Biology and conservation of overwintering monarch butterflies in Mexico

MISSING IMAGE

Material Information

Title:
Biology and conservation of overwintering monarch butterflies in Mexico
Physical Description:
xv, 164 leaves : ill. ; 29 cm.
Language:
English
Creator:
Alonso-Mejia, Alfonso, 1963-
Publication Date:

Subjects

Subjects / Keywords:
Monarch butterfly -- Mexico   ( lcsh )
Monarch butterfly -- Wintering   ( lcsh )
Zoology thesis, Ph. D
Dissertations, Academic -- Zoology -- UF
Genre:
bibliography   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1996.
Bibliography:
Includes bibliographical references (leaves 146-163).
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by Alfonso Alonso-Mejia.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 022859206
oclc - 34944564
System ID:
AA00022249:00001

Table of Contents
    Title Page
        Page i
        Page ii
    Dedication
        Page iii
    Acknowledgement
        Page iv
        Page v
        Page vi
        Page vii
    Table of Contents
        Page viii
        Page ix
        Page x
    List of Tables
        Page xi
    List of Figures
        Page xii
        Page xiii
    Abstract
        Page xiv
        Page xv
    Chapter 1. Introduction and background
        Page 1
        Page 2
        Page 3
        Page 4
        Page 5
        Page 6
        Page 7
        Page 8
        Page 9
        Page 10
        Page 11
        Page 12
        Page 13
        Page 14
        Page 15
        Page 16
    Chapter 2. Microclimatic differences between closed and opened forest and their consequences for the survival of monarch butterflies overwintering in Mexico
        Page 17
        Page 18
        Page 19
        Page 20
        Page 21
        Page 22
        Page 23
        Page 24
        Page 25
        Page 26
        Page 27
        Page 28
        Page 29
        Page 30
        Page 31
        Page 32
        Page 33
        Page 34
        Page 35
        Page 36
        Page 37
        Page 38
        Page 39
        Page 40
        Page 41
        Page 42
        Page 43
        Page 44
    Chapter 3. Use of lipid reserves by monarch butterflies overwintering in Mexico: Implications for conservation
        Page 45
        Page 46
        Page 47
        Page 48
        Page 49
        Page 50
        Page 51
        Page 52
        Page 53
        Page 54
        Page 55
        Page 56
        Page 57
        Page 58
        Page 59
        Page 60
        Page 61
        Page 62
        Page 63
        Page 64
        Page 65
        Page 66
        Page 67
        Page 68
        Page 69
        Page 70
        Page 71
        Page 72
        Page 73
        Page 74
        Page 75
        Page 76
        Page 77
        Page 78
    Chapter 4. Conservation implications of flowering plant availability for overwintering monarch butterflies in Mexico
        Page 79
        Page 80
        Page 81
        Page 82
        Page 83
        Page 84
        Page 85
        Page 86
        Page 87
        Page 88
        Page 89
        Page 90
        Page 91
        Page 92
    Chapter 5. Bird predation on overwintering monarch butterflies: Conservation implications
        Page 93
        Page 94
        Page 95
        Page 96
        Page 97
        Page 98
        Page 99
        Page 100
        Page 101
        Page 102
        Page 103
        Page 104
        Page 105
        Page 106
        Page 107
        Page 108
        Page 109
        Page 110
        Page 111
        Page 112
        Page 113
        Page 114
    Chapter 6. Mechanisms of cardiac glycoside loss as monarch butterflies age
        Page 115
        Page 116
        Page 117
        Page 118
        Page 119
        Page 120
        Page 121
        Page 122
        Page 123
        Page 124
        Page 125
        Page 126
        Page 127
        Page 128
        Page 129
        Page 130
        Page 131
        Page 132
        Page 133
        Page 134
    Chapter 7. Conclusions and future research
        Page 135
        Page 136
        Page 137
        Page 138
        Page 139
        Page 140
        Page 141
    Appendix. Lipid and lean contents in the annual cycle of the monarch butterfly
        Page 142
        Page 143
        Page 144
        Page 145
    Literature cited
        Page 146
        Page 147
        Page 148
        Page 149
        Page 150
        Page 151
        Page 152
        Page 153
        Page 154
        Page 155
        Page 156
        Page 157
        Page 158
        Page 159
        Page 160
        Page 161
        Page 162
        Page 163
    Biographical sketch
        Page 164
        Page 165
        Page 166
Full Text








BIOLOGY AND CONSERVATION OF OVERWINTERING MONARCH
BUTTERFLIES IN MEXICO








BY



ALFONSO ALONSO-MEJIA


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY


UNIVERSITY OF FLORIDA


1996



















Copyright 1996


by

Alfonso Alonso-Mejfa


















To my parents from whom I received the
principles that I use in my life



To my brother and sister with whom I shared the joy of being a kid


To my wife from whom I receive constant love,
support, and understanding



To Eneida and Eduardo from whom I learned the art of dedication to the
conservation of monarch butterflies












ACKNOWLEDGMENTS


Lincoln Brower has been a good advisor and supportive friend, not
only during the trials and tribulations of this dissertation, but also during
my M. S. research as well. I sincerely thank him for being my leader for
the past eight years of my life.
Crawford Holling, Douglas Levey, Thomas Emmel, and Thomas
Walker showed enthusiasm when I discussed my research ideas with
them and offered good advice when I needed it. They were excellent
committee members.
Very special thanks to Eneida Montesinos and Eduardo Rend6n for
their invaluable help, friendship, and support during the time we spent
hiking and working in the mountains of Central M~xico. I was very lucky
that their lives crossed mine at the time when they were searching for a
project for their bachelor's degree, and I was looking for field assistants.
This work could have not been done without their help.
My wife, Leeanne Tennant de Alonso, kindly revised every word that
was written for this dissertation. As you see, she improved my writing
with her critical comments and great knowledge of the English language.
I thank Leeanne for her time, patience and guidance. Her love of the study
of ants has inspired me to be a better biologist. We have formed a good team
for the study and conservation of arthropods. I thank Leeanne for these
and many projects that we will do in the future.







The Secretaria del Medio Ambiente, Recursos Naturales y Pesca
(Semarnap) gave me an official permit to work at the overwintering site in
the Sierra Chincua, Michoacdn. The personnel of the Instituto Nacional de
Ecologfa, especially Lucila Neyra and Mauricio Trejo, were extremely
helpful in facilitating many aspects of the field work. I also thank
members of the Ejido Cerro Prieto for giving me permission to work on
their land.
Many people facilitated my field work in M6xico. My parents met me
at the Laredo border and helped me obtain an import permit to take into
M6xico the microclimatic stations and La Paloma (a GMC, 1980 Suburban).
They also accompanied me from the U. S. Mexican border to the
overwintering sites and back again when I returned to Florida. Daniel
Pifieiro and Alicia Rojas of the Centro de Ecologia, UNAM, kindly sent me
an urgent letter to Laredo when I needed it the most. J. Manuel Maass and
the personnel of his laboratory at the Centro de Ecologia gave me logistical
support. Many thanks also to Dofia Cecilia del Rio for allowing the
transformation of one of the rooms of her house into a laboratory. She and
her friendly family made my stay very pleasant in Michoacdn. Abel Cruz,
Mario Dominguez, Eligio Garcia, Alejandro Mondrag6n, Martin
Mondrag6n, and Ruben Tellez also provided field assistance. My thanks to
them are warm and sincere.
The microclimatic data used in Chapter 2 required the constant
supervision of the machines and of the batteries that ran the stations.
Many thanks to Eneida, Eduardo, Eligio, and Mario for helping me on this
task. Thomas Walker helped in the design of the vaporimeters used in the
study. William Calvert collected monarchs in Texas, and Steve Malcolm
and Lincoln Brower gave me access to unpublished data used in Chapter 3.








This Chapter was also read by Natalia Arango, Ron Edwards, Ray Moranz,
and Jose Luis Osorno. Eduardo, Eneida, Eligio, and Mario also helped me
gather the data used in Chapters 4 and 5. Richard Kiltie, Jack Putz, and
Mark Yang gave me statistical advice to analyze data presented on Chapter
5. Many people were involved in the realization of Chapter 6. Eduardo,
Eneida, Eligio, and Mario help me with the collection of the fecal samples of
overwintering monarchs. My sister Guadalupe Alonso provided the tubes
where we stored the samples. Heather Howes kindly ran the samples on
the spectrophotometer. Natalia Arango reared the butterflies that I used in
one of the experiments. Leeanne spent many hours counting the number of
missing scales from the monarch wings. Many thanks to all of you.
Many people have made my stay in the United States very pleasant. I
thank all members of Dr. Brower's laboratory in 123 Bartram Hall for their
friendly company. They are Natalia Arango, Cristina Dockx, Shannon
Gibbs, Cara Gildar, Heather Howes, Amy Knight, Ray Moranz, Elizabeth
Rutkin, and Tonya Van Hook. I also thank Tonya, Ray, and Cristina for
the use of their CDs. Their fantastic music collection made my life much
easier during the long hours of laboratory analyses. My sincere thanks
also go to the personnel of the Department of Zoology, at the University of
Florida, for their unconditional support during my graduate education.
They provided me with teaching assistantships, support to attend scientific
meetings, as well as letters of reference when I borrowed equipment from
the Department.
Many thanks go to two professors that encouraged me to study the
fascinating ecology and behavior of butterflies. Jorge Sober6n pointed me in
such direction, and Lincoln Brower focused me on the study of monarchs.








Both of them have influenced my graduate education and professional
success in a very positive way.
Financial support was provided by the Biodiversity Support Program
(a consortium of the World Wildlife Fund, The Nature Conservancy and the
World Resources Institute with funding by the United States Agency for
International Development), The Wildlife Conservation Society, the Centro
de Ecologfa at the National University of M6xico, the Sistema Nacional de
Investigadores (M6xico), and the Department of Zoology at the University of
Florida. The laboratory research was supplied by NSF grant DEB 922091
with Lincoln Brower as principal investigator. The opinions expressed
herein are those of the author and do not necessarily reflect the views of the
U. S. Agency for International Development.












TABLE OF CONTENTS


ACKNOWLEDGMENTS .................................................................. iv

LIST OF TABLES ............................................................................... xi

LIST OF FIGURES ............................................................................ xii

A B ST R A C T ..................................................................................... xiv

CHAPTERS
1 INTRODUCTION AND BACKGROUND .............................. 1

Monarch Butterfly Migration ............................................... 1
Overwintering Biology ......................................................... 2
The Oyamel Fir Forest ......................................................... 5
Description .................................................................... 5
Forest Degradation ............................................................ 7
The Monarch Butterfly Special Biosphere Reserve .................... 9
T his Study ............................................................................ 12

2 MICROCLIMATIC DIFFERENCES BETWEEN CLOSED
AND OPENED FOREST AND THEIR CONSEQUENCES
FOR THE SURVIVAL OF MONARCH BUTTERFLIES
OVERWINTERING IN MEXICO .................................... 17

Introduction ...................................................................... 17
M eth ods ............................................................................... 20
S tu dy Site ......................................................................... 20
Microclimatic Conditions in Closed and Opened Areas ......... 20
Experiment 1. Estimates of Water Evaporation ................. 21
Experiment 2. Monarch Lipid Consumption in Closed
and Opened Areas ........................................................ 22
Statistical Analysis ........................................................... 23
R esu lts ................................................................................ 24
Microclimatic Conditions in Closed and Opened Areas ......... 24
Experiment 1. Estimates of Water Evaporation .................... 25
Experiment 2. Monarch Lipid Consumption in Closed
and Opened Areas ........................................................ 25
Discussion ....................................................................... 26








3 USE OF LIPID RESERVES BY MONARCH BUTTERFLIES
OVERWINTERING IN MEXICO: IMPLICATIONS
FOR CONSERVATION ................................................. 45

Introduction ..................................................................... 45
M ethods ........................................................................... 49
W inter Study Site .......................................................... 49
Collections of Inactive Monarchs Clustered on Trees ...... 50
Collections of Flower-visiting Monarchs ......................... 51
Collections of Migrating Monarchs ................................. 51
Estimates of Lipid Utilization by Clustered Monarchs ........... 53
Statistical Analysis ....................................................... 54
Results ..............................................................56
Clustered and Flower-visiting Monarch Comparisons ..... 56
Comparisons to Migrating Monarchs ............................. 57
Non-migratory, Reproductive Active Generations ................ 57
Estimates of Lipid Utilization During the 1993-94
Overwintering Period ............................................... 58
Rates of Lipid Utilization in Mxico, California and
A ustralia ................................................................ 59
Proportion of Clustered Monarchs That May Need to
Visit Flowers .......................................................... 60
D iscussion ........................................................ .......61
Clustered and Flower-visiting Monarch Comparisons ..... 61
Detrimental Effects of Logging on Monarch Survival ...... 64
Management Recommendations and Conclusions ............... 65

4 CONSERVATION IMPLICATIONS OF FLOWERING
PLANT AVAILABILITY FOR OVERWINTERING
MONARCH BUTTERFLIES IN MEXICO ............................ 79

Introduction ..................................................................... 79
M ethods ........................................................................... 81
Study Site ..................................................................... 81
Plant Species in Flower and Monarch Butterfly Use ............. 82
Determination of Forest Openness ...................................... 82
Statistical Analysis ....................................................... 83
R esults ............................................................................ 84
D iscussion ....................................................................... 85
5 BIRD PREDATION ON OVERWINTERING MONARCH
BUTTERFLIES: CONSERVATION IMPLICATIONS ........... 93

Introduction ......................................................................... 9
M ethods ........................................................................... 96
Study Site ..................................................................... 96
Determination of Closed and Opened Areas ..................... 97
Collections of Monarchs Preyed Upon by Birds ................. 98
Effects of Temperature on Bird Predation .......................... 100







Estimates of Monarch Density .......................................... 100
Comparisons of Clustered vs. Preyed upon Monarchs ....... 100
Statistical Analysis ......................................................... 101
R esu lts .............................................................................. 102
D iscu ssion ......................................................................... 103
Bird Predation in Closed and Opened Areas ...................... 103
Effects of Temperature on Bird Predation .......................... 106
Estimates of Monarch Mortality Due to Bird Predation ........ 107
Oriole and Grosbeak Predation ......................................... 108
Implications for Management of the MBSBR ..................... 109
6 MECHANISMS OF CARDIAC GLYCOSIDE LOSS AS
MONARCH BUTTERFLIES AGE ..................................... 115

Introduction ....................................................................... 115
M ethods ............................................................................. 117
Experiment 1. Study of CGs in Monarch Feces .................. 117
Experiment 2. CG Denaturation in Monarch Wings .......... 119
Esperiment 3. Scale Loss in the Wings of Reared
M onarchs ................................................................. 120
Statistical Analysis ......................................................... 121
R esu lts .............................................................................. 122
Experiment 1. Study of CGs in Monarch Feces .................. 122
Experiment 2. CG Denaturation in Monarch Wings .......... 122
Experiment 3. Scale Loss in the Wings of Reared
M onarch s ................................................................. 123
D iscussion ..................................................................... 123

7 CONCLUSIONS AND FUTURE RESEARCH ......................... 135

APPENDIX LIPID AND LEAN CONTENTS IN THE ANNUAL
CYCLE OF THE MONARCH BUTTERFLY ....................... 142

LITERATURE CITED ....................................................................... 146

BIOGRAPHICAL SKETCH ............................................................... 164












LIST OF TABLES


Table
1.1 Total area for the core and buffer zones for each of the five
protected overwintering sites of the Monarch Butterfly
Biosphere Reserve ................................................................ 16

2.1 Analysis of covariance (ANCOVA) comparing slopes and
y-intercepts of several microclimatic variables registered in
opened and closed areas ........................................................... 32

3.1 Coefficients of determination (r2) for correlations between
lipid mass, water content, lean mass and wing length data
for monarch butterflies ........................................................ 67

3.2 Comparisons between migrant monarchs and butterflies at
the overwintering sites in M 6xico ............................................. 68

4.1 Frequency of occurence of the 21 plant species found
flowering at the Sierra Chincua monarch butterfly colony in
March 1994 in relation to forest openness ............................... 89

5.1 Comparisons of bird predation in closed and open areas for
monarch butterflies overwintering at Sierra Chincua,
M ichoacdin, M 6xico ............................................................... 111

5.2 Sex ratios of female and male monarchs in closed and opened
areas at the Sierra Chincua monarch butterfly overwintering
site ...................................................................................... 112











LIST OF FIGURES


Figure page
2.1 Comparisons of average daily minimum and maximum
ambient temperatures recorded in closed and opened areas ......... 34
2.2 Comparisons of average daily minimum and maximum
percent relative humidity recorded in closed and opened
a rea s ..................................................................................... 36
2.3 Comparisons of average daily maximum wind speed
recorded in closed and opened areas ..................................... 38

2.4 Diurnal changes in the average wind velocity ........................4.... 0

2.5 Comparison of the amount of water evaporated (ml) in closed
and opened areas ................................................................ 42
2.6 Monarchs held in experimental enclosures consumed in
their lipid reserves at a faster rate when they were exposed to
the microclimatic conditions opened areas ............................. 44
3.1 Lipid mass frequency distributions of clustered and
flower-visiting monarchs during the overwintering period ..... 70
3.2 Mean lipid, water, and lean masses, and wing length of
clustered and flower-visiting monarch butterflies during 4
months of the overwintering season ...................................... 72

3.3 Lipid mass changes of 5211 monarch butterflies collected by
many researchers over several years at different times
during their annual life cycle ............................................... 74

3.4 Comparison between expected and observed lipid mass loss
for monarchs overwintering in M6xico ...................................... 76
3.5 Lipid indices for overwintering monarch butterflies from
Australia, California, and M6xico ......................................... 78

4.1 Diurnal range of minimum and maximum ambient
temperatures recorded in closed forest at the Sierra Chincua
overw intering site ...................................................................92








5.1 Comparisons of daily rates of bird predation (average number
of monarch butterflies killed/m2/day) in closed and opened
areas at the Llano del Toro monarch butterfly aggregation
during the 1993-1994 overwintering season ............................... 114

6.1 Cardiac glycoside concentration in monarch butterflies feces ..... 128
6.2 Cardiac glycoside concentration in monarch wings under
two experimental treatments as a function of time .................... 130
6.3 Relationship of the number of wing scales that adult
monarch butterflies lose while aging in a outdoor flight cage
in Gainesville, Florida ........................................................... 132

6.4 Relationship of the number of missing scales from monarch
butterflies wings and CG concentration ................................... 134


xiii













Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

BIOLOGY AND CONSERVATION OF OVERWINTERING MONARCH

BUTTERFLIES IN MEXICO

By

Alfonso Alonso-Mejia

May 1996

Chairman: Lincoln P. Brower
Major Department: Zoology
Millions of monarch butterflies (Danaus p1ipiu. overwinter in
M6xico in forests dominated by the Oyamel fir tree (Abies r.ligiQ.a).
Logging practices and clearing for agricultural fields represent major
threats to these forests. The Monarch Butterfly Special Biosphere Reserve
(MBSBR) was created in 1986 to protect the monarch butterfly. The reserve

experienced varying degrees of tree extractions before its creation such that
monarchs currently form their overwintering aggregations in forests that

contain both closed and opened areas. In an attempt to link ecological

research with conservation strategies for the MBSBR, I investigated how
forest microclimate and degree of disturbance affect the survival of
monarch butterflies in the Sierra Chincua, one of the most pristine sites

within the MBSBR. I defined closed and opened areas by using a new
method that considered tree density, total basal area, and forest overstory
density. Monarch butterflies that clustered in opened areas during the








1993-94 overwintering period experienced lower ambient temperatures
during the night, higher wind velocities, higher rates of water evaporation,
higher rates of lipid use, and higher rates of bird predation than monarchs
clustered in closed areas. I detected high rates of bird predation possibly
because monarch butterflies lose their chemical protection as they age. It
appears that scale loss and denaturation are the most important factors
explaining the decrease in cardiac glycoside concentration in adult
monarch butterflies. Contrary to recent suggestions that logging is needed
to provide more nectar sources for the monarchs, I found that the core area
already has many artificial openings. In addition, the species composition
of plants in flower is similar in artificially-opened and naturally closed
areas. My research shows that there is no need to create additional
artificial openings and that further thinning of the forest would be

extremely harmful for the monarchs since it would increase the risk of
mortality due to freezing, bird predation, and depletion of lipids needed for
the monarchs' return migration to the southern United States. Future
studies should address forest management plans and the improvement of
current agricultural practices that would provide alternatives to logging for
the local inhabitants.












CHAPTER 1
INTRODUCTION AND BACKGROUND




Monarch Butterfly Migration

The monarch has landed back in the land of
the free and the brave. M. Zalucki wrote on e-mail 1995


Monarch butterflies (Danaus l1ippu L.) are well-studied insects
with amazing life history traits. Monarch caterpillars feed exclusively on
milkweed plants (Asclepias spp.) from which they sequester chemical
compounds that are toxic to several vertebrate predators when ingested
(review in Brower 1984). The bright yellow and black colored stripes of the
caterpillar and the orange and black colors of the adult serve to advertise
the toxicity of the monarch. In addition to being one of the classical
examples of aposematic coloration, monarchs are exceptional as the only
insect species that performs long distance migrations over thousands of
kilometers (Brower 1985, 1995a).
According to Kitching et al. (1993), the subgenus Danaus probably
evolved in South America during the Pliocene from an Old World ancestor.
After colonizing Central America, subsequent monarch ancestors reached
North America and found a highly diverse milkweed flora (Woodson 1954).
During the Pleistocene, the alternating glacials and interglacials caused
contractions and expansions of the geographic ranges of both fir forest and
milkweed flora that may have caused migratory movements in monarch








butterflies (Graham 1973; Brower 1985, 1995a). Migration and aggregation
behaviors occur in other genera of the subfamily Danainae in the Old World

(Danaus (Salatura), Euploea, Ideopsis, Parantica, and Tirumala), and in
the New World (Anetia, Ivie et al. 1990; Wang and Emmel 1990;
Scheermeyer 1993). Furthermore, the high numbers of monarchs that now
migrate to Mexico seem to be the result of large scale deforestation that
occurred in eastern North America in the 1800's (Vane-Wright 1993).
According to this hypothesis, the distribution and abundance of the
milkweed plant Asclenias svriaca L. increased in newly cleared forest. The
higher availability of food increased the density of monarch butterflies
migrating to M~xico.




Overwinterin Biology


Each year during the autumn, the North American population of
monarch butterflies east of the Rocky Mountains migrates to several
mountain ranges in central Mexico. As their larval food source of
milkweed plants diminishes in North America at the end of the summer,
monarchs escape the cold northern winter and migrate to the cool, moist
environment of the Oyamel fir forest (Abies reliiosa H. B. K.). Monarchs
arrive in early November and form tightly packed aggregations of up to 10
million butterflies per hectare in several areas in the Transverse
Neovolcanic Belt in the states of Michoacdn and Mexico (Urquhart and
Urquhart 1978a, b; Brower 1985; Calvert and Brower 1986; Calvert et al.
1989; Calvert and Lawton 1993). There, they remain largely inactive,
clustered on tree trunks and tree branches, and maintain a state of








reproductive diapause until March when they migrate back to the southern
United States to exploit freshly emerging milkweed plants (Herman 1985;
Brower and Malcolm 1991; Malcolm et al. 1993).
In the high altitude Oyamel fir forests, monarchs find cool ambient
temperatures where they remain quiescent for most of the five months of
the overwintering period (Brower and Calvert 1985; Calvert et al. 1989).
This reduces the burning of the lipid reserves that are needed for their
return migration to the southern United States (Chaplin and Wells 1982;
Masters et al. 1988). Monarchs are periodically active when they are
dislodged from tree clusters by predatory birds and winter storms, or when
they fly out of the aggregation to drink water (Calvert and Cohen 1983;
Calvert et al. 1983; Brower and Calvert 1985; Alonso-M. et al. 1993; Calvert
1994). Downslope colony movement to more humid areas also occurs as the
dry season progresses (Calvert and Brower 1986; Calvert 1994). Lipid
reserves remaining at the end of the overwintering period are used for
migration and reproduction, and are probably supplemented by nectar
feeding along the migration route to the southern United States (Heitzman
1962; Brower 1985; Urquhart 1987).
These high concentrations of monarchs are prime targets for several
vertebrate predators. Monarch butterflies are a high quality resource for
predators because low ambient temperatures make them largely inactive,
and they posses large amounts of lipids (Brower 1985; Masters et al. 1988;
Calvert et al. 1989; Malcolm and Brower 1989). Moreover, two- to six-
month-old monarch butterflies overwintering in Mexico are poorly
protected chemically since the concentration of the chemicals that provide
them protection decreases with age (Alonso-M. and Brower 1994). Thus,
high rates of predation have been recorded at the overwintering sites in








M6xico (Calvert et al. 1979; Brower and Calvert 1985; Glendinning et al.
1988; Arellano et al. 1993). Black-backed orioles (Icterus grkula abeillei
Lesson) and black-headed grosbeaks (Pheucticus melanocephalus

Swainson) consumed close to 9% of the Sierra Chincua monarch
aggregation during the 1978-79 overwintering season (Brower and Calvert
1985). The scansorial black-eared mouse (Peromyscus melanotis J. A.
Allen and Chapman) consumed close to 5% of an aggregation that formed
at the same site in 1986 (Glendinning et al. 1988).
Most overwintering aggregations known to date form on the
southwest facing slopes of the Mexican mountains (Calvert and Brower
1986). Southwestern slopes are usually wetter than northern and eastern
slopes because moist rich air masses from the Pacific coast move into the
mountains during the winter (Mosifio-Alemdn and Garcia 1974; Calvert et
al. 1989). Monarchs also consistently form their aggregations at certain
altitudes. At the beginning of the period, they almost never aggregate below
3,100 m. As the aggregations move to more humid areas at the middle-to-
end of the overwintering period, they re-group at altitudes as low as 2,900 m
(Calvert and Brower 1986). The overwintering period overlaps with the dry
season which extends from November to April. The area receives more
than 1,000 mm of rain during the summer wet season (Rzedowski 1983).








The Oyamel Fir Forest

When the axe came into the forest,
the trees said: the handle is one of us!
(On a bumper sticker, 1995)


Description
Most overwintering aggregations of monarch butterflies form in
forests dominated by the Oyamel fir tree, an endemic species to the
mountains of Central M6xico (Rzedowski 1991). Oyamel fir forests are
relicts of more extensive boreal-like forests which advanced during the
glacial and interglacial periods of the Pleistocene (Graham 1973).
Currently, these forests have island-like distributions on mountain peaks at
elevations between 2,400 and 3,600 m where colder climates prevail and
other tree species such as Pinus L. (Pinaceae), Quercus L. (Fagaceae), and
Buddeia L. (Loganiaceae) occur at low densities (Loock 1947; Madrigal
1976; Manzanilla 1974; Rzedowski 1983). Cupressus lindleii Klotzsch
(Pinaceae) occurs in pure stands near the lower limits of the oyamel forest
(Calvert et al. 1989; Soto and Vazquez 1993).
The understory vegetation consists primarily of herbaceous and

bushy plant species in the Asteraceae (Senecio spp, E3atorium spp, Steia
spp) and Lamiaceae (Salvia spp), with a diverse assortment of ascomycetes,
basidiomycetes and bryophytes (Espejo et al. 1992). Ground cover includes
Acaene lonaa L., Alchemilla rm bens Rose (Rosaceae), and in some
areas a carpet of mosses including species in the genera Thuidium and
Mnium (Calvert et al. 1986). High altitude meadows (i.e. llanos) occur in
some flat areas where the drainage is restricted, the soils freeze, and the
vegetation is dominated by grasses (Potentilla candicans H. & B., Rosaceae)








and forbs. Llanos are usually bordered by the bush-sized Juninerus

monticola var. compacta Martinez (Pinaceae) and by Baccharis conferta H.
B. K. (Asteraceae, Snook 1993; Soto and Vazquez 1993). Attempts to reforest
these llanos with fir trees have failed, and should not be encouraged since
the habitat is not appropriate for tree survival (Snook 1993).
Oyamel trees have small needle-like leaves. This facilitates the
ability of monarchs to cluster close to one another. Moreover, the
architecture of fir trees allows them to support heavy loads of ice and snow
(Heinrich 1996) such that the branches of the Oyamel fir trees support large
numbers of monarch butterflies without breaking (Alonso-M. 1996).
Unfortunately, the ecology of the Oyamel fir tree has not been studied
in detail. For example, little is known about the appropriate field conditions
for seed germination and seedling establishment and survival. It is known
that fir trees can be infected by bark beetles (Scolytidae: Scolvtus hermosus,
S. mundus, a. ventralis Pitvohthorus blackmanii, and Pseudohvlesinus
~ Herrnndez and Cibridn 1981), mistletoes (Arceuthobium
abietis-reli-iosae, Rodriguez 1983), and periodic outbreaks of geometric
moths (Ejita hvalinaria blandaria Dyar Geometridae, Carbajal and L6pez
1987), and that any of these infections can lead to tree mortality (Snook
1993). However, further studies on the distribution, abundance, and
vertical and horizontal transmission of these diseases are needed.
Most Oyamel trees die standing by parasitism or by lightning. Thus,
small scale disturbances such as tree fall gaps are not observed in the area.
During the snow storm of 1981, Calvert et al. (1983) reported that one tree
was uprooted but such events are rare in the forest. Little is known of the
effects (positive or negative) of forest fires in Oyamel trees. According to
Guti6rrez (1983) most forest fires are set intentionally by cattlemen and








farmers to promote new spring grass growth for their animals and to clear
forested areas for planting as the agricultural frontier expands up the
mountains. The conversion of forest land to agricultural use is a key factor
in forest destruction, leading to soil erosion and poor productivity (Snook
1993).


Forest Degradation
Oyamel trees are extracted from the forest for commercial, industrial
and domestic purposes (Snook 1993). Commercial exploitation provides raw
materials for local sawmills and neighboring conglomerate board and
papermaking industries. Legal exploitations require the issuing of permits
for extraction and transportation of authorized volumes of wood and they
are carried out in accordance with the Mexican Method of Forest Control
(M6todo Mexicano de Ordenaci6n de Montes, Musalem 1979). This method
of forest management involves a low-intensity, periodic selective cutting of
35-40% of the volume of the desired tree species. It attempts to increase the
growth of the remaining trees, enhancing production, and regeneration by
reforestation. However, lumbering operations in the Oyamel forests have
not shown the expected results, since the area is not large enough to allow a
rotation that will permit forest regeneration. Moreover, logging
temporarily destroys large areas of the understory vegetation where trees
are felled and logs are transported to the trucks. The method also requires
reforestation practices. However, free-ranging livestock are commonly
found in the area and seem to have a negative effect on the survival of
oyamel fir tree seedlings (Calvert et al. 1989; Snook 1993).
According to Snook (1993), the uncontrolled timber cutting for
domestic purposes can have an even higher negative effect on the








degradation of the forest. She argues that significant quantities of wood are
extracted for construction materials (beams and shingles "tejamanil") and
fuelwood (for cooking and heating purposes) by the large populations of
local peasants that live in the surrounding areas near the overwintering
sites (i.e. about 15 000 people live in the Sierra Chincua, Campanario, and

Chivati-Huacal region). She estimated that approximately 75,000 m3/yr, or
about 40,000 fir trees of average size (1-3 m3), are taken each year. Most of
the local peasant population lives at a subsistence level deriving their food
from agriculture and grazing, their fuel and construction materials from
the forest, and their income from the sale of agricultural goods and wood
products (Chapela and Barkin 1995). The amount, origin and destination of
wood taken by each ejido for domestic purposes needs to be evaluated. The
rate of population growth of these communities also needs to be studied and
considered in future forest planning. Emigration to Mexico city seems to be
high but needs to be quantified (Chapela and Barkin 1995).
In addition to logging, local inhabitants obtain a limited number of
non-timber products from the oyamel forests. These include the collection
of flowers for religious rituals, herb plants for medicinal purposes (e.g. "te
de monte" from Satureja macrostema, Lamiaceae), the extraction of resin
from pine trees, and the harvesting of mushrooms during the rainy season.
The restricted distribution of the Oyamel forest to high altitude
mountain peaks, and the increasing pressure from logging and clearing
for agricultural fields make it more vulnerable to deforestation than any
other forest type in M6xico (Calvert et al. 1989; Snook 1993). The
degradation of the forest endangers the migratory phenomenon of the
monarch butterfly because the method of selective cutting of trees to
maximize timber production reduces the density and canopy coverage of the








forest needed to insulate the butterflies against extreme cold temperatures
that occasional winter storms bring to the area (Calvert et al. 1983; Wells et
al. 1983; Brower and Malcolm 1991; Culotta 1992; Anderson and Brower
1996; E. Rend6n unpublished data). Thus, a reserve was created to protect
the forest of five areas from logging where monarchs overwinter.




The Monarch Butterfly Special Biosphere Reserve


The overwintering sites in Mexico were first discovered by F. A.
Urquhart and colleagues after several decades of research on monarch
migration (Urquhart 1976; Urquhart and Urquhart 1976). Concerned about
the preservation of the spectacular migratory phenomenon, the Urquharts
decided not to share the exact location of their findings (Brower 1995a).
Despite this, the publication of popular and scientific articles on monarchs
overwintering in M6xico soon made the sites known (Urquhart 1976;
Brower 1977, 1985; Barthelemy 1978; Urquhart and Urquhart 1978a, b;
Calvert et al. 1979; Calvert and Brower 1986; review in Brower 1995a).
W. Calvert and L. Brower, from the University of Florida, conducted
pioneering research on the biology of monarch butterflies overwintering in
Mexico. In collaboration with Leonila Vazquez and Hector Perez,
professors of entomology at the National University of Mexico, Calvert and
Brower soon learned that the Oyamel fir trees were being commercially
exploited at a fast rate. They determined that the survival of overwintering
monarchs was closely related to the microclimate registered in closed
canopy forests, such that creating opened areas by logging enhanced








monarch mortality (Calvert and Brower 1981; Calvert and Cohen 1983;
Calvert et al. 1982, 1983).
It was not until 1986 that five forested mountain tops in the States of
Michoacin (Cerro Altamirano, Sierra Chincua, Sierra Campanario, and

Cerros Chivati-Huacal) and M6xico (Cerro Pel6n) were designated as
protected through the creation of the Monarch Butterfly Special Biosphere
Reserve (MBSBR) by the signing of a decree by Mexican President Miguel de
la Madrid (Diario Oficial 1986). Further investigations by Calvert
determined that monarchs form an additional four to seven overwintering
areas depending on the year (Calvert and Lawton 1993). These include
Cerro San Andrds, Mil Cumbres, Cerro Picacho in the State of Michoacdn,
and Cerro las Palomas, Oxtotilpan, Cerro Piedras Chinas, and Piedra
Herrada, in the State of Mdxico. Most of these overwintering aggregations
form along an arc stretching from the western slopes of Volcano Nevado de
Toluca in the state of Mdxico, northwest to the city of Zitdcuaro and north to
the Altamirano mountain in the state of Michoacdn (Figure 2 in Calvert
and Brower 1986; Calvert et al. 1989; Calvert and Lawton 1993). The total
length of the arc from Palomas, the most easterly site, to Altamirano, the
most northerly aggregation, is approximately 150 km.
The MBSBR was classified as special because of the relative small
area that it protects (Halffter 1984). It includes 16,110 ha of which 11,600 ha
are classified as buffer zones where forest extractions are permitted (Diario
Oficial 1986; Table 1.1). Therefore, logging-free wilderness areas consist
only of 4,500 ha. Core areas were created for protection of the animal and
plant species found in the relict Oyamel fir forests, and to serve as sources
of species to recolonize logged areas in the buffer zones. In the Chincua
area, 700 ha were purchased by the federal government and 80 ha were








expropriated by the State of Michoacdn for protection. The remaining land,
nearly 15,500 ha, is communally owned land (called ejidos) granted to more
than 30 groups of organized peasants. Before 1994, when President Carlos
Salinas modified the 27th article of the Mexican Constitution (Diario Oficial
1994), ejidos could not be bought, sold or transferred. With the new law,
ejidos can be divided and each ejidatario (owner), can sell his part.
However, the decision to divide an ejido has to be accepted by the majority of
the ejidatarios. Ejidos affected by the MBSBR have kept the same
organization as if the law had not been changed. They democratically
select a leader (comisario) and a treasurer every two years. Each can be re-
elected only once.
The different areas of the MBSBR experienced varying degrees of tree

extractions before the creation of the reserve such that monarchs currently
form their overwintering aggregations in forests that contain both opened
and closed areas. Calvert et al. (1989) found the tree density in the
overwintering sites ranges from 90 to 620 trees/ha and the basal areas
varies from 12.1 to 43.8 m2/ha. They compared these data to previous
studies on Oyamel forests in Mexico (Madrigal 1967; Manzanilla 1974), and
to forest stands of Abies amabilis (Dougl.) Forbes in the Cascade Mountains
of the northwest, and A. balsamea (L.) Mill. in the Adirondack Mountains
from the northeastern United States (Grier et al. 1981; Sprugel 1984). They
found that the forest in these previous studies had higher densities in
younger and mature stands, and higher, age-related basal areas than the
Mexican fir forests. They concluded that with the possible exception of the
Chincua and Herrada overwintering sites, the Oyamel fir forest stands
studied have being heavily exploited.








Core areas of the monarch reserve are already subject to numerous
human activities. Out of the 4,500 logging-free hectares of the reserve, 1,000
are being used as centers for ecotourism, one of the major alternative
incomes to ejidatarios (Campanario and Peldn). Fifteen hundred hectares
have been illegally logged and the remaining 2,000 ha have a mosaic of
closed and open forest patches where most of the scientific research on the
migrating monarchs and on the oyamel forest is conducted (A. Alonso-M.
unpublished data). Since the core areas the MBSBR are small, for them to
serve as sources of plant and animal species for recolonization of the 11,600
ha of the buffer zone, they should be increased in size and maintained as
logging-free areas. Ideally, unprotected overwintering aggregations

should be incorporated into the MBSBR so that the risk of extinction of the
endangered phenomenon of the monarch butterfly migration will decrease.




This Study


In an attempt to link ecological research with conservation strategies

for the MBSBR, I investigated how forest microclimate and degree of
disturbance affect the survival of monarch butterflies in the Sierra
Chincua, one of the most pristine overwintering sites in Mdxico (Calvert et
al. 1989; Calvert and Lawton 1993). I first designed an objective and
practical method to determine closed and open areas in these forests. I
then monitored daily changes in temperature, humidity, and wind speed in
closed and opened areas during most of the 1993-1994 overwintering season.
I also experimentally compared the rates of water evaporation and the rate
of lipid consumption for monarchs that were experimentally exposed to








closed and opened climatic conditions (Chapter 2). I found that monarchs
in opened areas experienced lower ambient temperatures during the night,
higher ambient temperatures during the day, higher wind velocities, and
higher rates of water evaporation. This indicates that butterflies clustering
in opened areas have a higher risk of freezing mortality, higher rates of
dehydration, and higher rates of lipid use. The higher rate of lipid loss in
monarchs clustered in opened areas is consistent with the hypothesis that
intact, closed forest is necessary for successful overwintering because this
permits them to conserve their lipid reserves for the spring migration back
to the United States. In fact, the San Andres overwintering monarch
aggregation that forms outside the limits of the MBSBR protected area (map
in Calvert and Brower, 1986), and the highly disturbed Altamirano and
Chivati-Huacal monarch aggregations have become smaller, ephemeral,
unstable and in several years have failed to form (W. H. Calvert,
unpublished data). These changes seem to be in response to low tree
densities resulting from current and past forest extractions.
I followed how inactive monarchs clustered on trees utilized their
lipid reserves throughout the overwintering period. I compared their rate
of lipid use to monarchs that were collected while visiting flowers in the
surroundings of the overwintering aggregation. I also compared lipid
amounts in autumn migrants collected in Texas, spring migrants collected
in the southern United States, and non-migratory, reproductively active
summer generations collected in Wisconsin and Minnesota (Chapter 3). I
found that clustered butterflies had significantly higher lipid mass, water
content, lean mass, and larger wings than did monarchs collected from
flowers. These differences were consistent throughout the overwintering
period. A high proportion of flower-visiting monarchs had lipid mass close








to zero, and very few of the butterflies had medium to high lipid levels. This

data suggests that flower-visiting monarchs may be a derived group from
the clustered monarchs and may represent a continual recruitment of
those individuals that are approaching starvation.
Despite the mounting evidence indicating that logging is detrimental
to the survival of overwintering monarchs (Calvert and Brower 1981;
Calvert et al. 1982, 1986, 1989; Calvert and Cohen 1983; Brower and Malcolm
1991; Alonso-M. et al. 1992; Snook 1993; Anderson and Brower 1996), Hoth
(1993) and Chapela and Barkin (1995) argue that logging should be
permitted in the core areas of the reserve. Their argument to justify
logging is that tree extraction would benefit monarchs by creating forest
openings in which more plants would flower. This reasoning maintains
that the increased availability of nectar could translate into fewer
monarchs depleting their lipid contents, and therefore, more monarchs
surviving the overwintering period. In Chapter 4, I looked at the frequency
of opened and closed areas and the distribution of flowering plants in both
closed and opened areas. I found that the core area of the Sierra Chincua
MBSBR already has a substantial number of areas with human-induced
disturbances, that the plants in flower were common during the winter at
this site, and that natural and artificially-opened areas had the same plant
species composition.
In order to investigate how increased logging may impact monarch
survival in the MBSBR, I also studied how monarch butterfly mortality
caused by bird predation is currently affected by logging extractions that
occurred in the early 1980s (Chapter 5). I monitored daily bird predation in
closed and opened areas and found that monarch butterflies overwintering








in areas with low tree density, low basal area, and low canopy coverage had
higher rates of bird predation than monarchs in areas with closed forest.
I also investigated three mechanisms by which adult monarchs may
lose their toxic cardiac glycoside as they age, since this may help explain
the high rates of bird predation observed at the overwintering sites (Chapter
6). Based on the data from a series of experiments, it appears that scale
loss and denaturation of the toxic compounds are the most important
factors explaining cardiac glycoside decrease in adult monarch butterflies.
The observed decrease in cardiac glycoside concentration results in a lower
level of chemical defense for adult monarch butterflies as they age. Since
monarchs are more active in opened areas, they may age and lose scales
faster than those in closed areas.
These new data provide further evidence against issuing logging
permits in the core areas of the MBSBR because logging opens the forest
and an opened forest is detrimental for monarch survival. A very
important aspect of the conservation efforts relates to the socioeconomic
situation of the local inhabitants. In Chapter 7, I discussed alternative
sources of forest and agricultural management that could provide
alternatives sources of income to the local people.








Table 1.1. Total area for the core and buffer zones for each of the five
protected overwintering sites of the Monarch Butterfly Biosphere Reserve.
Data from Diario Oficial (1986).


CORE ZONE


BUFFER ZONE


Altamirano

Chincua

Campanario

Chivati-Huacal

Pelon


TOTAL 4490 ha 11619 ha 16110 ha


TOTAL


244


1060


900

940

687


1133

1635

988

1074

6787


1377

2695

1888

2014

7474












CHAPTER 2
MICROCLIMATIC DIFFERENCES BETWEEN CLOSED AND OPENED
FOREST AND THEIR CONSEQUENCES FOR THE SURVIVAL OF
MONARCH BUTTERFLIES OVERWINTERING IN MEXICO



Introduction


Microclimate is a major factor in the survival of monarch butterflies

(Danaus plexippus L.) overwintering in M6xico. Every year, monarchs
migrate from eastern North America to high altitude, cool, humid Oyamel
fir (Abie relip-iosa H. B. K.) forests in the mountains of central Mexico
(Mosifio-Alemdn and Garcia 1974; Brower 1985; Calvert and Brower 1986;
Calvert et al. 1989). Cool ambient temperatures benefit monarchs in at least
two ways. Since temperatures during the day are usually below flight
threshold (Masters et al. 1988; Alonso-M. et al. 1993), the majority of
butterflies remain quiescent for most of the five month overwintering
period. This reduces burning of their lipid reserves, which are needed for
the return migration to the southern United States (Chaplin and Wells
1982; Masters et al. 1988). Monarchs are periodically active during their
overwintering period when dislodged from tree clusters by predatory birds
and winter storms and when they fly out of the aggregation to drink water
(Calvert and Cohen 1983; Calvert et al. 1983; Brower and Calvert 1985;
Alonso-M. et al. 1993; Calvert 1994). Furthermore, at the low temperatures
prevailing in the overwintering aggregations, monarchs remain in a
physiological state of reproductive diapause that represses the maturation








of their gonads, controlling lipid utilization and aging (Barker and Herman
1976; Dallman and Herman 1978; Lessman and Herman 1983; Herman
1985; Herman et al. 1989).
In 1986, five limited areas of the Oyamel fir forest where monarchs
overwinter were designated as protected through the creation of the
Monarch Butterfly Special Biosphere Reserve (MBSBR; Diario Oficial 1986).
These protected areas had experienced different degrees of tree extraction

before the creation of the reserve so that monarchs that form their
overwintering aggregations there do so in both closed and opened areas of
forest. Calvert et al. (1982) reported that colder temperatures during the
night were registered in areas that had been thinned by logging, and
suggested that monarchs perched in opened areas would have higher risks
of freezing mortality. Overwintering monarchs have the capacity to
acclimate rapidly to low ambient temperatures (i.e. cold-hardening) and to
survive temperatures moderately below freezing. However, when the
ambient temperature drops below the subzero temperature at which
spontaneous tissue freezing occurs (i.e. the supercooling point), the
butterflies die since they are freeze-susceptible insects (Salt 1961; Lee 1989;
Lee at al. 1987; Alonso-M. et al 1992; Larsen and Lee 1994; Anderson and
Brower 1996).
During a snow storm in January 1981, Calvert et al. (1983) found that
42% of a monarch aggregation perished in the Sierra Chincua, Michoacdn
(see map in Calvert and Brower 1986). In 1992, Brower (in Culotta 1992)
reported that 83% of the Herrada overwintering aggregation in the State of
M6xico were killed by freezing. In Michoac~n, 6% of the Mojonera Alta
aggregation perished at the end of 1995 after a snow storm hit the
mountains (E. Rend6n unpublished data). On these occasions, monarchs








that remained dry survived ambient temperatures as low as -8C. Wet
monarchs, however, died at ambient temperatures of -40C, when ice
crystals formed on their body and triggered internal ice nucleation via their
spiracles (Salt 1961; Alonso-M. et al. 1992; Larsen and Lee 1994; Anderson
and Brower 1993, 1996). Anderson and Brower (1996) have recently shown
that monarchs perched in areas of closed forest and those inside and on the
bottom of bough clusters are better protected from accumulating water on
the surface of their bodies than are those on the outside of a cluster and
those perched in opened areas. Monarchs may select perching sites within
the overwintering aggregation where they can remain dry.
In this chapter, I extended the study by Calvert et al. (1982) by closely
monitoring microclimatic conditions in well-defined closed and opened
areas within a monarch aggregation over the entire overwintering period.
I first used an objective method to locate closed and opened areas within the
forest. I then monitored and compared daily changes in temperature,
humidity, and wind speed in these areas during most of the overwintering
period. I hypothesized that closed areas would be more buffered (i.e.
narrow range in variation) than opened areas.
The overwintering period of monarch butterflies overlaps with the
dry season of the area (November to April, Rzedowski 1983; Calvert et al.
1989). Under such dry conditions, it may be important for monarchs to
select perches that minimize body water loss. I designed an experiment to
test if evaporation was higher in opened than in closed areas. In a second
experiment, I tested if monarchs clustered in opened areas consumed their
lipid reserves at higher rates than those in closed areas. I predicted that
since monarchs perched in opened areas would be exposed to more
sunlight, they could experience higher body temperatures and exhibit a








higher rate of lipid consumption (Chaplin and Wells 1982; Masters et al.
1988). Data gathered in this study will help us better understand the
relation between the structure of the oyamel fir forest and the biological
needs of monarch butterflies overwintering in M6xico.



Methods


Study Site

This study was conducted at the Llano del Toro overwintering
monarch aggregation, located on the southwestern facing slope of Zapatero
Canyon at the Sierra Chincua, a mountain range in the Transverse
Neovolcanic Belt in northeastern Michoacdn (1940'48"N and 10017'54"W,
Anonymous 1976). This overwintering aggregation has formed almost
every year since the overwintering sites were discovered in 1975 (Urquhart
and Urquhart 1976; Calvert and Brower 1986; Calvert et al. 1989).


Microclimatic Conditions in Closed and Opened Areas
I used on-site weather loggers (OWL, Electronically Monitored
Ecosystems, 2229 Fifth St., Berkeley, CA 94710) to record ambient

temperature (C), relative humidity (%), and wind velocity (m/sec) from
December 12, 1993 to March 24, 1994. OWLs were programmed to record
measurements every minute and to save the average of those
measurements over two hr periods. I had four recording stations, two in
closed and two in opened areas. Each recording station consisted of three
temperature probes, three relative humidity probes, and three
anemometers. All probes were placed near clustered butterflies at three








meters above the ground. One of the stations in a closed area did not have
an anemometer.
I determined if an area was closed or opened based on a
Closed/Opened Index. Six parallel transects were established and each
was subdivided into 14 contiguous 10 X 10 m quadrats. Tree density
(number of trees/quadrat), total basal area (based on the diameter at breast
height for all trees found per quadrat), and forest overstory density (based
on spherical densiometer readings, Lemmon 1957) were estimated for each
quadrat. Using data from all quadrats, I obtained a relative value based on
the maximum datum recorded for each variable. The three relative values
estimated for each quadrat were added to obtain a quadrat-index value. The
average of all quadrat-index values was used to obtain a Closed/Opened
Index (average = 56.9, S.E. = 1.39, range = 32.7-97.0, n = 75). Quadrats with
higher quadrat-index values than the Closed/Opened Index were classified
as closed, while quadrats with lower quadrat-index values than the
Closed/Opened Index were classified as opened. Using this method, I
classified 39 of the quadrats as closed and 36 as opened. Nine of the 84
quadrats were excluded from the analysis because they occurred on wide
hiking trails or abandoned logging roads. I randomly selected two closed
(Closed/Opened Indices = 85.5 and 71.2) and two opened quadrats
(Closed/Opened Indices = 50.3 and 49.3) in which to set the recording
stations.


Experiment 1. Estimates of Water Evaporation
I compared rates of water evaporation in closed and opened areas by
estimating the amount of distilled water that evaporated from plastic tubes
(8.5 cm height X 3 cm diameter). I put 40 ml of distilled water in each tube








and sealed it with a one-hole rubber stopper. A dental cotton wick (15 cm
long X 1 cm diameter) extended from the bottom of the tube, to about four
cm above the rubber stopper. A small plastic net covered the cotton to
prevent insects from drinking the water. The tubes were hung three
meters above the ground, in the shade, near clustering monarchs. I set
one tube in each of 10 closed quadrats and 10 opened quadrats. The rate of
water evaporation (in ml/week) was measured from 20 December, 1993, to
24 March, 1994 (n = 13 weeks).


Ex eriment 2. Monarch Lipid Consumption in Closed and Opened Areas
I performed an experiment with netted butterflies to determine if
monarchs had higher rates of lipid use when they were exposed to
conditions of opened areas. On February 2, 1986, I collected 500 monarchs
from three clusters that were three to four meters above the ground. I
estimated the wet mass of each butterfly to the nearest milligram using a
Sartorius 1205 MP electronic balance and selected 240 butterflies that
weighed between 500 650 mg. Forty butterflies (20 females and 20 males)
were collected as controls and were frozen the same day I started the
experiment for subsequent lipid analysis. Fifty butterflies (25 females and
25 males) were placed in each of four cylindrical nets (120 cm length X 30
cm diameter). Two nets were placed in closed and two in opened areas and
hung two and a half meters above the ground. Here, opened areas
consisted of sites that had no standing trees within a 200 m2 area (approx.
15 X 15 m), while the closed areas had standing trees that prevented
sunlight from reaching the understory vegetation.
All butterflies were fed water every other day by extending their
probosces into a wet sponge. On February 22, I removed five of each sex








from each net (n = 40) and froze them. On March 8, I removed and froze
eight females and seven males from each net (n = 60). I ended the
experiment on March 23 when I collected and froze all individuals that
were still alive (32 females and 21 males). Monarchs were placed in
individual glassine envelopes and stored in a freezer at -4C until lipid
analysis was performed. I recorded ambient temperatures in both areas
with Max-Min Taylor thermometers. Maximum ambient temperatures in
closed areas ranged from 8 17'C, while in opened areas they ranged from
10 250C.
I extracted lipids using the method of Walford (1980) and May (1992).
Each monarch was dried at 60'C for 16 hr, weighed, and ground in a
centrifuge tube in 20 ml petroleum ether with a Janke & Kunkel SDT Ultra
Turrax tissuemizer. Each sample was then placed in a medium speed
shaker bath at 35C for 30 min, with vortexing every 10 min. Tubes were
centrifuged in a Dynac Clay Adams centrifuge at 1000 RPM for seven
minutes, and the supernatant was decanted into preweighed aluminum
pans on a 30'C hot plate. I added 15 ml of ether to the remaining solids in
the centrifuge tube, and repeated the procedure described above. The
weighing pans with the supernatant were allowed to evaporate overnight. I
then weighed the remaining residue (mg), which was composed mainly of
lipids.


Statistical Analyses
I compared the slopes and the y-intercepts of the regression lines of
the sampling date (covariant) against the minimum ambient temperatures
(response variable) registered in closed and opened quadrats (nominal
variables) by using analysis of covariance (ANCOVA; Zar 1984). I used a







similar analysis to test the effects of the maximum ambient temperatures,
the minimum and maximum relative humidities (arcsin transformed),
and the maximum wind velocity. I used a repeated measures ANOVA to
test if water evaporation was higher in opened areas. I used two-way
ANOVA (treatment X date) to compare monarch lipid consumption in
closed and opened areas.




Results


Microclimatic Conditions in Closed and Opened Areas
As the overwintering season progressed, ambient temperatures
increased in both closed and opened areas. The lowest temperatures
occurred at 0600 0800 hr, and maximum temperatures at 1600 1800 hr in
both areas. Minimum temperatures registered in opened areas were
consistently lower during the night than those in closed areas (Table 2.1;
Fig. 2.1). The average daily minimum temperature registered in opened
areas was 3.8C (S.E. = 0.14, n = 96), compared to 4.90C in closed areas (S.E.
= 0.14, n = 103). Contrary to my expectations, the average daily maximum
temperature of 12.91C (S.E. = 0.26, n = 95) in opened areas did not differ
from that in closed areas (average = 12.31C, S.E. = 0.22, n = 103). Likewise,

average daily minimum relative humidity (opened: 55.7%, S.E. = 1.25, n =
75; closed: 55.6%, S.E. = 1.44, n = 83), and maximum relative humidity
(opened: 84.3%, S.E. = 0.81, n = 75; closed: 83.4%, S.E. = 1.02, n = 83) did not
differ significantly between closed and opened areas (p > 0.05). Relative
humidities remained high during most of the overwintering period (Fig.
2.2). I did not find correlations between minimum or maximum








temperatures and minimum or maximum relative humidities (Coefficients
of determination, r2, p > 0.05).
Wind velocity increased significantly throughout the overwintering
period in both closed and opened areas (Fig. 2.3). In closed areas, wind
velocity increased from an average maximum of 0.91 m/s (S.E. = 0.06, n = 20
days) at the beginning of the season to 2.01 m/s (S.E. = 0.12, n = 20) at the
end. Opened areas increased from 1.76 m/s (S.E. = 0.14, n = 20) to 2.60 m/s
(S.E. = 0.07, n = 20). As expected, wind speed was consistently higher in
opened areas throughout the season (Fig. 2.3; Table 2.1). During most of
the season, wind was consistently stronger between 1200 1800 hr in both
closed and opened areas (Fig. 2.4).


Experiment 1. Estimates of Water Evaporation
The rate of water evaporation significantly increased as the
overwintering season progressed both in closed (ANOVA F(i, 128) = 16.5, p
< 0.001) and opened areas (ANOVA F(1, 128) = 68.9, p < 0.001; Fig. 2.5).
However, the overall rate of water loss was significantly higher in opened
areas (Repeated Measures ANOVA F(i, 216).= 5.99, p < 0.025). There was

high variability in the amount of water that evaporated each week in both
closed and opened areas. In opened areas, an average of 24.7 ml of distilled
water evaporated per week (S.E. = 0.72, n = 130), while 21.9 ml evaporated

per week in closed areas (S.E. = 0.66, n = 130).


Experiment 2. Monarch Lipid Consumption in Closed and Onened Areas
At the beginning of the experiment, monarchs had an average lipid
mass of 68 mg (S.E. = 5.3, n = 40). Using a two-way ANOVA, I found that
both treatment (ANOVA F(1,145) = 37.2, p < 0.001) and date (F(1,145) = 11.8,








p < 0.001) had a significant effect on lipid content. I used the Ryan's Q test
for multiple comparisons (Day and Quinn 1989) to compare the average
lipid mass of butterflies in opened and closed areas on each date. I found
that by March 8 and March 23, monarchs held in closed areas had
significantly higher lipid contents than monarchs held in opened areas
(Fig. 2.6). After six weeks, monarchs in opened areas had lost an average
of 56.8 mg (S.E. = 2.3, n = 26, 83.5%) of their initial lipid weight, while
monarchs in closed areas only lost 27.6 mg (S.E. = 5.0, n = 27, 40.6%).




Discussion


In this study, monarch butterflies overwintering in closed areas
within the Oyamel fir forest at the Sierra Chincua found cool ambient
temperatures, high relative humidities, and low wind velocities. These
microclimatic conditions helped monarchs to remain quiescent for most of
the overwintering period, maintaining low rates of lipid and water loss.
Monarchs clustered in opened areas within the forest, however,
experienced lower ambient temperatures during the night and higher wind
intensities during the day. I also recorded higher rates of water
evaporation in opened areas. Thus, monarchs clustered in opened sites
had a higher risk of freezing mortality (Calvert et al 1983, Anderson and
Brower 1996), they could have had higher rates of water loss, and were
likely to deplete their lipid reserves before March, when they need them to
migrate back to the United States. Monarchs may seek perching sites in
areas of closed forest with better climatic conditions. In Mexico, about 10
million monarch butterflies aggregate in less than 500 trees per hectare







(Calvert and Brower 1986, see chapter 5). At such high densities, preferred
sites may be filled quickly, forcing many monarchs to perch in more opened
areas. By enhancing Oyamel fir seedling regeneration and the preventing
of illegal logging, opened areas would eventually become more closed,
providing preferable microclimatic conditions and reducing monarch
mortality.
Lower temperatures were registered in opened areas during the
night than in closed areas. In general, objects cool at night because they
lose heat by radiation to the atmosphere and by convection to surrounding
air and other objects (Geiger 1965; Calvert and Brower 1981; Calvert et al.
1982, 1986; Anderson and Brower 1996). In a closed forest, the radiation,
absorption, and re-radiation of heat from ground to foliage, foliage to
foliage, and foliage to ground reduce the amount of heat lost to the
atmosphere, and result in warmer temperatures. In contrast, opened
areas have fewer plants from which heat can be re-radiated and more gaps
through which infrared radiation escapes to the open sky. During cloudy
or foggy nights, the water present in the atmosphere absorbs and radiates
heat back to the emitting objects, retarding radiation loss to the night sky,
and thus maintaining similar ambient temperatures in closed and opened
areas (Geiger 1965; Calvert et al. 1982). Overwintering monarch butterflies
cluster on the middle and lower portions of the trees and avoid the canopy
where, in a closed forest, most of the radiational cooling occurs. Monarchs
thus may cluster very close to each other not only to reduce bird predation
(Brower and Calvert 1985; Arellano et al. 1993; Chapter 5), or to reduce the
risk of getting wet during a storm (Anderson and Brower 1996), but also to
reduce heat loss during the night.







Previous field studies on the cryobiology of overwintering monarchs
butterflies in Mexico have used naturally opened areas (i.e. meadows)
where high rates of radiational cooling occur and where unusually cold

temperatures have been recorded (Calvert and Brower 1981; Calvert and
Cohen 1983; Calvert et al. 1986; Alonso-M. et al. 1992). However, during the
1981 snow storm, Calvert et al. (1983) registered average ambient
temperatures of -4.1C, with readings as low as -5.0'C at several locations
within the Oyamel fir forest. Since many monarchs became wet during the
rainy period prior to the snow storm, their supercooling capacity was
greatly reduced. Their supercooling point was raised to -4'C from -8'C (the
temperature at which spontaneous tissue freezing occurs). This caused an
estimated mortality of 2.5 million monarchs. Calvert et al. (1982) also
showed that thinned logged areas were 1.190C colder than areas of closed
forest (compared to a 1.18'C temperature difference in my study). I propose
that during the 1981 and other snow storms (Calvert et al. 1983; Culotta
1992; E. Rend6n-S. unpublished data), monarchs clustered in closed areas
were better protected against freezing mortality than monarchs clustered
in opened areas. I hypothesize that a single, warmer degree in the ambient
temperature, and the reduced accumulation of water on the body of
monarchs clustered in closed areas (Anderson and Brower 1996), could
greatly decrease the susceptibility of monarchs to death by freezing.
Maximum ambient temperatures recorded in the shade did not differ
between closed and opened areas (Fig. 2.1). I found, however, that after six
weeks of experimentally holding monarch butterflies in opened areas, they
consumed 83.5% of their initial amount of lipids. Monarchs held in closed
areas only consumed 40.6% (Fig. 2.6). Adult monarchs are well suited to
gain heat such that they can rapidly bring up their thoracic temperature







well above the ambient (Masters et al. 1988). Their relatively large body

size, their coat of dark hair-like scales on the thorax, and the large wing
mass near the thorax facilitates rapid heating (Church 1960; Douglas 1978;
Kingsolver and Koehl 1985; Masters et al. 1988; Masters 1993). Douglas
(1978) showed that monarchs in resting position (i.e. with their wings
closed) increased their body temperature 100C above the ambient when
exposed to a source of light in the laboratory. Thus, the morphological
characteristics that usually favor activity at cooler temperatures caused
monarchs perched in opened areas to increase their body temperature and
consequently increase their rate of lipid use (Chaplin and Wells 1982;
Masters et al. 1988; Masters 1993). These results agree with published
observations by Leong (1990), who found that overwintering monarchs in
California avoid trees that have direct sun exposure and bright
illumination.
Results from this study strongly suggest that wind velocity is an
important environmental factor for monarch butterflies, as has been
suggested for monarchs overwintering in California (Leong 1990; Leong et
al. 1991). I found that monarchs clustered in opened areas were
consistently exposed to stronger winds than monarchs in dosed areas (Fig.
2.3). In addition, the highest wind velocities were recorded at the time of
maximum dryness during the day (Fig. 2.4). Thus, the faster movement of
dry air in the opened areas may explain the faster rates of water
evaporation that I recorded in the evaporation experiment (Fig. 2.5).
Stronger winds in opened areas may have also caused monarchs to be
dislodged from clusters more frequently than in closed areas. I have found
significantly higher numbers of live butterflies underneath clusters in
opened areas compared to closed areas (see Chapter 5).








In spite of the prevailing dryness during the winter months in
Central Mdxico (Rzedowski 1983; Calvert et al. 1989), data gathered in this
study showed that relative humidities remained high at the Sierra Chincua
overwintering site (Fig. 2.2). Monarch butterflies form their overwintering
aggregations primarily on the southwest facing slopes of the Mexican
mountains (Calvert and Brower 1986). Southwestern slopes are usually
wetter than northern and eastern slopes because moisture laden air
masses from the Pacific coast move into the high altitude mountains
during the winter (Mosifio-AlemAn and Garcia 1974; Calvert et al. 1989). I
observed that monarchs seem to be sensitive to changes in the relative
humidity of the forest. When the relative humidity dropped below 35% for
several days (Fig. 2.2), the monarch aggregation abandoned its location and
re-formed in a more humid area with an average of 60% relative humidity
(Alonso-M. unpublished data).
Previous studies also documented a massive movement of a monarch
aggregation to a new location. As the overwintering period progressed, the
entire aggregation departed from the original site and reformed on trees at
lower altitudes (Calvert and Brower 1986; Calvert et al. 1989; Calvert 1994).
By monitoring temperature, humidity, and rates of water evaporation
throughout the overwintering period, I found three distinct climatic
characteristics that were correlated with the movement of the monarch
aggregation. I noted that the Llano del Toro overwintering aggregation
departed from the study site when (1) the ambient temperature in the shade
was above 16'C (Fig. 2.1), which is the upper limit of the monarch's
thermal flight threshold (Masters et al. 1988; Alonso-M. et al. 1993); (2) the
minimum relative humidity was below 35% for several days (Fig. 2.2); and
(3) the rate of water evaporation was higher than 30 ml/week (Fig. 2.5). I








also observed that monarchs abandoned opened areas first, followed by
monarchs clustered in closed areas. Monarchs usually reform their
aggregations in sites with high relative humidity, such as above a stream
with running water (Calvert and Brower 1986; A. Alonso-M. unpublished
data).
Despite the mounting evidence indicating that logging is detrimental
to the survival of overwintering monarchs (Calvert and Brower 1981;
Calvert et al. 1982, 1986, 1989; Calvert and Cohen 1983; Brower and Malcolm
1991; Alonso-M. et al. 1992; Snook 1993; Anderson and Brower 1996), some
argue that logging should be permitted in the core areas of the reserve
(Hoth 1993; Chapela and Barkin 1995). This pressures government leaders
to consider changing the presidential decree. In this chapter, I analyzed
several microclimatic characteristics recorded in closed and opened areas
of the Oyamel fir forest. I conclude that monarch butterflies need cool, but
not freezing temperatures to preserve their lipid reserves. They also
require relatively high humidities, but at the same time need to avoid the

accumulation of water on their body surfaces which would cause death at
temperatures below freezing. I recommend that the core areas of the
MBSBR should be managed to provide a closed canopy with a natural
understory vegetation. Further logging should not be permitted in the core
areas of the MBSBR.








Table 2.1. Analysis of covariance (ANCOVA) comparing slopes and y-
intercepts of several microclimatic variables registered in opened and
closed areas during the 1993-1994 overwintering season at the Llano del
Toro monarch butterfly aggregation, located at Sierra Chincua,
Michoacan, M6xico. The rate of change was not different in any of the
comparisons (i.e. slope p > 0.05). In two comparisons I found a significant
difference in the y-intercept. Warmer temperatures were registered in
closed areas and higher wind speeds were registered in opened areas.

F-value d. f. p-value

Ambient Temperature

Minimum

slope 0.30 1, 195 p = 0.58
y-intercept 58.0 1, 196 p < 0.001
Maximum
slope 1.55 1, 194 p = 0.22
y-intercept 3.82 1, 195 p = 0.06

Relative Humidity
Minimum

slope 0.02 1, 194 p = 0.89
y-intercept 0.03 1, 195 p = 0.88

Maximum

slope 0.29 1, 194 p = 0.59
y-intercept 0.70 1, 195 p = 0.40

Wind Velocity
Maximum
slope 3.44 1, 124 p = 0.07
y-intercept 66.1 1, 125 p < 0.001


















Figure 2.1. Comparisons of average daily minimum and maximum
ambient temperatures recorded in closed and opened areas at the Llano del
Toro monarch butterfly aggregation during the 1993-1994 overwintering
season. Day 40 refers to December 10 and day 140 refers to March 21, 1994.
Differences between closed and opened daily minimum and maximum
temperatures were calculated by subtracting temperatures recorded in
opened areas from those recorded in closed areas. Data in C indicate that
minimum temperatures were consistently colder in opened areas (average
difference between closed and opened areas = 1.18'C). Data in D show that
opened areas were increasingly hotter later in the winter (average difference
between closed and opened areas after day 100 of overwintering = -0.60'C).












Figure 2.1

















20 A: CLOSED AREAS 20-

a: 18 18-
I-
< 16 16"
LU
14- 14

. 12 12
LIJ 10 10"

,< 8 8"

9 6 6-


Z 4 4-

2 2
. 04-
0 40 60 80 100 120 140 40







25- C: MINIMUM TEMPERATURES
(CLOSED OPENED VALUES)
Uj 2.0-0
o20



w 2..0
Cl)




0
<
U
a- 1
2 10


60 80 100 120 140


D: MAXIMUM TEMPERATURES
(CLOSED -OPENED VALUES)


40 60 80 100 120 140


DAYS OF OVERWINTERING


40 60 80 100 120


DAYS OF OVERWlNTE RING


















Figure 2.2. Comparisons of average daily minimum and maximum
percent relative humidity recorded in closed and opened areas at the Llano
del Toro monarch butterfly aggregation during the 1993-1994 overwintering
season. Day 40 refers to December 10 and day 140 refers to March 21, 1994.
Note that the relative humidity remained high during most of the
overwintering season. When it dropped below 35% for a few days (dashed
line in A and B), the monarch aggregation abandoned the location and
reformed in an area near running water. Data in C and D indicate that
minimum and maximum relative humidities were consistently higher in
opened than in closed areas (average difference between minimum and
maximum relative humidities in closed and opened areas = -2.1, and -2.2%
respectively).












Figure 2.2


A. CLOSED AREAS


6: OPENED AREAS


40 60 80 100 120 140


C: MINIMUM RELATIVE HUMIDITY
(CLOSED-OPENED AREAS)


40 60 80 100 120 140


D: MAXIMUM RELATIVE HUMIDITY
(CLOSED-OPENED AREAS)


DAYS OF OVERWINTERING


10"
.R
cn

w
0

z 5*
uJ
CC
w
IL
U-
U-
_5 0

-5-





J -10,
IL
r

S -15,


40 60 80 100 120 140


DAYS OF OVERWINTERING

















Figure 2.3. Comparisons of average daily maximum wind speed
recorded in closed and opened areas at the Llano del Toro monarch butterfly
aggregation during the 1993-1994 overwintering season. Day 40 refers to
December 10 and day 140 refers to March 21, 1994. Differences between
closed and opened daily maximum wind speed were calculated by subtracting
wind speed recorded in closed areas from those recorded in opened areas.
Values below zero indicate that lower wind speed values were recorded in
closed quadrats.












Figure 2.3




CLOSED FOREST


40 60 80 100 120 140


OPENED FOREST


40 60 80 100 120 140


1.0 MAXIMUM WIND SPEED DIFFERENCES
BETWEEN CLOSED MINUS OPENED VALUES
0-5-





-0.5"








-2.0-



40 60 80 100 120 14


DAYS OF OVERWINTERING


40


















Figure 2.4. Diurnal changes in the average wind velocity during 64, 24
hr cycles in closed (dark circles) and opened (opened circles) areas. Data were
gathered during the 1993-1994 overwintering season at the Llano del Toro
monarch butterfly aggregation, located at Sierra Chincua, Michoacan,
Mxico. Each point is the average of 2-hourly values, with standard errors.





40


Figure 2.4












2.0- CLOSED AREAS
-G OPENED AREAS
()
I)
E .5
w

a
z 1.0

w


w


0 2 4 6 8 10 12 14 16 18 20 22

HOUR


















Figure 2.5. Comparison of the amount of water evaporated (ml) in
closed and opened areas during the 1993-1994 overwintering season. Each
bar (dark = closed; clear = opened) represents the average of the water
evaporated at 10 sites/week (+ standard error). Day 57 refers to the week of
December 20 to 27, 1993, and day 144 refers to the week of March 18 to 24,
1994. Water evaporated at a higher rate in opened (24.7 ml/week) than in
closed (21.9 ml/week) areas. At 123 and 130 days, the monarch aggregation
abandoned the study location and reformed down slope in a more humid area.






42

Figure 2.5










-40-
(-Monarchs abandoned Iocatio

E CLOSED
Z 30E OPENED
0





I-
uJ

010
0



IL
0
w- 0_


n


DAYS OF OVERWINTERING


57 65 73 81 88 95 102 109 116 123 130 137 144


















Figure 2.6. Monarchs held in experimental enclosures consumed their
lipid reserves at a faster rate when they were exposed to the microclimatic
conditions in opened areas. At the beginning of the experiment, on February
7, 1986, monarchs had an average lipid mass of 68 mg ( 5.3). Monarchs in
open areas metabolized lipids faster than monarchs in closed areas. I used
the overall mean square error of the two way (treatment X date) ANOVA to
compare means per date. Samples with different letters are significantly
different (Ryan Q-test). Each bar represents the average of 20-30 butterflies
with its standard error.






44


Figure 2.6


FEB 22


CLOSED




a








b





MAR 23


80


70


- 60
E
ci,
(, 50-


- 40-

w
(5 30-

w
> 20-


10-


0-


MAR 8












CHAPTER 3
USE OF LIPID RESERVES BY MONARCH BUTTERFLIES
OVERWINTERING IN MEXICO: IMPLICATIONS FOR
CONSERVATION




Introduction


Each year during the autumn in North America, tens of millions of
monarch butterflies (anaus plexippus L.) migrate to the mountains of
central M6xico. As their larval food source of milkweed plants (Asclepias
spp.) diminishes in North America at the end of the summer, monarchs
escape the cold northern winter and migrate to the cool, moist environment
of high altitude mountain peaks. They arrive in early November and form
dense aggregations in several areas in the Transverse Neovolcanic Belt in
the states of Michoac~n and Mexico (Urquhart and Urquhart 1978a, b;
Brower 1985; Calvert and Brower 1986; Calvert et al. 1989; Calvert and
Lawton 1993). There, they remain largely inactive and maintain a state of
reproductive diapause until March when they migrate back to the southern
United States to exploit freshly emerging milkweeds (Herman 1985; Brower
and Malcolm 1991; Malcolm et al. 1993).
Autumn migrant monarchs differ in behavior and physiology from
monarchs during the rest of the annual cycle. Spring and summer
generations form extensive breeding populations and individuals do not
accumulate large lipid reserves (Beall 1948; Tuskes and Brower 1978;
Brower 1985; James 1984; Walford 1980). In contrast, late summer and








autumn generations do not mature sexually and build up extensive lipid
reserves as they migrate to overwintering areas in Mexico (Beall 1948;
Brown and Chippendale 1974; Walford 1980; Brower 1985; Masters et al.
1988; Calvert and Lawton 1993). These differences have also been found in
monarch populations in California (Tuskes and Brower 1978; Chaplin and
Wells 1982; Wells et al. 1993), and eastern Australia (James 1984, 1986,
1993). As they migrate southward, monarchs stop frequently to obtain
nectar from wild flowers, which is then converted to lipids and stored for
use during the overwintering period (Cenedella 1971; Brown and
Chippendale 1974; Turunen and Chippendale 1980). Lipids are an efficient
energy source for flying insects due to their relatively low weight and high
energy content (Kozhantshikov 1938; Beenakkers et al. 1981). The
accumulation of lipids before overwintering has been reported in other
insects, including flies (Valder et al. 1969; Adedokun and Denlinger 1985),
beetles (Hodek and Cerkasov 1961; Lambremont et al. 1964; Dortland and
Esch 1979), other butterflies (Pullin 1987), and moths (Chippendale 1973;
Gunn and Gatehouse 1986).
Lipid mass in monarchs declines slowly during the overwintering
period due to the inactivity of clustered monarchs at low ambient
temperatures (Chaplin and Wells 1982; Masters et al. 1988; Calvert et al.
1989; Leong 1990, Leong et al. 1991; James 1993). A small percentage of
overwintering monarchs show active behaviors such as flying and gliding
within the colony, as well as flights to nearby water and nectar sources
(Masters et al. 1988). Downslope colony movement to more humid areas
also occurs as the dry season progresses (Calvert and Brower 1986; Calvert
1994; Chapter 2). Since these behaviors consume the monarchs' limited
lipid reserves, Masters et al. (1988) hypothesized that monarch butterflies







may try to avoid them in order to conserve energy. Lipid reserves
remaining at the end of the overwintering period are used for migration
and reproduction, and are probably supplemented by nectar feeding along
the migration route (Heitzman 1962; Brower 1985; Urquhart 1987).
In Mexico, monarchs overwinter in forests dominated by the Oyamel
fir (Abies Jjgji H.B.K.) where microclimatic conditions are suitable for
their five month overwintering period (Calvert et al. 1989). These forests
have island-like distributions on mountain peaks at elevations between
2,400 and 3,600 m (Rzedowski 1983). The restricted distribution of the
Oyamel forest and increasing logging and clearing for agricultural fields
make it more vulnerable to deforestation than any other forest type in
M6xico (Calvert et al. 1989; Snook 1993). Degradation of the forest
endangers the migration phenomenon of the monarch butterfly (Brower
and Malcolm 1991; Malcolm 1993; Anderson and Brower 1996). In 1986, the
Monarch Butterfly Special Biosphere Reserve (MBSBR) was created by a
presidential decree for protection of the monarch butterfly. It includes
16,110 ha of which 11,600 ha are classified as buffer zones where forest
extractions are permitted (Diario Oficial 1986). Therefore, logging-free
areas consist of only 4,500 ha. Notwithstanding the small size and
protection status, the federal government is under pressure from land
owners to change the presidential decree to approve logging permits in the
core areas.
Hoth (1993) has recently advanced an argument to justify logging in
the core areas of the monarch reserves. He hypothesized that tree
extraction would create open areas where understory plants may produce
more flowers than in closed areas. In theory, such increased availability of
nectar resources upon which monarchs may feed could translate into fewer








monarchs depleting their lipid reserves and, therefore, more monarchs
surviving the winter in Mexico and successfully migrating the following
spring. Brower and Malcolm (1991), however, found that monarchs
visiting flowers during the overwintering period in January 1981 had lower
lipid mass than did inactive monarchs clustered on trees. Their results

suggest that flower-visiting monarchs may differ from clustered monarchs
in their ability to overwinter and migrate successfully. Moreover, Brower
(1995b) pointed out that the amount of nectar available within the area
utilized by the butterflies during the winter would be inadequate even if the
entire core areas were thinned.
In this chapter, I extended Brower and Malcolm's 1991 study to
compare several other physical characteristics between flower-visiting
monarchs and inactive monarchs clustered on trees. I compared 1) lipid
mass, an indicator of monarch energy reserves; 2) water mass, which is
important for survival and may drive monarchs to become active for
rehydration; 3) lean mass, which is a reflection of body size and protein
content; and 4) wing length, which is another indicator of body size. I
hypothesized that the cohort of overwintering monarchs that visit flowers
do so because they are attempting to replenish low lipid and/or water
reserves. I also hypothesized that monarchs clustered on the trees beneath
the fir forest canopy would have high lipid mass at the beginning of the
overwintering period and then use those lipid reserves through the period
at a rate not different from that expected based on their basal metabolic
rate. This would indicate that monarchs do not move very much during the
overwintering period and that nectar feeding is probably not important for
clustered butterflies.








I compared the lipids reserves of the flower-visiting monarchs and
the inactive monarchs clustered on trees to 1) autumn migratory monarchs
collected in Texas on their way to the overwintering sites in M6xico, 2)
migrating monarchs that had successfully returned from the
overwintering sites in M6xico to the southern United States, and 3)
reproductively active spring and summer generations collected in
Wisconsin and Minnesota. I hypothesized that these monarch groups
would differ in the quantity of their lipid reserves due to contrasting activity
levels and reproductive states.




Methods


Winter Study Site
In M6xico, monarchs form 9 12 overwintering aggregations near
the summits of mountains in an area from the western slopes of Volcano
Nevado de Toluca in the state of M6xico, northwest to the city of Zitdcuaro
and north to the Altamirano mountain in the state of Michoacdin (Figure 2
in Calvert and Brower 1986; Calvert et al. 1989; Calvert and Lawton 1993).
The overwintering period overlaps with the dry season which extends from
November to April. The area receives more than 1,000 mm of rain during
the summer wet season (Rzedowski 1983).
This study was conducted on the southwestern facing slope of
Zapatero Canyon at the Sierra Chincua, a mountain range in northeastern
Michoacdn (1940'48"N and 100017'54"W, Anonymous 1976). In the winter
of 1993-1994, I studied the Llano del Toro overwintering monarch
aggregation, a colony that has formed almost every year since the








overwintering sites were discovered in 1975 (Urquhart and Urquhart 1976).
The forest is dominated by A. religiosa trees; other tree genera found in the
area include Pinus L., Quercus L., and Buddleia L. (Soto and Vazquez
1993). The understory vegetation consists primarily of herbaceous and
bushy plant species in the Asteraceae and Lamiaceae, with a diverse
assortment of ascomycetes, basidiomycetes and bryophytes (Espejo et al.
1992).


Collections of Inactive Monarchs Clustered on Trees
Early in the morning, before ambient temperature was high enough
to permit flight (9.9-16.1C, Masters et al. 1988; Alonso-M. et al. 1993), I
collected monarchs that were inactive, hanging immobile on the oyamel fir
branches in overwintering clusters. Since no monarchs departed from the
clusters before I made my collections, my samples are butterflies that spent
the night clustered on the tree branches. With a pole attached to a standard
butterfly net, I netted all butterflies within a cluster about four meters above
the ground. I haphazardly selected 50 females and 50 males from the
sample. Collections were made monthly on November 8 and December 5,
1993, and on January 10, February 15, and March 13, 1994. The mean
number of butterflies per cluster was 265 (S.E. = 33.5, n = 5, range = 171 -
349). Butterflies were placed in individual glassine envelopes. They were
immediately transported to a laboratory where I recorded right fore wing
lengths, measured to the nearest 0.5 mm along the costal margin from base
to apex, and wet mass to the nearest milligram, using a Sartorius 1205 MP
electronic balance. Monarchs were then stored in a freezer at -41C until
chemical analysis was performed.








I dried the butterflies at 601C for 16 hr, weighed them, and extracted
lipids using the following method of Walford (1980) and May (1992). Each
monarch was ground in a centrifuge tube in 20 ml petroleum ether with a
Janke & Kunkel SDT Ultra Turrax tissuemizer. Each sample was then
vortexed and placed in a medium speed shaker bath at 35-380C for 30 min,
with vortexing every 10 min. Tubes were centrifuged in a Dynac Clay
Adams centrifuge at 1000 RPM for seven min, and the supernatant was
decanted into preweighed aluminum pans on a 301C hot plate. I added 15
ml of ether to the remaining solids in the centrifuge tube, and repeated the
procedure described above. The weighing pans with the supernatant were
allowed to evaporate for four hours to constant mass. The lipids were then
weighed. I estimated water content as wet mass minus dry mass, and lean
mass as dry mass minus lipid mass for each butterfly.


Collections of Flower-visiting Monarchs
I collected 50 females and 50 males visiting flowers on each of the
following dates: December 12, 1993, January 13, February 12, and March 13,
1994. All butterflies were collected within a 300 m radius of the center of the
2.01 ha Llano del Toro overwintering colony. A monarch was considered to
be visiting a flower only when it was perched on a flower or flower head and
its proboscis was inserted into a flower. Lipid mass, lean mass, water
mass, and wing length were recorded for each butterfly. I used the same
methods as described above for the clustered butterflies.


Collections of Mierating Monarchs
For comparisons between overwintering and other monarch groups,
I measured and analyzed the same characteristics in monarch butterflies








that were migrating through Texas to the overwintering sites in Mdxico.
In October, 1993, 61 migrating females and 73 males were collected from
transient clusters at Eagle Pass, Crystal City, and Castroville in south

Central Texas.
I also compared these characteristics to monarchs that successfully
migrated from M6xico to the southern United States. Migrating monarchs
can be identified by their cardiac glycoside fingerprint patterns, which
reflect the specific cardiac glycoside content of the milkweed plants they fed
on as larvae. Malcolm et al. (1993) found that 92% of monarch butterflies
overwintering in M6xico had fed as larvae on Ascle~ias vraca, a
milkweed species with a distribution in the northern United States. Any
monarchs collected in the spring in the southern United States with
similar cardiac glycoside fingerprint patterns would almost certainly be
migrants returning to the United States from the overwintering sites in
M6xico. In April and May 1985, Malcolm et al. (1993) analyzed 134
monarchs collected in Texas, Louisiana, and Florida and found 110
migrant butterflies with the A. svra fingerprint pattern (49 females, 61
males). In April and May 1986, Malcolm et al. (unpublished data) analyzed
225 butterflies collected in Texas, Louisiana and Oklahoma and found 162
migrant monarchs with the A. svraca pattern (75 females, 87 males).
From Malcolm's et al. data (1993, unpublished), I selected spring migrants
with the A. syriaca pattern that were collected before April 21 (83 females
and 80 males) to evaluate lipid, water, and lean mass depletion during
migration from M6xico to the southern United States. I did not include
later collections since data from those monarchs may show reproductive
and aging effects. By comparing monarchs collected in different years, I
am assuming that nectar available during the spring migration was the








same. I did not find differences in the mean fore wing length between
clustered monarchs collected in March in 1985 and 1986 (Van Hook 1993),
and clustered monarchs collected in March 1994 (p > 0.05).
I also compared the lipid and lean masses of my collections to several
other published and unpublished samples of monarch butterflies,
including freshly eclosed, summer breeders, and autumn and spring
migrant monarchs (see Appendix).


Estimates of Lipid Utilization by Clustered Monarchs
Chaplin and Wells (1982) estimated resting metabolic rates for
monarch butterflies. They studied energy budgets based on oxygen
consumption at different ambient temperatures, ranging from five to 22'C,
and determined that the resting metabolic rate for overwintering monarchs
was related to thoracic temperature by the function


loglo E = 0.048Tth 0.368


where E = energy expenditure in joules per hour, and Tth = thoracic

temperature in degrees Celsius. Following Masters et al. (1988), I used this
function to estimate the expected energy consumed by inactive monarchs
clustered on trees during the overwintering period. Since thoracic
temperatures of inactive monarch butterflies beneath the forest canopy are
not different from ambient temperatures (Chaplin and Wells 1982; Alonso
et al. 1993), I used ambient temperatures that I recorded at the site as Tth.
I used on-site weather loggers (OWL, Electronically Monitored
Ecosystems, 2229 Fifth St., Berkeley, CA 94710) to record ambient
temperatures from December 12, 1993 to March 24, 1994. OWLs were








programmed to register the ambient temperature every minute and to
record the mean of those measurements over two hr periods. I recorded the
ambient temperature at six different closed canopy locations within the
monarch colony. Temperature probes were placed near clustered
butterflies at three meters above the ground. Data from the six locations
were used to obtain mean ambient temperatures for the monarch colony
every two hr. I then estimated the energy consumed by resting monarchs
over two hour periods. I converted the energy spent to milligrams of lipid
burned based on the energy yield from the oxidation of lipids as 37.66
joules/mg (i.e. 9.01 cal/mg; 4.18 joules/cal; Gordon 1977). From this, I
calculated an expected amount of lipid that clustered monarchs burned

each day.
I then compared my results to published rates of lipid consumption
in two other overwintering populations of monarch butterflies. These
included a population of western monarchs overwintering in California
(Chaplin and Well 1982; Wells et al. 1993), and a population of monarchs
overwintering in eastern Australia (James 1984).


Statistical Analysis
For each of my monarch collections, lipid mass, water content, lean
mass, and wing length size data were first tested for normality using the
Shapiro-Wilk W test (Zar 1984). I tested the means for homogeneous
variances using Levene's test (Snedecor and Cochran 1980). Coefficients of
determination (r2, Zar 1984) were computed to detect possible correlations
between lipid mass, water content, lean mass, and wing length data.
Contrary to clustered monarchs (p > 0.05), lipid mass for flower-
visiting monarchs was not normally distributed (p < 0.001; Fig. 3.1). The








variances were, however, homogeneous for both clustered and flower-
visiting monarchs (Levene-median's test F(7, 786) = 1.83, p > 0.05).

Nonparametric Kruskal-Wallis one-way analyses of variance were used to
test for differences among lipid mass and lipid index medians. The lipid
index is computed as the lipid mass (mg) divided by the lean mass (mg) X
100 (Chaplin and Wells 1982; James 1984). It was determined to estimate
the proportion of lipid mass to lean mass in each monarch and to facilitiate
comparisons to published studies. When comparisons yielded a significant
Kruskal-Wallis statistic (p < 0.05), the lipid data were further analyzed by
the Joint-Rank Ryan nonparametric test for stepwise unplanned multiple
comparisons. For this test, I obtained the absolute value of the differences
between the mean ranks of samples, and calculated the critical value by
using the large-sample approximation (Day and Quinn 1989).
Parametric two-way analyses of variance (clustered and flower-
visiting monarchs x month) were performed on water content, lean mass,
and wing length data, after verifying that the means had homogeneous
variances (Levene's mean test, p > 0.05, Snedecor and Cochran 1980).
Comparisons yielding significant F values (p < 0.05) were further analyzed
by the parametric Ryan's Q test for stepwise unplanned multiple
comparisons. I used the Joint-Rank Ryan and the Ryan's Q test for
multiple comparison tests because they are the best to control the
experimentwise type I error rate (Day and Quinn 1989).
Based on the ambient temperatures that clustered monarchs
experienced during the overwintering period, I calculated the amount of
lipids consumed each day. I then compared this rate to the observed rate
found for inactive and flower-visiting monarchs. I used analysis of
covariance (ANCOVA, Zar 1984) to compare the slopes and the y-intercepts








of the rate of lipid loss between the rate of the expected and the observed
lipid loss found for inactive butterflies clustered on trees. I also compared
the observed and expected rate of lipid loss in flower-visiting monarchs.
All tests of significance were two tailed with a probability level for
significance of 0.05, while all comparisons tested a null hypothesis of no
difference. Statistics were computed using SuperAnova, Statview II, and
JUMP statistical packages on a SE/30 Macintosh computer.




Results


Clustered and Flower-visiting Monarch Comparisons
Inactive monarchs clustered on trees had significantly higher
amounts of lipid mass, water content, lean mass, and larger wings than
the flower-visiting monarchs (p < 0.001; Figs. 3.1 and 3.2). The lipid index
was also significantly higher for clustered monarchs, indicating that lipid
mass differences between the two groups were not entirely due to monarch
size (Kruskal-Wallis test, p < 0.001).
The physical condition of the clustered monarchs changed as the
overwintering season progressed. At the beginning of the season, they had
higher amounts of lipid mass, water content, and lean mass than at the
end of the period (Figs. 3.1 and 3.2). Flower-visiting monarchs had low
lipid mass values during the season, with water mass being lowest in the
middle of the overwintering period (p < 0.001). No lean mass decline was
found through time for flower-visiting monarchs (p = 0.10). Since I did not
find wing length differences among individuals for either clustered (p =
0.31) or flower-visiting monarchs through time (p = 0.09), these results








suggest that the observed mass loss through time was not the result of size-
related mortality or to the emigration of larger butterflies.
While the lipid mass of clustered monarchs was significantly
correlated to water mass, lean mass, and wing size data, the coefficients of
determination (r2) accounted for less than 10% of the variability in those
data (Table 3.1). For both clustered and flower-visiting monarchs, the
highest coefficients of determination were found for wing size and lean
mass data: the larger the butterfly, the higher the lean mass.


Comparisons to Migrating Monarchs
Monarchs migrating southward through Texas in October had lower
amounts of water than did monarchs clustered on trees at the
overwintering sites in November (Table 3.2). No differences in lipid mass (p
= 0.07), lean mass (p = 0.42), or wing size (p = 0.32) were detected.

Comparisons of monarchs collected at the end of the overwintering
period in March in Mexico and monarchs collected in April in the southern
United States, showed that clustered monarchs had higher lipid and lean
masses, but not larger wings than migrant butterflies (p < 0.001; Table 3.2).
Moreover, migrant monarchs had higher lipid mass and larger wings
than flower-visiting monarchs (p < 0.05). No differences in water mass
were found between the three groups (p = 0.36).


Non-migratory. Reproductive Active Generations
Lipid mass recorded from monarchs of the eastern and western
populations in North America changed throughout the year (see Appendix;
Fig. 3.3). Non-migratory, reproductively active generations of monarchs of
the eastern population collected from April to August had mean lipid








amounts of about 20 mg, which is similar to mean values recorded for
flower-visiting monarchs at the overwintering site in M6xico (Fig. 3.1). Fall
migrants in September and October increased their lipid amounts from
about 60 and 80 mg respectively, to very high amounts (135 mg) when they
arrived to the overwintering sites in November (Fig. 3.3). Both the eastern
and western populations of clustered monarchs had high lipid masses at
the beginning of the overwintering period that progressively decreased
throughout the season.


Estimates of Linid Utilization During the 1993-94 Overwintering Period
Using Chaplin and Wells' (1982) equation for basal metabolic rate
and the ambient temperatures recorded at the overwintering colony during
1993-1994, I estimated that the resting metabolic rate of inactive monarchs
within the winter clusters was on average 0.695 mg of lipid mass per day (n
= 103 days, S.E. = 0.02). On December 5th, clustered monarchs had a mean
lipid mass of 113 mg (S.E. = 4, n = 100). Since I started recording ambient
temperatures on December 12th (i.e., day 42 of overwintering, November 1 =
day 1), I used the calculated mean daily loss to estimate the expected lipid
mass for that date based on the December 5th datum (113 mg (0.695 mg X 7
days) = 108.1 mg). I then used 108.1 mg of lipids as the initial datum from
which I subtracted the expected amount calculated for each day (Fig. 3.4).
For example, I expected monarchs to have 107.5 mg of lipids by day 43 of
overwintering (108.1 0.596, estimated amount consumed on day 43). The
expected rate of lipid loss of clustered monarchs followed the linear
regression model


lipid mass (mg) = 138.2 0.669 (days of overwintering),









with an r2 = 0.99 (ANOVA F(i, 102) = 196.9, p < 0.001). I then regressed the
observed monthly mean lipid content of clustered monarchs, and obtained

the following linear function


lipid mass (mg) = 138.4 0.653 (days of overwintering),


with an r2 = 0.95 (ANOVA F(i, 4) = 61.2, p < 0.01). The slopes of the expected
and observed functions are not significantly different from each other
(ANCOVA F (1, 104) = 0.6, p = 0.44; Fig. 3.4). From these data, it is clear

that inactive monarchs clustered on trees lost their lipids passively in
relation to the ambient temperature, as would be expected based on their
resting metabolic rate.


Rates of Lipid Utilization in Mdxico. California and Australia
I compared the rate of lipid utilization found for clustered monarchs
in M~xico to published rates of lipid use in two other overwintering
locations, in California (Chaplin and Wells 1982; Wells et al. 1993) and
Australia (James 1984). I found that the rates of lipid use were not
significantly different among the three sites (ANCOVA F (2, 8) = 0.08, p =

0.92; Fig. 3.5). These data suggest that monarchs from these three distant
locations remain largely inactive during the overwintering period and that
overwintering temperatures are similar. I also found that monarchs in
Mexico have higher lipid mass (133 mg) at the beginning of the season,
than butterflies in California (84 mg) and eastern Australia (94 mg), and
that monarchs departed at the end of the season from the three sites with
similar mean lipid masses: M6xico 56, California 57, and Australia 53 mg








(see Appendix). Note that the mean lipid mass for flower-visiting
monarchs in March was only 21 mg.


Proportion of Clustered Monarchs That May Need to Visit Flowers
I estimated the proportion of inactive monarchs clustered on trees
that may need to visit flowers as they deplete their lipid reserves during the
overwintering period. For each monthly collection (n = 100 butterflies), I
determined the number of clustered monarchs that had lipid amounts
lower than the average lipid mass plus one standard deviation recorded for

flower-visiting monarchs. For example, in December, this value for flower-
visiting monarchs would be 97 mg (average = 53 + 44, S.D.). For monarchs
clustered on trees, 41% had lipid masses lower than 97 mg.
I also determined the number of flower-visiting monarchs that had
higher lipid amounts than the average lipid mass found for clustered
monarchs, minus one standard deviation. In December, clustered
monarchs had on average 113 mg of lipid mass (S.D. = 44). I thus
determined that 34 flower-visiting monarchs (34%) had a higher lipid mass
than 69 mg. I performed a similar analysis for January, February, and
March. I combined data for the 4 months and obtained an average of 26%.
In other words, of 400 flower-visiting monarchs, 104 had lipid amounts
greater than the monthly average recorded for clustered monarchs minus
one standard deviation. Of 394 clustered monarchs, 120 (30.5%) had lipid
amounts lower than the monthly average plus one standard deviation
recorded for flower-visiting monarchs. I thus propose that about 30% of the
overall population of inactive monarchs clustered on trees from December
to March may depart from roosting locations to visit flowers.









Discussion


Clustered and Flower-visiting Monarch Comparisons
Fir forests provide the microclimate needed for successful
overwintering in M6xico. In November, inactive monarchs clustered on
trees had an average lipid mass of 133 mg. These monarchs consumed
their lipid reserves in relation to the ambient temperature (Chaplin and
Wells 1982), such that by the middle of March, just prior to departure from
the overwintering site, they had an average of 56 mg of lipids, a 57.9% lipid
loss in 5 months (Fig. 3.4). Monarchs that successfully migrated to the
southern United States in April had on average 26 mg of lipid mass, a
further loss of 22.6%, indicating that migration to the United States is
energetically expensive.
In contrast, monarchs that visited flowers at the overwintering sites
had highly depleted lipid reserves. In December, they had an average lipid
mass of 53 mg, 47% of that of clustered butterflies, and the same average as
clustered monarchs in March (56 mg). Moreover, during the last three
months of the overwintering period, the frequency distributions of the lipid
mass data for flower-visiting monarchs showed a high proportion of
butterflies with lipid masses close to zero (Fig. 3.1). Since monarchs
convert carbohydrates from nectar into lipids within a few hours after
ingestion (Cenedella 1971), and I found very few butterflies with medium or
high lipid levels visiting flowers, I suggest that either (1) flower-visiting
monarchs maintained their lipid reserves at low levels by visiting flowers
but were unable to reach levels found in clustered monarchs, or (2) that
flower-visiting monarchs starved to death, and were replaced by clustered







monarchs that departed from their roosting trees to visit flowers when their
lipid reserves dwindled to critical levels. The frequency distribution of

clustered monarchs shifted toward lower lipid mass categories as they
consumed their lipid reserves during the season (Fig. 3.1). Clustered
monarchs with low lipid reserves of about 20 mg ( 20) may have departed
from their roosting trees to visit flowers. Monarchs that recently departed
from clusters could still have high levels of water mass, as I found for
flower-visiting monarchs in February and March (Fig. 3.2). I then suggest
that as the season progressed and clustered monarchs consumed their
lipid reserves, monarchs with low lipid masses departed from clusters to
visit flowers. I also acknowledge that some flower-visiting monarchs could
cluster. However, only 26% had lipid mass (minus one standard deviation)
of that recorded for clustered monarchs. My data suggest that about 30% of
the overwintering population may become flower-visiting monarchs.
I found consistent physical differences between clustered and flower-
visiting monarchs (Fig. 3.2). Clustered butterflies had higher lipid mass,
higher water content, higher lean mass, and larger wings than flower-
visiting monarchs throughout the season. Since few flower-visiting
monarchs had characteristics similar to clustered monarchs, they either
(1) stayed near the overwintering aggregation but did not cluster in trees,
(2) returned to the overwintering aggregation but clustered in higher
positions within the fir trees where I could not collect them, (3) departed
early from the monarch aggregation, or (4) died at the overwintering sites.
Several lines of evidence support my hypothesis that flower-visiting
monarchs in March are in such poor condition that they may not be able to
migrate to breeding areas in the southern United States. First, migrating
monarchs collected in April in Texas and Louisiana had significantly







higher amounts of lipid mass than flower-visiting monarchs collected in
March at the overwintering site, an indication that flower-visiting
monarchs may not have enough lipid reserves to migrate back to the United
States successfully (Table 3.2). Second, migrating monarchs arrived in the
southern United States with less than 50% of the lipid mass and less than
10% of the lean mass found in clustered butterflies in March, but they had

the same wing length size, suggesting that migration is energetically
expensive, and that this collection of butterflies came from clustered
monarchs in M6xico (Table 3.2). Third, flower-visiting monarchs may be in
a different physiological state than clustered monarchs. They had lipid
reserves (20 mg) similar to those found in non-migratory summer breeders
(Fig. 3.3). Moreover, Van Hook (1993) described physical characteristics of
male monarchs that became reproductively active during the overwintering
period. She found that they had low wet mass, and small wings that were
in poor condition (i.e. less scales, broken wings), which are characteristics
that closely resemble those of flower-visiting monarchs. Reproductively
active monarchs have high levels of juvenile hormone, a hormone that
controls development of reproductive organs, and utilization of lipids, and
accelerates aging in monarch butterflies (Barker and Herman 1976;
Dallman and Herman 1978; Lessman and Herman 1983; Herman 1985;
Herman et al. 1989).
It therefore appears that flower-visiting monarchs do not have
enough lipid reserves to migrate back to the breeding areas of the southern
United States. My analysis supports the hypothesis that this cohort of
butterflies is derived from the clustering butterflies that arrived in Mexico
with low lipid levels. By opening the forest, ambient temperatures will
increase and hence so will the metabolic rate of clustered monarchs








(Chapter 2). Thus, a higher percentage of the population would enter the
cohort of butterflies with low lipid contents. My data argue that there is no
need to create open areas in the core zones of the MBSBR to promote the
production of flowers for overwintering monarch butterflies. All
indications are that this would make matters worse, not better.


Detrimental Effects of Logging on Monarch Survival
Data presented in this chapter strongly support the hypothesis that
the overwintering success of monarch butterflies is not a matter of
replenishing their lipid reserves by visiting flowers. Instead, success is
dependent upon how well monarchs conserve their lipid reserves under the
appropriate microclimatic conditions. There is abundant evidence for
detrimental effects that logging has on overwintering monarch survival in
relation to microclimate. Calvert and Brower (1981) and Calvert et al. (1982)
showed that clustered monarchs freeze to death more often in thinned
forest than monarchs in closed canopy forest. In 1981, 2.5 million
butterflies died during a period of extreme cold weather (Calvert et al. 1983).
Similarly, in 1992, about 83% of the Herrada monarch colony perished
(Culotta 1992). Overwintering monarchs can survive temperatures several
degrees below freezing because their body fluids can supercool (Lee 1989;
Anderson and Brower 1993, 1996; Larsen and Lee 1994). However, colder
temperatures registered in thinned forests promote the accumulation of
dew on the body surface of monarchs and wetting during storms (Calvert
and Brower 1981; Anderson and Brower 1996). When the external moisture
freezes, ice crystals penetrate the cuticle and promote ice nucleation in the
haemolymph, killing the butterflies (Alonso-M. et al. 1992; Anderson and
Brower 1993; Larsen and Lee 1994). Thinned forests also increase the







intensity of bird predation on monarchs (Brower and Calvert 1985; Chapter
5), and greater exposure to sunlight increases butterfly activity and lipid
use (Chaplin and Wells 1982; Masters et al. 1988). In fact, monarch
colonies such as those at Altamirano, San Andres and Chivati-Huacal
(map in Calvert and Brower 1986) have most likely changed in response to
low tree densities resulting from current and past forest extractions. Over
time, these colonies have become smaller, ephemeral, unstable and in
several years absent (W. H. Calvert unpublished data). Similar
observations have been recorded from degraded forest groves where
monarchs overwinter in California (Weiss et al. 1991).


Management Recommendations and Conclusions
Core areas of the monarch reserve are already subject to numerous
human activities. Out of the 4,500 logging-free hectares of the reserve, 1,000
are being used as centers for ecotourism, one of the major alternative
incomes to land owners. Fifteen hundred hectares have been illegally
logged and the remaining 2,000 ha have a mosaic of closed and open forest
patches where most of the scientific research on the migrating monarchs
and on the oyamel forest is conducted (A. Alonso-M. unpublished data).
Since the core areas the MBSBR are small, for them to serve as sources of
plant and animal species for recolonization of the 11,600 ha of the buffer
zone, they should be increased in size and maintained as logging-free
areas. Ideally, unprotected overwintering aggregations should be
incorporated into the MBSBR so that the risk of extinction of the endangered
phenomenon of the monarch butterfly migration will decrease.
The understory vegetation of the oyamel fir forest also plays an
important role in the survival of overwintering monarch butterflies. Many








clustered monarchs are dislodged from tree trunks and tree branches
during storms and by predatory birds at ambient temperatures when
monarchs are unable to fly (Brower and Calvert 1985; Masters et al. 1988;
Arellano et al. 1993). Monarchs on the ground shiver and crawl up onto
nearby vegetation (Alonso-M. et al. 1993) to avoid freezing ground
temperatures (Calvert and Cohen 1983; Alonso-M et al. 1992) and mouse
predation (Glendinning et al. 1988). To date, conservation and
management strategies have not emphasized protecting the understory
vegetation. It is currently being damaged by logging practices, cattle
grazing, and trampling by tourists (Calvert et al. 1989; Snook 1993). Future
management must also minimize impact on the oyamel tree seedlings and
on the understory vegetation.
Results from this study demonstrate that it is ill-advised to thin fir
forests to create nectar sources. Based on the rate of lipid loss throughout
the overwintering period, monarch butterflies need intact closed forest for
successful overwintering. The overall low rate of lipid mass loss by inactive
monarchs clustered on trees does not support the argument favoring
artificial openings by logging in the monarch reserves. Management
strategies for the conservation of the endangered migration phenomenon of
the monarch butterfly should totally preclude logging practices in core
areas of the reserve.








Table 3.1. Coefficients of determination (r2) for correlations between lipid
mass, water content, lean mass and wing length data for monarch
butterflies overwintering in M6xico. Correlations for inactive monarchs
clustered on trees are shown in the right part of the matrix, while flower-
visiting monarchs are in the left.

LIPID WATER LEAN WING LENGTH

LIPID - 0.10* 0.08* 0.04*

WATER 0.01ns -- 0.33* 0.31*

LEAN 0.01ns 0.30* --- 0.50*

WING LENGTH 0.02ns 0.19* 0.65* - -


* significant positive correlations p < 0.001








Table 3.2. Comparisons between migrant monarchs through Texas in
October (1993) and early arrivals at the overwintering sites in November
(1993) in M6xico. I also compared migrant monarchs collected in the
southern United States in Texas-Louisiana-Oklahoma-Florida before April
21 (1985, 1986) to monarchs at the overwintering sites in M6xico about to
migrate back to the United States in March, 1994. Mean lipid mass, water
content, and lean mass (mg), and right fore wing length (mm) data are
shown, with standard errors in parentheses. Samples with different letters
in columns are significantly different (Joint-Rank Ryan test for lipid
samples and Ryan Q-test for other samples).


N LIPID 1 WATER2


LEAIN2 WING LENGTH2


MIGRANTS VS
Texas Oct
M6xico Nov


OVERWINTERING
132 120 (6)a
100 133 (5)a


OVERWINTERING VS
Inactive Mar 94
Flowers Mar 100
Southern USA 163


MIGRANTS
56 (3)a
21 (2)b
26 (2)c


281 (4)a
303 (4)b



261 (4)a
246 (4)a
252 (6)a


172 (3)a
170 (3)a



168 (2)a
157 (2)b
148 (2)c


52.4 (0.2)a
52.3 (0.2)a



52.1 (0.2)a
51.3 (0.2)b
51.9 (0.2)a


1= Kruskal-Wallis test; 2 = ANOVA.


-----------------------------------------------------------------------------------------------------------

















Figure 3.1. Lipid mass frequency distributions of clustered and flower-
visiting monarchs during 4 months of the 1993-1994 overwintering season in
the Sierra Chincua, Michoacdn, Mexico. Mean values (mg), standard
deviations (S.D.), and sample sizes (N) are given for each distribution. The
intervals of the histograms are 0-9, 10-19, etc.












Figure 3.1


INACTIVE MONARCHS CLUSTERED ON TREES


FLOWER-VISITING MONARCHS


25-
DECEMBER DECEMBER
mean =113 20 mean = 53

SDS = 44
n = 100 n=100
5-0


JANUARY
mean =101
SD= 43
n = 100


65-
60-
55"
50-
45-
40-
35-
30-
w
S25-

~20-
15
10-
Z 5,


z
04

0' 45-
a:4-
Wj 40-
0 35-

30-
25-

20-
15-
10-
5-

0-



40-
35-

30-
25-

20-
15-
10-
5-

0-


MARCH
mean = 56
SD = 28





0 30 60 90 120 150 180 210

LIPID MASS (mg)


JANUARY
mean = 24
SD =35
n = 100


FEBRUARY
mean = 58
SD = 33
n = 100


FEBRUARY
mean = 21
SO = 21
n=100














MARCH
meon = 21
SD = 20
n =100


LIPID MASS (mg)


















Figure 3.2. Mean lipid, water, and lean masses (mg), and wing size
length (mm) of clustered (dark bars) and flower-visiting monarch butterflies
(light bars) during 4 months of the 1993-1994 overwintering season. Samples
with different letters are significantly different (Joint-Rank Ryan test for
lipid samples and Ryan Q-test for other samples). Each bar represents the
average of 100 butterflies with its standard error.












Figure 3.2


100"



& 80
E

(n 60



a 40


20


DEC JAN FEB MAR


a abc ab


DEC JAN FEB MAR


300-


250-


E 200-
I.-
z
w
z 150"
0
r
LU 100-


50-


0-


DEC JAN FEB MAR


'a
E

100

z
LJ
--j


1 a


DIEC JAN FEB VAR


















Figure 3.3. Lipid mass changes of 5211 monarch butterflies collected
by many researchers over several years at different times during their annual
life cycle. Data for the eastern North American population, and flower-
visiting monarchs are presented. Numbers indicate the total number of
sample collections used to obtain the average (S.E.) for each month. Refer to
the Appendix to consult original source of data. Note that non-migratory
spring and summer generations have a mean value of 20 mg of lipid mass,
the same value recorded for flower-visiting monarchs.








Figure 3.3


FLOWER-VISITING
MONARCHS AT THE OVERWINTERING SITE


JUL AUG SEP OCT NOV DEC JAN FEB MAR APR MAY JUN


















Figure 3.4. Comparison between expected and observed lipid mass loss
for monarchs overwintering in M6xico. Points with standard errors (straight
bars) are the average of 100 butterflies (50 females and 50 males). Day 1
refers to November 1, 1993 and day 150 to March 31, 1994. The regression
line of lipid mass for clustered monarchs (dark dots) against time is plotted.
The expected lipid loss was estimated from the ambient temperatures that
monarchs experienced during the season (see text). Notice that the 2 lines
virtually overlap. At day 42, flower-visiting monarchs (clear dots) had less
than half of the lipid resources of the clustered monarchs, and by day 74
through day 140, lipids stabilized at about 20 mg. Clustered butterflies lost
their lipids as predicted but did not drop below 50 mg, 250% higher than the
butterflies that visited flowers.









Figure 3.4


INACTIVE MONARCHS IN CLUSTERS


EXPECTED FAT LOSS


FLOWER-VISITING
MONARCHS

0 2 40 60 80 100 120 1 4
o 20 40 60 80 100 120 140


DAYS OF OVERWINTERING


140-


120-


100-


















Figure 3.5. Lipid indices for overwintering monarch butterflies
through time. Monarchs from Australia (James 1984), California (Chaplin
and Wells 1982), and Mkxico (this chapter) have the same rate of lipid loss
during the overwintering period.








Figure 3.5


y = -0.36d + 79.2


0 MEXICO 1993-94
O3 CALIFORNIA 1971-72
A AUSTRALIA 1983


y = -0.38d + 58.4


y = -0.34d + 71.2


0 20 40 60 80 100 120 140 160


DAYS OF OVERWINTERING












CHAPTER 4
CONSERVATION IMPLICATIONS OF FLOWERING PLANT
AVAILABILITY FOR OVERWINTERING MONARCH BUTTERFLIES IN
MEXICO




Introduction


Each year the North American population of monarch butterflies

(Danaus plexippus L.) east of the Rocky Mountains migrates to several
mountain peaks in central M6xico. These overwintering sites were first
described by F. A. Urquhart and colleagues after several decades of
research on monarch migration (Urquhart 1976; Urquhart and Urquhart
1976). Interested in the preservation of the spectacular monarch
aggregations, the Urquharts decided not to share the exact location of their
findings (Brower 1995a). Despite this, the publication of popular and
scientific articles on monarchs overwintering in M6xico soon made the
sites known. Ten million monarchs per hectare were difficult to hide
(Urquhart 1976; Brower 1977, 1985; Barthelemy 1978; Urquhart and
Urquhart 1978a, b; Calvert et al. 1979; Calvert and Brower 1986).
W. H. Calvert and L. P. Brower, from the University of Florida,
conducted pioneering research on the biology of monarch butterflies
overwintering in M~xico. They soon realized that the Oyamel fir trees

(Abies reliiosa H. B. K.), upon which monarchs roost during the
overwintering period, were being commercially exploited at a fast rate.
They determined that the survival of overwintering monarchs was closely








related to the microclimate registered in closed canopy forests, such that
creating open areas by logging enhanced monarch mortality (Calvert and
Brower 1981; Calvert and Cohen 1983; Calvert et al. 1982, 1983).
In 1986, the Monarch Butterfly Special Biosphere Reserve (MBSBR)
was created by the president of M6xico for protection of the monarch
butterfly (Diario Oficial 1986). Five reserves located on mountain peaks in
the Transverse Neovolcanic Belt in the States of M6xico and Michoacin
constitute the MBSBR (Calvert and Brower 1986; Calvert et al. 1989). The
MBSBR includes 16,110 ha of which 11,600 ha are classified as buffer zones
where forest extractions are permitted (Diario Oficial 1986). Therefore,
wilderness logging-free areas consist only of 4,500 ha. Core areas were
created for protection of the animal and plant species found in the relict
Oyamel Fir forests, and to serve as sources of species to recolonize logged
areas in the buffer zones. The restricted distribution of the Oyamel forest
and increasing pressure from logging and clearing for agricultural fields
make it more vulnerable to deforestation than any other forest type in
Mexico (Calvert et al. 1989; Snook 1993).
In this chapter, I present data to refute one of the arguments

currently used to promote logging in the core areas of the MBSBR. Hoth
(1993) has recently suggested that monarchs need to feed on nectar during
the overwintering period, and that the core areas of the reserve do not have
open areas where the butterflies can find plants in flower. He then
hypothesized that logging would be beneficial for monarchs, since
understory plants growing in logged-opened areas may produce more
flowers than in closed areas. He assumed that with nectar readily
available during the overwintering period, fewer monarchs would deplete








their lipid reserves. This hypothesis has been rapidly accepted by
politicians and rural developers (Chapela and Barkin 1995).
To investigate whether increased logging is needed in the core area of
one of the most pristine overwintering sites of the MBSBR (Calvert et al.
1989; Calvert and Lawton 1993), I looked at the distribution of flowering
plants in existing closed and opened areas.



Methods


Study Site

I studied the Llano del Toro 1993-1994 overwintering monarch
aggregation located in Sierra Chincua, a mountain range in northeastern
Michoacdn, M6dxico (19'40'48"N and 100'17'54"W, Anonymous 1976; see
map in Calvert and Brower 1986). Monarchs select high altitude oyamel fir
forests (3,000 m) where they find cool temperatures and high relative
humidity needed for successful overwintering (Masters et al. 1988; Calvert
et al. 1989). The understory vegetation consists primarily of plants in the
Asteraceae and Lamiaceae plant families (Espejo et al. 1992; Soto and
Vazquez 1993).
In addition to logging, local inhabitants obtain a limited number of
non-timber products from the oyamel forests. These include the collection
of flowers for religious rituals, herb plants for medicinal purposes (e.g. "te
de monte" from Satureia macrostema, Lamiaceae), the extraction of resin
from pine trees, and the harvesting of mushrooms during the rainy season.
Non-commercial, domestic use of wood in the area includes fuelwood and
charcoal production, beams for housing construction, and the production of







shingles. Free-ranging livestock are commonly found in the area and seem
to have a negative effect on the survival of oyamel fir tree seedlings (Calvert
et al. 1989; Snook 1993).


Plant Species in Flower and Monarch Butterfly Use
Since butterflies require a proboscis as long as the corolla tube to
successfully obtain nectar from a flower (May 1992), I measured the
proboscis length of 50 female and 50 male monarchs to the nearest 0.5 mm
by extending their proboscis on a 15 cm ruler. I compared these monarch
proboscis measurements to data on flower corolla length from Sanchez
(1986). I also followed Sanchez' nomenclature to determine plant species.


Determination of Forest Openness
I first determined the degree of openness within the forest where
monarchs formed their overwintering aggregation. From the center of the
monarch aggregation, I ran four line transects (505 m each) directed 90
apart from one another. During most of the overwintering period,
monarchs did not fly further than a 500 m radius from the studied 2.01 ha
colony. Data were gathered at the end of March 1994. Each line transect
was subdivided into 101, five meter stations that I defined as my sampling
units (n = 404). Each station was subdivided into quarters from which I
estimated the degree of canopy coverage. I classified the stations into six
canopy openness categories: A) Naturally Closed: the station was 75-100%
covered with canopy vegetation; B) Partially Closed: 50% of the station was
covered with canopy and 50% was opened by natural causes. I considered
natural causes of canopy openings to be dead standing or fallen trees,
streams and creeks, rocky outcrops, and/or natural meadows dominated by







Potentilla candicans H. and B. Rosaceae (Soto and Vazquez 1993); C)
Partially Opened: 50% of the station was covered with canopy and 50% was
opened by artificial causes. Opening by artificial causes was evidenced by
human activities such as logging roads, tree stumps, and/or property
boundary delimitations (see figures in Brower and Calvert 1985; Snook
1993); D) Naturally Qn: more than 50% of the station was opened by
natural causes; E) _n: 50% of the station was opened by natural causes
and 50% was opened by artificial causes; F) Artificially Opened: more than
50% of the station was opened by artificial causes. At each station I
recorded the most abundant flowering plant species that monarchs can use

for nectaring.


Statistical Analysis
I determined the number of stations and the number of plant species
in flower for each of the six canopy openness categories (Table 4.1). For the
statistical analysis I combined categories A-C and called them "closed",
while the combined data of categories D-F were called "opened". Using a G-
test (Sokal and Rohlf 1995), I examined whether there was an association
between the species of plants in flower and the habitat where there were
found (i.e. closed versus opened sites). I also used a G-test to examine if
each of the six most common flowering plant species (Table 4.1) were found

more frequently either in closed or opened sites.











Based on the matching of flower morphology and corolla length to the
proboscis of overwintering monarchs, I found that there were six species of
plants that had longer corolla lengths than the proboscis of monarchs.
These include Salia eleeans Vahl. (corolla length = 20-35 mm), E.
cardinalis H. B. K. (25-40 mm), Satureia macrostema (Benth.) Briq. (25
mm), Stachvs coccinea Jacq. (20-25 mm) (Lamiaceae), Castilleja avnsis
Benth. (30-35 mm, Scrophulariaceae), and Lupinus e (nectar is
inaccessible due to flower morphology, Fabaceae; Sanchez 1986). These six
species were found in 25% of the stations. The average proboscis length of
overwintering monarchs was 15.9 mm (S.E. = 0.09, n = 100). Male
monarchs had significantly longer probosces than females (male average =
16.2 mm, S.E. = 0.11, n = 50; female average = 15.6, S.E. = 0.11, n = 50, p <
0.001). Further analysis did not include these species.
I found that the core area of the Sierra Chincua reserve has many
canopy openings (Table 4.1). Forty-three percent of the stations had some
degree of human disturbance (categories C, E, and F), 41% of the stations
were naturally closed (A), and 16% were open or partially opened by natural
causes (B and D). These percentages are based on data from 399 stations. I
originally had 404 stations but three stations did not have plants in flower;
one station could not be placed in any of the categories since it was 1/4
naturally open, 1/4 artificially opened and 50% closed canopy.
I found a total of 21 plant species flowering in the area, including 14
in the Asteraceae and four in the Lamiaceae. Ninety-nine percent of the
stations had plants in flower and most of the species were found in both
closed (17 species) and opened (18 species) sites (Table 4.1). There was not a







significant difference between closed and opened sites in the number of
plants observed flowering (G-test = 0.01, d. f. = 1, p = 0.90). I used the eight
most common species (i.e. those that were found more than 10 times in the
stations, Table 4.1) to test if they occurred more frequently in closed or
opened sites. I found that Viola grahami Benth. (Violaceae) and Senecio

Qrenantoides A. Rich (Asteraceae) occurred more frequently in closed sites
(A-C), while Bidens trinlinervia H. B. K. (Asteraceae) occurred more
frequently in opened stations (D-F; G-test = 41.0, d. f. = 7, p < 0.001). No
significant difference was found for the other five species between closed
and opened sites.




Discussion


This study showed that the core area of the Sierra Chincua MBSBR
already has a substantial number of areas with human-induced
disturbances. Logging extractions that occurred before the MBSBR was
created produced artificial openings in the forest such that 43% of the
stations in this study had some degree of disturbance (Table 4.1). Thus,
contrary to Hoth's (1993) observation that the core areas of the MBSBR do
not have evidence of human disturbances, in this chapter I showed that the
core area of the Sierra Chincua already has many artificial openings, even
though it is considered one of the most pristine overwintering sites (Calvert
et al. 1989; Calvert and Lawton 1993). Additional field observations indicate
that it may take several decades for the Oyamel forest to recover from
human perturbations. High altitude ecosystems tend to have low primary
productivity compared to middle elevation pine-oak forests due to either low