamp mam— mmm— vamp wan, mmm— vam—

 

 

 

 

 

 

 

 

 

 

28.5053

85088:; «as... 8§8ch E...— 88925 Em
mam. mam. cam—

oaa— mma— vam—

 

 

mmegu $3.3 3.58m 5:qu

o 1 o
V
w 8
e m w
9
s w m.
a
n w 8
mw nu.
W6
8. R 8
mm
D
8. w _d_ 8.
.. 8.

 

 

 

Edvn t
3.8 U
“:0 I

 

 

58 0133 32505.0
85088:... at: 85088:... at: 85985 Em
32 can: 32 :3— omm— 3.: 3mm—
] I. II I o
o... w
r. m
8— ms
9
. 8N n w.
w 0
SN m5
m
. an w

 

 

 

wzohyfl mtmEQEE<

uonelnumaov Keg
33.1630 Bugmmg 363.131“,

 

 

39
E EE’ | l E E E' | . E]
l I. E l l . [f If] and °|l . {f If]
were monitored for Papaver hultenii and Saxifraga punctata, two erect forb

species. The phenophase W leaf did not occur

significantly earlier in the OTCs, in any of the three monitored growing
seasons. In 1994 the W occurred significantly
earlier in the OTC for Papaver hultenii (p < 0.001) when examining Julian
date of occurrence, and in 1995 for date since snow free(p < 0.01). The
phenophase of WW occurred earlier in OTCs in 1996 for Julian
date of occurrence for the species of Saxifraga punctata. OTCs did not have a
significant effect for the phenophase of W. In 1996, the Julian
date of occurrence of the WW occurred earlier
in the OTCs (p < 0.001). Figures 13 and 14 represent the Julian date of
occurrence, and the days since the plot was snow free at which time the
phenophases happened for both of these species. Again there was a general
trend for most phenophases to occur earlier in the OTCs, although most of
these results were not significant (p<0.01). Although Figure 13 shows a trend
for there to be more degree days accumulated in the OTCs before undergoing
phenophases, most of these differences were not significant. The only

phenophase for Papaver hultenii to require a significantly larger amount of

accumulated growing days was the stage of the WW

4o

 

Papa ver hultenii

 

 

 

 

 

 

 

 

 

 

 

 

a)
I OTC
D Control
* p<0.01 *
zn I'Ll
>~.
(B
D
C m
5..“
3
fl
to
no * 0
1994 1995 1996 1994 1995 1996 1995 1995 1994 1995 1996 1994 1995 1996
Ermrgenoe of Emergence of Elongation First flower First flower

fest green leaf first bud of peduncles open wither

Saxifraga puncta ta

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

a)
I OTC
[3 Control
an * p<0.01 7
>
3 FL
I) I
C
.9
'5
fi
0
m . ;
1994 1995 1996 19941995 1996 1995 1996 1994 1995 1996 1994 1995 1996
Emergence of Emergence of Elongation First flower First flower
first ween leaf first bud of pedmcles open wither

 

 

 

Figure 13. Effects of OTC warming on Julian Date of occurrence of forb
phenophases.

41

 

Papaver hultenii

 

 

 

 

 

so
4S .0“:
a: 4o [3me
e * p<0.0‘|
“- 35 ’
g 30
m 25
o
8 20
i7)
m 15
is“
D

 

 

 

 

 

 

   

1994 1995 1996 1994 1995 was 1995 1996 1994 1995 1996 1994 1995 ms
Emergence of Emergence of Elmgation First flower First flower
first ween leaf first bud of perhncles open wither

Saxifraga puncta ta

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

so
Iorc

q, so DControl

0 * p<0.01

h ,

LL 4o

5

6%
30

m

U

.E

U) 20

m

g _

D 10 , l
o

194 I” 1% 1994 1%!) I“ 1995 19% I”! 1% I“ ‘94 15 I”

Emergenoeof Emergenceof Elongation Flrstflower Fistfiower
Mtueenleaf ihtbud ofpedncles open wither

 

 

 

Figure 14. Number of days snow free prior to the occurrence of forb
phenophases.

42

 

300

Average Growrng
Degree Day Acuumulation

300

200

Average Growmg
Degree Day Acuumulation

 

Papaver hultenii

 

 

 

- OTC
D Control
* p<0.01

 

 

 

 

 

 

 

1994 1995 1996 1994 1995 1996
Emergece of First bud

firstweenleaf visible

 

 

 

 

 

1995 1996 19941995 1996
Elongation First flower
of peduncles open

Saxifraga puncta ta

1994 1995 1995
First flower
wither

 

 

 

 

I OTC
D Control
* p<0.01

 

   

 

1995 1996 19941995 1936
Elongation Firstflower

ofpeduncles open

 

 

1994 1995 1996
Fist flower
wither

 

Figure 15. Number of accumulated Growing degree days prior to the
occurrence of forb phenophases.

 

 

 

43

Many Saxifraga punctata phenophases seemed to need significantly larger
amounts of growing degree days within the OTCs. In 1995 the phenophase of
W195. required a significantly larger amount of
Growing degree days (p = 0.01). The phenophase of first withering of a flower
required significantly more growing degree days within the chambers in all
three years (p < 0.01, p < 0.000001, p < 0.000001). Both forb species also show a
general trend of equal or larger amounts of growing degree days within the
OTCs to undergo phenophases, although most phenophases show a trend of

occurring earlier within the chambers.

For Salix rotundifolia emergence of W W
mm mm W and Wm
were monitored. In 1994 small wires were placed within the plots to mark off
small areas of Sal ix that were determined to have male or female catkins.
These became the monitored areas for the proceeding growing seasons, and
only stages that occurred within these units were monitored (also stated in

ITEX Manual, 1996).

In 1994, and 1996 for Sal ix rotundifolia the Julian date of occurrence of
W took place significantly earlier in the OTCs. The Julian
date of occurrence for the phenophase of W occurred earlier
in the OTCs in 1994, and 1995 (p < 0.00001; p < 0.01). The male stage of“
129mm took place significantly earlier in the OTCs in 1994, 1995, and 1996

for the Julian date of occurrence (p = 0.01; p < 0.01; p<0.01), and for the days

4.4

since the plots were snow free in 1994 and 1995 (p < 0.01; p<0.01). The next
male phenophase,_all_pg11gn_§hgd, took place significantly earlier in OTCs in
1994, and 1995 for Julian date of occurrence (p < 0.01; p < 0.01), and in 1995,
and 1996 for days since snow free (p < 0.01; p < 0.01). The OTCs did not have
a significant effect for the next female stage of W. In
1996 the vegetative stage of mm: had a significantly earlier Julian
date of occurrence, and fewer days since the snow melted off the plots (p =
0.001; p < 0.01). Figures 16 and 17 again show the general trend of
phenophases occurring earlier in OTCs than in controls. These trends agree
with those found on other Salix species (as well as S. rotundi fol ia) within the

ITEX community (Jones et a1. 1997).

Sal ix had several phenophases that consistently appeared to need
significantly more energy in the OTCs than in the control plots. Significantly
more energy accumulated within the OTCs in all three years (p < 0.01; p <
0.000001; p< 0.00001) before the male phenophase of aflpgllmshgd was
reached, even though the phenophase occurred earlier. Significantly more
energy amassed within the chambers in 1995, and 1996, before the female
phenophase MW was reached, even though the phenophase

did not occur significantly earlier.

.mmmazaocozm 3.3.5.352 3:3 «0 35:33 .0 Son— 523 .3 859m

 

45

 

855 sense .85 3.» 28s .3. :85 fie

8.3 7...... too» at: :38 -< :88 ans... gm 32E .0 cocoocgm
323232 82323.: 323233. 823232 mam—mmflga. 323232

_ w o:
i on.
a on—

  

 

 

 

 

 

 

 

 
 
 

 

 

_H_ WII
m.
U
_ _ Oa—
¢ 4 as m
.d. 8... MA:
O—N
ONN
.4.
H_ on...
033.

 

£83232 £3

 

 

62.580 omenmocosm a 283 3.: 393 :25 me: .03 2: gap .0 «3:52 .2 0.5mm—

 

 

  

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

855 328% 8.3 no... sans .8. =85 a»...
3.8 E: 28 E: 8.8 .< 5.8 6: Sue: E: .o 8:255
loam. man. 33 can. 33 v3— wmm— mmm— vamp cam. 33 vom— mma. manl— vma— 32 man. vamp o
,. - _ a—
G
3 e
,A
.L _.1_ S
a. a. On 8
U
3
ov a
S
w
an M
d
on m
f 3
.1
as
on
23:33 8 £8

47

 

 

 

 

 

 

 

 

 

 

 

 

 

 

§
$9
"“5
gs
§
'7‘, —:
r: i ii;
‘2 l—gi
3 ‘_35
0 :—§§3
x
-: r-Igg.
6'3 "'m
-§g
s§§ ”ii
In. 'M.
‘ .ggt
§. 5 3 5 .3. § § :3 .93 3 °

uouemwnoov Keg
99.1690 5UlM019 SBBJGAV

 

Figure 18. Accumulated Growing Degree Days prior to Salix roiundifolr'a phenophases.

48

These results show that the earlier phenophases are perhaps more dependent
upon the date of occurrence, or time since snow free, while the latter
phenophases are less dependent on temperature, since the OTCs accumulated
more energy, although for the most part the phenophases did not occur

earlier

In Cassiope tetragona the phenophases of W first
elongationflpednncles ficsLflmmpm andflrstscmllafimp were
monitored on a specific marked ramet for the growing seasons of 1994, and
1995, and the stage of W in 1996. Because in the 1996 growing
season there were not any marked ramets that had buds emerge, the entire
plot (control or OTC) then became the monitored unit, and any stages that

any of the Cassiope within the plot underwent were then recorded.

The first phenophase of W was significantly earlier in the
OTCs in 1995, than it was in the control plots for the Julian day (Figure 19) at
which it occurred (p < 0.01), and earlier in the OTCs than in the control plots
for the days since snow free (Figure 20, p < 0.01). The next stage of first
elongafigngflmdfi was significantly earlier in the OTCs in 1994, and 1995 for
the Julian date at which they elongated (p < 0.01) and in 1995, for the days
from which the plot had been snow free at which they elongated (p = 0.01). In

1994, 1995, and 1996 W significantly earlier in the OTCs for the

.egomutu. “£9.85 .8. 88,399.23 .9 8:99.980 .9 32. 5::— da 953m

 

 

49

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

do... :30 «3052“ do «53> van
«=28 6...”. .026: “PE 5.69.0.0 “2.“. Bio: «2E
mam. mm? 3m. camp 39 $9 mam. mam. vamp mam. mm? 39 o
- 3
W 8.
W PI
_ _ m
9 w.
com
_ _ G
H m
_ _
H.
Edvn « CNN
.228 U
05 I
O¢N
mcomm: 3 muofimmu

 

 

.omasmoconm 3393.8.“ 5 3:38 28.826 823 8.: 395 $3 8E 95 35 9:3 .9 .6953: 2:. .8 893m

 

 

 

 

 

 

 

  

 

 

 

 

no... coco «.863 .o 0363 van
«=98 «2...— ..030: it .c0_uauco_o 35E .330: ambu—
wmmp mam. vam— mmm. mam. camp 89 39 «a? camp 33 gm.

. c
a G
B
K
S

B
B
U
3
9. 9
S
U
a m
H
a a
8

R

mcommb 8 30.983

 

 

 

51

.omacmocosm a owumvca 3 K0330 .5. 89—mon— wEBouu .o «.595 «mango of. Au 953m

 

89 5% 2893 .0 «33> .93
2.28 92.“. .250: 92: 5:85? «mi 326: 55
mag mmm— vmm— mmm— mmm— 32 mmm— mmm— vamp cam. .33 vamp

 

 

  

 

 

 

 

 

.Qova «,
.228 U
0.5 I

 

 

 

 

 

 

«comm: 3 mqoummmu

Om—

OO~

OnN

can

'uogieInwnoov Keg 991690

5U!MOJ[-) 95919Av

 

 

52

Julian date at which the phenophase occurred (p = 0.000001; p = 0.00001; p =
0.001). In 1994 filmed significantly fewer days after the plots had been
snow free in the OTCs than in the controls (p = 0.01). The last phenophase of
Wm occurred earlier in the chambers in 1994 and 1995 for the Julian
date at which the flowers dropped their corollas (p < 0.001 ; p = 0.001). Coronas
W significantly fewer days since the plot had been snow free, from the
flowers in the chambers , than in the control plots for all three years (p <

0.001 ; p < 0.0001; p = 0.01). Figures 19 and 20 show the responses of Cassiope
plants to warming at the canopy level. This species also shows the general
trend of phenophases occurring earlier in the OTCs, but again many of these
differences between chambers and control plots are not significant.
Significantly more growing degree days accumulated in 1996 before the
developmental stage of Wimp occurred (p < 0.01). Except for this
one phenophase in 1996, all other phenophases show a trend of similar

amounts of accumulated energy within years.

Figures 10 - 21 and Table 7 all show trends of the phenophases
occurring earlier in the OTCs , however, because of the variability of the
system these trends were not found to be significant for most species
phenophases. Measurement of the Julian day at which a particular stage
occurred for a plant, or the days the plot has been snow free before a stage
takes place both determine that the woody species, Cassiope tetragona, and

Sal ix rotundi fol in are more responsive to the effects of warming (Appendix 1

53

and 2). Table 7 contains a summary of the significant responses to warming
that occurred during the three monitored growing seasons. The total number
of significant responses per year indicate that regardless of the year’ 5
environmental conditions the chambers had the same effect (Table 7). Julian
date of occurrence of phenophase graphs indicate a trend for phenophases to
occur at earlier Julian dates each year, but since the control plots are also
undergoing phenophases at earlier dates, this indicates that it is not a
cumulative effect from the chambers that is being observed, and is likely an
effect of time of snow melt. Comparing Julian date of occurrence to number
of days the plot is snow free before W occurs indicates
that this stage is dependent upon when the plots melt out. All species
undergo this stage at a later date in 1995, than in 1996, but the number of days
since snow free until the Wages determines that 1995 takes the least
amount of snow free days until a leaf emerges. This indicates that it is a
complex series of processes by which the WWW since 1995
was the coldest year, with the least amount of heat accumulation. Also, since
each of the species utilized a significantly different amount of energy each

year, W is not likely to be dependent upon the amount of

heat that is accumulated.

Graphs 12, 15, 18, and 21 show the total amount of energy that is
accumulated before a phenophase occurs. These graphs determine that it is

not temperature alone that is the determining factor of a species

54

development. If the hypothesis that it was only temperature that was causing
the accelerated phenology, the total amount of energy for the different
phenophases would be expected to be similar regardless of year, Julian date of
occurrence, or the time period since snow free. However, some of the
examined species show that some phenophases, although they do occur
earlier within the chambers, seem to use significantly larger amounts of
energy than the corresponding control response. This suggests that there is a
more complex reason than just increased temperature as to why the
phenophase is occurring earlier. The amount of growing degree days
accumulated before a phenophase is induced, is also very dependent upon the
species. Some species are more responsive to temperature treatments, and
therefore, similar amounts of energy are accrued in the OTCs and in the

control plots.

Both species of Luzula show a trend that suggests that a step-wise
pattern in the Julian date of occurrence of a phenophase for both reproductive
stages measured, with 1994 having WM; and open at the latest
date, and 1996 with the earliest date (Figure 10). However, the number of days

since the plot was snow free indicates that the plants must be snow free for a

similar (or minimum) length of time before an W
.ngn(Figure 11). In Luzula confusa themgrgengufjnjnflorfigenge and
the WW may also be dependent upon the amount of

heat accumulated within the plot, since more time has elapsed in the control

55

plots than in the chamber plots. However, this relationship becomes more
complex (Figure 12) as the chambers in 1994 accrue a significantly larger
amount of energy before the inflorescence can open. Lindskog and Jonsdottir
(1997) found in their study of the graminoid sedge Carex bigelowii that the
vegetative efforts of plants within the chambers and in control plants are
similar, although reproductive efforts within the chambers are accelerated.
This trend is similar to that of the graminoid phenology on the dry heath,
although the differences between OTCs and controls in Barrow are not
significant. The examination of the amount of accumulated growing days
shows that this trend of the phenophases occurring earlier in the OTCs
requires even more energy than the plants in the controls, which suggests a
trade off between occurring earlier, and needing more accumulated energy to

undergo a developmental stage.

The two forb plants, Papaver hultenii and Saxifraga punctata, also
show this decreasing step pattern of the Julian date of occurrence of when
phenophases occurred. Again, the number of days that the plot has been
snow free reverses this pattern, and instead indicates that for at least the first
three phenophases, fewer days occurred in 1995 before phenophases
developed than in 1994, or 1996. In 1996 the average number of days before
WWW in Papaver hultenii within control
plots was similar to the number of days in 1995 that produced the same

phenophases. This suggests that it is not the specific date (or corresponding

56

environmental conditions such as light intensity) that triggers the plants to
open their flowers. However, significantly different amounts of energy each
year are used to produce the phenophases of these species. It is likely that
other environmental conditions are playing a role in allowing these species
to develop. Molgaard and Christensen (1997), and Alatolo and Totland (1997)
found in studies of Papaver radicatum (which is related to Papaver hultenii),
and Silene acaul is (a forb), that both had an earlier eneeLeLflmefing in
response to increased warming. Papaver hultenii , and Saxifraga punctata
follow this trend, although the differences between the OTCs and control are

not large enough to be significant.

The two most responsive species, of those that were examined, were
the woody species, Cassiope tetragona, and Salix rotundifolia. In 1994 and
1995 all reproductive stages for Julian date occurred significantly earlier in
OTCs than in control plots, except for the phenophase of We.
The number of days since the plot was snow free, displayed a similar result
except Webb, and befimbndebngefing, which did not
occur significantly earlier in the controls. In 1994 and 1995 the Julian date at
which Sal ix rotu ndi fol ia underwent reproductive phenophases all occurred
earlier in the chambers than in the controls. The stage of firethenehed,
and gnmuenehed in 1995, and the stage of W in 1994, took
place a significantly fewer number of days from the plots being snow free in

the OT Cs, than did the control plots of Sal ix. Unlike the trends shown in the

57

graminoid and forb plants, the patterns of response in Cassi ope do not
change when examining the Julian date at which a phenophase occurred, or
the number of days since the plot had been snow free that the stages took
place. Also, all stages of Cassi ope development, except for eqmllefimp in
1996, accumulate similar amounts of energy to undergo phenophases, which
suggests that Cassiope is sensitive to changes in temperature. This
contradicts what was found in a similar study in the Swedish Lapland site of
Latnjajaure (Molau, 1997). However, because Barrow and Latnjajaure
represent two different areas within the species distribution, the findings of
Wookey et a1. (1993) support that species found in the High Arctic respond
more to an increase temperature. The latter phenophases that Sal ix
undergoes show very little difference between treatment, or the way in which
they were analyzed, which suggests that a possible ”catch up” effect is taking
place as reported for alpine species by Bock, 1976 and Webber et al., 1976. This
means that at the end of the season plants within the control plots have
reached the developmental advancement level of the plants in the chambers.
However, Figure 18 shows that the amount of energy trapped during the time
at which these latter phenophases is significantly larger, which suggests that
perhaps a different factor than that of temperature, is responsible for the

occurrence of these phenophases.

Effects on growth and stature:

Two different types of growth measurements were recorded to
determine a quantitative response of the tundra plants to the experimental
warming. At the end of each of the three monitored growing seasons the
stature of each species was ascertained. In 1994, 1995, and 1996 the length of
the reproductive shoots (graminoid inflorescence, or forb peduncles) were
measured. In each of these three growing seasons, the tallest three
individuals of a species were measured for each of the plots that contained
the species. In 1996 the stature of vegetative shoots (for graminoids) were also
measured following this same procedure. Cassiope tetragona was measured
for annual growth increment, by measuring the amount of growth that had
been produced at the end of each growing season. The length of the longest
Salix leaf within a marked unit was also measured for each of the three
years. Data was again transformed using a dynamic Box-Cox method. All
growth data was transformed to 0.0 level, except for Cassiope, which was
transformed to the -0.5 level. All end of the season stature and growth
measurements were analyzed using a three-way ANOVA with the main
effects of species, year, and plot type (OTC or control plot), and Scheffe Post-

Hoc tests to look at specific non-orthogonal contrasts for each of the
interactions. Again, the data was examined at a significance level of 99% (p s

0.01).

59

In 1994 and 1995, two species, Luzula confusa, and Saxifraga punctata,
had a significantly larger length of reproductive shoots inside the OTCs, than
plants in the control plots (Figures 22 and 23 , p < 0.01, p < 0.01, p < 0.01, p <
0.01 respectively by species, and year). Figure 24 shows that Cassiope did not
have a significant increase in annual growth increment. This lack of
significant differences in yearly growth is also supported by Molau (1997 ), who
suggests that because Cassiope and other cushion and dwarf evergreen plants
show a similar lack of response to increased temperature that they could be a
possible grouping of plants for further analyses. In 1996 the only species to
produce taller reproductive shoots within the chambers, as compared to those
within the control plots was Luzula arctica. Further studies are needed to
determine the possible case of this observed increase in growth. In 1996,
Luzula confusa and Luzula arctica had significantly larger vegetative shoots
inside the chambers (p < 0.01). However, since increased length is a common
response among graminoids to shelter (Lindskog and Jonsdottir, 1997), it is
impossible to determine if these increases in growth are in response to the

increase in temperature.

In 1996 growth measurements were also made throughout the season
to track the differences between plants of each species within the chambers, to
plants within the control plots. This was accomplished by placing a small
wire circlet around the first emerging individual of each species, in each plot.

This allowed for the same marked plant to be measured every other day for

$55.33 633.65 2 335:5 Manganetucx can .anaxgoo £3.25 €9.32: 3335 go 8259.2 539.0 .NN gamma

 

 

 

 

 

 

 

 

 

 

 

 

 
 

 

 

 
 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

89.» 39.» woos» 39.9 805 ~09? goon»
ozufiooo> 028329.. 0526953. ix 3:380) 3.52.9.9: 0593850..
camp 82 mam. .32 0 camp 03— mam—
H
..c m.
5
m
.w
5 .
2 :1.
r « NP
«macoo «ENS 838w «35:
~85
25833.: . ,
@Om— F0.0VQ a.
o .228 U
.L
Lm . .
H
m.
6
i.— \ml
:8
3
£352 $395393.

 

 

61

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Saxifraga punctata
r—L. r.lL.
12
5, a
H
.C
.9
£3 4
0‘ are cn. arc CTL 07c cn.
1994 1995 1996
Papaver hultenii
20 * I
16
5, .2
E
.9 8
0
I
4
0‘
1995

 

 

 

 

Figure 23. Growth responses of Papaver and Saxifiaga to increased warming.

62

 

Cassiope te traggna

 

—a -o
.fi 05

.0
N

n.
0

Length (mm)

fi

N

 

 

 

 

   

 

 

 

 

 

OTC CTL OTC CTL OTC CTL
1994 1995 1996

 

 

 

Figure 24. Annual growth increment of Cassiope tetragona.

the entire growing season. The largest leaf of the marked plants of
Arctagrostis latifolia, Luzula arctica, and Luzula confusa plants were
monitored throughout the entire season. Data was plotted using a scatter
plot, at which time a true-ess smoothing curve (Data Desk, 1988) was fit to
thedata, so that the approximate line model that the data fit could be
determined. A second order polynomial curve was determined to have the
best fit for all scatterplots. A comparison of curves tat (Potvin et al., 1990)
allowed the differences of the growth patterns of plants in the OT Cs to be

compared to the patterns of plants grown in control plots. The comparison of

63

curves test uses an F-test to determine if the fitted-curves of each of the
different treatments (OTC and control plants) are significantly different from

a curve of the same model that includes both of the treatments together.

Second order best-fit polynomial lines fit to the Arctagrostis lati fol ia
scatter plots (Figure 25) indicated that the plants in the OTCs were always
larger throughout the entire season, and at no time did the height of plants in
control plots equal that of plants in experimental plots, although the curving
slope of the line suggests that the actual growth rates were similar. The
comparison of curves test determined that the lines were significantly

different (F4565 20.99; p<0.001).

Second order polynomial lines fit to Luzula arctica and Luzula confusa
scatterplots (Figures 26, and 27). Luzula arctica lines are extremely different
for the OTC and the control. The first half of the growing season, the rates
and the amount of growth are approximately equal. However, during the rest
of the season, the plants in the chambers seem to have a growth rate that
reaches a critical level and then levels off. The Luzula arctic control plants
seemed to have a growth rate that continued to increase, forming an almost
straight line. The scatter of the points for the graph also differ. Plants within
the OTCs have a much smaller range than that of the plants within the
control plots. The comparison of curves test indicates that the lines are
significantly different (F4304 21.35; p<0.001). The best fit second order

polynomial curves that are fit to the Luzula confusa data show that the

64

experimental and control plants are almost identical. This measurement of
growth rate indicates that an increase in warming does not cause an increase
of height for this species, with the comparison of curves test indicating that
the lines were not significantly different (F4,929 1.057; p<0.01). The monitoring
of the growth patterns of these co-genors indicates that the plants vegetative

growth patterns do not respond similarly.

The differences that are seen in the results from measuring the three
largest plants at the end of the season, and monitoring plants throughout the
season can be explained by the fact that the number of points on the scatter
plot indicates that there were many measurements taken in order to reach
that point. The points reflect the data from many chambers. Within a plot,
there is only one plant being measured per species, and the monitored plant
was selected based on emergence, and not on size. These graphs also
demonstrate the need to monitor the growth of the plants throughout the
entire season and not just at the end, as important trends could be missed.
The end of the season height measurements yield data that suggest that it is
reproductive growth that is most responsive to experimental warming. This
trend is also supported by the phenological data. In the phenology data, only
one species in two years (and only for the Julian date of occurrence) happened
significantly earlier in the chambers. The rest of the species demonstrated
little response in the vegetative stage of first green leaf. Throughout the

grong season there were however significantly earlier occurrences of

Arctagrostis Iatifolia

 

8

8

Vegetative shoot length (cm)

 

 

o OTC o
><Cn_ o

o

000 00000
Xxx X

X

. .1." mm. x

Q...
' ‘ .4
.» w ~v » J,» .
x .‘o". o‘o‘o

§.&%&°xm§xx
f . . s

 

o l .1 % at l t
15) 1a» rm um “D an mo Z5
Julian Days
Figure 25. The effects of warming on Arctagrostis latijblia stature during

1996.

 

Luzula arctica

 

12
E oOTC
810‘“. XCTL XXX
x
5 "x
X x
g 8* x xxxxiifixxxxxxxxxxx
.‘2 x" X
OEXOogaxQEXQM
H O X 39<>< xgé x3
8 6"“— 3X X
.C
in
0
.2
H
N
4.:
d)
O?
G)
>

 

 

 

 

Julian Days

Figure 26. The effects of warming on Luzula arctica stature during 1996.

 

220

  

X 0 X . o O.
O ...o . x9 0
.. a 3 w. .
X X X
m Vm ....o E o o 10
we Xo 80 .1? o o J m
x .8
o o 8%500 0.5;? 00 o
x m o o o
m x 0an oo 2
x ooxo 3..., O
x O X 0 EOXxoo o fflgQO O 0 11m
x x O Ox 3.0x O

x x0 OX 3.009 O

0.. D cows: aha?

67

1
l
170

Luzula con fusa

 

o OTC
x CT L
>0

ex
.3:

i
160

_ _ b b
d J _ 4

L F

_ a W
2 0 8 6 4 2 o
1 1

AES 598. “occm o>33omo>

 

14

Days

Julian

Figure 27. The effects of warming on Luzula confusa stature during 1996.

68

reproductive phenophases, most notable in Cassiope and Sal ix.. These
response patterns could be an important tool in examining how theses species
will react to a possible global warming. Sexual reproduction could
becomemore important in this established perennial community that has

little seedling recruitment under present conditions.

Final discussion

Fossil records show that vegetation responses have lagged behind
climatic changes (Davis, 1989). Nevertheless, the present arctic ecosystem is
more limited by physical stresses, than by competition, and therefore these
ecosystems are sensitive to environmental changes (Roots, 1989). This project
was designed to examine responses to an increase in temperature that is
within the predicted magnitude of global change. To date, only short-term
responses of one environmental factor (i.e. temperature change) have been
examined at the Barrow site. Long-term studies of impacts of environmental
change upon vegetation have shown that short-term responses are relatively
poor predictors of plant responses after a longer time period (Chapin et al.
1995). However, this study lays the ground work for continued monitoring of

plant responses to increased warming.

A number of authors (for example: Webber, 1978; Shaver and

Kummerow, 1992) have shown that tundra species behave individualistically

L5

69

to, and have uniqe requirements and responses to environmental variation.
This study shows that the seven species examined at the dry heath site at
Barrow exhibit very differing responses to the experimental warming.
Among all of the herbaceous species none have consistent significant
responses to the increased warming, either between species, or between years.
However, the trends (Table 7) indicate in the data that most responses tend to
occur earlier within the OTCs, however these differences are not, for the most
part statistically significant. Most species undergo phenophases equal or
earlier in terms of Julian Days, and days since snow free, and use more
accumulated GDD within the chambers before undergoing a phenophase.
Papaver is the exception to using more accumulated energy, since the trends
for this species indicate that the control plots accumulate more energy than
the OTCs before some phenophases can occur. Growth responses are also not

consistent within species or within years.

Although the current year may have no effect on flower buds, as many
arctic species pre-form buds in previous years, the success of these buds is
dependent upon the conditions of the current year (Shaver and Kummerow,
1992). The two woody species that were studied were both the most
responsive in terms of the number of significantly earlier occurrences of
phenophases. More stability in the occurrence of a significant response is
evident in Cassi ope. In 1995 all phenophases occurred earlier within the

chambers, and since the total amount of accumulated energy is not

70

significantly different within the chambers and controls, this suggests that
Cassiope is dependent on temperature. Salix consistently utilizes more
energy that has accumulated within the chambers to undergo phenological
development, even though the phenophases are mostly occurring at the
same time (Julian date, or days since snow free) in both the control and OTC
plots. This suggests that the process by which Sal ixx undergoes its latter
phenophases is an extremely complex process, that is likely not dependent on

temperature.

Although temperature is an important factor in any arctic ecosystem,
the plants on the dry heath tundra showed few consistent year to year
responses to an increase in temperature. Most species show some response in
phenophase occurrence, even though these responses between the OTCs and
the control plots are not consistently significant. Most of the phenophases
that occur significantly earlier within the chambers are reproductive stages,
which suggest that increased temperature could provide some evolutionary
advantage in seed production and set, or perhaps a shift from predominantly
vegetative growth that dominates the arctic to a more sexually dominated life
cycle. However, the nature of the phenophase monitoring is biased towards
recording more reproductive stages than vegetative, which means that the
trend for reproductive phenophases to be more likely to occur earlier within
chambers could be biased. A continuation of monitoring in the same detailed

fashion is necessary to determine the long-term effects of increased warming,

U

 

71

so that this information could be potentially useful to global change modelers

that would like to include vegetation responses, as a part of their predictions.

APPENDICES

APPENDIX A

72

APPENDIX A

Raw data and metadata for all species are stored on floppy disk at Michigan
State University (in care of Dr. Patrick J. Webber). Data are in the ITEX '
recommended format used for compiling all of the circumpolar data for the
N CEAS (National Center for Ecology and Synthesis) conference in December

of 1996.

Key
The following key applies to Tables 8 - 13 (Appendices A and B), for

AN OVA and Post-Hoe results. ANOVA tables are set up in standard format.

Const = Constant term

Spc = Species

yr = year

Spc*yr = interaction between species and year

pt = plot type (OTC or control)

Spc*pt = interaction between species and plot type

yr*pt = interaction between year and plot type

Spc*yr*pt = interaction between species, year and plot type

Error = the error term for the ANOVA

Total = the total amount of degrees of freedom and sums of squares that are
accounted for

Prob = probability

73

APPENDIX A

Table 8. Analysis of Variance result tables - Julian Date of Occurrence.

 

Dependent variable: Emergence of First Green Leaf (-0.50 transformed)

 

 

Source df Sums of Mean Square F-ratio Prob
Squares
Const 1 1595.27 1595.27 510860417 5. 0.0001
Spc 4 0.001209 0.000302 96.820 s 0.0001
yr 2 0.006424 0.003212 1028.6 5 0.0001
Spc*yr 8 0.000111 0.000014 4.4280 5 0.0001
pt 1 0.000039 0.000039 12.432 0.0005
Spc*p t 4 0.000013 0.000003 1.0355 0.3884
yr*pt 2 0.000018 0.000009 2.8664 0.0580
Spc*y r*pt 8 0.000016 0.000002 0.64149 0.7428
Error 438 0.001368 0.000003
Total 467 0.009792

 

Dependent variable: First Inflorescence Visible (-0.50 transformed)

 

 

Source df Sums of Mean Square F-ratio Prob
Squares
Const 1 1007.33 1007.33 245818824 5 0.0001
Spc 4 0.000939 0.000235 57.268 s 0.0001
yr 2 0.003023 0.001512 368.88 5 0.0001
SPC*yr 8 0.000210 0.000026 6.4086 5 0.0001
Pt 1 0.000113 0.000113 27.497 s 0.0001
Spc*pt 4 0.000059 0.000015 3.5785 0.0073
WM 2 0.000038 0.000019 4.6080 0.0108
SPC*y:-*pt 8 0.000062 0.000008 1.8998 0.0603
Error 264 0.001082 0.000004
Tota1 293 0.006963

 

 

74

APPENDIX A

Table 8 (Cont’ d)

‘13-‘35 .

 

Dependent variable: First Inflorescence Visible (-0.50 transformed)

 

 

 

 

 

Source df Sums of Mean Square F-ratio Prob
Squares
Const 1 359.315 359.315 77603367 5 0.0001
Spc 1 0.000153 0.000153 33.053 5 0.0001
yr 1 0.000416 0.000416 89.947 5 0.0001
Spc*yr 1 0.000004 0.000004 0.79534 0.3747
pt 1 0.000032 0.000032 7.0021 0.0095
Spc*pt 1 0.000001 0.000001 0.13359 0.7155
yr*pt 1 0.000020 0.000020 4.2867 0.0411
Spc*yr*pt 1 0.000000 0.000000 0.03344 0.8553
Error 0.000449 0.000005
Total 0.001134
Dependent variable: First Elongation of Pedicels (-0.50 transformed)
Source Sums of Mean Square F—ratio Prob
Squares
Const 1 359.315 359.315 77603367 5 0.0001
Spc 1 0.000153 0.000153 33.053 s 0.0001
yr 1 0.000416 0.000416 89.947 5 0.0001
Spc*yr 1 0.000004 0.000004 0.79534 0.3747
pt 1 0.000032 0.000032 7.0021 0.0095
Spc*pt 1 0.000001 0.000001 0.13359 0.7155
yr*pt 1 0.000020 0.000020 4.2867 0.0411
Spc*yr*pt 1 0.000000 0.000000 0.03344 0.8553
Error 97 0.000449 0.000005
Total 104 0.001134

 

75

APPENDIX A

Table 8 (Cont’d)

 

Dependent variable: First Inflorescence Open (-0.50 transformed)

 

 

Source df Sums of Mean Square F-ratio Prob
Squares
Const 1 1159.59 1159.59 250029400 5 0.0001
Spc 4 0.001220 0.000305 65.761 5. 0.0001
yr 2 0.000917 0.000459 98.915 5 0.0001
Spc*yr 6 0.000072 0.000012 2.5988 0.0180
pt 1 0.000034 0.000034 7.3401 0.0071
Spc*pt 4 0.000055 0.000014 2.9615 0.0200
yr*pt 2 0.000006 0.000003 0.61779 0.5398
Spc*yr*pt 6 0.000072 0.000012 2.5906 0.0183
Error 311 0.001442 0.000005
Total 336 0.005671

 

Dependent variable: First Flower Wither (-0.50 transformed)

 

Source df Sums of Mean Square F-ratio Prob

 

Squares
Const 1 515.355 515.355 122594098 s 0.0001
Spc 1 0.000034 0.000034 8.1459 0.0050
yr 2 0.000821 0.000410 97.625 5 0.0001
Spc*yr 2 0.000027 0.000013 3.1979 0.0439
pt 1 0.000064 0.000064 15.243 0.0001
Spc*pt 1 0.000006 0.000006 1.3955 0.2395
yr*pt 2 0.000019 0.000009 2.2043 0.1142
Spc*yr*pt 2 0.000016 0.000008 1.9144 0.1514
Error 137 0.000576 0.000004
Total 148 0.001565

 

J

Table 8 (Cont’ d)

76

APPENDIX A

 

Dependent variable: First Green Leaf (Sal ix rotundifolia, -0.50 transformed)

 

 

 

 

 

 

 

 

 

Source df Sums of Mean Square F-ratio Prob
Squares
Const 1 477.314 477.314 205882044 5 0.0001
yr 2 0.001411 0.000706 304.35 5 0.0001
pt 1 0.000069 0.000069 29.874 s 0.0001
yr*pt 2 0.000020 0.000010 4.3536 0.0147
Error 134 0.000311 0.000002
Total 139 0.001827
Dependent variable: First Inflorescence visible (-0.50 transformed)
Source df Sums of Mean Square F-ratio Prob
Squares
Const 1 668.011 668.011 322450333 ' 5 0.0001
Spc 1 0.000016 0.000016 7.5098 0.0067
yr 2 0.000497 0.000249 119.99 5 0.0001
Spc*yr 2 0.000018 0.000009 4.3357 0.0145
pt 1 0.000095 0.000095 45.965 5 0.0001
Spc*pt 1 0.000008 0.000008 3.8480 0.0513
yr*pt 2 0.000016 0.000008 3.8774 0.0224
Spc*yr*pt 1 0.000032 0.000032 15.601 0.0001
Error 184 0.000381 0.000002
Total 194 0.001997
Dependent variable: First Elongation of buds (CassiOpe tetragona, -0.50
transformed)
Source df Sums of Mean Square F-ratio Prob
Squares
Const 1 281.631 281.631 159526344 5 0.0001
yr 2 0.000185 0.000093 52.425 5 0.0001
pt 1 0.000043 0.000043 24.502 5 0.0001
yr*pt 2 0.000025 0.000012 6.9554 0.0017
Error 76 0.000134 0.000002

Total 81 0.000511

11"

Table 8 (Cont’ d)

77

APPENDIX A

 

Dependent variable: First Flower Open(Cassiope tetragona, -0.50

 

 

transforemd)

Source df Sums of Mean Square F-ratio Prob

Squares
Const 1 361.298 361.298 122149789 5 0.0001
yr 2 0.001403 0.000702 237.22 5 0.0001
pt 1 0.000217 0.000217 73.295 5 0.0001
yr*pt 2 0.000012 0.000006 ' 2.0116 0.1392
Error 99 0.000293 0.000003
Total 104 0.001947

 

Dependent variable: First Corolla Drop (-0.50 transformed)

 

 

 

 

 

 

Source df Sums of Mean Square F-ratio Prob
Squares

Const 1 311.034 311.034 100909295 5 0.0001

yr 2 0.000880 0.000440 142.70 5 0.0001

pt 1 0.000116 0.000116 37.672 5 0.0001

yr*pt 2 0.000011 0.000005 1.7534 0.1795

Error 84 0.000259 0.000003

Total 89 0.001451

Dependent variable: First Pollen Shed (Sal ix rotund i fol in, -0.50 transformed)

Source df Sums of Mean Square F-ratio Prob
Squares

Const 1 397.488 397.488 226227454 5 0.0001

yr 2 0.001402 0.000701 398.97 s 0.0001

pt 1 0.000142 0.000142 80.612 5 0.0001

yr*pt 2 0.000013 0.000006 3.6201 0.0300

Error 110 0.000193 0.000002

Total 115 0.001761

 

Table 8 (Cont’d)

78

APPENDIX A

 

Dependent variable: A11 Pollen Shed (Sal ix rotu ndi fol in, -0.50 transformed)

 

 

 

 

 

 

 

 

Source df Sums of Mean Square F-ratio Prob
Squares

Const 1 395.600 395.600 290920245 5 0.0001

yr 2 0.002689 0.001345 988.91 s 0.0001

pt 1 0.000772 0.000772 567.89 s 0.0001

yr*pt 2 0.001077 0.000538 395.95 s 0.0001

Error 109 0.000148 0.000001

Total 114 0.004876

Dependent variable: First Seed Dispersal (Sal ix rotundifolia, -0.50

transformed)

Source df Sums of Mean Square F-ratio Prob
Squares

Const 1 264.414 264.414 42205448 5 0.0001

yr 2 0.000153 0.000077 12.211 5 0.0001

pt 1 0.000010 0.000010 1.5642 0.2152

yr*pt 2 0.000013 0.000006 1.0235 0.3647

Error 70 0.000439 0.000006

Total 75 0.000622

Dependent variable: First Color Change (Sal ix rotundifolia, -0.50

transformed)

Source df Sums of Mean Square F-ratio Prob
Squares

Const 1 299.424 299.424 207002150 5 0.0001

yr 2 0.000057 0.000028 19.610 5 0.0001

pt 1 0.000003 0.000003 1.7488 0.1898

yr*pt 2 0.000020 0.000010 6.9672 0.0016

Error 80 0.000116 0.000001

Total 85 0.000186

 

Table 9. Analysis of Variance result tables - Days since snowfree.

79

APPENDIX A

 

Dependent variable: Emergence of First Green Leaf (0.0 transformed)

 

 

Source df Sums of Mean Square F-ratio Prob
Squares
Const 1 2323.53 2323.53 10068 5 0.0001
Spc 5 75.8993 15.1799 65.775 5 0.0001
yr 2 132.616 66.3078 287.31 5 0.0001
Spc*yr 10 11.2257 1.12257 4.8641 s 0.0001
pt 1 3.30157 3.30157 14.306 0.0002
Spc*pt 5 2.11399 0.422799 1.8320 0.1047
yr*pt 2 0.708412 0.354206 1.5348 0.2164
Spc*yr*pt 10 1.36011 0.136011 0.58934 0.8232
Error 566 130.625 0.230786
Total 601 360.977

 

Dependent variable: First Inflorescence Visible (0.0 transformed)

 

Source

Const
Spc

yr

Spc*yr

pt

Spc*pt
yr*pt
Spc*yr*pt
Error
Total

df

HNOHHNONi—l
§§H N

Sums of
Squares
3914.22
15.6660
15.7228
8.81384
1.52972
1.24067
0.594247
4.87514
35.0688
95.7982

Mean Square

3914.22
2.61100
7.86139
0.734487
1.52972
0.206779
0.297124
0.443194
0.078630

F-ratio

49781

33.206
99.980
9.341 1
19.455
2.6298
3.7788
5.6365

Prob

s 0.0001
5 0.0001
5 0.0001
5 0.0001
5 0.0001

0.0162

0.0236
5 0.0001

[11

 

‘

APPENDIX A

Table 9 (Cont’ d)

 

Dependent variable: First Elongation of Pedicels

 

 

 

 

 

Source df Sums of Mean Square F-ratio Prob
Squares
Const 1 462.043 462.043 212009 5 0.0001
Spc 3 0.124892 0.041631 19.102 s 0.0001
yr 2 0.035307 0.017653 8.1003 0.0004
Spc*yr 2 0.007697 0.003849 1.7660 0.1741
pt 1 0.011998 0.011998 5.5054 0.0201
Spc*pt 3 0.000972 0.000324 0.14865 0.9304
yr*pt 2 0.003614 0.001807 0.82906 0.4382
Spc*yr*pt 2 0.012925 0.006462 2.9653 0.0541
Error 175 0.381388 0.002179
Total 190 0.854681
Dependent variable: First Inflorescence Open
Source df Sums of Mean Square F-ratio Prob
Squares
Const 1 4931.47 4931.47 136199 s 0.0001
Spc 4 10.6812 2.67029 73.749 5 0.0001
yr 2 2.66137 1.33068 36.751 5 0.0001
Spc*yr 8 1.64143 0.205179 5.6667 _<. 0.0001
pt 1 2.62776 2.62776 72.574 5 0.0001
Spc*pt 4 0.484602 0.121150 3.3460 0.0103
yr*pt 2 0.018973 0.009486 0.26200 0.7696
Spc*yr*pt 8 0.502602 0.062825 1.7351 0.0885
Error 408 14.7728 0.036208
Total 437 33.9431

 

Table 9 (Cont’ d)

8]

APPENDIX A

 

Dependent variable: First Withering of Flowers

 

 

 

 

 

Source df Sums of Mean Square F-ratio Prob
Squares

Const 1 2109.36 2109.36 113649 5 0.0001

Spc 1 0.115308 0.115308 6.2126 0.0139

yr 2 0.290888 0.145444 7.8363 0.0006

Spc*yr 2 0.060030 0.030015 1.6172 0.2022

pt 1 0.255712 0.255712 13.777 0.0003

Spc*pt 1 0.009330 0.009330 0.50270 0.4795

yr*pt 2 0.042290 0.021145 1.1392 0.3231

Spc*yr*pt 2 0.049070 0.024535 1.3219 0.2700

Error 137 2.54276 0.018560

Total 148 3.41369

Dependent variable: First Pollen Shed (Sal ix rotind i fol in)

Source df Sums of Mean Square F-ratio Prob
Squares

Const 1 275.947 275.947 139984 s 0.0001

yr 2 0.041731 0.020865 10.585 5 0.0001

pt 1 0.095309 0.095309 48.349 5 0.0001

yr*pt 2 0.020850 0.010425 5.2884 0.0064

Error 110 0.216840 0.001971

Total 115 0.375407

 

 

Table 9 (Cont’ d)

82

APPENDIX A

 

Dependent variable: All Pollen Shed (Sal ix roti ndi fol ia)

 

 

Source df Sums of Mean Square F-ratio Prob
Squares

Const 1 297.164 297.164 268635 5 0.0001

yr 2 0.111522 0.055761 50.408 3 0.0001

pt 1 0.174663 0.174663 157.89 s 0.0001

yr*pt 2 0.244697 0.122348 110.60 3 0.0001

Error 109 0.120576 0.001106

Total 114 0.668153

 

Dependent variable: First Seed Dispersal (Sal ix roti ndi fol in, 0.0 transformed)

 

 

 

 

Source df Sums of Mean Square F-ratio Prob
Squares

Const 1 1297.27 1297.27 67930 5 0.0001

yr 2 0.227015 0.113508 5.9437 0.0041

pt 1 0.041475 0.041475 2.1718 0.1450

yr*pt 2 0.042975 0.021488 1.1252 0.3304

Error 70 1.33679 0.019097

Total 75 1.75185 V

Dependent variable: First Color Change (Sal ix roti nd i fol ia)

Source df Sums of Mean Square F-ratio Prob
Squares

Const 1 1470.04 1470.04 277542 3 0.0001

yr 2 0.374763 0.187382 35.378 s 0.0001

pt 1 0.004423 0.004423 0.83506 0.3636

yr*pt 2 0.050612 0.025306 4.7778 0.0110

Error 80 0.423730 0.005297

Total 85 0.923840

 

Table 9 (Cont’ d)

APPENDIX A

 

Dependent variable: First Corolla Drop (Cassiope tetragona, 0.0 transformed)

 

 

Source df Sums of Mean Square F—ratio Prob
Squares

Const 1 1242.47 1242.47 99092 5 0.0001

yr 2 0.298055 0.149027 11.886 5 0.0001

pt 1 0.586721 0.586721 46.794 s 0.0001

yr*pt 2 0.042869 0.021434 1.7095 0.1872

Error 84 1.05323 0.012539

Total 89 2.00540

 

85

APPENDIX A

Table 10. Analysis of Variance result tables - Growing Degree Day Accumulation.

‘12; .

 

 

Dependent variable: Emergence of First Green Leaf (-0.50 transformed)

 

 

Source df Sums of Mean Square F-ratio Prob
Squares
Const 4801.53 4801.53 8232.5 s 0.0001
Spc 149.761 29.9522 51.355 5 0.0001
yr 46.6995 23.3498 40.035 5 0.0001
Spc*yr 31.3325 3.13325 5.3721 5 0.0001
pt 1.65198 1.65198 2.8324 0.0929
Spc*pt 9.63210 1.92642 3.3030 0.0060
yr*pt 3.08044 1.54022 2.6408 0.0722
Spc*yr*pt 12.3037 1.23037 2.1095 0.0221
Error 571 333.030 0.583240
Total 606 584.662

 

Dependent variable: First Inflorescence Visible (-0.50 transformed)
Source df Sums of Mean Square F-ratio Prob

 

Squares
Const 8225.00 8225.00 59684 s 0.0001
Spc 31.9460 5.32434 38.636 5 0.0001
yr 9.60967 4.80484 34.866 5 0.0001
Spc*yr 16.0710 1.33925 9.7182 5 0.0001
pt 0.023532 0.023532 0.17076 0.6796
Spc*pt 4.16240 0.693734 5.0340 s 0.0001
yr*pt 0.364041 0.182020 1.3208 0.2680
Spc*yr*pt 6.80214 0.618377 4.4872 5 0.0001
Error 61.6006 0.137809
Total 139.687

 

85

APPENDIX A

Table 10. (Cont’d)

 

Dependent variable: First Elongation of Buds (-0.50 transformed)

 

 

Source df Sums of Mean Square F-ratio Prob
Squares
Const 1 3524.37 3524.37 40306 5 0.0001
Spc 2 3.18156 1.59078 18.193 5 0.0001
yr 2 1.46594 0.732971 8.3824 0.0003
Spc*yr 2 0.109237 0.054619 0.62463 0.5367
pt 1 0.037717 0.037717 0.43134 0.5122
Spc*pt 2 0.174781 0.087391 0.99942 0.3702
yr*pt 2 0.268396 0.134198 1.5347 0.2185
Spc*yr*pt 2 0.725790 0.362895 4.1502 0.0174
Error 172 15.0399 0.087441
Total 185 25.4887

 

Dependent variable: First Inflrescence Visible (-0.50 transformed)

 

 

Source df Sums of Mean Square F-ratio Prob
Squares
Const 1 10373.6 10373.6 160130 3 0.0001
Spc 5 22.5999 4.51997 69.771 5 0.0001
yr 2 1.48834 0.744171 11.487 5 0.0001
Spc*yr 8 2.31858 0.289823 4.4738 s 0.0001
pt 1 0.279891 0.279891 4.3205 0.0383
Spc*pt 5 1.04958 0.209917 3.2403 0.0070
yr*pt 2 0.143326 0.071663 1.1062 0.3318
Spc*yr*pt 8 1.29295 0.161619 2.4948 0.0118
Error 410 26.5609 0.064783
Total 441 65.3455

 

APPENDIX A

Table 10. (Cont’d)

 

 

Dependent variable: First Flower Wither (-0.50 transformed)

 

 

 

 

 

Source df Sums of Mean Square F-ratio Prob
Squares

Const 1 4253.21 4253.21 182786 5. 0.0001

Spc 1 0.211919 0.211919 9.1074 0.0030

yr 2 1.33697 0.668487 28.729 s 0.0001

Spc*yr 2 0.191190 0.095595 4.1083 0.0185

pt 1 2.15048 2.15048 92.419 5 0.0001

Spc*pt 1 0.030154 0.030154 1.2959 0.2570

yr*pt 2 0.114832 0.057416 2.4675 0.0886

Spc*yr*pt 2 0.081973 0.040986 1.7614 0.1757

Error 136 3.16455 0.023269

Total 147 8.27370

Dependent variable: First Corolla Drop (Cassiope tetragona, -0.50 transformed)

Source df Sums of Mean Square F-ratio Prob
Squares

Const 1 2609.64 2609.64 139240 3 0.0001

yr 2 0.997671 0.498836 26.616 5 0.0001

pt 1 0.455112 0.455112 24.283 5 0.0001

yr*pt 2 0.184111 0.092055 4.9117 0.0096

Error 84 1.57433 0.018742

Total 89 2.99423

 

Dependent variable: First Pollen Shed (Sal ix rotundzfolia, 0.50 transformed)

 

 

Source df Sums of Mean Square F-ratio Prob
Squares

Const 1 2109.37 2109.37 21246 5 0.0001

yr 2 0.279019 0.139510 1.4052 0.2497

pt 1 0.098303 0.098303 0.99012 0.3219

yr*pt 2 0.234307 0.117154 1.1800 0.3111

Error 110 10.9212 0.099283

Total 115 11.5582

 

APPENDIX A

Table 10. (Cont’ d)

 

 

Dependent variable: All Pollen Shed (Sal ix rotund i fol ia, 0.50 transformed)

 

 

 

 

 

Source df Sums of Mean Square F-ratio Prob
Squares

Const 1 2109.37 2109.37 21246 5 0.0001

yr 2 0.279019 0.139510 1.4052 0.2497

pt 1 0.098303 0.098303 0.99012 0.3219

yr*pt 2 0.234307 0.117154 1.1800 0.3111

Error 110 10.9212 0.099283

Total 115 11.5582

Dependent variable: A11 Pollen Shed (Sal ix rotu ndi fol ia, 0.50 transformed)

Source df Sums of Mean Square F-ratio Prob
Squares

Const 1 2536.56 2536.56 59720 5 0.0001

yr 2 1.89356 0.946779 22.291 5 0.0001

pt 1 0.275728 0.275728 6.4917 0.0122

yr*pt 2 8.28525 4.14263 97.533 5 0.0001

Error 109 4.62968 0.042474

Total 114 15.7699

 

Dependent variable: First Seed Dispersal (Sal ix rotu nd i fol id, 0.50 transformed)

 

 

Source df Sums of Mean Square F-ratio Prob
Squares

Const 1 2530.68 2530.68 104653 s 0.0001

yr 2 1.22447 0.612234 25.318 s 0.0001

pt 1 1.14007 1.14007 47.146 5 0.0001

yr*pt 2 0.271101 0.135550 5.6055 0.0055

Error 70 1.69271 0.024182

Total 75 3.98836

 

1
i I
1

APPENDIX A

Table 10. (Cont’d)

 

Dependent variable: First Color Change (Sal ix rotu nd i fol in, 0.50 transformed)

 

 

 

Source df Sums of Mean Square F-ratio Prob
Squares

Const 1 2841.24 2841.24 372984 5 0.0001

yr 2 3.44215 1.72108 225.93 5 0.0001

pt 1 2.25344 2.25344 295.82 5 0.0001

yr*pt 2 0.057176 0.028588 3.7529 0.0277

Error 80 0.609408 0.007618

Total 85 6.01709

 

89

APPENDIX B

Table 11. Scheffe Post Hoc test results - Julian Date of Occurrence

 

Dependent variable: Emergence of First Green Leaf (-0.5 transformed)

 

 

Difference std. err. Prob
Arclat,1994,0TC - Arclat,1994,CT L -0.000222 0.0007 1.00000
Luzarc,1994,0T C - Luzarc,1994,CT L -0.000533 0.0006 0.999507
Luzcon,1994,0T C - Luzcon,1994,CTL -0.000931 0.0005 0.926870
Paphul,1994,0TC - Paphul,1994,CT L -0.001002 0.0008 0.992737
Saxpun,1994,0T C - Saxpun,1994,CT L -0. 001292 0.0006 0.749999
Arclat,1995,0T C - Arclat,1995,CT L -0.000141 0.0007 1.00000
Luzarc,1995,0T C - Luzarc,1995,CTL 0.000057 0.0006 1.00000
Luzoon,1995,0T C - Luzcon,1995,CT L -0.000162 0.0005 1.00000
Paphul,1995,0TC - Paphul,1995,CT L -0.000075 0.0008 1.00000
Saxpun,1995,0T C - Saxpun,1995,CTL 0.000159 0.0006 1.00000
Arclat,1996,0T C - Arclat,1996,CT L -0.000458 0.0007 0.999892
Luzarc,1996,0T C — Luzarc,1996,CTL -0.000014 0.0006 1.00000
Luzcon,1996,0'l~ C - Luzcon,1996,CTL -0.000852 0.0006 0.973636
Paphul,1996,0TC - Paphul,1996,CT L -0.002701 0.0008 0.201658
Saxpun,1996,0T C - Saxpun,1996,CTL -0.000834 0.0006 0.985747

 

Dependent variable: First Inflorescence Visible (-0.5 transformed)

 

 

Difference std. err. Prob

Arclat,1994,0TC - Arclat,1994,CT L -0.003320 0.0018 0.918038
Luzarc,1994,0T C - Luzarc,1994,CF L -0.001132 0.0012 0.998926
Luzcon,1994,0T C - Luzcon,1994,CTL -0.001492 0.0008 0.894657
Paphul,1994,0TC - Paphul,1994,CT L -0.004885 0.0010 0.004497
Saxpun,1994,0T C - Saxpun,1994,CT L -0.001397 0.0007 0.856332
Arclat,1995,0TC - Arclat,1995,CT L -0.003412 0.0013 0.557513
Luzarc,1995,0T C - Luzarc,1995,CTL -0.000955 0.0008 0.995678
Luzcon,1995,0T C - Luzcon,1995,CT L -0.000097 0.0007 1.00000

Paphul,1995,0TC - Paphul,1995,CI‘ L 0.000733 0.0009 0.999628
Saxpun,1995,0TC - Saxpun,1995,CT L 0.001512 0.0007 0.825675
Arclat,1996,0T C - Arclat,1996,CT L -0.003026 0.0017 0.909199
Luzarc,1996,0T C - Luzarc,1996,CT L 0.000795 0.0011 0.999828
Luzcon,1996,0TC - Luzcon,1996,Cl” L -0.001701 0.0009 0.895485
Paphul,1996,0TC - Paphul,1996,CT L -0.002763 0.0010 0.446759
Saxpun,1996,0TC - Saxpun,1996,CT L -0.000982 0.0009 0.996770

 

Table 11. (Cont’ (1)

APPENDIX B

 

 

Dependent variable: First Elongation of Pedicels (-0.5 transformed)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Difference std. err. Prob
Paphul,1995,0TC - Paphul,1995,CT L -0.000012 0.0010 0.990822
Saxpun,1995,0TC - Saxpun,1995,CTL -0.000496 0.0007 0.497458
Paphul,1996,0TC - Paphul,1996,CTL 0001999 0.0010 0.040404
Saxpun,1996,0TC - Saxpun,1996,CT L -0.002160 0.0007 0.004863
Dependent variable: First Flower Wither (-0.5 transformed)

Difference std. err. Prob
Paphul,1994,0TC - Paphul,1994,CT L -0.001265 0.0011 0.531564
Saxpun,1994,0TC - Saxpun,1994,CT L -0.001288 0.0008 0.277899
Paphul,1995,0TC - Paphul,1995,CT L -0.000502 0.0009 0.862207
Saxpun,1995,0T C - Saxpun,1995,CT L -0.000702 0.0007 0.612467
Paphul,1996,0TC - Paphul,1996,CT L -0.003727 0.0009 0.000606
Saxpun,1996,0T C - Saxpun,1996,CTL -0.000951 0.0007 0.426382
Dependent variable: First Emergence of Green Leaf (Sal ix Rotu nd i fol in, -0.5

. transformed)

Difference std. err. Prob
1994,0I' C — 1994,CTL -0.002194 0.0004 0.000011
1995,0T C - 1995,CTL -0.000399 0.0004 0.663670
1996,0T C - 1996,CTL -0.001633 0.0005 0.002398
Dependent variable: First Stigma Visible (Sal ix Rotundi fol in, -0.5

transformed)

Difference std. err. Prob
Castet,1994,0T C - Castet,1994,CT L -0.000968 0.0004 0.020855
Salrot,1994,0TC - Salrot,1994,CT L -0.002000 0.0004 0.000003
Castet,1995,0T C - Castet,1995,CT L -0.004320 0.0007 0.000000
Salrot,1995,0TC - Salrot,1995,CT L -0.001251 0.0005 0.008945
Salrot,1996,0T C - Salrot,1996,CT L -0.000822 0.0005 0.118488
Dependent variable: First Pollen Shed (Sal ix Rotu nd i fol in, -0.5

transformed)

Difference std. err. Prob
1994,01" C - 1994,CT L -0.002167 0.0004 0.000002
1995,01~ C - 1995,CTL -0.003091 0.0004 0.000000
1996,0T C - 1996,CTL -0.001411 0.0004 0.008694

91 . -_ J

APPENDIX B
Table 11. (Cont’d)

 

Dependent variable: A11 Pollen Shed (Sal ix Rotu nd i fol in, -0.5

 

 

transformed)
Difference std. err. Prob
1994,0T C - 1994,C'I'L -0.001108 0.0004 0.008633
1995,0T C - 1995,CTL -0.014013 0.0004 0
1996,0TC - 1996,CTL -0.000575 0.0004 0.363890 ‘

 

Dependent variable: First Seed Dispersal (Sal ix Rotu nd i fol in, -0.5

 

 

transformed)
Difference std. err. Prob
1994,0T C - 1994,CT L -0.001938 0.0010 0.148047
1995,0T C - 1995,CI' L -0.000110 0.0012 0.995532
1996,0T C - 1996,CT L -0.000220 0.0010 0.975845

 

Dependent variable: First Color Change (Sal ix Rotu nd i fol ia, -0.5

 

 

transformed)
Difference std. err. Prob
1994,0T C - 1994,CT L 0.000608 0.0005 0.458801
1995,0TC - 1995,CTL 0.000017 0.0005 0.999290
1996,0T C - 1996,CTL -0.001670 0.0004 0.000951

 

Dependent variable: First Corolla Drop (Cassi ope tetragona, -0.5 transformed)
Difference std. err. Prob

1994,0T C - 1994,CT L -0.002502 0.0006 0.000958
1995,01" C - 1995,CT L -0.004401 0.0011 0.000516
1996,01" C - 1996,CTL -0.001993 0.0007 0.025710

 

 

Table 12. Scheffe Post Hoc test results - Days since snow free

92

APPENDIX B

 

Dependent variable: Emergence of First Green Leaf (0.0 transformed)

 

 

Difference std. err. Prob
Arclat,1994,0TC - Arclat,1994,CT L 0052225 0.1861 1.00000
Luzarc,1994,0TC - Luzarc,1994,CT L -0.120219 0.1754 0.999995
Luzcon,1994,0T C - Luzcon,1994,CTL -0.189213 0.1435 0.997937
Paphul,1994,0TC - Paphul,1994,CT L -0.146854 0.2232 0.999997
Salrot,1994,0TC - Salrot,1994,Cl' L -0.512537 0.1387 0.192432
Saxpun,1994,0TC - Saxpun,1994,CTL -0.146696 0.1561 0.999900
Arclat,1995,0TC - Arclat,1995,CTL -0.080048 0.1907 1.00000
Luzarc,1995,CT C - Luzarc,1995,CT L 0.029746 0.1727 1.00000
Luzcon,1995,0T C - Luzcon,1995,CI'L 0.015813 0.1454 1.00000
Paphul,1995,0TC - Paphul,1995,CT L -0.176166 0.2232 0.999981
Salrot,1995,0TC - Salrot,1995,CT L -0.095661 0.1387 0.999995
Saxpun,1995,0TC - Saxpun,1995,CT L -0.019492 0.1674 1.00000
Arclat,1996,0T C - Arclat,1996,CT L -0.276404 0.1820 0.993283
Luzarc,1996,0T C — Luzarc,1996,CT L 0.128292 0.1727 0.999989
Luzcon,1996,0T C - Luzcon,1996,CTL -0.069016 0.1559 1.00000
Paphul,1996,0TC - Paphul,1996,CTL -0.544687 0.2207 0.806831
Salrot,1996,0TC — Salrot,1996,CT L 0364036 0.1450 0.788341
Saxpun,1996,0TC - Saxpun,1996,CT L -0.183145 0.1679 0.999611

 

Dependent variable: First Inflorescence Visible (0.0 transformed)

 

 

Difference std. err. Prob

Arclat,1994,0TC - Arclat,1994,CT L 0256529 0.2560 0.999946
Castet,1994,0T C - Castet,1994,Cl" L -0.124256 0.0809 0.996692
Luzarc,1994,0T C - Luzarc,1994,CTL -0.127600 0.1685 0.999997
Luzcon,1994,0T C - Luzcon,1994,CT L -0.152944 0.1098 0.998650
Paphul,1994,0T C - Pa phul,l 994,C1'L -O.610419 0.1413 0.071448
Salrot,1994,0TC - Salrot,1994,CT L -0.266548 0.0809 0.458397
Saxpun,1994,0T C - Saxpun,1994,CT L -0.125509 0.0969 0.999317
Arclat,1995,0T C - Arclat,1995,CT L 0234993 0.1983 0.999713
Castet,1995,0T C - Castet,1995,CT L -0.723047 0.1389 0.005351
Luzarc,1995,0T C - Luzarc,1995,CT L -0.239890 0.1170 0.962942
Luzcon,1995,0T C - Luzcon,1995,CTL 0.010624 0.0910 1 .00000

Paphul,1995,0TC - Paphul,1995,Cl‘ L 0.501546 0.1288 0.180141
Salrot,1995,0TC - Saxpun,1995,CTL 0.226453 0.0980 0.912613
Saxpun,1995,0TC - Saxpun,1995,CTL 0.287960 0.1008 0.697822
Arclat,1996,0T C - Arclat,1996,CT L -0.196584 0.2290 0.999989
Luzarc,1996,0T C - Luzarc,1996,CTL 0.067268 0.1514 1.00000

Luzcon,1996,0T C - Luzcon,1996,CI' L -0.118521 0.1254 0.999970
Paphul,1996,0TC - Paphul,1996,CT L -0.261137 0.1363 0.978008
Salrot,1 996,0T C - Salrot,1996,CT L -0.110523 0.1021 0.999882
Saxpun,1996,0TC - Saxpun,1996,CTL -0.179509 0.1254 0.998254

Table 12 (Cont’ d)

93

APPENDIX B

 

Dependent variable: First Elongation of Pedicels (-0.50 transformed)

 

 

Difference std. err. Prob
Castet,1994,0TC - Castet,1994,CT L -0.029945 0.0136 0.092230
Castet,1995,CTC - Castet,1995,CT L -0.110873 0.0231 0.000020
Paphul,1995,0TC - Paphul,1995,CTL 0.012542 0.0227 0.858374
Saxpun,1995,0TC - Saxpun,1995,CT L -0.024656 0.0158 0.297910
Castet,1996,0T C - Castet,1996,CT L -0.024660 0.0381 0.811371
Paphul,1996,0TC - Paphul,1996,CT L -0.031926 0.0209 0.313011
Saxpun,1996,0TC — Saxpun,1996,CT L -0.038563 0.0163 0.062795

 

Dependent variable: First Inflorescence Open (-0.50 transformed)

 

 

 

 

 

Difference std. err. Prob
Castet,1994,0TC - Castet,1994,CT L -0.268993 0.0603 0.012140
Luzarc,1994,0T C - Luzarc,1 994,CT L -0.175415 0.0777 0.746348
Luzoon,1994,0T C - Luzcon,1994,CT L -0. 073164 0.0581 0.990978
Paphul,1994,0TC - Paphul,1994,CT L -0.288628 0.0983 0.377106
Saxpun,1994,0TC - Saxpun,1994,CT L -0.l37724 0.0721 0.886423
Castet,1995,0T C - Castet,1995,CTL -0.373487 0.0943 0.049890
Luzarc,1995,CT C - Luzarc,1995,CTL -0.114915 0.0733 0.963159
Luzcon,1995,0T C - Luzcon,1995,CT L -0.197354 0.0595 0.204391
Paphul,1995,0TC - Paphul,1995,CT L -0.063842 0.0855 0.999794
Saxpun,1995,01' C - Saxpun,1995,CT L -0.076419 0.0654 0.994609
Castet,1 996,0T C - Castet,1996,CT L -O.204770 0.0645 0.263017
Luzarc,1996,0T C - Luzarc,1996,CTL 0.024675 0.1059 1.00000
Luzcon,1996,0T C - Luzcon,1996,Cl' L -0.265722 0.0653 0.037422
Paphul,1996,0TC - Paphul,1996,CI L -0.252317 0.0874 0.404232
Saxpun,1996,0TC - Saxpun,1996,CT L -0.074485 0.0673 0.996305
Dependent variable: First Flower Wither (0.0 transformed)
Difference std. err. Prob

Paphul,1994,0TC - Paphul,1994,CT L -0.090277 0.0746 0.482895
Saxpun,1994,0T C - Saxpun,1994,CI'L -0.087887 0.0532 0.259261
Paphul,1995,0TC - Paphul,1995,CT L —0.028469 0.0612 0.897633
Saxpun,1995,0T C - Saxpun,1995,CTL -0.065509 0.0471 0.382025
Paphul,1996,0TC - Paphul,1996,CT L -0.198600 0.0626 0.007776
Saxpun,1996,0T C - Saxpun,1996,CTL -0.062157 0.0483 0.438501

APPENDIX B
Table 12. (cont’d)

 

Dependent variable: First Pollen Shed (0.0 transformed)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Difference std. err. Prob
1994,0TC - 1994,CT L -0.058702 0.0134 0.000142
1995,0T C — 1995,CT L -0.091268 0.0146 0.000000
1996,01” C - 1996,CT L -0.023055 0.0150 0.311521
Dependent variable: A11 Pollen Shed (0.0 transformed)

Difference std. err. Prob
1994,01" C - 1994,CT L -0.022106 0.0100 0.092812
1995,0TC - 1995,CTL -0.211159 0.0110 0.000000
1996,0T C - 1996,CI' L -0.002792 0.0115 0.970906
Dependent variable: First Seed Dispersal (0.0 transformed)

Difference std. err. Prob
1994,0TC - 1994,CT L -0.117653 0.0540 0.100521
1995,0TC — 1995,CTL -0.008719 0.0640 0.990756
1996,01" C - 1996,CTL -0.021118 0.0549 0.928651
Dependent variable: First Color Change (0.0 transformed)

Difference std. err. Prob
1994,01' C - 1994,CTL 0.033647 0.0293 0.520495
1995,0T C - 1995,CT L 0.003377 0.0275 0.992492
1996,0T C - 1996,CT L -0.080742 0.0259 0.010289
Dependent variable: First Corolla Drop (0.0 transformed)

Difference std. err. Prob
1994,0T C - 1994,CT L 0.033647 0.0293 0.520495
1995,0T C - 1995,CT L 0.003377 0.0275 0.992492
1996,0'1' C — 1996,CTL -0.080742 0.0259 0.010289

Table 13 Scheffe Post Hoc test results - Growing Degree Day Accumulation

95

APPENDIX B

 

Dependent variable: Emergence of First Green Leaf (0.0 transformed)

 

 

Difference std; err. Prob
Arclat,1994,0T C - Arclat,1994,CT L 0.845230 0.2958 0.612740
Luzarc,1994,0T C - Luzarc,1994,CT L 0.655101 0.2745 0.839170
Luzcon,1994,0T C - Luzcon,1994,CTL 0.398307 0.2282 0.979984
Paphul,1994,0TC - Paphul,1994,CT L 0.507094 0.3549 0.995921
Salrot,1994,0TC - Salrot,1994,C1' L -0.824181 0.2205 0.177419
Saxpun,1994,0TC - Saxpun,1994,C1' L 0.256815 0.2481 0.999758
Arclat,1995,0T C — Arclat,1995,CT L 0040042 , 0.3032 1.00000
Luzarc,1995,0T C - Luzarc,1995,CT L 0.049951 0.2660 1.00000
Luzcon,1995,0T C - Luzcon,1995,CT L 0095525 _ 0.2312 1.00000
Paphul,1995,0TC - Paphul,1995,CT L -0.320149 0.3433 0.999907
Salrot,1995,0TC - Salrot,1995,CT L -0.064091 0.2205 1.00000
Saxpun,1995,0TC - Salrot,1995,C1' L 0.359836 0.2568 0.996541
Arclat,1996,0T C - Arclat,1996,Cl" L 0.336902 0.2887 0.999286
Luzarc,1996,0T C - Luzarc,1996,CTL 0.227634 0.2660 0.999958
Luzcon,1996,0T C - Luzcon,1996,CT L -0.005543 0.2478 1.00000
Paphul,1996,0TC - Paphul,1996,CT L -0.527410 0.3509 0.993816
Salrot,1996,0TC - Salrot,1996,CTL -0.007040 0.2305 1.00000
Saxpun,1996,0T C - Saxpun,1996,CTL 0.423662 0.2705 0.991387

 

Table 13. (Cont’d)

APPENDIX B

 

Dependent variable: First Inflorescence Visible (0.0 transformed)

 

 

Difference std. err. Prob
Arclat,1994,0T C - Arclat,1994,CT L -0.029097 0.3389 1.00000
Castet,1994,0T C - Castet,1994,Cl' L 0.326364 0.1072 0.596798
Luzarc,1994,0T C - Luzarc,1994,Cl” L 0.311655 0.2231 0.998608
Luzcon,1994,0T C - Luzcon,1994,CTL 0.201328 0.1454 0.998720
Paphul,1994,0TC - Paphul,1994,Cl‘ L 0551135 0.1871 0.651236
Salrot,1994,0TC - Salrot,1 994,CT L 0.060365 0.1072 1.00000
Saxpun,1994,0TC - Saxpun,1994,CTL 0.241750 0.1298 0.982518
Arclat,1995,0TC - Arclat,1995,CTL 0.013698 0.2396 1.00000
Castet,1995,0TC - Castet,1995,CTL -0.415099 0.1839 0.925420
Luzarc,1995,0T C - Luzarc,1995,CT L 0.182417 0.1550 0.999731
Luzcon,1995,0T C - Luzcon,1995,CTL 0.378417 0.1204 0.542834
Paphul,1995,0TC - Paphul,1995,CT L 0.417033 0.1669 0.855043
Salrot,1995,0TC - Salrot,1995,CT L -0.061266 0.1221 1.00000
Saxpun,1995,0T C - Saxpun,1995,CT L 0.599497 0.1334 0.046096
Arclat,1996,0TC - Arclat,1996,CT L 0.143215 0.3031 1.00000
Castet,1996,0T C - Luzarc,1996,CTL 1.63254 0.2396 0.000006
Luzarc,1996,0TC - Luzarc,1996,Cl' L 1.02097 0.2005 0.007800
Luzcon,1996,0T C - Luzcon,1996,CTL 0.044553 0.1660 1.00000
Paphul,1996,0TC - Paphul,1996,CT L -0.190433 0.1804 0.999909
Salrot,1996,0TC - Salrot,1996,CT L 0.204737 0.1351 0.997059
Saxpun,1996,0TC - Saxpun,1996,CT L 0.153435 0.1660 0.999976

 

Dependent variable: First Elongation of Pedicels (0.0 transformed)

 

 

Difference std. err. Prob
Castet,1994,0TC - Castet,1994,CT L 0.226862 0.0863 0.033752
Castet,1995,0TC -Castet,1995,CT L -0.275205 0.1465 0.174377
Paphul,1995,0TC - Paphul,1995,CT L 0.351718 0.1437 0.052607
Saxpun,1995,0TC - Saxpun,1995,CT L 0.300671 0.1000 0.012219
Castet,1996,0T C - Castet,1996,CT L 0.353280 0.2414 0.345113
Paphul,1996,0TC - Paphul,1996,CT L 0.001388 0.1322 0.999945
Saxpun,1996,0TC - Saxpun,1996,CT L 0.176913 0.1045 0.241727

APPENDIX B
Table 13. (Cont’ d)

 

Dependent variable: First Inflorescence Open (0.0 transformed)

 

 

Difference std. lerr. Prob

Castet,1994,0TC - Castet,1994,CT L 0.000143 0.0807 1
Luzarc,1994,0T C - Luzarc,1994,CT L 0.193271 0.1039 0.901489
Luzcon,1994,01' C - Luzcon,1994,CTL 0.325912 0.0777 0.026443
Paphul,1994,0TC - Paphul,1994,Cl' L -0.058009 0.1314 0.999996
Saxpun,1994,0TC - Saxpun,1994,CTL 0.173059 0.0964 0.918998
Castet,1995,0TC - Castet,1995,CT L -0.015125 0.1261 1.00000
Luzarc,1995,0TC - Luzarc,1995,Cl" L 0.285998 0.0980 0.387567
Luzcon,1995,0T C - Luzcon,1995,CTL 0.071566 0.0795 0.999169
Paphul,1995,0TC - Paphul,1995,CTL 0.272516 0.1144 0.683391
Saxpun,1995,0TC - Saxpun,1995,CT L 0.293068 0.0875 0.192798
Arclat,1996,0TC — Arclat,1996,CTL 0.259540 0.2939 0.999277
Castet,1996,0TC - Castet,1996,CT L 0.220600 0.0863 0.587734
Luzarc,1996,0T C - Luzarc,1996,CTL 0.614507 0.1416 0.017483
Luzcon,1996,0T C - Luzcon,1996,CTL 0.076792 0.0873 0.999297
Paphul,1996,0TC - Paphul,1996,CT L 0.092996 0.1169 0.999668
Saxpun,1996,0TC - Saxpun,1996,Cl' L 0.368338 0.0900 0.035266

 

Dependent variable: First Flower Wither (0.0 transformed)

 

 

Difference ‘ std. err. Prob
Paphul,1994,0TC - Paphul,1994,CT L 0.200403 0.0836 0.059759
Saxpun,1994,0TC - Saxpun,1994,CTL 0.190384 0.0596 0.007296
Paphul,1995,0TC — Paphul,1995,CT L 0.340917 0.0686 0.000012
Saxpun,1995,0T C - Saxpun,1995,CT L 0.334159 0.0527 0.000000
Saxpun,1996,0TC - Saxpun,1996,CTL 0.341499 0.0551 0.000000

 

Dependent variable: First Pollen Shed (Sal ix rotundifolia, 0.0 transformed)

 

 

Difference std. err. Prob
1994,0TC - 1994,CTL -0.056329 0.0950 0.839044
1995,0TC — 1995,CTL -0.173976 0.1039 0.250782
1996,0TC - 1996,CTL 0.054583 0.1066 0.877197

 

Dependent variable: All Pollen Shed (Sal ix rotundifolia, 0.0 transformed)

 

 

Difference std. err. Prob
1994,0TC - 1994,CT L 0.216223 0.0621 0.003212
1995,01" C - 1995,CT L —0.869398 0.0680 0.000000

1996,0TC - 1996,CT L 0.356585 0.0712 0.000012

APPENDIX B

Table 13. (Cont’d)

 

Dependent variable: First Seed Dispersal (Sal ix rotu nd i fol ia, 0.0 transformed)

 

 

Difference std. err. Prob
1994,0T C - 1994,CTL 0.087510 0.0608 0.359749
1995,0TC - 1995,CTL 0.380260 0.0720 0.000008
1996,0TC - 1996,CTL 0.305509 0.0617 0.000027

 

Dependent variable: First Color Change (Salix rotu ndi fol in, 0.0 transformed)

 

 

Difference std. err. Prob
1994,0TC -1 994,CT L 0.314583 0.0352 0.000000
1995,0TC -l995,CT L 0.397439 0.0330 0.000000

1996,0TC —1996,CT L 0.274766 0.0311 0.000000

LITERATURE CITED

Alatolo, J. M. and O. Totland. 1997. Response to simulated climatic change in
an alpine and sub-arctic pollen-risk strategist, Silene acaulis. Global
Change Biology [In press]

Bliss, L.C. Adaptation of arctic and alpine plants to environmental
conditions. Arctic 15:117-144.1962

Bock, J.H. 1976. The effects of increased snowpack on the phenology and seed
germinability of selected alpine species. In: Steinhoff, H.W., J.D. Ives
(eds.,) Ecological impacts of snowpack augmentation in the San Juan
Mountains, Colorado. US Department of the Interior, Bureau of
reclamation, Denver. pp. 265- 280.

Chapin, RS. 1987. Environmental controls over growth of tundra plants.
Ecological Bulletins 38:69-76.

Chapin, F. S. 1995. Responses of arctic tundra to experimental and observed
changes in climate. Ecology 76:694-711.

Chapman, W.L. and IE Walsh. 1993. Recent variations of sea ice and air
temperatures in high latitudes. Bulletin of the American
Meteorological Society 74:33-47.

Cohen, 5.]. 1990. Bringing the global warming issue closer to home: the
challenge of regional impact studied. Bulletin of the American
Meteorological Society 71:520-526

Davis, MB. 1989. Lags in vegetation response to greenhouse warming.
Climatic Change 15:75-82.

Fennessey, M. J. and Y. Xue. 1997. Impact of USGS vegetation map on GCM
simulations over the United States. Ecological Applications 7:22-33.

Foster, J. L. 1989. The significance of the date of snow disappearance on the
arctic tundra as a possible indicator of climate change. Arctic and
Alpine Research 21:60-70.

Grotc

Haag.

Henr}

Hultel

Kenne

K0ma1

Lindsk

NIaCCI‘

8.1.. .

100

Grotch, S.L. 1991. A statistical intercomparison of temperature and
precipitation predicted by four general circulation models with
historical data. In: M. Schlesinger. Greenhouse-gas-induced climatic
change: a critical apprasial of simulations and observations. Elsevier
Science Publishers, New York, NY pp. 3-16.

Haag. R. W. 1974. Nutrient limitations to plant production two tundra
communities. Canadian Journal of Botany 52:103-116.

Henry, G. H. R. and U. Molau. 1997. Tundra plants and climate change:
International Tundra Experiment (ITEX) - Introduction. Global Change
Biology (In Press).

Hulten, E. 1968. Flora of Alaska and Neighboring Territories. Stanford
University Press.

Kennedy, AD. 1995. Simulated climate change: are passive greenhouses a
valid microcosm for testing the biological effects of environmental
perturbations? Global Change Biology 1:29-42.

Komarkova, V. 1979. Alpine vegetation of the Indian Peaks area, Front
Range, Colorado Rocky Mountains. J. Cramer, Vaduz, Liechtenstein.

Lindskog, A. and LS. Jonsdottir. 1997. Responses of Carex bigelowii, a clonal
sedge, to simulated climate change. Global Change Biology [In press].

MacCracken, MC. 1995. The evidence mounts up. Nature 376:645-646.

MacCracken, M.C., U. Cubasch, W.L. Gates, L.D. Harvey, B. Hunt, R. Katz, E.
Lorenz, S. Manabe, B. McAvaney, N. McFarlane, G. Meehl, V.
Meleshko, A. Robock, G. Stenchikov, R. Stouffer, W.C. Wang, W.
Washington, R. Watts, S. Zebiak. 1991. A critical appraisal of model
simulations. In: M. Schlesinger. Greenhouse-gas-induced climatic
change: a critical apprasial of simulations and observations. Elsevier
Science Publishers, New York, NY pp. 583-590.

Manabe, S. 1997. Earliy development in the study of greenhouse warming:
the emergence of climate models. Ambio 26:47-51.

Marion, G.M., G.H.R. Henry, D.W. Freckman, J. Johnstone, G. Jones, M.H.
Jones, E. Levesque, U. Molau, P. Molgaard, A.N. Parsons, J. Svoboda,
and RA. Virginia. 1997. Open-top designs for manipulation field
temperature in high-latitude ecosystems. Global Change Biology [In
press].

Marion.
\

[1
l
C
l

Maxwe
I
(

Maxwe

McGra

MolaL

3101611
h-lolg

N100.

I)Otv

RO(

101 - 43

Marion, G. M., G.H.R. Henry, P. Molgaard, W.C. Oechel, M.H. Jones, and G.L.
Vourlitis. 1993. Open-Top Devices for manipulating field
temperatures in tundra ecosystems. Proceeding of the Fourth
International Symposium on Thermal Engineering and Science for
Cold Regions, US Army Cold Regions Research and Engineering
Laboratory, Hanover, NH 28 September- 1 October, 1993.

Maxwell, B. 1992. Arctic Climate: Potential for Change under Global
Warming. In: Chapin, RS. 9 ed.), Arctic Ecosystems in a Changing
Climate. Academic Press, p.11-17.

Maxwell, J.B. and LA. Barrie. 1989. Atmospheric and climatic change in the
Arctic and Antarctic. Ambio 18:42-49.

McGraw, J.B. and N. Fetcher. 1992. Response of tundra plant populations to
climatic change. In: Chapin, F.S., R.L. Jefferies, J.F. Reynolds, G.R.
Shaver, and J. Svoboda. Arctic Ecosystems in a Changing Climate: An
Ecophysiologi cal Perspective. Academic Press, San Diego, CA pp. 359-
376.

Molau, U. 1997. Responses to natural climatic variation and experimental
warming in two tundra plant species with contrasting life forms:
Cassiope tetragona and Ranunculus nivalis. Global Change Biology
[In Press].

Molau, U. and P. Molgaard. 1996. ITEX Manual. Second Edition. Danish
Polar Center, Copenhagen.

Molgaard, P. and K. Christensen. 1997 . Response to experimental warming
in Papaver radicatum. Global Change Biology [In press].

Moore, PD. 1995. Opening time by degrees. Nature 375:186-187.

Potvin, C., M.J. Lechowicz, and S. Tardif, 1990. The statistical analysis of
ecophysiological response curves obtained from experiments
involoving repeated measures. Ecological Applications 71:1389-1400.

van Minnen, J. G., K.K. Goldewijk, and R. Leeman. 1995. The importance of
feedback processes and vegetation transition in the terrestrial carbon
cycle. Journal of Biogeography 22:805-814.

Roots, ER 1989. Climate change: high latitude regions. Climatic Change
15:223-253.

Shave:

SokaL

Webbe

[Vookl

Zhanc

102

Shaver, G. R. and J. Kummerow. 1992. Phenology, resource allocation, and
growth Arctic vascular plants. In: Chapin, F .S., R.L. Jefferies, J.F.
Reynolds, G.R. Shaver, and J. Svoboda. Arctic Ecosystems in a
Changing Climate: An Ecophysiological Perspective. Academic Press,
San Diego, CA pp. 193-211.

Sokal, RR. and F]. Rohlf. 1995. The principles and practice of statistics in
biological research. Third edition. W.H. Freeman and Company, New
York

Webber, P. J., J.C. Emerick, D.C. Ebert May, and V. Komarkova. 1976. The
impact of increased snowfall on Alpine vegetation. In: Steinhoff, H.W.,
J.D. Ives (eds.), Ecological impacts of snowpack augmentation in the
San Juan Mountains, Colorado. US Department of the Interior, Bureau
of reclamation, Denver. pp. 201-264.

Wookey, P.A., A.N. Parsons, J.M. Welker, J.A. Potter, T.V. Callaghan, J.A.
Lee, and MC. Press. 1993. Comparah've responses of phenology and
reproductive development to simulated environmental change in sub-
arctic and high arctic plants. Oi kos 67:490-502.

Zhang, T., T.E. Osterkamp, K. Stamnes. Some characteristics of the climate in
Northern Alaska, USA Arctic and Alpine Research 28:509-518.

 

"JJJJJJJJJJJJJJJJI