Entomology M.S. Thesis - Uelmen

!
EFFECTS OF TEMPERATURE REGIMES AND POPULATION SOURCE ON THE
SUCCESS, TIMING, AND HOST PLANT SYNCHRONY OF EGG HATCH BY
MALACOSOMA DISSTRIA
by
Johnny A. Uelmen Jr.
A thesis submitted in partial fulfillment of
the requirements for the degree of
Master of Science
(Entomology)
at the
UNIVERSITY OF WISCONSIN-MADISON
2014

!
TABLE OF CONTENTS
Table of Contents…………………………………………………………………………......i
Thesis Acknowledgments………………………………………………………..………….iii
Thesis Abstract…………………………………………………..…………………..…........vi
Thesis Introduction…………………………………………………..………………...….....1
Thesis References…………………………………………....………….……………………4
Chapter 1: Effects of simulated temperature increases and forest tent caterpillar
population source on synchrony of egg hatch with aspen and birch budbreak.............…7
Abstract……………………………………………..…………………………………7
Introduction………………………………………………………..…………………..9
Materials and Methods…………………………..……............................…………...12
i. Experimental design……………………………………………..….………12
ii. Insect populations…………………………………………...........................13
iii. Partitioning and enclosure of overwintering egg bands…..…………..…….14
iv. Plant phenology determination…………………………...……..…………..15
v. Statistical analyses…………………………….……………….……………15
Results………………………………………………......………..…………………..19
i. What factors affect survival and timing of egg hatch by forest tent
caterpillar?......................................................................................................19
ii. How does temperature affect budbreak phenology of trembling aspen
and paper birch?.............................................................................................22
iii. How do insect population source, plant species, and temperature interact
to structure phenological synchrony?.............................................................23
Discussion………………………………………………………...………………….25
Acknowledgments……………………………………………………………………30
References…………………………………………………………………..…...…...31
Tables……………………………………………………………………..………….38
Figure legends……………………………………………………………..………....43
Figures……………………………………………………………………..…………46
Supplemental materials………………………………………………………..….….55
i. Tables……………………………………………………………………….56
ii. Figure legends………………………………………………………...….....59
iii. Figures………………………………………………………..…….….……62
Chapter 2: Supercooling points of diapausing forest tent caterpillar eggs………….….70
Abstract…………………………………………………………………..…………..70
Introduction………………………………………………………..…………………72
i

!
Materials and Methods….........…………………………………………..…….…….74
i. Insect population sources……………………………………………....…...74
ii. Overwintering of eggs.........…………………………………..…………….75
iii. Freeze tolerance and supercooling points………………………………......75
iv. Sampling periods……………………………………..................…..………75
v. Statistical analysis………………………………………………….....…….76
Results……………………………………………………………..………......……..76
Discussion…………………………………………………………………..….…….77
Acknowledgments……………………….…………………………………..……….79
References……………………………..…………………………………..…………81
Tables………………………………..……………………………………….………86
Figure legends…………………….......…………………………………………..….89
Figures………………………………..…………………………...……...……...…...90
Thesis Conclusions………………………………………………………………...…..……91
Thesis Appendices..................................................................................................................92
ii

!
THESIS ACKNOWLEDGMENTS
My journey to graduate school began as an undergraduate at Wisconsin. As a senior, I
was still looking for a major that I truly loved. It was my fall semester when I discovered my
passion for Entomology, enrolling in Dan Young’s Entomology 302: Introduction to
Entomology class. He opened up my eyes to an amazing biological subject that included so
many diverse and unique characteristics spanning an immense array of topics, few I knew
even existed. I was so excited to continue learning about entomology and couldn’t wait to
enroll in another course the next semester. Well, it was now my last semester of my
undergraduate career, and I decided to take Entomology 473: Plant-Insect interactions. Little
did I know, the professor, Kenneth Raffa, would be my future advisor.
Ken’s course truly put the lid on my search for research and career paths. Learning from
one of the most prominent forest entomology minds has truly been an honor, and I am
forever thankful for the opportunity. Ken is truly a master of his work, mentoring me on the
balance of communication, hard work, and honesty. Thank you for fostering a learning
environment that constantly challenged me to think outside the box and only strive for the
best. Most importantly, thank you for opening my mind and helping me find a career that I
love.
I thank Rick Lindroth for making this all happen, agreeing to take a chance and co-advise
an admittedly sub-par writer. From the first few weeks of meeting, your emphases on
literature reviews and quality of writing from the start has, without a doubt, made me a more
critical and cognizant thinker and writer. I like to thank my committee members Eric Kruger
and Ezra Schwartzberg for their always-positive suggestions and support. Ezra, thank you
iii

!
for being a great mentor and friend. From the days of insect collections to our time in
northern Minnesota gathering data, you were always providing a guiding light and were there
for me. I also like to thank Patrick Tobin for your frequent and incredibly insightful
conversations and suggestions. You saved me a lot of grief! Thanks to Jun Zhu and Peter
Crump for your support and encouragement with statistics and modeling. Thank you to the
technicians and people of B4WarmED, especially Artur Stefanski and Karen Rice for often
saving the day, using your expertise to deal with technical site-specific information.
Thanks to the members of the Raffa and Lindroth labs. Many of you served as friends,
colleagues, and sources of inspiration. I especially thank Mary Jamieson, Ned “Kennedy”
Rupert-Nason, Charlie Mason, Jesse Phammatter, Adam Krause, Rachel Arango, and Todd
Johnson. Charlie and Jesse, the two of you consistently provided brilliant comments and
reviews of my presentations and paper drafts. I truly appreciated every suggestion you two
gave me. Thanks to both the Entomology and Forest and Wildlife Ecology departments and
staff for always being friendly and fun to be around. Finally, I thank the UW Zoology
department and the PEOPLE program for not only funding the remainder of my research, but
also for allowing me the opportunity to represent your organizations. It was truly a blast!
I owe all of this to my loved ones. First, to my family back in Fond du Lac, WI. My
little sister, Nikki Pikki, thank you for always understanding me and being there when I
needed you. I’m proud of you and for being the beautiful young woman you are today.
Mom and Dad, you two have pushed me to be the best since I can remember. Thank you for
your unconditional love and support and believing in everything I do. My morals,
personality, work ethic, and everything else that makes me is a testament of the two of you.
Thank you to Barry and Linda Fox. You both came into my life and have pushed me to
iv

!
strive for even more. Thank you for your understanding, love, and devotion with all things in
life, and especially for allowing me to share your daughter with you two. Lana, you have
been my rock and #1 fan throughout the last couple years. The words “thank you” barely
graze my love and appreciation for you. The last three years have experienced many
emotions, both good and bad. The last 12 months have been extraordinarily difficult, losing
two of the closest people in my life: my Grandpa DeLaRosa and my best friend and brother,
Gerardo “Jerry” Medina Jr. The hard work and dedication that were put into this thesis, and
what I will put in for the rest of my life, is dedicated to you both. I will try my best to
become as accomplished, honorable, and loved as the two of you. I love you all always and
forever.
v

!
THESIS ABSTRACT
The effects of anticipated warming temperatures on plant-herbivore interactions are
poorly understood. In particular, we have little information on how climate change will
affect phenological synchrony between herbivores and their host plants, a key element
affecting insect population dynamics, tree defoliation, and forest health. Our ability to
predict phenological relationships is further complicated by the substantial differences in
dispersal rates between insects and trees as both expand their northern ranges.
We investigated how interactions among four population sources of forest tent caterpillar
(Malacosoma disstria Hübner), two of its major hosts, trembling aspen (Populus tremuloides
Michaux) and paper birch (Betula papyrifera Marshall), and three controlled temperature
regimes at two mesoscale outdoor sites influence the relative timing of egg eclosion and
budbreak. Because M. disstria adults that migrated northward would also encounter variable
winter temperatures, we also subjected egg masses to three overwintering regimes. In
addition, cold tolerances were estimated by measuring supercooling points of diapausing M.
disstria eggs. Population source, time of winter, and overwintering location were assessed as
potential sources of variation.
Egg masses were collected across a 552-km latitudinal gradient. Supercooling points
varied by population source, time of winter, overwintering location and time of
winter*overwintering location interaction. Even though population source affected
supercooling points, it did not follow a strict latitudinal order. All M. disstria populations
within this range showed high survival and tolerance of extreme overwintering temperatures.
Egg mortality was uniformly low, and unrelated to any of the above factors.
vi

!
Timing of egg hatch varied among population sources, overwintering locations, and
controlled spring temperature regimes. As expected, all plants and insects developed more
rapidly under warmer conditions. However, aspen and birch advanced more rapidly than did
M. disstria, which altered synchrony. Additionally, under temperature regimes simulating
future conditions, some insect populations currently south of the test plants became more
synchronous with these hosts than currently coexisting insect populations. We developed an
accumulated degree-day model of M. disstria egg hatch to extend the robustness of these
results across its broad geographic range and future conditions. These findings can assist the
prediction and management of forest insect defoliators in a warming climate.
vii

!
THESIS INTRODUCTION
Climate is undergoing pronounced anthropogenic change, and is expected to warm most
dramatically at high latitudes (Sala et al. 2000). For example, mean temperatures in the
boreal forest regions may increase an additional 5.4°C by 2070 (Beaumont et al. 2011). One
of the least understood consequences of climate change is how new temperature regimes will
influence interactions between forest trees and insect herbivores. In particular, we know
little about how phenological synchrony between spring egg hatch and budbreak will be
affected. Phenological synchrony is particularly critical to leaf feeders, as the relative
availability and suitability of newly developing host tissues plays an important role in
whether their populations remain at low densities or erupt into outbreaks (van Asch and
Visser 2007).
Increases in the boreal forest’s average annual temperatures will facilitate range
expansions by populations of native and introduced insect species. These expansions and
migrations have been estimated to be up to 100 km per decade (Climate Change 2014:
summary report). In addition to potential differences in development rates between plants
and insects, the enormous differences in migration ability add further complexity to our
predictions of how climatic changes will affect insect biodiversity, forest health, and
socioeconomic benefits derived from forests.
Insect survival within an area depends on cold tolerances that can vary in response to
seasonal temperature fluctuations (Hanec 1966). Climate change is projected to raise mean
winter temperatures, increasing the likelihood of higher survival for overwintering insects
1

!
(Ayres and Lombardero 2000; Hayhoe et al. 2006). However, as average annual
temperatures (and climatic variability) increase, the extent to which sudden extreme cold
periods in newly established populations may become a high source of egg mortality is
poorly understood (Overpeck 1996; Karl and Trenberth 2003).
The forest tent caterpillar, Malacosoma disstria Hübner (Lepidoptera: Lasiocampidae), is
an early season, univoltine folivore that feeds on a broad range of angiosperm trees. It is one
of the most widely distributed tree-feeding insects in North America, and undergoes
intermittent outbreaks that cause large-scale defoliation of its major host trees. Egg hatch
synchrony with budbreak of its host plants is critical to reproduction by M. disstria (Parry et
al. 1998; Jones and Despland 2006). Adults of both sexes are strong fliers, and this insect
has the ability to move up to 19 km a year (Evenden et al. In review). Previous literature by
Brown (1965) has recorded M. disstria adults flying 300 miles in 12 hours with the assistance
of turbulent cold front air masses. Thus, M. disstria serves as an ideal model for studying
climate change effects on egg hatch phenology with host trees.
As with most early spring-feeding folivores of perennial plants, access to young leaves is
crucial for optimal growth and survival of M. disstria (Dukes et al. 2008). If eggs hatch too
late, the host plants will confront larvae with increased leaf toughness and secondary
metabolite concentrations, and decreased nitrogen and water contents (James and Smith
1978; Hunter and Lechowicz 1992; Lindroth and Hwang 1996). Newly hatched larvae will
therefore take longer to develop and be exposed to predation for longer periods (Parry et al.
1998). If eggs hatch too early, the larvae will starve. Under increased temperatures, both
host plants and insect development rates accelerate, but their relative changes are poorly
understood.
2

!
The southern boreal forest is among the biomes most affected by global warming
worldwide (Ruckstuhl et al. 2007). Investigating the complex dynamics that climate change
brings to the establishment, success, and northern range expansion of early spring folivores
in this region can improve the understanding needed for effective forest management,
prevention of outbreaks, and conservation.
This thesis comprises two chapters. Chapter 1 addresses the question, “How does
elevated temperature affect synchrony between M. disstria egg hatch and host plant
budbreak?” Specifically, the date, success, and timing of egg hatch among four M. disstria
population sources along a latitudinal gradient subjected to three controlled spring
temperature regimes in outdoor mesocosms are evaluated. Chapter 2 addresses the question,
“What are the supercooling points of diapausing M. disstria eggs, and how are they affected
by population source, time of winter, and overwintering location?” The two chapters are
formatted in the styles of the Canadian Journal of Forest Research and the Canadian
Entomologist, respectively. In addition to primary tables and figures, Chapter 1 also has
Supplemental Tables and Figures that provide additional details and are intended to
accompany submission. Items that are not intended for publication are included in: 1) The
Thesis Appendix, which includes distributions of test subjects, overall design and assignment
information for the study, and 2) Electronic Supplemental Materials, which includes raw
data, photos, and colored versions of all figures, not intended for submission. The former,
but not the latter, are included in the bound thesis.
3

!
THESIS REFERENCES
Ayres, M. P. and Lombardero, M. J. 2000. Assessing the consequences of global change for
forest disturbance from herbivores and pathogens. The Science of the Total
Environment 262:263-286.
Beaumont, L. J., Pitman, A., Perkins, S., Zimmermann, N. E., Yoccoz, N. G., and Thuiller,
W. 2011. Impacts of climate change on the world’s most exceptional ecoregions.
Proceedings of the National Academy of Sciences of the United States of America
108:2306-2311.
Brown, C. E. 1965. Mass transport of forest tent caterpillar moths, Malacosoma disstria
Hübner, by a cold front. The Canadian Entomologist 97:1073-1075.
Climate Change 2014: Impacts, adaptation, and vulnerability. IPCC WGII AR5 Summary
for Policymakers. Figure SPM.5. http://ipcc-
wg2.gov/AR5/images/uploads/IPCC_WG2AR5_SPM_Approved.pdf.
Dukes, J. S., Pontius, J., Owig, D, Garnas, J. R., Rodgers, V. L., Brazee, N., Cooke, B.,
Theohardies, K. A., Stange, E. E., Harrington, R., Ehrenfeld, J., Gurevitch, J., Lerdau,
M., Stinson, K., Wick, R., and Ayres, M. 2009. Responses of insect pests, pathogens,
and invasive plant species to climate change in the forests of northeastern North
America: What can we predict? Canadian Journal of Forest Research 39:231-248.
Evenden, M. L., Whitehouse, C. M., and Jones, B. C. Resource allocation to flight in an
outbreaking forest defoliator, the forest tent caterpillar, Malacosoma disstria.
Environmental Entomology (In review).
4

!
Hanek, W. 1966. Cold-hardiness in the forest tent caterpillar, Malacosoma disstria Hübner
(Lasiocampidae: Lepidoptera). Journal of Insect Physiology 12:1443-1449.
Hayhoe, K., Wake, C. P., Huntington, T. G., Luo, L., Schwartz, M. D., Sheffield, J., Wood,
E., Anderson, B., Bradbury, J., DeGaetano, A., Troy, T. J., Wolfe, D. 2006. Past and
future changes in climate and hydrological indicators in the U.S. Northeast. Climate
Dynamics 28:381-407.
Hunter, A. F. and Lechowicz, M. J. 1992. Predicting the timing of budburst in temperate
trees. Journal of Applied Ecology 29:597-604.
James, T. D. W. and Smith, D. W. 1978. Seasonal changes in the major ash constituents of
leaves and some woody components of trembling aspen and red osier dogwood.
Canadian Journal of Botany 56:1798-1803.
Jones, B.C. and Despland, E. 2006. Effects of synchronization with host plant phenology
occur early in the larval development of a spring folivores. Canadian Journal of
Zoology 84:628-633.
Karl, T. R. and Trenberth, K. E. 2003. Modern global climate change. Science 302:1719-
1723.
Lindroth, R. L. and Hwang, S-Y. 1996. Clonal variation in foliar chemistry of quaking aspen
(Populus tremuloides Michx.). Biochemical Systematics and Ecology 24:357-364.
Overpeck, J. T. 1996. Warm Climate Surprises. Science 271:1820-1821.
Parry, D., Spence, J. R., Volney, W. A. 1998. Budbreak phenology and natural enemies
mediate survival of first-instar forest tent caterpillar (Lepidoptera: Lasiocampidae).
Environmental Entomology 27:1368-1374.
5

!
Ruckstuhl, K. E., Johnson, E. A., Miyanishi, K. 2007. The boreal forest and global change.
Philosophical Transactions Of The Royal Society 363:2245-2249.
Sala, O. E., Chapin III, F. S., Armesto, J. J., Berlow, E., Bloomﬁeld, J., Dirzo, R., Huber-
Sanwald, E., Huenneke, L. F., Jackson, R. B., Kinzig, A., Leemans, R., Lodge, D. M.,
Mooney, H. A., Oesterheld, M., Poff, N. L., Sykes, M. T., Walker, B. H., Walker, M.,
Wall, D. H. 2000. Global biodiversity scenarios for the year 2100. Science
287:1770-1774.
van Asch, M. and Visser, M. E. 2007. Phenology of forest caterpillars and their host trees:
the importance of synchrony. Annual Review Entomology 52:37-55.
6

!
Chapter 1: Effects of Simulated Temperature Increases and Forest Tent Caterpillar
Population Source on Synchrony of Egg Hatch with Aspen and Birch Budbreak
Abstract
The effects of anticipated warming temperatures on phenological synchrony between
herbivores and their host plants are poorly understood. Controlled temperature
manipulations, followed by measures of insect and plant development, can help address this
knowledge gap. However, the ability of many insects to disperse substantial distances over a
few generations, in contrast to the relative immobility and long lifespans of trees, adds
significant complexity to our interpretations and impairs our predictive abilities. The fact
that eggs may overwinter on plants far from the ones on which parents fed adds still further
complexities. We investigated how interactions among four population sources of forest tent
caterpillar (Malacosoma disstria Hübner), two of its major hosts, trembling aspen (Populus
tremuloides Michaux) and paper birch (Betula papyrifera Marshall), and three controlled
temperature regimes at two mesoscale outdoor sites, influence the relative timing of egg
eclosion and budbreak. Egg masses were collected across a 552 km latitudinal gradient in
Minnesota and Wisconsin, and subjected to three overwintering and three spring temperature
regimes in a split plot nested design. Timing of egg hatch varied among population source,
overwintering location, and spring temperature regime. Egg mortality was very low, and
unrelated to any of the above factors. As expected, all plants and insects developed more
rapidly under warmer conditions. However, aspen and birch advanced more rapidly than did
M. disstria. Additionally, under temperature regimes simulating future conditions, some
insect populations currently south of the test sites became more synchronous with the
7

!
manipulated hosts than did currently coexisting insect populations. We developed an
accumulated degree-day model of M. disstria egg hatch to extend the robustness of these
results across its broad geographic range. These findings can improve the prediction and
management of forest insect defoliators responding to a warming climate.
Keywords: climate change, synchrony, phenology, forest, insect
8

!
Introduction
One of the least understood consequences of climate change is how new temperature
regimes will influence phenological synchrony between forest trees and insect herbivores
(Schwartzberg et al. 2014). Phenological synchrony is particularly critical to leaf feeders, for
whom the relative availability and suitability of newly developing host tissues determine
whether herbivore populations remain at low densities or erupt into outbreaks (van Asch and
Visser 2003). Some lepidopteran species exhibit dramatic population fluctuations that occur
simultaneously over large geographical areas, suggesting that climate plays a major role in
the timing and occurrence of their outbreaks (Post and Forchhammmer 2002; Liebhold et al.
2004; Haukioja 2005). In some cases, warmer springs and winters have been associated with
higher frequencies and durations of forest lepidopteran outbreaks (Roland 1998; Volney and
Fleming 2000). Altered plant-herbivore relationships are of concern from the perspectives of
both plant injury and arthropod biodiversity (Ayres and Lombardero 2000; Parmesan and
Yohe 2003).
Boreal forests are dominated by a relatively small number of tree species, including
aspen, birch, fir, spruce, pine, cedar, and larch (Friedman and Reich 2005), and covers 30%
of the earth’s forested land (Volney and Fleming 2000). As the largest intact terrestrial
biome remaining on Earth, the boreal forest (Thesis App. 1A) stores enormous amounts of
carbon. Pan et al. (2011) estimates it contains 32% of the world’s forest carbon stock, or 272
± 23 petagrams of carbon. In the southern boreal forest (Bale et al. 2002; Battisti et al. 2005;
Netherer and Schopf 2010), mean annual temperatures have risen ~1.5°C since 1940, and are
expected to increase an additional 3-7° C in winter and 3-11°C in summer by 2100 (Kling et
9

!
al. 2003). Boreal ecosystems are also quite vulnerable to climate change, however, and even
slight increases in mean annual temperature may be critical along the current lower
latitudinal limits of the southern boreal forests, potentially contributing to forest die-back and
regime shifts to new ecosystems (Olsson 2010; Michaelian et al. 2011; Fisichelli et al. 2013).
Synchrony of egg hatch with host plant budbreak in the spring is crucial to the success of
many folivorous Lepidoptera in temperate forests. Development outside a relatively narrow
phenological window can have severe fitness consequences for early spring folivores (Feeny
1970; Mitter et al. 1979; Raupp et al. 1988; Stoyenoff et al. 1994; Quiring 1994; Mattson et
al. 1996; Lawrence et al. 1997; van Asch and Visser 2007). Larvae that hatch too late will
face increased leaf toughness and secondary metabolite concentrations, as well as decreased
nitrogen and water contents (James and Smith 1978; Hunter and Lechowicz 1992; Lindroth
and Hwang 1996). Hatching too early can result in starvation. The forest tent caterpillar
(Malacosoma disstria Hübner), eastern tent caterpillar (Malacosoma americanum Fabricius),
European gypsy moth (Lymantria dispar dispar Linnaeus), eastern spruce budworm
(Choristoneura fumiferana Clemens), and jack pine budworm (Choristoneura pinus
Freeman) are examples of early season, eruptive defoliators, whose ranges overlap in the
ecotonal boundaries of the southern boreal forests in the Great Lakes region. These insects
overwinter either as pharate (1st
instar within eggs) or 2nd
instar larvae, and begin feeding in
synchrony with budbreak of their primary hosts (Parry et al. 1998). Each of these species
also undergoes intermittent landscape-scale outbreaks that both influence fundamental
ecosystem processes and exert high socioeconomic impacts.
Current approaches to estimating future phenological synchronies between plants and
herbivores are based on either direct experimentation or simulation modeling (Schwartzberg
10

!
et al. 2014). However, we currently lack fundamental knowledge about how an additional
important factor, dispersal, will affect future regimes. Specifically, the much higher
migration rates of insects than of trees will create novel interactions between more southerly
insect populations and extant tree populations in northern latitudes (Climate Change 2014:
summary report). Potential migration of lepidopteran folivores with warming temperatures
may have important ramifications for their population dynamics (Hodson 1941; Blais et al.
1955; Fitzgerald and Costa 1986).
Malacosoma disstria is a gregarious leaf-feeding insect that has a relatively broad
host range, and can cause severe defoliation of deciduous trees throughout much of Canada
and the United States (Trudeau et al. 2010). Its distribution (Thesis App. 1B) extends from
Nova Scotia to California, and northern Alberta to south Texas. Females oviposit all of their
eggs in a single band encircling distal ends of host plant twigs from late June to early July
(Parry et al. 2001). Egg masses range from 120 to 240 eggs in northern Minnesota (latitude
48°N) (Witter et al. 1975). The number of eggs in a clutch is influenced by female size,
latitude, population densities, and successive years of outbreak (Ives 1971; Witter et al. 1975;
Smith and Goyer 1986; Parry et al. 2001).!!The pharate larvae hatch in early April and May.
Larvae undergo five instars, each 7-10 days long, before pupating in early to late June. After
another seven to ten days, adult moths emerge, mate, and oviposit. Adults of both sexes are
strong fliers, with the ability to move up to 19 km a year (Evenden et al. In review), although
most fly shorter distances. Some M. disstria have been known to fly hundreds of kilometers
(Fullard and Napoleone 2001), with a maximum range of 300 miles in 12 hours with
assistance from turbulent cold air masses (Brown 1965).!
11

!
Two of the most preferred hosts of M. disstria across its northern range are trembling
aspen (Populus tremuloides Michaux) (Thesis App. 1C) and paper birch (Betula papyrifera
Marshall) (Thesis App. 1D). These species have similar ranges, and are two of the most
widely distributed trees in North America, extending from Nova Scotia, Maine and New
Jersey on the Atlantic coast to Washington, British Columbia, and Alaska on the Pacific
coast. Both species are present in all Canadian provinces except Nunavut, and have isolated
populations in high elevations in the southern United States and northern Mexico (Thesis
App. 1 C, D). Aspen is one of the first species to establish after a disturbance and is clonal.
Phenology, morphology, and chemical constituents within a clone show low variance
(Barnes 1969, Chilcote et al. 1992, Lindroth and Hwang 1996), offering an ideal system to
study the varying hatch time on caterpillar performance (Parry et al. 1998). Birch grows on
most soils in the boreal forest, and can tolerate a wide variation in precipitation (Grant and
Thompson 1975).
I investigated the effects of population source and temperature on the extent, timing, and
duration of egg eclosion, and on the phenological synchrony of egg hatch with two of M.
disstria’s major host plants, trembling aspen and paper birch, under manipulated temperature
regimes. These manipulations included both winter and spring conditions, by using
movement of egg masses and an outdoor mesocosm, respectively. The manipulated
temperatures mimicked those anticipated within the region during the next 75 to 100 years.
The specific objectives were to: 1) Quantify the relative effects of increased temperature on
timing of egg hatch by M. disstria populations collected along a latitudinal gradient, and 2)
determine the effects of elevated temperatures, insect population source, and overwintering
12

!
regime to develop a robust model of synchrony between insect egg hatch and host plant
budbreak.
Materials and Methods
Experimental design
We measured egg hatch among four M. disstria populations, which were subjected to
three overwintering regimes, and three spring temperature treatments that were applied in
replicated plots at two sites. Each plot contained two host tree species, trembling aspen and
paper birch. The two manipulated-treatment sites were in northeastern Minnesota along the
ecotone of southern boreal and temperate zones (Fig. 1), at the Cloquet Forestry Center
(CFC) near Cloquet and the Hubachek Wilderness Research Center (HWRC) near Ely. This
research was conducted using an outdoor, mesoscale infrastructure managed by the
University of Minnesota, termed ‘B4WarmED’ (http://forestecology.cfans.umn.edu/!
Research/B4WARMED), which simulates temperatures predicted in climate change models.
This infrastructure includes replicated heated plots that contain juvenile stands of ten native
tree species. Underground heating coils (Danfoss GX, Devi A/B, Denmark) and
aboveground ceramic lamps (Salamander Model FTE-1000) simultaneously manipulate soil
and plant surface temperatures.
Three temperature treatments were administered: ambient controls, +1.7°C above
ambient, and +3.4°C above ambient. These temperatures were selected to bracket the
anticipated warming in the region during the next 75-100 years (Kling et al. 2003; Wuebbles
and Hayhoe 2004; Parry et al. 2007). Each experimental ring (plot) received one temperature
treatment. There were six rings per block, and three blocks per site, thus a total of 18 rings
13

!
per site. Each block contained two rings per treatment, providing a total of 12 replicates per
treatment across the two sites.
The experimental rings were 3 meters in diameter. The heated rings have three ceramic
lamps to administer +1.7°C and four lamps to administer +3.4°C, evenly distributed around
the perimeter. The control plots have 3 similarly distributed mock lamps. Experimental
rings were activated on 21 March 2012 in Cloquet and 27 March 2012 in Ely, and treatments
continued nonstop for 248 days in Cloquet (stop date: 24 November 2012) and 234 days in
Ely (stop date: 16 November 2012). Each ring contained a thermal sensor that recorded
aboveground temperature every fifteen minutes. Each ring contained four seedlings each of
B. papyrifera, P. tremuloides, and ten other tree species.
Insect populations
Malacosoma disstria egg masses were collected in the fall of 2011 from naturally
occurring populations in four regions across a latitudinal gradient from southern Wisconsin
to northern Minnesota, located (from south to north) near Prairie du Chien, WI, Baraboo, WI,
Mille Lacs Lake, MN, and Bemidji, MN (Fig. 1). The Prairie du Chien egg bands were
collected mid to late-December, southeast of the Wisconsin River along 180 m of a steep
slope (42°58'27.98"N, 90°59'10.34"W). The Baraboo egg bands were collected mid-October
through early November, within an 800 m radius in the Baraboo Hills (43°25'12.53"N,
89°38'8.69"W). The Mille Lacs Lake egg bands were collected 23 October 2012, south of
Isle, MN (46° 8'30.15"N, 93°27'33.71"W), along 275 m of a recreational trail. The Bemidji
egg bands were collected 24 October 2012, from two locations 32 km apart: Potato Lake (47°
0'14.05"N, 95° 2'39.18"W) and Elbow Lake (47° 7'31.56"N, 95°34'8.55"W). Egg bands
14

!
were collected using pruning poles and hand pruners to remove the distal ends of branches
from host trees.
The number of egg bands collected ranged from 400 at Mille Lacs Lake to 100 at Prairie
du Chien. Each egg band was inspected thoroughly, and those with damage (many eggs
missing and/or spumulin chipped off) or unusual form were discarded. The remaining egg
bands were tabulated for total number of eggs, number of missing eggs, and total number of
hatched eggs (Thesis App. 2). Of the egg bands used in this study, 52 had some missing eggs
(mean of 2.28 per clutch). The egg bands were sterilized by applying an aqueous solution of
1% Tween-20 and 5% bleach.
Partitioning and enclosure of overwintering egg bands
To account for the potential influence of winter environment on how egg bands respond
to variable degree-day accumulation in the spring, and to test whether populations might
differ in this regard, we also administered three overwintering regimes. In early December,
half (36 per population, 144 total) of the egg bands were brought to Madison, WI, and half
were taken immediately to the B4WarmED sites, distributed evenly between Cloquet and Ely
(Fig. 1). Egg masses were stored outdoors in mesh bags 152 cm above the ground.
In mid-March, just prior to ring activation, the 144 egg bands that had overwintered in
Madison were taken to the B4WarmED sites (72 each to Cloquet, Ely). At this time, all egg
bands were placed in individual 7.5 x 12.5 cm mesh sleeves. Sleeves were fastened with
binder clips, and suspended from 183 cm tall poles on a 60 cm radius rebar tie wire arm in
the center of the experimental plot. Each of the 36 experimental plots contained 8 M. disstria
egg bands: 1 per overwintering treatment (on B4WarmED site vs. in Madison) per population
(Thesis App. 3A). Individual insect egg band assignments are shown in Thesis App. 3B.
15

!
Once egg bands were placed in their respective experimental plots, the egg masses
were monitored daily. It was typical for a few pharate larvae to hatch one to two days before
the remainder of the clutch hatched in a much more uniform fashion. Because of this initial
skewing, the time at which 50 eggs had hatched was designated as the date of hatch for each
egg mass. Observations on each egg mass continued until five days after the last observed
hatch. The duration of egg hatch was designated as the interval between when the first 50
and final larvae had emerged.
Plant phenology determination
Two aspen and two birch trees per plot were evaluated daily for budbreak phenology.
We assigned the following stages (Thesis App. 4): 1- dormant bud, darkened; 2- bud
darkened to light green color near the base; 3- budbreak, leaf tips visible.
Statistical analyses
We evaluated the effects of manipulated temperature, overwintering regime, hatch site,
and population source on the timing and duration of egg hatch by M. disstria. We evaluated
the effects of manipulated temperature and site on budbreak by trembling aspen and paper
birch. We also assessed phenological synchrony, by calculating the difference between M.
disstria hatch and host plant budbreak dates within the same temperature conditions. We
employed both discrete and continuous models, as they serve complementary purposes. The
discrete analyses characterize and quantify all sources of variation. The continuous analyses
integrate and expand the generality of results across a common temperature scale, utilizing
the full power of the within-ring temperature recordings.
In the discrete models, we employed a split-plot nested experimental design. For the
success, onset, and duration of egg hatch, an analysis of variance was applied, using Yijkl = !
16

!
+ Oi + Pj + Sk + Tl + OPij + OTil + PTjl + eijkl, where ! is the overall mean response, O
represents overwintering location i, P is population source j, S is site k, T is temperature
treatment l, and eijkl is the whole-plot error. Yijkl is the average response of overwintering
location i, population source j, site k, and temperature treatment l. Temperature was analyzed
as three categorical treatments (ambient, +1.7°C and +3.4°C) across both hatch sites. The
main factors were site, population source, and overwintering regime, unless otherwise stated.
For budbreak, we used: Yij = ! + Si + Tj + STij + eij, where ! is the overall mean response, S
is site i, T is temperature treatment j, and eij is the whole-plot error. Yij is the average
response of site i and temperature treatment j.
In the continuous models, the temperature measured within each ring (N=36) was
treated as a continuous variable across both sites. These provide a common temperature
scale derived from plot-specific data, and also account for the latitudinal difference between
Cloquet and Ely. We used two methods, based on mean ring temperature and accumulated
degree-days, respectively. Mean ring temperatures were calculated by taking the grand mean
from daily averages of 36 plot-specific temperatures between the date of heat activation and
date of last egg hatch (8 May 2012). The accumulated degree-days model makes full use of
the fifteen-minute intervals at which temperature was recorded, rather than averaging over a
season. The accumulated degree-day model employed a baseline threshold of 4.4°C based
on previous studies (Ives 1973; Parry et al. 1998). Degree-days were calculated from
recorded data every 15 minutes from each ring, between 1 January at 00:00 to 31 May at
23:45, 2012. We used the ‘trapezoidal integration’ method (Tobin et al. 2001) to capture the
smallest temperature increment (to 0.01 °C) above threshold.
17

!
A preliminary continuous analysis indicated that egg hatch varies mainly by one
covariate, temperature treatment. Therefore, we constructed our analysis of covariance to
employ a common slope for all populations: Yij = µ + αi + !Xj + αiX + εi + εij. In this mixed-
effects model, plot nested within site is incorporated as a random factor, µ is the overall
mean response, α is the effect of population source i, X is the covariate term for temperature
treatment (either accumulated degree-days or mean ring temperature) at a common slope !,
εij represents errors associated within plot and εijk represents random or unexplained error
associated with plot nested within site. Yij represents the average egg hatch response of
population source i and temperature treatment j. Regression models were used to determine
the relationships of date of egg hatch to mean ring temperature, and proportion of hatch to
accumulated degree-days.
We analyzed host plant budbreak with a balanced two-way analysis of variance, using
Yij = µ + αi + βj + (αβ)ij + εij. In this fixed-effects model, µ is the overall mean response, α is
the effect of the site i, βj refers to the effect of mean ring temperature j and εijk represents
random or unexplained error. Yij represents the average budbreak response of site i and
mean ring temperature j. Regression models assessed the relationships between day of
budbreak and mean ring temperature. Site was retained as a factor for plants because local
conditions (i.e. soil) could potentially influence budbreak, in contrast to egg masses in mesh
bags suspended from poles.
An analysis of covariance was used to assess the influence of increased temperature on
phenological synchrony between M. disstria and host plant budbreak, using Yij = µ + αi +
!Xj + αiX + εi + εij. In this mixed-effects model, plot nested within site is incorporated as a
random factor, µ is the overall mean response, α is the effect of population source i, X is the
18

!
covariate term for temperature (either mean ring temperature or accumulated degree-days) at
a common slope !, εij represents errors associated within plot and εijk represents random or
unexplained error associated with plot nested within site. Yij represents the average
difference between egg hatch and budbreak of population source i at temperature treatment j.
Regression models were used to determine the relationships between the differences in
phenological events (egg hatch to birch/aspen budbreak) and temperature.
Statistical analyses were performed using Proc GLM, Proc MIXED (general linearized
models) and Proc REG (regression) in SAS 9.3 and JMP 11.0.0 (SAS Institute Inc., Cary,
NC). Results are reported (type III error) with corresponding P values of < 0.05 as
‘significant’ and 0.05 < P ≤ 0.10 as ‘marginally significant’.
Results
What factors affect survival and timing of egg hatch by forest tent caterpillar?
Overwintering survival of eggs
The proportion of eggs that hatched was uniformly high, with an overall mean of 81.5%
(SE= 0.83%) and an overall median of 84.5% (N= 284). There were no significant effects of
population source, overwintering regime, site, or experimentally manipulated temperature on
egg hatch success (Table 1). There was a marginally significant and biologically slight
tendency for lower egg hatch at Ely than Cloquet (79.7% vs. 83.3%; P=0.055). Five egg
bands showed very few or no insect hatch and were discarded from the study. The
proportions of successful egg hatch for each population and overwintering regime are in
Table S1.
Timing of egg hatch
19

!
The timing of M. disstria egg hatch was strongly influenced by hatch site, temperature
treatment, overwintering location, and population source (Table 2A). There were no
significant interactions, however, among those factors. Overall, egg hatch ranged from 11
April to 8 May among all treatments, with a grand mean of 28 April (SE = 0.4 days) (Fig. 2),
and a median of 1 May (N = 141). The median dates of egg hatch show similar trends, with
second quartiles ranging from 17 April to 6 May (Fig. S1).
As expected, hatch occurred first among egg bands exposed to +3.4 °C, then +1.7 °C,
then ambient conditions. Across all populations, overwintering regimes, and hatch sites,
mean egg hatch at +3.4°C occurred an average of 2.5 days before those at +1.7°C, which
were 4 days before those at ambient temperatures. When populations and overwintering
locations were pooled, egg hatch occurred earlier at Cloquet than Ely by 2 days at ambient
temperature and +1.7 °C, and by 5 days at +3.4°C (Table S2). There was a general trend for
egg bands collected from southern populations to hatch before those from northern
populations, but this pattern was not absolute. Across the study, populations from Baraboo
and Mille Lacs Lake hatched earlier than those from Bemidji and Prairie du Chien (Table
S2). Likewise, insects that overwintered in northeastern Minnesota hatched 3 days later than
those that overwintered at the lower latitude, Madison, when all other factors were pooled.
This overwintering effect held even though the egg bands that overwintered in Madison were
brought to northeastern Minnesota prior to any degree-day accumulation based on current
threshold estimates, and thereafter were exposed to the same accumulation of degree-days
within rings.
Both continuous models, mean ring temperature and accumulated degree-days,
likewise indicated a strong response of M. disstria egg hatch date to temperature (Table 3).
20

!
Results are shown based as Day of Hatch and Proportion Cumulative Hatch, respectively.
Population source had a significant effect on egg hatch, for both locally overwintered egg
masses and those overwintered in Madison, in the model based on mean ring temperature.
Population had a significant effect on egg hatch on egg masses that overwintered in Madison
in the accumulated degree-day model. Interactions were not analyzed, as they largely
comprised the random effects ANCOVA model error. The r2
values were substantially
higher in the accumulated degree-days than mean ring temperature models.
Under both continuous temperature models, egg masses that overwintered in Madison
(mean date = 26 April, SE = 0.4; mean degree-days = 217.9, SE = 3.5; median date = 01
May; median degree-days = 213.2; N = 143) hatched earlier than locally overwintered egg
masses (mean date = 28 April, SE = 0.5; mean degree-days = 236.8, SE = 3.3; median date =
26 April; median degree-days = 233.2; N = 141) by 2.5 days, or 19 degree-days. The slopes
of insect development versus temperature were very similar between Madison and locally
overwintered eggs in both the accumulated degree-day (Fig. 3) and mean ring temperature
(Fig. S2) continuous models.
The hatch order of egg masses from various M. disstria populations differed among
wintering sites in the continuous models based on mean ring temperature (Fig. S3).
However, the order in which the end of hatching occurred among populations was not
affected by wintering site. In the continuous models based on accumulated degree-days,
populations under both wintering regimes had the same initial and end of hatch orders (Fig.
S4). As in the other analyses, r2
values in the models employing accumulated degree-days
were consistently and substantially higher than those employing mean ring temperatures.
Duration of egg hatch
21

!
Overall, the duration of egg hatch within a band averaged 5.0 (SE = 0.2) days, with a
median of 5.0 days (N = 141). The duration of egg hatch was strongly influenced by
treatment site, overwintering location, and population source (Table 2B). Temperature
treatment did not affect hatch duration, and there were no significant interactions among
factors. The mean duration of egg hatch ranged from 3.7 days (Mille Lacs Lake population,
locally overwintered, Ely, +1.7 °C) to 8.7 days (Baraboo, Madison overwintered, Cloquet,
+1.7 °C) (Fig. 4). Egg bands that overwintered in Ely had the shortest duration of egg hatch
(mean days = 4.6, SE = 0.2; median days = 4.0; N = 70), followed by Cloquet (mean days =
5.3, SE = 0.2; median days = 6.0; N = 71) and Madison (mean days = 5.9, SE = 0.2; median
days = 6.0; N = 142).
The onset and duration of egg hatch were inversely related (Fig. 5). Populations whose
eggs hatched sooner also exhibited greater variability in their hatch durations. For example,
egg masses that started to hatch on 12 April had an average hatch duration of 16 days,
compared with egg masses that started to hatch on 04 May, which averaged 2 days.
Populations from Baraboo and Mille Lacs Lake had the earliest egg hatches, but they also
had the longest mean hatch durations (6.5 and 5.5 days, respectfully). The two latest
hatching populations, Bemidji and Prairie du Chien, had the shortest mean durations of hatch
(5 and 4.8 days, respectfully) (Fig. S5). Insect populations demonstrated higher variability in
hatch durations in the continuous models. For both continuous models, Baraboo eggs were
the first to complete hatching, but exhibited the greatest variability in hatch duration (11 days
for degree-days and 6.5 days for mean ring temperature). Mille Lacs Lake eggs experienced
the next highest variability in hatch duration, followed by Bemidji, and concluding with
Prairie du Chien.
22

!
How does temperature affect budbreak phenology of trembling aspen and paper birch?
In the discrete analysis, the onset of budbreak by both aspen (grand mean = 25 April,
SE = 1.0 days; median = 25 April; N = 68) and birch (grand mean = 19 April, SE = 0.9 days;
median = 16 April; N = 69) were strongly influenced by temperature treatment (Table 4).
Site was a significant factor for aspen budbreak (P<0.0001) and marginally significant for
birch budbreak (P=0.087). As expected, warmer experimental temperatures advanced host
plant budbreak, which occurred first at +3.4 °C, then at +1.7 °C, then at ambient (Fig. 6).
Also as expected, aspen budbreak occurred earlier in Cloquet than Ely. However, birch
showed a marginally significant opposite trend, breaking bud earlier at Ely. Overall,
budbreak occurred over 35 days for aspen (7 April to 12 May) and 30 days for birch (9 April
to 9 May), among all treatments (Table S3).
As temperatures increased, the interval between mean budbreak by these two tree species
decreased. Birch budbreak occurred an average of ten days earlier than aspen budbreak at
ambient temperatures, seven days earlier at +1.7 °C, and five days earlier at +3.4 °C. Median
host plant budbreak showed the same trend (Table S3).
In the continuous analyses, increases in mean ring temperature likewise advanced
budbreak by both aspen and birch (Table 4). Site had a significant effect on aspen budbreak,
but not on birch budbreak. Aspen budbreak and birch budbreak each showed negative linear
relationships with increasing mean ring temperatures (r2
= 0.38 and 0.27, respectively).
Birch broke bud 10 days earlier than aspen at ambient conditions, but this difference
decreased to 5 days under the warmest simulated temperature conditions. Aspen responded
more to temperature increases than birch (slopes = -3.6 and -2.7, respectively), and their
phenologies converged.
23

!
Accumulating degree-days also advanced budbreak in both aspen and birch (Table 4),
although site was not a significant factor for either plant. Aspen budbreak and birch
budbreak each showed stronger positive linear relationships (r2
= 0.69 and 0.67, respectively)
than in the mean ring temperature continuous model.
How do insect population source, plant species, and temperature interact to structure
phenological synchrony?
Discrete Analysis
Manipulated spring temperatures advanced the timing of insect egg hatch and host plant
budbreak, but with both plant species advancing more than M. disstria (Fig. 7). At ambient
temperatures, M. disstria egg hatch occurred earlier than aspen budbreak. Aspen budbreak
occurred before egg hatch at +1.7°C, and advanced further at +3.4°C (Tables S2, S3). Birch
budbreak occurred before M. disstria egg hatch at all temperatures.
With all factors pooled at ambient temperatures, mean birch budbreak occurred 5.5 days
before mean egg hatch, and 9 days before mean aspen budbreak. At +1.7°C, birch budbreak
advanced to 10 days before egg hatch, but decreased to 7 days before aspen budbreak. At
+3.4°C, the interval between birch budbreak and egg hatch remained the same. However,
aspen was only 5 days behind birch budbreak at the warmest temperature conditions.
Both insect egg hatch and aspen budbreak occurred earlier at Cloquet than Ely, consistent
with a latitudinal effect. Host plant budbreak advanced more rapidly than insect egg hatch
when subjected to increasing temperatures, with birch budbreak occurring earlier than aspen.
Continuous Analysis
Both continuous models, mean ring temperature and accumulated degree-days, indicate
strong temperature effects on the phenological relationship between M. disstria egg hatch
24

!
and host plant budbreak (Table 5). Accumulated degree-days indicated strong responses
between egg hatch and both host plant budbreak. Mean ring temperature indicated a strong
response between egg hatch and aspen budbreak, but not birch budbreak. Population source
did not affect this interval for either continuous temperature model. Interactions were not
analyzed under the ANCOVA model, as the error composed largely from random effects.
Both continuous models indicated that host plant budbreak responded more strongly than
M. disstria egg hatch to high temperatures, altering phenological synchrony (Fig. 8). Host
plant budbreak had steeper slopes than M. disstria (accumulated degree-days: birch r2
=
0.718, aspen r2
= 0.683, M. disstria r2
= 0.501, mean ring temperature: birch r2
=!0.267,!
aspen!r2
=!0.382,!M. disstria r2
=!0.195).
With the exception of aspen budbreak under ambient mean ring temperatures, host plant
budbreak occurred earlier than M. disstria egg hatch at all temperatures. Birch broke bud
three degree-days before mean egg hatch, or 5 days as manipulated by mean ring temperature
under ambient temperatures, and 12.5 degree-days, or 12 days under the warmest simulated
temperatures. Aspen broke bud 3 degree-days after mean egg hatch, or 7.5 days after as
manipulated by mean ring temperatures, and 4.5 degree-days, or 5 days under the warmest
simulated temperatures (Fig. 9). Malacosoma disstria population-specific synchrony data are
in Figure 9C.
Both continuous models adequately explain host plant budbreak, insect egg hatch,
and the phenological relationship between the two. The accumulated degree-day model
provided a better goodness of fit, as indicated by much stronger correlation coefficient values
(Table 3, Figs. 9, S3, and S4). This is also true for host plant budbreak (Table 4). Overall,
the phenological relationship between M. disstria egg hatch and host plant budbreak is better
25

!
explained by the accumulated degree-day model, as indicated by the more consistently
significant p-values (Table 5) when compared to mean ring temperature.
Discussion
These results identified winter temperatures, spring temperatures, plant species and
insect population as sources of variation in phenological relationships between a univoltine
early-season herbivore and its host trees, and quantified how these relationships may respond
to changing climate. Experimental warming treatments that simulate predicted temperatures
over the next 75-100 years advanced both host plant budbreak and insect egg hatch
phenologies. However, host plants generally showed greater advances. At ambient
temperatures, average insect egg hatch occurred nearly in synchrony with, or shortly before
or after, host plant budbreak depending on species, which agrees with prior work (Gray and
Ostaff 2012; Parry et al. 1998; Schwartzberg et al. 2014). As temperatures increased, the
magnitude of difference between host plant budbreak and egg hatch enlarged. Some other
systems have shown the opposite trend (Parmesan 2007), further illustrating the complexity
of ecological consequences to climate change, especially in trophic interactions.
Several key processes, operating at various spatiotemporal scales, will determine how
these changes in phenological synchrony influence the stability of plant–herbivore
interactions. At the scale of individual plants and insects, an advance in budbreak relative to
egg hatch typically lowers insect performance, due to reduced growth resulting from poor
foliar nutritional quality, and prolonged exposure to predators (Parry et al. 1998; Jones and
Despland 2006). At a regional spatial scale, M. disstria might compensate for altered
phenological synchrony through migration. That is, more southern populations might
26

!
advance northward within the lifespan of individual trees, thus restoring prior temporal
relationships between egg hatch and budbreak. The likelihood of this occurring is supported
by the strong flight abilities of both genders of M. disstria, and the orders of magnitude in
generation times and longevities between these hosts and herbivores (Brown 1965, Peltonen
2002, Climate Change 2014: summary report). We investigated two components of this
scenario. First, we demonstrated that populations along a latitudinal gradient vary in their
responses to temperature. Moreover, this latitudinal gradient was of a sufficiently small scale
(552 km) that all of these populations could access the trees in the experimental mesocosms
within the time frame being simulated. Second, for migration to be successful, the eggs
deposited by females that had flown northward in late spring would need to survive winter.
This also appears likely, as egg survival was uniformly high (81%), and was not affected by
the latitude of overwintering site, population source, or their interactions. On a larger
temporal scale, local populations of M. disstria might adapt to the altered phenologies of
their major host plant species. This would require strong selection against individuals that
hatched too late, a prospect that seems reasonable given the high degree of phenological
synchrony that current populations evolved. Also, our data show high variation within
populations and substantial variation within individual egg masses, both of which would
facilitate local genetic adaptation. Additionally, the duration of hatch within an egg mass
spans some of the plant phenological shifts we observed, and also increases with warming
temperatures, further providing opportunities for the herbivore to undergo genetic shifts
within the span of the climatic changes we simulated.
At a finer spatial scale, oviposition on or larval movement to later-developing plants
that provide more suitable foliage, could facilitate survival by current insect populations
27

!
(Gray and Ostaff 2012). Numerous later-developing tree species are suitable hosts in this
region, such as red oak (Quercus rubra Linnaeus), cherry (Prunus spp.), black ash (Fraxinus
nigra Marshall), white ash (Fraxinus americana Linnaeus), and basswood (Tilia americana
Linnaeus). Some alternate tree species would advance from a less to more synchronous
relationship with this polyphagous herbivore. Also, aspen shows strong variation in
budbreak among clones, so some might become more synchronous. This raises the
possibility that clones that currently avoid herbivory through late budbreak may be less well
chemically defended, due to trade-offs arising from metabolic costs, and hence become more
susceptible with changing climate. Possibly slowing local adaptation is M. disstria’s
behavior of usually confining their feeding to one plant (Despland and Noseworth 2006), and
the weak dispersed ability of early instars.
The results of this study provide an opportunity to integrate anticipated phenological
and migratory shifts to climate change. Bemidji is the closest population in latitude to the
B4WarmED sites, so using hatching patterns of Bemidji egg bands under ambient conditions
provides the best baseline for comparison with other populations at various temperatures.
Interestingly, Bemidji was the only population to successfully explain the difference between
egg hatch and budbreak for both aspen and birch, and also showed the highest r2
values
(Table 5C). To construct these estimates in a generalized fashion, we employed data from
each population’s response to degree-day accumulation (Table 5C), using only locally
overwintered egg masses. We made these calculations using both mean and median egg
hatch, as the results can sometimes differ substantially (Table 6). The number of
accumulated degree-days required to become synchronous with host tree budbreak varies by
population, but does not follow a strict latitudinal gradient. Based on the mean data, two of
28

!
the three southern insect populations require fewer degree-days than Bemidji. Future
research should include additional M. disstria populations, both from similar and more
distant latitudes, to expand the applicability of this approach.
Among-population variation in hatch order among populations showed a general
relationship to latitude (Parmesan 2007), but this sequence was not absolute (Figs. S3, S4).
In particular, eggs from the southernmost population, Prairie du Chien, hatched later than
expected. This may be partially attributable to microclimate. For example, the Prairie du
Chien eggs were mostly collected along a steep riverbank, facing east-northeast, and so were
shaded during the warmest period of each day.
The combined use of discrete and continuous temperature analyses provided powerful
assessments for measuring the effects on anticipated climatic changes on egg hatch,
budbreak, and phenological synchrony. The discrete analysis identified the main sources of
variability, while the continuous analysis identified more robust and generalizable
relationships. Mean ring temperature provided useful relationships, but accumulated degree-
days generally provided better goodness of fit. These results also provide a useful template
for understanding and predicting trophic interactions arising from climate change, and for
suggesting proactive forest management strategies. Future studies should 1) identify and
incorporate species-specific thresholds for aspen and birch, 2) include more insect
populations, both within migratory access of the southern boreal forest ecotone and across
the full latitudinal range of M. disstria, 3) test mature trees for possible differences with
seedlings, 4) administer additional temperatures to better calibrate linear versus curvilinear
zones of insect and plant response (Berggren et al. 2009), 5) include more tree species from
within the relatively broad8 host range of M. disstria, 6) include more genotypes within these
29

!
tree species, and 7) include more insect species to better estimate how climate change will
influence trophic interactions in temperate forests.
This study demonstrates increased asynchrony in response to warming temperatures,
in a direction that may favor the host plants. Once budbreak occurs, late hatching larvae
have a small window for successful insect development. These results indicate high disparity
within birch (lag time of 12 days), and moderate disparity with aspen (lag time of 5 days) at
the warmest temperatures (Fig. 9A). This research provides real-scenario mesoscale
examples for understanding and predicting responses to climate warming.
Acknowledgments
We thank the University of Minnesota for use of the B4WarmED site, materials, and
technicians who assisted in data collection and field bioassays. Site managers Artur
Stefanski and Karen Rice provided technical expertise. Dr. Jun Zhu and Peter Crump
(University of Wisconsin) provided valuable statistical modeling assistance. The Wisconsin
DNR, Minnesota DNR and Jana Albers, Christine Buhl, as well as local citizens provided
assistance locating and collecting insects. Dr. Patrick Tobin (USFS) provided extensive help
with degree-day calculations. This research was funded by the USDA NIFA AFRI Grant No.
2011-67013-30147, the UW College of Agricultural and Life Sciences, a UW Zoology
Department teaching assistantship, and a UW PEOPLE project assistantship.
30

!
References
Ayres, M. P. and Lombardero, M. J. 2000. Assessing the consequences of global change for
forest disturbance from herbivores and pathogens. The Science of the Total
Environment 262:263-286.
Bale, J. S., Masters, G. J., Hodkinson, I. D., Awmack, C., Bezemer, T. M., Brown, V. K.,
Butterfield, J., Buse, A., Coulson, J. C., Farrar, J., Good, J. E. G., Harrington, R.,
Hartley, S., Jones, T. H., Lindroth, R. L., Press, M. C., Symrnioudis, I., Watt, A. D.,
and Whittaker, J. B. 2002. Herbivory in global climate change research: direct effects
of rising temperature on insect herbivores. Global Change Biology 8:1-16.
Barnes, B. V. 1969. Natural variation and delineation of clones of Populus tremuloides and
P. grandidentata in northern lower Michigan. Silvae Genetica 18:130-142.
Battisti, A., Stastny, M., Netherer, S., Robinet, C., Schopf, A., Roques, A., and Larsson, S.
2005. Expansion of geographic range in the pine processionary moth caused by
increased winter temperatures. Ecological Applications 15:2084-2096.
Berggren, A., Bjorkman, C., Bylund, H., and Ayres, M. P. 2009. The distribution and
abundance of animal populations in a climate of uncertainty. Oikos 118:1121-1126.
Brown, C. E. 1965. Mass transport of Forest Tent Caterpillar moths, Malacosoma disstria
Hübner, by a cold front. The Canadian Entomologist 97:1073-1075.
Chilcote, C. A., Witter, J. A., Montgomery, M. E., and Stoyenoff, J. L. 1992. Intra- and
inter-clonal variation in gypsy moth larval performances on bigtooth and trembling
aspen. Canadian Journal of Forest Research 22:1676-1683.
31

!
Climate Change 2014: Impacts, adaptation, and vulnerability. IPCC WGII AR5 Summary
for Policymakers. Figure SPM.5. http://ipcc-wg2.gov/AR5/images/uploads/IPCC_
WG2AR5_SPM_Approved.pdf.
Department of Forest Resources: College of Food, Agricultural Natural Resource Sciences,
University of Minnesota, http://forestecology.cfans.umn.edu/B4WARMED.html.
August 9, 2010.
Despland, E. and Noseworthy, M. 2006. How well do specialist feeders regulate nutrient
intake? Evidence from a gregarious tree-feeding caterpillar. Journal of Experimental
Biology 209:1301-1309.
Etilé, E. and Despland, E. 2009. Developmental variation in the forest tent caterpillar: life
history consequences of a threshold size for pupation. Oikos 117:135-143.
Evenden, M. L., Whitehouse, C. M., and Jones, B. C. 2014. Resource allocation to flight in
an outbreaking forest defoliator, the forest tent caterpillar, Malacosoma disstria.
Environmental Entomology (In review).
Fisichelli, N. A., Frelich, L. E., and Reich, P. B. 2014. Temperate tree expansion into
adjacent boreal forest patches facilitated by warmer temperatures. Ecography
37:152-161.
Fitzgerald, T. D., and Costa, J. T. 1986. Trail based communication and foraging behavior of
young colonies of the forest tent caterpillar, Malacosoma disstria (Lepidoptera:
Lasiocampidae). Annals of the Entomological Society of America 79:999-1007.
Friedman, S. K., and Reich, P. B. 2005. Regional legacies of logging: departure from pre-
settlement forest conditions in northern Minnesota. Ecological Applications 15:726-
744.
32

!
Fullard, J. H. and Napoleone, N. 2001. Diel flight periodicity and the evolution of auditory
defences in the Macrolepidoptera. Animal Behaviour 62:349-628.
Grant, W. F. and Thompson, B. K. 1975. Observations of Canadian birches, Betula
cordifolia, B. neoalaskana, B. populifolia, B. papyrifera, and B. x caerulea. Canadian
Journal of Botany 53:1478-1490.
Gray, D. R. and Ostaff, D. P. 2012. Egg hatch of forest tent caterpillar (Lepidoptera:
Lasiocampidae) on two preferred host species. Canadian Entomologist 144:790-797.
Haukioja, E. 2005. Plant defenses and population fluctuations of forest defoliators:
mechanism-based scenarios. Annales Zoologici Fennici 42:313-325.
Hicke J. A., Allen, C. D., Desai, A. R., Dietze, M. C., Hall, R. J., Hogg, E.T., Kashian, D.
M., Moore, D., Raffa, K. F., Sturock, R. N., and Voglemann, J. Effects of biotic
disturbances on forest carbon budgets of the United States and Canada. Global
Change Biology. In press.
Hodson, A.C., 1941. An ecological study of the forest tent caterpillar, Malacosoma disstria
Hbn. University of Minnesota Agricultural Experiment Station Technical Bulletin
148.
Ives, W. G. H. 1971. The forest tent caterpillar in Alberta. Environmental Canada, Canadian
Forest Service, Northern Forestry Centre Internal Report, NOR-4, Edmonton,
Alberta.
Ives, W. G. H. 1973. Heat units and outbreaks of the forest tent caterpillar, Malacosoma
disstria (Lepidoptera: Lasiocampidae). Canadian Entomologist 105:529-543.
33

!
Jones, B. C. and Despland, E. 2006. Effects of synchronization with host plant phenology
occur early in the larval development of a spring folivores. Canadian Journal of
Zoology 84:628-633.
Kling, G. W., Hayhoe, K., Johnson, L., Magnuson, J., Polasky, S., Robinson, S., Shuter, B.,
Wander, M., Wuebbles, D., and Zak, D. 2003. Confronting Climate Change in the
Great Lakes Region. Union of Concerned Scientists, Ecological Society of America;
Cambridge, MA, Washington, D.C. 92 pp.
Lindroth, R. L. and Hwang, S-Y. 1996. Clonal variation in foliar chemistry of quaking aspen
(Populus tremuloides Michx.). Biochemical Systematics and Ecology 24:357-364.
Leibhold, A., Koenig, W. D., and Bjornstad, O. N. 2004. Spatial synchrony in population
dynamics. Annual Review of Ecology and Systematics 35:467-490.
Logan, J. A., Régnière, J., and Powell, J. A. 2003. Assessing the impacts of global climate
change on forest pests. Frontiers in Ecology and the Environment 1:130-137.
Michaelian, M., Hogg, E. H., Hall, R. J., and Arsenault, E. 2011. Massive mortality of aspen
following severe drought along the southern edge of the Canadian boreal forest.
Global Change Biology 17:2084-2094.
Natural Resources Canada: Insects and diseases of Canada’s forests, Forest Tent Caterpillar;
http://imfc.cfl.scf.rncan.gc.ca/insecte-insect-eng.asp?getID=9374. March 8, 2011.
Netherer, S. and Schopf, A. 2010. Potential effects of climate change on insect herbivores in
European forests-general aspects and the pine processionary moth as specific
example. Forest Ecology and Management 259:831-838.
Olsson, R. 2010. Boreal Forest and Climate Change. Air Pollution & Climate Secretariat.
http://www.airclim.org/sites/default/files/documents/borealforest-regional.pdf.
34

!
Pan, Y., Birdsey, R. A., Fang, J., Houghton, R., Kauppi, P. E., Kurz, W. A., Phillips, O. L.,
Shvidenko, A., Lewis, S. L., Canadell, J. G., Ciais, P., Jackson, R. B., Pacala, S. W.,
McGuire, A. D., Piao, S., Rautiainen, A., Sitch, S., and Hayes, D. 2011. A large and
persistent carbon sink in the world’s forests. Science 333:988-993.
Parmesan, C., and Yohe, G. 2003. A globally coherent fingerprint of climate change impacts
across natural systems. Nature 421:37-42.
Parry, D., Spence, J. R., and Volney, W. A. 1998. Budbreak phenology and natural enemies
mediate survival of first-instar forest tent caterpillar (Lepidoptera: Lasiocampidae).
Environmental Entomology 27:1368-1374.
Parry, D., Goyer, R. A., and Lenhard, G. J. 2001. Macrogeographic clines in fecundity,
reproductive allocation, and offspring size of the forest tent caterpillar Malacosoma
disstria. Ecological Entomology 26:281-291.
Parry, M. L., Canziani, O. F., Palutikof, J. P., van der Linden, P. J., and Hansen, C. E. 2007.
IPCC, 2007: climate change 2007: impacts, adaptation, and vulnerability.
Contribution of working group II to the fourth assessment report of the
intergovernmental panel on climate change.
Parmesan, C. 2007. Influence of species, latitudes and methodologies on estimates of
phenological response to global warning. Global Change Biology 13: 1860-1872.
Peltonen, M., Liebhold, A. M., Bjornstad, O. N., and Williams, D. W. 2002. Spatial
synchrony in forest insect outbreaks: roles of regional stochasticity and dispersal.
Ecology 83:3120-3129.
Post, E., and Forchhammer, M. C. 2002. Synchronization of animal population dynamics by
large-scale climate. Nature 420:168-171.
35

!
Qi, X., Wang, X., Kang, L. 2007. Influence of soil moisture on egg cold hardiness in the
migratory locust Locusta migratoria (Orthoptera: Acridiidae). Physiological
Entomology 32:219-224.
Roland, J. 1993. Large-scale forest fragmentation increases the duration of tent caterpillar
outbreak. Oecologia 93:25-30.
Schwartzberg, E. G., Jamieson, M. A., Raffa, K. F., Reich, P. B., Montgomery, R. A., and
Lindroth, R. L. 2014. Simulated climate warming alters phenological synchrony
between an outbreak insect herbivore and host trees. Oecologia 175:1041-1049.
Smith, J. D. and Goyer, R. A. 1986. Population fluctuations and causes of mortality for the
forest tent caterpillar, Malacosoma disstria (Lepidoptera: Lasiocampidae), on three
different sites in Southern Louisiana. Environmental Entomology 15:1184-1188.
Smith, G. J. and Raske, A. G. 1968. Starvation experiments with first instar forest tent
caterpillar larvae. Canadian Department of Northern Affair and National Resources,
Forestry Branch, Annual Report on Forest Research Notes 24:39.
Tobin, P. C., Nagarkatti, S., and Saunders, M. C. 2001. Modeling Development in Grape
Berry Moth (Lepidoptera: Torticidae). Environmental Entomology 30:692-699.
van Asch, M. and Visser, M. E. 2007. Phenology of forest caterpillars and their host trees:
the importance of synchrony. Annual Review Entomology 52:37-55.
Volney, W. J. A. and Fleming, R. A. 2000. Climate change and impacts of boreal forest
insects. Agriculture, Ecosystems, and Environment 82:283-294.
Witter, J. A., Mattson, W. J., and Kulman, H. M. 1975. Numerical analysis of a forest tent
caterpillar (Lepidoptera: Lasiocampidae) outbreak in northern Minnesota. Canadian
Entomologist 107:837-854.
36

!
Wuebbles, D. J. and Hayhoe, K. 2004. Climate change projections for the US Midwest.
Mitigation and Adaptation Strategies for Global Change 9:335-363.
37

!
Source DF F P
Model 47 0.82 0.784
Error 235
Site 1 3.74 0.055
Pop. 3 1.55 0.204
Site*Pop. 3 0.27 0.849
Overwinter 2 0.23 0.791
Overwinter*Pop. 6 0.83 0.550
Temp. (°C) 2 0.06 0.943
Temp. (°C)*Site 2 0.84 0.432
Temp. (°C)*Pop. 6 0.42 0.869
Temp. (°C)*Site*Pop. 6 0.97 0.450
Temp. (°C)*Overwinter 4 0.17 0.954
Temp. (°C)*Overwinter*Pop. 12 1.25 0.251
Table&1:&Results!of!ANOVA,!relating!proportion!of!M.#disstria!egg!bands!that!hatched!to!population!source,!temperature,!
overwintering!regime,!and!site!(N=!288!egg!bands).!
38

!
!
Source DF F P
A Day of Hatch
! ! !
!
Model 27 11.18 <0.0001*
!
Error 256
! !
! ! ! ! !
!
Site 1 33.18 <0.0001*
!
Temp. (°C) 2 61.13 <0.0001*
!
Overwinter 2 19.32 <0.0001*
!
Pop. 3 19.02 <0.0001*
!
Temp. (°C)*Pop. 6 0.34 0.822
!
!
B Hatch Duration !! !! !!
!
Model 27 3.62 <0.0001*
!
Error 256
! !
! ! ! ! !
!
Site 1 29.43 <0.0001*
!
Temp. (°C) 2 2.15 0.119
!
Overwinter 2 11.57 <0.0001*
!
Pop. 3 3.92 0.0092*
!
Temp. (°C)*Pop. 6 0.79 0.575
!
!
!
Site*Pop. 3 1.33 0.264
Table 2: Results of ANOVA, relating the onset (A) and duration (B) of hatch of M. disstria egg bands under manipulated
temperature regimes. An asterisk denotes significant factor at P ≤ 0.05 (N = 288 egg bands).
!
39

!
Source Overwintering a b r2
DF F P
A! Day of Hatch
!
! ! !
!
Model Locally -1.685 33.532 0.195 4 14.78 <0.0001*
!
Error
!
136
! !
! ! ! ! ! ! ! ! !
!
Mean Ring Temp. (C)
!
1 38.65 <0.0001*
!
Pop.
!
3 7.04 0.0002*
!
Model Madison -1.478 28.767 0.112 4 15 <0.0001*
!
Error
!
138
! !
! ! ! ! ! ! ! ! !
!
Mean Ring Temp. (C)
!
1 21.65 <0.0001*
!
Pop. 3 12.61 <0.0001*
B Proportion Cumulative Hatch
!
! ! !
!
Model Locally 0.0049 -0.659 0.426 4 26.05 <0.0001*
!
Error
!
136
! !
! ! ! ! ! ! ! ! !
!
Accumulated Degree-days
!
1 82.4 <0.0001*
!
Pop. 3 0.6 0.617
!
Model Madison 0.0051 -0.604 0.533 4 43.73 <0.0001*
!
Error
!
138
! !
! ! ! ! ! ! ! ! !
!
Accumulated Degree-days
!
1 104.74 <0.0001*
!
Pop. 3 2.75 0.045*
Table 3: Results of ANCOVA relating the onset of egg hatch to mean ring temperature (°C) (A) and the proportion cumulative
hatch to accumulated degree-days (B) for M. disstria. An asterisk denotes significant factor at P ≤ 0.05. Interactions not
assessed as these largely comprise the error terms.
40

!
A ANOVA: Discrete model
Source DF F P
Aspen Budbreak
Model 5 26.58 <0.0001*
Error 62
Temperature Treatment (°C) 2 56.22 <0.0001*
Site 1 19.97 <0.0001*
Temperature Treatment (°C)*Site 2 0.28 0.7597
Birch Budbreak
Model 5 10.16 <0.0001*
Error 63
Site 1 3.02 0.087
Table 4: Results of discrete (A) and continuous (B) models of aspen and birch budbreak in
response to manipulated early spring temperatures. An asterisk denotes significant factor at P
≤ 0.05 (N=68 for aspen, 69 for birch).
41

!
Aspen Budbreak
Model 0.141 -12.816 0.688 7 1.49 <0.0001*
Error 65
Accumulated Degree-days 1 96.54 <0.0001*
Site 1 0.79 0.378
Birch Budbreak
Model 0.15 -15.75 0.667 2 65.85 <0.0001*
Error 66
Accumulated Degree-days 1 127.5 <0.0001*
Site 1 0.02 0.892
B ANOVA: Continuous model
Source a b r2
DF F P
Aspen Budbreak
Model -3.614 52.511 0.382 2 35.44 <0.0001*
Error 65
Mean Ring Temp. (°C) 1 59 <0.0001*
Site 1 14.72 0.0003*
Birch Budbreak
Model -2.664 34.924 0.267 2 14.56 <0.0001*
Error 66
Mean Ring Temp. (°C) 1 27.09 <0.0001*
Site 1 1.38 0.244
42

!
!
A ANOVA : Discrete model
! ! ! ! !
!
Source DF F P
!
Egg hatch to aspen budbreak
! ! !
!
Model 5 24.23 <0.0001*
!
Error 278
! !
! ! ! ! !
!
!
Site 1 10.78 0.0012*
!
!
Egg hatch to birch budbreak
! ! !
!
Model 5 16.26 <0.0001*
!
Error 278
! !
! ! ! ! !
!
!
Site 1 56.87 <0.0001*
!
Table 5: Results of discrete (A) and continuous (B) models on the interval between M.
disstria egg hatch and host plant budbreak. Results of continuous models on the interval
between M. disstria population-specific egg hatch and host plant budbreak (C). An
asterisk denotes significant factor for P ≤ 0.05.
43

!
B ANCOVA : Continuous models
! ! !
!
Source a b r2
DF F P
!
Egg hatch to aspen budbreak
!
Model -2.13 19.355 0.199 4 8.75 <0.0001*
!
Error
! ! !
138
! !
! ! ! ! ! ! ! !
!
Mean Ring Temp. (°C)
! ! !
1 32.25 <0.0001*
!
Population
! ! !
3 0.89 0.449
!
! !
!
!
Model -0.705 -1.274 0.02 4 0.87 0.484
!
Error
! ! !
138
! !
! ! ! ! ! ! ! !
!
Mean Ring Temp. (°C)
! ! !
1 1.52 0.219
!
Population 3 0.65 0.587
Egg hatch to aspen budbreak !! !! !! !! !! !!
Model -0.374 8.439 0.05 4 2.64 0.036*
Error
! ! !
138
! !
! ! ! ! ! ! !Accumulated Degree-days
! ! !
1 8.23 0.0048*
! ! !
! ! !Model -0.045 2.742 0.067 4 5.2 0.0006*
Error
! ! !
138
! !
! ! ! ! ! ! !Accumulated Degree-days
! ! !
1 18.61 <0.0001*
44

!
C Population a b r2
DF F P
!
Egg hatch to aspen
budbreak
Bemidji
!
Model -0.06923 0.02188 0.1252 1 10.01 0.0023*
!
Error 70
!
Mille Lacs Lake
!
Model -0.01532 2.05523 0.0075 1 0.51 0.476
!
Error 68
!
Baraboo
!
Model -0.00339 0.56161 0.0004 1 0.03 0.871
!
Error 69
!
Prairie du Chien
!
Model -0.02701 4.44557 0.034 1 2.43 0.124
!
Error 69
!
Egg hatch to birch
budbreak
Bemidji
!
Model -0.08185 10.05625 0.1691 1 14.25 0.0003*
!
Error 70
!
Mille Lacs Lake
!
Model -0.0343 -1.47063 0.0334 1 2.35 0.130
!
Error 68
!
Baraboo
!
Model -0.01201 -5.11005 0.0042 1 0.29 0.593
!
Error 69
!
Prairie du Chien
!
Model -0.04752 2.30867 0.0836 1 6.29 0.015*
!
Error 69
45

!
Population
km South of
Sitesb
Number of Added Degree-Days that will Reconstruct Local Conditionsa
To Become Synchronous with Aspen To Become Synchronous with Birch
Based on Median Based on Mean Based on Median Based on Mean
Bemidji 0 0 0 0 0
Mille Lacs Lake 140.33 -15.50 -13.00 -19.32 -13.54
Baraboo 451.82 -32.96 -24.22 -33.60 -26.88
Prairie du Chien 496.57 3.35 19.02 15.51 15.48
Values calculated from standardized degree-day calculation across both B4WarmED sites for duration of study. Values
calculated as the difference between mean or median accumulated degree-days for each host plant budbreak and M. disstria
population-specific egg hatch at ambient temperature.
a
At local conditions, birch breaks bud a mean of 54.04 and a median of 53.39 degree-days before mean M. disstria egg hatch,
and aspen breaks bud a mean of 19.39 and a median of 24.6 degree-days before mean M. disstria egg hatch.
b
Midpoint between Ely and Cloquet, MN.
Table 6: Integration of anticipated climate change, phenological synchrony, and insect population-specific responses to degree-
day accumulation. Values calculated from Bemidji egg bands as a baseline to recreate local conditions. An asterisk denotes
significant factor for P ≤ 0.05.
46

!
Figure Legends
Figure 1: Locations of experimental sites at which M. disstria egg bands and host trees were
subjected to manipulated spring temperatures (solid stars), population sources (circles), and
overwintering sites of egg bands (open and solid stars). Experimental design is illustrated on
accompanying flow diagram on left.
Figure 2: Mean dates of egg hatch among four M. disstria populations subjected to three
manipulated temperature regimes and three overwintering regimes, during Spring 2012.
Differences among sites, overwintering treatments, temperature treatments, and populations
of M. disstria are significant (ANOVA, df = 83, P < 0.0001) (±1 standard error).
Figure 3: Relationship of proportion of cumulative M. disstria egg hatch to accumulated
degree-days (threshold = 4.4°C) at the B4WarmED sites, Spring 2012. Data are shown
separately for three overwintering locations. Widths of trendlines indicate 95% C.I. Statistics
are in Table 3. Population source was significant for Madison overwintered egg bands (Fig.
S4).
Figure 4: Mean duration of hatch of M. disstria egg bands subjected to three controlled
spring temperature regimes and three overwintering regimes, 2012. Differences among sites,
overwintering treatments, and populations of M. disstria are statistically significant
(ANOVA, df = 27, P = 0.0001). Temperature treatment was not significant (±1 standard
error).
Figure 5: Inverse relationship between onset and within-band duration of egg hatch in M.
disstria. Data are shown based on (A) Hatch Date and (B) Accumulated Degree-Days. Data
47

!
collected at the B4WarmED sites, Spring 2012 (ANOVA (A) F1, 282 = 132.17, P < 0.0001, N
= 284) (a = -1.526, b = 24.79, r2
= 0.319); ANOVA (B) F1, 282 = 98.20, P < 0.0001, N = 284)
(a = -9.91, b = 281.36, r2
= 0.258). Widths of trendlines indicate 95% C.I.
Figure 6: Mean date of budbreak among birch and aspen trees subjected to three temperature
regimes at the B4WarmED sites, Spring 2012. Birch budbreak advanced earlier with
warming temperatures (site was moderately significant) (ANOVA, df = 5, P < 0.0001).
Aspen budbreak advanced earlier with both warming temperatures, and varied between sites
(ANOVA, df = 5, P < 0.0001) (±1 standard error).
Figure 7: Box and whisker plots displaying how synchrony of M. disstria egg hatch with
aspen (A) and birch (B) budbreak varies with M. disstria population source and temperature
treatment. Degree of synchrony calculated as difference (days) between median budbreak
and egg hatch at the B4WarmED sites, Spring 2012, under three temperature treatments.
Negative x-values denote earlier budbreak development (±1 quantile).
Figure 8: Mean distribution of M. disstria and host plant phenology subjected to
accumulated degree-days (threshold = 4.4°C) (A) and mean ring temperature (B) at the
B4WarmED sites, Spring 2012. Mean ring temperatures computed from the grand mean of
daily averages from 36 plot-specific temperatures between heat activation date to last egg
hatch. Egg hatch and budbreak always advance with increasing temperature, but host plants
advance more rapidly. Widths of trendlines indicate 95% C.I. All insects overwintered
locally (N = 144). Statistics are provided in Table 3 for M. disstria and Table 4 for host
plants.
48

!
Figure 9: Difference in timing between M. disstria egg hatch and host plant budbreak
subjected to accumulated degree-days (A) and mean ring temperature (B) at the B4WarmED
sites, Spring 2012. Negative values indicate plant budbreak occurred before insect egg hatch.
Widths of trendlines indicate 95% C.I. Insects overwintered locally (N=144). Mean birch
budbreak occurred before M. disstria egg hatch throughout the study. A and B: Pooled
populations; C: Separate populations subjected to accumulated degree-days. Statistics are
located in Table 5.
49

!
!Figure 1.
Collect egg bands from 4 populations along
550 km latitudinal gradient in November.
Distribute egg bands (early December) among 3 sites
along 600 km latitudinal gradient (Ely, Cloquet,
Madison) for overwintering.
Bring eggs that overwintered in Madison to Ely and
Cloquet in early March.
Distribute egg bands among 3 spring temperature
treatments per site: six rings per treatment per site, one
egg band per ring.
Activate treatments in late March.
Record onset and duration of egg hatch, budbreak of
aspen and birch.
Record temperature within each ring every 15
minutes.
Analyze as Split Plot Nested Design
Discrete: ANOVA
Continuous: ANCOVA, based on Mean Ring Temperature,
Accumulated Degree-Days
= Insect population source
!
= Madison overwintering location
!
= B4Warmed overwintering locations and spring
controlled-temperature treatment sites
!
50

!
Overwintering
Treatment
Proportion of Successful Egg Hatch
df F PInsect Population
Bemidji Mille Lacs Lake Baraboo Prairie du Chien
Ely 0.825 0.826 0.795 0.722 3 1.28 0.288
Cloquet 0.828 0.843 0.78 0.817 3 0.73 0.538
Madison 0.828 0.798 0.785 0.807 3 0.42 0.738
df 2 2 2 2 5 -- --
F 0 0.7 0.03 1.28 -- 0.82 --
P 0.995 0.5 0.967 0.284 -- -- 0.537
Table S1: Results of ANOVA, relating proportion of successful M. disstria egg hatch to population source and overwintering
treatment (N=288).
64

!
Temperature
Above
Ambient! (°C)
M. disstria Population
Hatch
Site
Overwintering
LocationBemidji
Mille Lacs
Lake
Baraboo
Prairie du
Chien
Pooled
Populations
0
5/2
(5/1-6)
5/3
(5/1-6)
5/1
(4/30-5/2)
5/5
(5/1-7)
5/2
(4/30-5/7)
Ely
Local
1.05 0.97 0.31 1.07 0.51
1.7
5/2
(4/29-5/7)
4/29
(4/24-5/1)
4/28
(4/19-5/6)
5/1
(4/29-5/7)
4/30
(4/19-5/7)
1.46 1.12 2.76 1.2 0.82
3.4
4/28
(4/24-5/4)
4/26
(4/20-5/1)
4/22
(4/19-5/1)
4/30
(4/29-5/4)
4/27
(4/19-5/4)
1.68 1.7 1.97 0.83 0.95
0
5/1
(4/26-5/4)
5/1
(4/24-5/8)
4/28
(4/24-5/1)
5/3
(5/1-7)
5/1
(4/24-5/8)
Cloquet
1.28 2.4 1.33 1 0.83
1.7
4/28
(4/24-5/1)
5/1
(4/29-5/7)
4/25
(4/23-29)
5/1
(4/24-5/7)
4/29
(4/23-5/7)
1.17 1.36 0.92 1.81 0.83
3.4
4/22
(4/15-27)
4/21
(4/16-24)
4/22
(4/15-5/1)
4/25
(4/15-5/1)
4/23
(4/15-5/1)
1.69 1.43 2.33 2.42 1
Table S2: Overall distribution of Madison (A) and Locally (B) overwintered M. disstria hatch dates in 2012. Each cell contains
mean (earliest and latest) hatch date. Standard error (days) is beneath each hatch date (N = 288).
A"
"
65

!
Temperature
Above
Ambient! (°C)
M. disstria Population
Hatch
Site
Overwintering
LocationBemidji
Mille Lacs
Lake
Baraboo
Prairie du
Chien
Pooled
Populations
0
4/30
(4/24-5/7)
4/29
(4/25-5/1)
4/29
(4/27-5/2)
5/2
(5/1-6)
4/30
(4/24-5/7)
Ely
Madison
1.71 1.36 0.79 0.76 0.62
1.7
4/26
(4/24-5/1)
4/25
(4/19-5/1)
4/25
(4/24-29)
5/2
(5/1-6)
4/27
(4/19-5/6)
1.3 1.74 0.85 0.87 0.88
3.4
4/27
(4/24-5/1)
4/24
(4/19-29)
4/19
(4/11-29)
5/2
(4/29-5/7)
4/26
(4/11-5/7)
1.2 1.54 3.04 1.46 1.31
0
4/28
(4/24-5/7)
4/25
(4/24-27)
4/24
(4/16-5/2)
5/1
(5/1-4)
4/27
(4/16-5/4)
Cloquet
1.65 0.48 2.24 0.48 0.89
1.7
4/28
(4/24-5/7)
4/22
(4/15-30)
4/20
(4/16-25)
4/29
(4/24-5/8)
4/25
(4/15-5/8)
2.06 2.49 1.77 2.05 1.26
3.4
4/21
(4/15-25)
4/18
(4/15-27)
4/19
(4/15-24)
4/19
(4/15-26)
4/19
(4/11-27)
1.73 1.93 1.81 1.95 0.89
B"
"
66

!
Budbreak Date
Temperature Above
Ambient (°C)
Plant Species
Site
B. papyrifera P. tremuloides
Mean
0
4/23 (4/16-5/2) 5/6 (4/30-5/12)
Ely
2.17 1.33
1.7
4/16 (4/11-30) 4/27 (4/22-5/2)
2.07 1.16
3.4
4/13 (4/11-22) 4/22 (4/11-30)
0.96 1.86
0
4/26 (4/19-5/9) 5/1 (4/26-5/7)
Cloquet
1.72 0.97
1.7
4/19 (4/13-26) 4/22 (4/16-30)
1.33 1.15
3.4
4/14 (4/9-23) 4/16 (4/7-25)
1.26 1.53
Median
0
25-Apr 6-May
Ely
16-Apr 3-May
1.7
16-Apr 27-Apr
11-Apr 24-Apr
3.4
14-Apr 24-Apr
11-Apr 16-Apr
0
26-Apr 2-May
Cloquet
23-Apr 1-May
1.7
19-Apr 23-Apr
16-Apr 21-Apr
3.4
13-Apr 16-Apr
13-Apr 4/11!
Table S3: Overall distribution of host plant budbreak in 2012. Each cell contains mean (earliest and latest) budbreak with
standard error (days) located beneath. Each median budbreak cell contains median with 1st
quantile located beneath.
67

!
Figure Legends
Figure S1: Median dates of egg hatch among four M. disstria populations subjected to three
overwintering regimes and three spring temperature regimes, Spring 2012. Differences
among sites, overwintering treatments, temperature treatments, and populations of M. disstria
are statistically significant (ANCOVA, df = 83, P < 0.0001) (±1 quantile).
Figure S2: Effect of mean ring temperatures on mean hatch date of M. disstria at the
B4WarmED sites, Spring 2012. Data are shown separately for three overwintering locations.
Hatch date for M. disstria advanced with increasing mean ring temperature for both
overwintering locations. Widths of trendlines indicate 95% C.I. Population source was
statistically significant for each wintering regime (Fig. S3). Statistics are located in Table 3.
Figure S3: Mean hatch dates of egg masses from four M. disstria populations that
experienced local (A) and Madison (B) overwintering regimes. Hatch date for M. disstria
advanced with mean ring temperature, and varied among populations, at both wintering
regimes (ANCOVA, df = 4, P < 0.0001). Width of trendline indicated by 95% C.I.
Figure S4: Proportion of cumulative egg hatch among four M. disstria populations,
separated by locally overwintered (A) and Madison overwintered (B) regimes. Egg hatch
was influenced by accumulated degree-days for both wintering regimes (ANCOVA, df = 4, P
< 0.0001). Population source was not significant.
Figure S5: Relationship of duration of egg hatch among four M. disstria populations to
accumulated degree-days (A) and mean ring temperature (B). Duration of hatch was
68

!
influenced by overwintering regime, hatch site, and population source (ANOVA, df = 27, P
< 0.0001). Temperature treatment was not significant.
69

!
Chapter 2: Supercooling Points of Diapausing Forest Tent Caterpillar Eggs
Abstract
The forest tent caterpillar (Malacosoma disstria Hübner) is a widely distributed folivore
that causes large-scale defoliation during intermittent outbreaks. Malacosoma disstria is
univoltine, and overwinters as pharate larvae within egg bands. Populations in the northern
portions of its range are subjected to very cold winter temperatures. Little is known about
how extreme temperatures affect the winter survival and cold tolerances of M. disstria eggs,
the modality of cold hardiness, and how cold tolerances may vary over time and among
populations. I evaluated the supercooling points of M. disstria eggs from four populations
collected along a 552 km latitudinal gradient from southern Wisconsin to northern Minnesota
in the fall of 2011. To test for potential effects of winter environment on supercooling
points, we also administered three overwintering regimes. Measurements of supercooling
points were recorded at three times during the insect’s overwintering period, November,
February, and March. Supercooling points differed significantly with the time of season
tested, population source, and overwintering treatment. Average supercooling points from all
the entire study decreased from a maximum of -26.8°C (pooled populations, overwintering
location at population source, November) to a minimum of -40.3°C (Baraboo, Madison
overwintered location, February). This relationship likely explains how M. disstria can
survive the extreme cold temperatures that characterize winters in its northern range.
Keywords: Malacosoma disstria, cold tolerance, climate change
75!

!
Introduction
Insects in temperate and boreal forest regions are subject to prolonged periods of extreme
cold temperatures (Rochefort et al. 2011). Diapause is a genetically based mechanism for
escaping harsh environmental conditions, and is commonly triggered by environmental cues
such as photoperiod and temperature (Denlinger 1991; Bale and Hayward, 2010; Schiesari
and O’Connor 2013). However, the physiological mechanisms for tolerating cold
temperatures during diapause vary. Insect cold-hardiness mechanisms are commonly divided
into two main categories: freeze tolerance or freeze avoidance (Duman 2001; Voituron et al.
2002; Bale and Hayward 2009). Freeze tolerant insects are able to survive the formation of
ice in the haemolymph or gut (Zachariassen and Hammel 1976). Freeze avoidant insects are
able to lower the freezing point of their body fluids, and supercool (Doucet et al. 2009). Both
strategies are physiologically complex, and a comprehensive understanding of the insect’s
cold tolerance and acclimation ability are required to determine the mechanisms by which
these ectotherms survive in regions with low winter temperature (Zachariassen 1985; Duman
et al. 1991; Clark and Worland 2008; Doucet et al. 2009; Denlinger and Lee 2010).
Most folivorous insects and terrestrial arthropods in North America are freeze avoidant
(Somme 1982), using energy to produce antifreeze products such as glycerol, sugars, and
other polyols (Doucet et al. 2009; Sformo et al. 2011). The quality and abundance of
protective antifreeze products can directly affect overwintering survival, and may be
associated with host plant quality (Rochefort et al. 2011). Previous studies have shown that
insect cold tolerance and overwintering success can vary along a latitudinal gradient (Addo-
Bediako et al. 2000), while seasonal temperatures can also play an important role in
supercooling variation, inducing glycerol content (Somme 1964). Glycerol content in the
76

!
forest tent caterpillar (Malacosoma disstria Hübner) remains relatively low in the late fall
and early winter months (October and November), but triples during the next three months,
before returning to late fall levels in March (Hanec 1966).
Like many insects in the temperate/boreal forests regions, M. disstria is subjected to an
extended period of subzero temperatures (Trudeau et al. 2010). The range of M. disstria
extends from British Columbia to California on the west coast, and Maine to Florida on the
east coast (Thesis App. 1B). This folivore feeds in colonies and causes severe, large-scale
defoliation during intermittent outbreaks (Wood et al. 2010). Females oviposit in bands
around twigs and branches in the crowns of host trees. Egg bands are covered in spumulin, a
foamy coating that hardens when exposed to air. This foamy coating provides a small
amount of protection (Parry et al. 2001), but other than intermittent ice and snow, egg bands
are otherwise directly exposed to the harsh winter conditions. Pharate larvae emerge from
their eggs in early spring in synchrony with budbreak of their host plants. They undergo five
larval instars, at approximately one instar per week. Pupation occurs mid-June and adults
emerge in July, mate and oviposit. Adults of both sexes are strong fliers, with the ability to
move up to 19 km a year (Evenden et al. In review). Malacosoma disstria can fly long
distances (Fullard and Napoleone 2001), with a maximum of 480 km in 12 hours when
assisted by turbulent cold air masses (Brown 1965).
Although M. disstria causes substantial damage to forests in outbreak years, populations
in northern latitudes typically have less success, with extreme cold temperatures being
partially responsible (Daniel and Myers 1995). Winter mortality of eggs varies annually
within populations (Hanec 1966; Cooke and Roland 2003).
77!

!
Climate change is likely to facilitate northward expansion of many insect species in
north-temperate zones. Historically, M. disstria outbreaks in northern latitudes are less
common and less severe (Daniel and Myers 1995). As the average temperatures in their
northern ranges become more tolerable, insect folivores that overwinter as eggs may benefit
from higher survival rates (Ayres and Lombardero 2000). An IPCC summary report (2014)
estimates early spring folivores have the ability to migrate approximately 100 km per decade,
or 10 km per year. Previous literature has shown this range is well within M. disstria’s flight
capacity (Brown 1965; Fullard and Napoleone 2001; Evenden et al. In review).
I investigated the supercooling points of M. disstria eggs from populations collected
along a 552 km latitudinal gradient, in November, February, and March of 2011-2012. I
hypothesized that M. disstria eggs from northern latitudes have lower supercooling points
than those from more southern latitudes, and that supercooling points decrease as winter
proceeds.
Materials and Methods
Insect population sources
Malacosoma disstria egg masses were collected across a latitudinal gradient from
southern Wisconsin to northern Minnesota. Four naturally occurring populations were
collected in the fall of 2011 (from south to north) near Prairie du Chien, WI, Baraboo, WI,
Mille Lacs Lake, MN, and Bemidji, MN (Fig. 1). The ‘Prairie du Chien’ egg bands
(42°58'27.98"N, 90°59'10.34"W) were collected during early November to mid-December.
The ‘Baraboo’ egg bands (43°25'12.53"N, 89°38'8.69"W) were collected during mid-October
through early November. The ‘Mille Lacs Lake’ egg bands (46° 8'30.15"N, 93°27'33.71"W)
78

Entomology M.S. Thesis - Uelmen

Entomology M.S. Thesis - Uelmen

Recommended

Recommended

More Related Content

Similar to Entomology M.S. Thesis - Uelmen

Similar to Entomology M.S. Thesis - Uelmen (20)

Entomology M.S. Thesis - Uelmen