Citation
Challenging the Paradigm of Monarch Migrations

Material Information

Title:
Challenging the Paradigm of Monarch Migrations Behavioral Complexity and Isotopic Variation of the Eastern North American Population of Danaus Plexippus
Creator:
Chaffee, Carol L
Place of Publication:
[Gainesville, Fla.]
Florida
Publisher:
University of Florida
Publication Date:
Language:
english
Physical Description:
1 online resource (129 p.)

Thesis/Dissertation Information

Degree:
Doctorate ( Ph.D.)
Degree Grantor:
University of Florida
Degree Disciplines:
Zoology
Biology
Committee Chair:
WAYNE,MARTA L
Committee Co-Chair:
BURLEIGH,JOHN GORDON
Committee Members:
MACFADDEN,BRUCE J
DANIELS,JARET C
REED,DAVID LEE
BAKER,SHIRLEY M
Graduation Date:
12/13/2013

Subjects

Subjects / Keywords:
Breeding ( jstor )
Butterflies ( jstor )
Carbon ( jstor )
Genetic variation ( jstor )
Genetics ( jstor )
Isotopes ( jstor )
Monarchs ( jstor )
Overwintering ( jstor )
Population genetics ( jstor )
Population migration ( jstor )
Biology -- Dissertations, Academic -- UF
behavior -- butterfly -- genetic -- isotope -- migration -- monarch -- population -- stable -- structure
Miami metropolitan area ( local )
Genre:
bibliography ( marcgt )
theses ( marcgt )
government publication (state, provincial, terriorial, dependent) ( marcgt )
born-digital ( sobekcm )
Electronic Thesis or Dissertation
Zoology thesis, Ph.D.

Notes

Abstract:
Movement is one of the key factors that influences evolution. My dissertation uses the unique migration behavior of the monarch butterfly (Danaus plexippus) to investigate the effects of migratory behavior on population structure. Although monarchs are not yet endangered as a species, their migration is facing a variety of threats. Migration may be key to the stability of the species, for example by preserving or redistributing genetic variation. Understanding how movement influences the distribution of genetic variation across an organism’s range is particularly important in conservation genetics. I used an approach that integrated genetic and stable isotope analyses to answer three research questions: 1) To what degree is the resident population in South Florida connected with the main migratory population? 2) Has the summer breeding range expanded northward? 3) Do monarchs from different origins mix in the Mexican overwintering colonies? I found that migrants from throughout the summer range enter the resident South Florida population, but that the degree of connectivity is different for males and females. Males immigrating to South Florida also appear to be in worse nutritional condition than either butterflies originating in the resident population or migrants to Mexico, but are larger than resident males. My results indicate that monarchs are breeding well north of the published breeding limit, and that monarchs originating north of the putative breeding limit are successfully migrating to the overwintering colonies in Mexico. Finally, genetic variation is homogeneously distributed across the overwintering colonies in Mexico, but there are subtle differences between the overwintering colonies in origins of the butterflies at each colony. Thus, overwintering colonies are not equivalent. Overall, my dissertation shows that monarch movement produces patterns with greater fine-scale complexity than had previously been described. Because of the conservation implications this complexity, future work investigating the processes that produce these patterns is warranted. ( en )
General Note:
In the series University of Florida Digital Collections.
General Note:
Includes vita.
Bibliography:
Includes bibliographical references.
Source of Description:
Description based on online resource; title from PDF title page.
Source of Description:
This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Thesis:
Thesis (Ph.D.)--University of Florida, 2013.
Local:
Adviser: WAYNE,MARTA L.
Local:
Co-adviser: BURLEIGH,JOHN GORDON.
Electronic Access:
RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2014-12-31
Statement of Responsibility:
by Carol L Chaffee.

Record Information

Source Institution:
UFRGP
Rights Management:
Applicable rights reserved.
Embargo Date:
12/31/2014
Resource Identifier:
907646414 ( OCLC )
Classification:
LD1780 2013 ( lcc )

Downloads

This item has the following downloads:


Full Text
xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID E1C9U7DSK_AHZGSJ INGEST_TIME 2014-05-13T19:12:09Z PACKAGE UFE0046095_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES



PAGE 1

1 CHALLENGING THE PARADIGM OF MONARCH MIGRATIONS: BEHAVIORAL COMPLEXITY AND ISOTOPIC VARIATION OF THE EASTERN NORTH AMERICAN POPULATION OF DANAUS PLEXIPPUS By CAROL LYNN CHAFFEE 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 2013

PAGE 2

2 2013 Carol Lynn Chaffee

PAGE 3

3 To my wonderful family and friends for all their support

PAGE 4

4 ACKNOWLEDGMENTS It would have been impossible to complete my dissertation without an immense amount of assistance and support from many people. I wish to thank my advisor, Dr. Marta Wayne, for all her assistance in plann ing and executing my dissertation project. In particular, I must thank Marta for continually helping me get through the multitude of little crises that developed along the way to completing my dissertation, since this project would never have been finished without her optimistic assurance. She completely earned her title of "dissertation therapist"! In addition to Marta, I need to thank my wonderful committee Shirley Baker, Gordon Burleigh, Jaret Daniels, Bruce MacFadden, and David Reed for their support an d guidance throughout my graduate career. Collections at the Mexican overwintering colonies would have been nearly impossible, and much less fun, without the efforts of my collaborator Antonio Gonzal ÂŽ z Rodrigu ÂŽ z. Antonio drove back and forth across the mou ntains of central Mexico negotiated with the local ejidos at the colony sites, provided lab space for DNA extractions, and even arranged for us to stay in a lovely home. I also wish to thank representatives of the e jidos of Cerro Prieto, El CapulÂ’ n, Los S aucos, and Ocampo at the monarch butterfly overwinter ing colonies in Mexico for their advice and assistance in collecting specimens. Lincoln Brower also graciously provided additional specimens he collected at the Gota de Agua colony in Mexico advice on c ollection sites in South Florida, and immensely helpful discussions that shaped my dissertation project. For the South Florida specimens, I must thank Elane and Ron Nuehring of the Miami Blue Chapter of the North American Butterfly Association for letting me use their garden as a collection site. In Canada, I thank Valerie Minelga and the entire staff at Point Pelee National Park in Ontario for their assistance in collecting monarchs within the national

PAGE 5

5 park, and for allowing me to present my research to th e public with a talk at the park visitor's center. Despite unusually cold winter conditions in Houston, Nancy Greig at the Cockrell Butterfly Center, and Mary Anderson and Dora Valdez at Oak Forest Elementary School deserve thanks for helping me to find a small group of lab raised specimens that I could use for testing and evaluating the stable isotope protocol. I thank Engy Mageed for her consistent dedication to getting the DNA lab work completed, and Alex Pinter and Aishwarya Potdar for their persistence in transferring sections of monarch wing tissue to tiny metal capsules for stable isotope analysis. Sweta Pattanaik helped keep me going with endless runs for coffee, and her desire to learn For helping me to maintain my sanity with periodic dinners, gam e nights, and general awesome support, I thank my friends Kin Han, Melania Lopez Castro, Rosana Zenil, Ping Huang, and Natalie Wright. My parents, Maggie and Dick Chaffee, may not have completely understood my project, but I would never have made it throug h graduate school without their unwavering support. I thank all my family and friends who are too numerous to list, but who have helped me in ways both large and small during this journey. Finally, I never got to see them as often as I would have liked but I thank my niece and nephew, Meg and Will Chaffee, for continually reminding me why I wanted to become a scientist in the first place, demonstrating the importance of asking lots of questions, and inspiring me to search for the answers!

PAGE 6

6 TABLE OF CON TENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ ... 4 LIST OF TABLES ................................ ................................ ................................ ............. 8 LIST OF FIGURES ................................ ................................ ................................ ........... 9 LIST OF ABBREVIATIONS ................................ ................................ ............................ 11 ABSTRACT ................................ ................................ ................................ .................... 12 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ ..... 14 Importance of Movement ................................ ................................ ......................... 14 Relati onship Between Movement and Population Structure ................................ .... 16 Conservation Genetics ................................ ................................ ............................ 18 Tracking Movement Using Stable Isotopes ................................ ............................. 21 Study System: Monarch Butterflies ................................ ................................ .......... 23 Research Questions and Approach ................................ ................................ ......... 30 2 SOUTH FLORIDA RESIDENT POPULATION ................................ ........................ 33 Methods ................................ ................................ ................................ ................... 35 Specimen Collection ................................ ................................ .......................... 36 Natal Origin Analysis ................................ ................................ ......................... 36 Wing S ize Analysis ................................ ................................ ............................ 37 Larval Condition Analysis ................................ ................................ .................. 38 Results ................................ ................................ ................................ ..................... 38 Discussion ................................ ................................ ................................ ............... 39 3 CANADIAN POPULATION ................................ ................................ ...................... 57 Method s ................................ ................................ ................................ ................... 58 Specimen Collection ................................ ................................ .......................... 58 Natal Origin Analysis ................................ ................................ ......................... 59 Wing Size Analysis ................................ ................................ ............................ 60 Results ................................ ................................ ................................ ..................... 60 Discussion ................................ ................................ ................................ ............... 61 4 MEXICAN OVERWINTERING COLONIES ................................ ............................. 70 Methods ................................ ................................ ................................ ................... 74 Specimen Collection ................................ ................................ .......................... 75

PAGE 7

7 DNA Analysis ................................ ................................ ................................ .... 75 Natal Origin Analysis ................................ ................................ ......................... 77 Wing Size Analysis ................................ ................................ ............................ 78 Results ................................ ................................ ................................ ..................... 79 Discussion ................................ ................................ ................................ ............... 80 5 SYNTHESIS ................................ ................................ ................................ ............ 97 Resident Populations as Alternate Migratory Destinations ................................ ...... 97 Phenotypic Variation in the Mexican Overwintering Colonies ................................ 99 Differentiation of the Mexican Overwinte ring Colonies ................................ .......... 100 Future Directions ................................ ................................ ................................ ... 103 APPENDIX: ISOTOPE FRACTIONATIO N IN DIFFERENT WING COLORS OF MONARCH BUTTERFLIES ( DANAUS PLEXIPPUS ) ................................ ........... 105 LIST OF REFERENCES ................................ ................................ .............................. 114 BIOGRAPHICAL SKETCH ................................ ................................ ........................... 129

PAGE 8

8 LIST OF TABLES Table page 2 1 Collection site and specimen details ................................ ................................ ... 55 2 2 Stable isotope profile data ................................ ................................ ................... 55 2 3 Natal origin results ................................ ................................ ............................... 56 3 1 Sightings, by year of observation, from the Journey North project suggesting northward expansi on of the monarch breeding range. ................................ ........ 68 3 2 Collection site and specimen details ................................ ................................ ... 68 3 3 Stable isotope profile data ................................ ................................ ................... 69 4 1 Overwintering colonies at which specimens were collected. ............................... 95 4 2 Summary of specimens collected and analyzed. ................................ ................ 95 4 3 Summary of molecular diversity statistics ................................ ........................... 95 4 4 Pairwise F ST values for Mexican overwintering colonies ................................ ..... 95 4 5 AMOVA under the assumption that each overwintering colony is a separate population ................................ ................................ ................................ ............ 96 A 1 Stable isotope data for different colored wing sections. ................................ .... 113

PAGE 9

9 LIST OF FIGURES Figure page 1 1 Heavy isotope depletion gradients for carbon and hydrogen in monarch wing tissue ................................ ................................ ................................ ................... 31 1 2 Major migratory flyways of the fall monarch butterfly migration in eastern North America ................................ ................................ ................................ ...... 32 2 1 Major migratory flyways of the fall monarch butterfly migration in eastern North America ................................ ................................ ................................ ...... 47 2 2 Landmarks used to measure wing width ................................ ............................. 48 2 3 Distribution of F statistics generated under the null hypothesis for permutation analysis of wing size ................................ ................................ ........ 48 2 4 Distribution of F statistics generated under the null hypothesis for permutation analysis of nitrogen isotope profiles ................................ ................ 49 2 5 Appr oximate natal origins of specimens collected in South Florida .................... 50 2 6 Wing size differences between monarchs collected in South Fl orida ................. 51 2 7 Non resident males collected in Florida tend to be in worse condition than resident males ................................ ................................ ................................ ..... 52 2 8 Non resident males collected in Florida are in significantly worse condition than males collected at the Mexican overwintering colonies ............................... 53 2 9 Revised migratory flyway map illustrating the connectivity between each of the major migratory routes and the South Florid a population ............................. 5 4 3 1 Landmarks used to measure wing width ................................ ............................. 65 3 2 Map of approximate origins of monarch butterflies collected at Point Pelee National Park ................................ ................................ ................................ ....... 66 3 3 Size differences of specimens collected at Point Pelee and Mexico ................... 67 4 1 Major migratory flyway s of the fall monarch butterfly migration in eastern North America ................................ ................................ ................................ ...... 86 4 2 Heavy isotope depletion gradients for carbon and hydroge n in monarch wing tissue ................................ ................................ ................................ ................... 87 4 3 Map of overwintering colony sites ................................ ................................ ....... 88

PAGE 10

10 4 4 Landmarks used to measure wing width ................................ ............................. 88 4 5 Network illustrating the distribution of genotypes among the overwintering colonies in Mexico ................................ ................................ ............................... 89 4 6 Network illu strating the distribution of genotypes by sex ................................ ..... 90 4 7 Neighbor joining tree illustrating t he clear separation of monarchs from their sister species, the queen butterfly ................................ ................................ ....... 91 4 8 Origin differences at each overwintering colony site ................................ ........... 92 4 9 Distribution of wing sizes at Mexican overwintering colonies .............................. 93 4 10 Relationship between !D and wing size at the different overwintering colony sites ................................ ................................ ................................ ..................... 94 A 1 Heavy isotope depletion gradients for carbon and hydrogen in monarch wing tissue ................................ ................................ ................................ ................. 111 A 2 Mean isotope fractionation in different colors of monarch butterfly wings is significantly different for carbon ................................ ................................ ......... 112 A 3 Mean isotope fractionation in different colors of monarch butterfly wings is not significantly different for nitrogen ................................ ................................ 113

PAGE 11

11 LIST OF ABBREVIATIONS A MOVA Analysis of molecular variance is a statistical test that evaluates the degree to which genetic variation is explained among and between groups. A NOVA Analysis of variance is a statistical test that evaluates the degree to which variation in data is explained by different factors. I BD Isolation by distance describes the distribution of genetic variation where increasing geographic distance is positively correlated with incre asing genetic distance. P CR The polymerase chain reaction is a laboratory technique for amplifying many copies of a target segment of DNA. V PDB Vienna Pee Dee Belemnite is the international standard reference against which carbon stable isotopes are measur ed. V SMOW Vienna Standard Mean Ocean Water is the international standard reference against which hydrogen stable isotopes are measured.

PAGE 12

12 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfi llment of the Requirements for the Degree of Doctor of Philosophy CHALLENGING THE PARADIGM OF MONARCH MIGRATIONS: BEHAVIORAL COMPLEXITY AND ISOTOPIC VARIATION OF THE EASTERN NORTH AMERICAN POPULATION OF DANAUS PLEXIPPUS By Carol Lynn Chaffee December 2013 Chair: Marta L. Wayne Major: Zoology Movement is one of the key factors that influences evolution. My dissertation uses the unique migration behavior of the monarch butterfly ( Danaus plexippus ) to investigate the effects of migratory behavior on population structure. Although monarchs are not yet endangered as a species, their migration is facing a variety of threats. Migration may be key to the stability of the species for example by preserving or redistributing genetic variation. Understanding how moveme nt influences the distribution of genetic variation across an organism's range is particularly important in conservation genetics. I used an approach that integrated genetic and stable isotope analyses to answer three research questions: 1) To what degree is the resident population in South Florida connected with the main migratory population? 2) Has the summer breeding range expanded northward? 3) Do monarchs from different origins mix in the Mexican overwintering colonies? I found that migrants from throu ghout the summer range enter the resident South Florida population, but that the degree of connectivity is different for males and females. M ales immigrating to South Florida also appear to be in worse nutritional condition than either butterflies originat ing in the resident population or

PAGE 13

13 migrants to Mexico but are larger than resident males My results indicate that monarchs are breeding well north of the published breeding limit, and that monarchs originating north of the putative breeding limit are succ essfully migrating to the overwintering colonies in Mexico Finally, genetic variation is homogen e ously distributed across the overwintering colonies in Mexico but there are subtle differences between the overwintering colonies in origins of the butterfli es at each colony. Thus, overwintering colonies are not equivalent. Overall, my dissertation shows that monarch movement produces patterns with greater fine scale complexity than had previously been described. Because of the conservation implications this complexity future work investigating the processes that produce these patterns is warranted.

PAGE 14

14 CHAPTER 1 INTRODUCTION Movement is a key behavior that affects evolution. This behavior can be classified into three basic categories: daily movement, dispersal, and migration (Rabb 1979) Because movement drives gene flow in animal populations, each type of movement has different implications for evolva bility the degree to which a species is able to produce the genetic variation that drives the capacity for change (Houle 1992; Houle et al. 1996; Kirschner & Gerhart 1998). Animal movement is particularly important for determining how genetic variation is distributed across a species' range, since it is one factor governing whether the pool of potential mates comprises the entire species across the whole range or just a small group clustered in a limited area. My dissertation centers on investigating how m igratory movement in particular affects population structure. Importance of Movement Both daily and dispersal movement describe the behavior of individuals within or between population s whereas migration describes the behavior of an entire local population. On a daily basis, individuals move to find food, water, and shelter, and to avoid predators. During reproductively active periods they also search for mates, and reproduction sites. The se daily movement s determine a range compr ising an organism's habitat (Hughes 1979). Both the characteristics of the landscape in which suitable patches of habitat are found and individual variation determine the amount and type of movement by organisms of the same species (Bowler & Benton 2005; T urlure et al. 2011). Daily movement behavior determines the number of potential mates that an animal encount ers, and thus defines the demes local interbreeding group s of

PAGE 15

15 organisms of the same species (Gilmour & H eslop Harrison 1954) across a species' range In contrast, dispersal and migration behaviors determine gene flow between demes. The degree of movement in a population thus can affect the distribution of genetic variation in a population As a result of movement, variation may be distributed relative ly homogeneously through the population, or it may be partitioned among demes that are differentiated from each other. The literature on animal movement is filled with various views on how (or whether ) to differentiate between d ispersal and migration, and the debate on how to define these behaviors has been ongoing for decades. Heape (1931 ) defined migration as requiring that animal s return to the starting location, with one way movements categorized as dispersal; Williams (1957) held that migration was mov ement between habitats regardless of whether this move was temporary or permanent ; Taylor and Taylor (1977) argued that dispersal and migration are fundamentally similar r esponses to adverse conditions ; Fahrig and Merriam (1994) viewed dispersal as moveme nt of animals between local populations in different habitat patches. I will be following the definitions used by Dingle (1972) D ingle defined dispersal a s movement away from the emergence site and thus among demes D ispersal distance is often driven by population density, relative to the carrying capacity of the natal area. Thus, dispersal behavior influences evolvability by determining the amount of gene flow that occurs between demes in each generation. In contrast, Dingle (1972) defined seasonal migra tion a s a life history adaptation to recurring shifts in climate and resource availability that involves the movement an entire population to a refuge site or sites Migration behavior therefore may bring geographically separate demes into contact,

PAGE 16

16 althoug h the potential for gene flow between those demes depends on how reproductively active periods fit into the migration cycle In my dissertation, I am focused on the effects of migration and its intersection with the reproductive period on population struct ure. For insects in particular Rabb (1979) describes four major types of sites between which movement occurs: feeding, mating, reproduction, and refuge. Mating sites include the sites of courtship and mating, whereas r eproduction sites include the sites of oviposition, as well as lar val, pupal and adult emergence R efuge sites are either daily or seasonal sites. The types of sites, their number, and spatial distribution vary by species, and sometimes by populations in different geograph ic areas within a species' range. Daily movement may occur between feeding, mating, oviposition, and daily refuge sites. Dispersal involves movement beyond the daily range Though dispersal may similarly be between feeding, mating, oviposition, or refuge s ites, dispersal should be among populations or demes of a species Insect migration on the other hand, generally involves movement to or from a refuge site, although in some cases, such as with Mormon cricket populations that migrate from feeding site t o feeding site ( Srygley & Lorch 2011 ), it can also involve movement between other types of sites Relationship B etween Movement and Population Structure Movement patterns between demes within a geographically distributed population affect genetic populatio n structure. R andom movement between demes leads to panmixia. In contrast, population structure develops as d emes accumulate genetic differences with respect to one another when movement between them is either absent or consistently asymmetric. The null expectation for p opulation structure is a straightforward isolation by distance (IBD) pattern (Wright 1943) where the degree of

PAGE 17

17 genetic differentiation between demes is correlated with the geographic distance between them. Barriers to movement, however, m ay limit gene flow between some demes (reviewed by Manel et al. 2003) Vicariant barriers are geographic features that limit movement between demes; behavioral barriers, such as assortative mating or willingness to move between fragments of habitat patches can limit movement between demes even in the abs ence of vicariant barriers When barriers exist p opulation structure may show more complex patterns than IBD such as locally isolated demes, source and sink areas, and metapopulatio ns Depending on the degree of connectivity and ecological differences between demes, demes may accumulate differences in both the amount and character of genetic variation. Ecological differences, for example, may favor certain variants over others, so that a llele frequencies may differ across demes S maller, more isolated demes may also harbor less genetic variation compared to other larger or more highly connected demes In the most extreme case, when ecological differences are large and connectivity between demes is low, speciation may eventually occur ( Darwin 1859 ). Islands are classic examples of locally isolated demes ( MacArthur & Wilson 1967 ), but landscape features such as mountain ranges, rivers or fragmented habitats may also limit gene flow between demes on the same land mass ( Johnson et al. 1992 ) Source and sink patterns occur when movement consistently occurs from one set of demes into another set, but rarely occurs in the opposite direction (Pulliam 1988). Anothe r way to think about a collection of demes is as a metapopulation Rather than describing how genetic variation is distributed across a stable collection of demes, a metapopulation consists of a collection of demes across which the population is relatively stable, but the individual

PAGE 18

18 demes experience periodic local extinction and col onization events (reviewed by Hanski 1998 ). A variety of complex patterns in the distribution can thus develop. There is no clear association between specific complex patterns an d particular types of movement, so the relationship between movement behavior and the distribution of genetic variation for a species must be determined empirically. Both dispersal and migration are means by which gene flow occurs, but the two type s of mov ement ha ve distinct effects on population structure. For example, i n the absence of vicariant barriers to dispersal, unidirectional movement away from the natal area leads to clear patterns of IBD that are primarily driven by the physiological ability of a n organism to m ove between patches of habitat ( Wright 1943 ) For seasonal migra tion however, entire demes move between geographic areas, so IBD patterns created by dispersal are maintained only when behavior prevents or limits mixing between demes (Wright 1946). Examples of both intermingling (Combreau et al. 2011) and behavioral separation (Irwin et al. 2011) have been found in well studied migratory bird taxa. Conservation Genetics N ew alleles are added to individual demes in only two ways: mutation or immigration. Mutation rates can be expected to be consistent across all demes, barring environmental influences on mutation rate that vary among demes. T he number of new alleles introduced by imm igration will be a function of the degree of connectivity between a deme and the rest of the population. Thus, well connected demes will be more likely to add new variants than more isolated demes. The equilibrium fixation index ( F = 1/[ 1 + 4 Nm ] where N i s population (deme) size, and m is t he migration rate) is a measure of differentiation, since it indicates the probability that different neutral

PAGE 19

19 alleles will be fixed in two populations (demes). Since F approaches zero as the number of migrants approaches 1 migrant per generation ( Nm = 1), this index illustrates that even a relatively low level of migration keeps demes "well connected" (Hartl & Clark 2006 ) Genetic variation can be removed by non random mating, natural selection or genetic drift Non random mating occurs when individuals of a population do not have an equal probability of mating with the entire pool of potential mates in the population. This may occur due to assortative mating (either positive like with like or negative like with u nlike ), or to barriers within the population that prevent access to the complete pool of potential mates. Genetic variation is not necessarily removed by non random mating, but it reduces the mixing of variants within the population. Both genetic drift and natural selection directly remove genetic variation. Natural selection removes variants by reducing the number of offspring for carriers of deleterious alleles, and increasing the number of offspring for carriers of advantageous alleles. Genetic drift, in contrast to selection, is a neutral process in which alleles are removed from the population by random chance. All three processes are acting at any given time, in any population or deme. Selection and drift tend to operate in opposition, with their relat ive strengths varying depending on the population size ( N ), the effective population size ( N e ) a measure of the size of the population that is successfully contributing genetic variants to future generations (Wright 1938) and the strength of selection ( s ). Selection dominates when 4 N e s >> 1, whereas drift dominates when 4 N e s << 1 (Kimura 1983) For example, natural selection becomes less efficient as the size of the breeding pool decreases, so this process has a lower effect on isolated demes.

PAGE 20

20 In isolated demes, both genetic drift and non random mating are key processes by which genetic variants are removed. Because of the interplay between genetic drift, selection, and population size discussed earlier, in an isolated deme, genetic drift is equall y likely to remove both detrimental and small effect beneficial alleles. The net effect is that more isolated demes have a greater probability of less genetic variation as well as lower quality alleles, than their well connected counterparts. Inbreeding d oes not reduce genetic variation ( and thus evolvability ) per se but it does reduce the amount of heterozygosity by increasing the proportion of alleles that are identical by descent (Hšglund 2009). As more homozygous individuals appear inbreeding depress ion decreased overall fitness in the deme due to inbreeding can occur as rare deleterious alleles are expressed more frequently ( Bruce 1910 ), and advantageous heterozygotes appear less frequently ( East 1910 ). Thus, any negative effects of inbreeding accumu late over time as a smaller proportion of heterozygous individuals appear in each new generation. S mall, isolated demes are more susceptible to fixation of mildly deleterious alleles due to genetic drift and weak selection as well as inbreeding depression (Lynch & Gabriel 1990) These mildly deleterious alleles increase the genetic load of the deme, and, given time, may go to f ixation within the deme. Once a deleterious allele is fixed, there is no genetic variation on which selection can operate to give a n advantage to a proportion of the deme. Thus, a s mildly deleterious alleles fix within a deme, the principle of Muller's ratchet applies to the fitness of the deme: the maximum fitness possible in that deme is reduced with respect to the population as a w hole (Felsenstein 1974 ; Muller 1963 ). If a sufficient number of deleterious alleles accumulate within a

PAGE 21

21 deme, these mutations may affect the size of the deme due to the action of natural selection As the deme shrinks, the effects of inbreeding depression and genetic drift increase, which, in turn, increases the genetic load and the probability of fixing deleterious alleles (Lynch & Gabriel 1990) Lynch and colleagues (1995) termed this cycle "mutational meltdown", and argued that once this cycle crossed a threshold of genetic load that is correlated with reproductive rate and carrying capacity of the deme that extinction of the dem e was essentially unavoidable. Therefore, if a population as a whole contains a large number of relatively isolated demes, the n the population itself may go extinct as a result of processes acting independently on the individual demes Tracking Movement Using Stable Isotopes Direct observation of movement is often both labor intensive and expensive. Alternatively, one may attempt to indirectly observe movement by studies of genetic variation. However, genetic approaches have been criticized by Cockerham and Weir (1993 ). Although they supported the continued use of approaches that estimate population structure based on genetic diff erentiation such as the analysis of F ST their concern was that such measures rely heavily on an underlying migration model. A s with all models there is no way to differentiate between violation of underlyin g assumptions and real signal Stable isotope ana lysis provides a n alternat ive for measuring movement that requires neither an underlying migration model, nor the need to directly observ e movement Stable isotopes of carbon and hydrogen vary in a predictable pattern across North America (Hobson & Wassenaar 1997; Hobson et al. 1999) In contrast to radioactive isotopes, stable isotopes are variants of an element that do not degrade over time, but which vary in atomic weight. In the carbon and hydrogen isotope regimes

PAGE 22

22 used to study animal movements, the lighter isotopes (H and 12 C) are far more common than the heavy isotopes (Deuterium (D) has an abundance of 0.015% and 13 C has an abundance of 1.11% ) (Sharp 2007) Because the absolute quantity of rare isotopes is so low, using a ratio rather than an a bsolute quantity increases the precision of the measurement (Sharp 2007). T he stable isotope profile of a sample is measured as the relative difference in the ratio of heavy to light isotopes in the sample compared to the ratios of these isotopes in an int ernational standard e.g., for hydrogen, !D = (D sample /H sample D standard /H standard ) / (D standard /H standard ) x 1000 Hydrogen isotopes are measured relative to ocean water (Vienna Standard Mean Ocean Water, VSMOW); carbon isotopes are measured relative to a mineral standard (Vienna PeeDee Belemnite, VPDB) (Coplen 1994) The isotope ratios found in these international stand ards are used as the zero point. Samples with ratios of heavy to light isotopes that are greater than the standard will have positive delta values ( i.e. are enriched for the heavy isotope) while samples with ratios that are less than the standard will have negative delta values ( i.e. are depleted for the heavy isotope) The geographic pattern of hydrogen ratios (Figure 1 1) varies based on precipitation (meteoric water), although temperature, altitude, latitude, position on the continent, and amount of rainfall all influence the ratio of deuterium to hydrogen ( Sharp 2007 ) Because there is a seasonal effect on hydrogen stable isotope ratios, focusing on the ratios generated during the growing season is preferable when using hydrogen isotopes to study animal movements (Meehan et al. 2004).

PAGE 23

23 There is also predictable variation in carbon isotopes based on geographic region, but the geographic patterns vary by type of plant (Sharp 2007) Plants fractionate here meaning preferentially incorporate 12 C during photosynthesis. The degree of fractionation v aries by the type of photosynthesis (C 3 C 4 or CAM) and by the amount of carbon dioxide in the surrounding air. Hobson and colleagues (1999) demonstrated that the carbon ratios of milkweed plants vary in a geographically predictable manner (Figure 1 1) across eastern North America Due to their associations with geography, s table isotopes of hydrogen and carbon have been used since the late 1990s to investigate migratory behavior (e.g., Hobson 1999; Hobson & Wassenaar 1997; Pinnegar & Polunin 2000; Rube nstein & Hobson 2004). By selecting tissues with little to no turnover, such as bones, wing membranes and feathers, analysis of stable isotopes are useful intrinsic markers for tracking migration without the need to capture the same organism at both ends of the migration path. In contrast mark recapture methods which have been extensively used to study migration, require capturing the same individual twice. Accordingly, mark recapture can be limited by recovery of only a small fraction of the marked orga nisms (Lewis et al. 1997). Studies in a variety of taxa have combined genetic analyses with stable isotope analyses to address diverse issues, including the flight patterns of seabirds (GomŽz D’az & Gonz‡lez Sol’s 2007), recent connectivity patterns in end angered fish (Cook et al. 2007), and trophic partitioning by different pods of killer whales (Foote et al. 2009). Study System: Monarch Butterflies Monarch butterflies ( Danaus plexippus ) present a unique system in which to study the effects of seasonal migration on population structure. The phenomenon of the

PAGE 24

24 annual monarch butterfly migration in North America has attracted scientific interest since the mid 1800s. Despite this long histor y of scientific inquiry, however, there is still a great deal that is unknown about how these migration movements affect population structure in monarchs (Urquhart & Urquhart 1977; Wassenaar & Hobson 1998). While the monarch genus Danaus most likely split from its sister Tirumala in the late Pliocene (~4.9 million years ago), D. plexippus is believed to have diverged from its closest relative, the queen butterfly ( D. gilippus ) as recently as 40,000 years ago (Lushai et al. 2003). Native populations of mona rchs are found throughout the Americas and the Caribbean. Major introduced populations are found in Hawaii (Stimson & Kasuya 2000; Stimson & Meyers 1984), Australia (Hughes & Zalucki 1984; Zalucki & Suzuki 1987) and New Zealand (Pawson & Berndt 2004), with small, introduced populations also found scattered on a number of Pacific (Zalucki & Clarke 2004) and Atlantic islands (Hilburn 1989; Neves et al. 2001; Strecker & Wilkens 2000). The monarch life cycle involves five larval instars, pupation and an adult phase. While highly temperature dependent, monarchs in the wild generally live six to eight weeks (Urquhart 1987). Monarchs eclosing in the fall months, however, emerge from pupation and enter reproductive diapause for migration to overwintering colony sit es. The generally accepted paradigm is that monarchs i n the eastern United States and southern Canada roughly follow the major f lyway routes shown in Figure 1 2 (US Forest Service 2008) to overwinter at a small cluster of colonies in central Mexico (GarcÂ’a Serrano et al. 2004; Slayback et al. 2007; Urquhart & Urquhart 1977; Williams et al. 2007), while monarchs west of the Rocky Mountains migrate to a series of small colonies scattered along the California coast (Dingle et al. 2005; Leong 1990; Leong et

PAGE 25

25 al. 2004). These migrants remain in diapause for the duration of the winter until exiting diapause in the early spring. Monarchs then mate within the overwintering colonies, and begin the northward migration before laying eggs in northern Mexico or southern Texas to complete the life cycle. Thus, the overwintering generation may live nearly ten times longer than its spring and summer counterparts: six to eight months, rather than six to eight weeks. Monarch mating has often been described as a un ique breeding system involving forced copulation by males, rather than female choice ( Brower et al. 2007; Pliske 1975; Rothschild 1978 ). Males are ardent, and will attempt copulation with any butterfly including other males that they are able to take down ( Hill et al. 1976; Oberhauser & Frey 1999). Other butterflies of the Danainae subfamily use elaborate courtship rituals that involve pheromones produced by alar glands on the hindwings of male butterflies and distributed by hairpencil structures that can b e everted from the abdomen (BopprÂŽ 1993). These structures however, are reduced in Danaus plexippus compared with other Danainae, which some researchers have used as evidence that males do not court females (BopprÂŽ 1993 ; Pliske 1975 ). Although mating beha vior has been studied in the overwintering colonies of both Mexico and California, and in captive conditions, less is known about mating behavior throughout the summering range. Both Frey (1999) and S olensky (2004), however, demonstrated that f emales exert a degree of choice in the mating process by manipulat ing the position of their abdomen to control male mating success Females have been shown to vary the degree of these resistance behaviors based on male size (V an Hook 1993), wing

PAGE 26

26 coloration (Davis et a l. 2007), and previous mating (Frey 1999) suggesting that female choice does indeed exist in monarchs. Monarchs tightly cluster into colonies that cover a mere 75 km 2 of high altitude oyamel fir forest scattered across multiple peaks of the Transvolcanic Belt in central Mexico (Slayback et al. 2007 ). M ost estimate s indicate that over 10 million monarchs congregate at each colony in this small area (Brower et al. 1977; Calvert & Brower 19 86; Slayback & Brower 2007). Overwintering survival is tightly coupled with the specific microclimate found at each colony site ( Bojorquez Tapia et al. 2003 ). Key characteristics for successful overwintering include temperature, tree density, degree of canopy cover, and composition of understory vegetation (Alonso Mejia et al. 1992; Alonso Mejia et al. 1993; Alonso Mejia et al. 1998); t hese parameters vary at different colony sites, often due to differences in human activity (Ramirez et al. 2003 ). The forest characteristics are particularly important to protect overwintering monarchs from either freezing or desiccating (Bojorquez Taipa et al. 2003). Because these characteristics vary by site, some sites are better suited to overwintering survival than others. Although the mechanisms driving monarchs to choose particular sites are unknown, colonies consistently form at major sites, even when the forest characteristics are no longer optimal due to human activities such as logging (Garcia Serrano et al. 2004 ; Williams et al. 2007 ). It is unclear whether monarchs select colonies at random with respect to their natal origins. If monarchs do choose colonies non randomly with respect to geographic origin, then monarchs originating from areas associated with less desirable colony sites may be less likely to survive the overwintering s eason potentially leading to non random loss of particular summer demes If monarchs assort into colonies

PAGE 27

27 randomly with respect to origin, however, there is less chance of deterministic loss of demes based on selection of suboptimal habitat, and an opport unity for gene flow between different summer breeding area s or flyways Random mating across demes within the colonies has the same effect as migration with respect to increasing effective population size and hence action of selection relative to drift, th ereby shielding monarchs from Muller's ratchet or mutational meltdowns (Lynch et al. 1995) by increasing the amount of purifying selection. Since monarchs migrating northward in the spring are reproductively active, and thus live only six eight weeks, multiple generations are required to repopulate the entire range into southern Canada. Monarchs are therefore mating through out the northward migration. As successive generations move northward, the migrating population fans out across North America east o f the Rockies, so butterflies following different trajectories become increa singly geographically separated Since four to five generations may occur between overwintering generations, this geographic separation could lead to an IBD pattern that is positiv ely associated with increasing latitude. Although there have been major efforts to identify and track monarch migration through mass mark recapture and observational studies (Brindza et al. 2008; Garland & Davis 2002; Howard & Davis 2004; Urquhart & Urquha rt 1978; Urquhart & Urquhart 1979a, b), few studies have explicitly investigated how fine scale movement behavior affects population structure. One exception is the work of Eanes and Koehn (1978) who used allozymes to calculate F statistics for sites acro ss the summering range Contrary to expectations, their data support panmixia even in the summer populations rather than population structure ( i.e. of IBD) However, as discussed in the discussion of

PAGE 28

28 conservation genetics above, even occasional bouts of r andom mating across demes, such as are believed to be happening in Mexico, would be expected to obscure or even eliminate any genetic signals of population differentiation. If population structure is established along the migratory routes during the spri ng and summer breeding season it will only persist if there is some relationship between summer breeding sites, migratory routes, and the overwintering colonies. M onarchs generally follow the migra tory flyways shown in Figure 1 2 The flyways indicate the general direction of monarch butterfly flight, rather than specific routes followed by migrating butterflies, but they do highlight two major barriers that tend to separate summer breeding populations: the Great Lakes and the Appalachian Moun tains. Miller and colleagues (2011) demonstrated that the Appalachians form a porous barrier, as southward migrating monarchs can cross them. However, the degree to which monarchs breed across the Appalachians during the summer remain s unknown. Thus, summe r breeding sites seem to roughly correspond to migratory flyway s T here is mixed evidence however, regarding consistent associations between migratory routes and overwintering colony sites. Based on mass mark recapture data, Urquhart and Urquhart (1978) s uggested that western colonies were populated from summering grounds in the Great Plains, whereas central and eastern colonies were populated from summering grounds roughly east of the Great Lakes. In contrast, studies using allozymes (Eanes 1979; Eanes & Koehn 1978) and stable isotope analyses (Hobson et al. 1999 ; Wassenaar & Hobson 1998 ) found no evidence for a clear relationship between breeding grounds and overwintering colony sites. This project therefore seeks to clarify the relationship between sprin g and summer breeding grounds and

PAGE 29

29 overwintering colony sites by integrating stable isotope and genetic data for specimens collected across the full summer range and within colonies from throughout the overwintering range in Mexico In addition, non migrato ry populations are scattered across the southern United States, but neither the long term stability of these populations nor their connectivity with the main migratory population are known. Although poorly studied, there have been regular reports of cont inuously breeding, non migratory monarch populations scattered throughout the southern United States (Brower 1961, 1985; Cohen 1983; Funk 1968; Urquhart 1960). By using a combination of observations, morphometric measures, mark recapture efforts, and carde nolide fingerprinting (using chemical composition to analyze milkweed species) Knight (1998) characterized the southern Florida population and its relationship to the main migratory population. She found breeding monarchs throughout the year, with recaptu re data indicating a fairly stable number of monarchs at the two study sites, although insufficient data were collected to accurately estimate total population size. By using cardenolide fingerprinting to identify the milkweed species on which a captured m onarch was a larva, Knight also found no evidence of a spring influx of monarchs returning to North America from overwintering sites in the Caribbean or Central America, but did find that a sizeable number of southward migrants enter south Florida in the f all. Dockx and colleagues (2004) also found that some of these southward migrants continued down the Florida peninsula to cross over to Cuba, but replicated Knight's findings that monarchs did not return north once they had entered a resident population.

PAGE 30

30 R esearch Questions and Approach Even the most rapidly evolving genetic markers fail to detect most movement events occurring within a generation, but stable isotope analyses can detect such patterns. By using both genetic markers and stable isotopes, the mo vement patterns of monarch butterflies both within and across generations can be better evaluated than by using either approach alone This integrated approach can be applied to study the relationship between migratory routes and overwintering colonies, as well as the relationship between the migratory population and the resident population in south Florida. A proof of concept study showed that stable isotope techniques that h ad been successfully used to track migratory birds were equally successful in tracking migratory monarch butterflies (Hobson et al. 1999 ). Stable isotope analyses of monarchs have been used to determine monarch origins (Wassenaar & Hobson 1998, Miller et a l. 2011) but they have not been combined with genetic analyses to investigate pattern differences both within and across generations. My dissertation builds on the broad patterns that have been identified by others over the past decades to how well the f ine scale patterns that monarch movement behavior has generated fit with the generally accepted paradigm of monarch migration Chapter 2 focuses on the resident population in South Florida. Th is chapter addresses the key question of how and to what degree is this non migratory population connected with the main migratory population in eastern North America. Chapter 3 examines movement patterns along the northeastern edge of the monarch summer range in Canada, and evaluat es the degree to which range expansio n may be occurring. Chapter 4 looks at the distribution patterns of genetic variation that occur at the Mexican overwintering colonies Finally, Chapter 5 synthesizes my results to fill in some of the

PAGE 31

31 gaps in our knowledge of fine scale patterns of monarch movement behavior, and outlines future research directions suggested by my findings. Figure 1 1. Heavy isotope depletion gradients for carbon and hydrogen in monarch wing tissue. Arrows point in the direction of increasing depletion of heavy isotopes ( 13 C and D). Orange = carbon, blue = hydrogen. Based on data from Wassenaar and Hobson (1998) ; hydrogen values modified according to Meehan and colleagues (2004)

PAGE 32

32 Figure 1 2 Major migratory flyways of the fall monarch butterfly migration in eastern North America. Light orange arrows indicate the primary routes taken from summering locations throughout Canada and the United States to the overwintering colonies in central Mexico Width of the arrows is roughly correlated with number of monarchs migrating al ong a flyway. Gold arrow indicates hypothesized flyway connecting migratory population with resident South Florida population. Monarchs migrate in a general south or southwest orientation throughout the eastern range, so migrating monarchs are also found i n areas not directly covered by the arrows. Red diamonds indicate sampling locations. Due to size constraints of the map, the location of the overwintering colonies is shown northeast of its actual location in central Mexico Based on the United States For est Service ( 2008)

PAGE 33

33 CHAPTER 2 SOUTH FLORIDA RESIDENT POPULATION Alternative life history strategies are found in a wide variety of taxa, including birds (Gršnroos et al. 2012), butterflies (Velde & Van Dyck 2013), crickets (Judge et al. 2008), f rogs (Leary et al. 2005), grasshoppers (Greenfield & Shelly 1985), horseshoe crabs (Smith et al. 2013), lizards (Hews et al. 1994), and salmon (Paez et al. 2011). While the details between strategies vary as widely as the taxa in which they are observed, a key component in all cases is a trade off, often between an optimal strategy that requires "good" conditions, and alternate strategy that provides a fitne ss advantage under "bad" conditions. The relevant conditions may be environmental, physiological or a combination of both. Selection of a strategy therefore involves one or more environmental or physiological thresholds that lead to an individual following a particular strategy. For example, Velde and Van Dyck (2013) found clear trade offs in the alternative strategies employed by male speckled wood butterflies ( Pararge aegeria ) to locate mates. Territorial males perch in a sunlight patch that is defended f rom other males; non territorial males fly throughout a large area of forest to search for mates. The territorial strategy involves rapid launch and accelerate flight that is energetically costly, whereas the non territorial strategy relies on gliding flig ht that requires much less energy. In addition, food resources in a given territory are limited. Velde and Van Dyck demonstrated that only males that eclosed with sufficient lipid reserves would follow the territorial strategy, and males with insufficient lipid reserves would not follow this strategy, even when a suitable territory was available. Thus, to understand the evolutionary implications of the alternative life history strategies in a particular species, it is necessary to understand both the trade offs and the triggers involved.

PAGE 34

34 Monarch butterflies ( Danaus plexippus ) employ two contrasting life history strategies: migratory or resident. Individuals following the migratory strategy are found in the temperate zone. There are several generations throug h the spring and summer seasons. Although the average life span varies somewhat with temperature, butterflies in these generations usually have 6 8 week life spans. Monarchs that emerge in the late summer or fall, however, enter reproductive diapause immed iately after emergence, then undertake the migration to the overwintering grounds. In the spring, these butterflies exit diapause and breed before beginning the journey north. Th us, th e overwintering generation of monarchs have life spans nearly ten times as long as their spring and summer counterparts. In the eastern United States and southern Canada, these butterflies migrate during the fall along the major routes shown in Figure 2 1 to overwinter at a small cluster of colonies in central Mexico while mo narchs west of the Rocky Mountains migrate to a series of small colonies scattered along the California coast (Dingle et al. 2005; Leong 1990; Leong et al. 1995). Individuals following a resident strategy, however, neither enter diapause nor migrate, and a re reproductively active throughout the year. These individuals also have short life spans throughout the year similar the spring and summer animals from the migratory populations Monarchs following the resident strategy are found in isolated colonies sc attered across the southern United States (Dockx 2007; Knight & Brower 2009; Pyle 1999; Urquhart & Urquhart 1977). Decreasing photoperiod and lower temperature are environmental cues that trigger monarch diapause (Goehring & Oberhauser 2002), but the chang es in photoperiod and temperature at sub tropical and tropical latitudes are insufficient to trigger diapause. Thus, monarchs eclosing in a

PAGE 35

35 resident population in sub tropical or tropical latitudes are unlikely to enter diapause, and are therefore be unlik ely to switch to a migratory strategy. However, it is possible that some monarchs eclosing in to the main migratory population could enter resident populations and switch to a life history strategy that does not involve reproductive diapause. I hypothesized that natal location would be a key determinant of life history strategy: butterflies eclosing in a resident population would follow a resident strategy, while butterflies eclosing elsewhere would follow a migratory strategy. However, it is possible that m igrant animals of both sexes could join resident populations (immigration), or that northbound migratory females might mate with males from a resident population, and continue their migration (gene flow). Even for a well studied resident population in Sout h Florida (e.g., Brower 1961; Urquhart and Urquhart 1979; Knight 1998; Dockx 2002), the degree of connectivity with the migratory population is unknown External factors such as displacement via storms could bring migrating monarchs into conditions that would cause them to drop out of diapaus e and start mating with resident animals or with one another However, if internal factors such as nutritional condition are involved in determining migratory desti nation, then the trigger to switch to an alternate destination most likely occurs prior to or during migration toward the overwintering colonies. Therefore, I further hypothesized that this trigger would be condition dependent. Methods I used stable isotop e profiles to infer immigration from the main migratory population into the resident population in South Florida. Stable isotope ratios vary predictably with location, so these ratios can serve as intrinsic markers for tracing

PAGE 36

36 animal movement behavior (Hob son 1999). The carbon and hydrogen isotope ratios found in monarch wings are determined by the origin of the host plants on which animals fed as caterpillars The keratin proteins that make up the wing are stable following metamorphosis, so measuring the s table isotope ratios of wing tissue provides an estimate of the natal location of the butterfly (Wassenaar & Hobson 1998). Specimen Collection Specimens were collected from two sites in South Florida (18 specimens). For wing size comparison, specimens were collected from five overwintering colonies in Mexico (181 specimens). Live specimens were captured in South Florida, immediately placed in glassine envelopes, and stored at 20¡C for at least 12 hours prior to measurement and dissection. In Mexico specim ens were collected from the pool of dead butterflies littering the forest floor at overwintering colony sites, and sorted into glassine envelopes at the end of the day. Details regarding collection sites and specimens are shown in Table 2 1. Sex was determined b y presence or absence of the androconium on the dorsal side of the hindwing before separating wings from bodies, and wings were returned to the glassine envelopes for later stable isotope analysis. Bodies were preserved for later DNA extraction in 2.0 mL t ubes filled with 95% ethanol. Both wings and bodies were stored at 4¡C after specimens were brought to the lab. Natal Origin Analysis One hindwing from each South Florida specimen was washed in a 2:1 chloroform:methanol solution to remove surface impurities, lipids, and residual hemolymph then air dried. Small sections of hindwing (0.350 0.02 mg for hydrogen and 0.825 0.05 mg for carbon) were cut and placed into capsules (3x5mm silver capsules for hydrogen and 5x9mm tin caps ules for carbon) for stable isotope analysis.

PAGE 37

37 To reduce small fluctuations in stable isotope profiles due to varying degrees of wing pigmentation, sections were consistently excised from the black portion of the wing margin (see Appendix). Carbon and nitro gen isotope profiles were measured at the Light Stable Isotope Mass Spec Lab at the University of Florida, and hydrogen isotope profiles were measured at the Colorado Plateau Stable Isotope Laboratory at Northern Arizona University Approximate natal origi ns were assigned based on the combined isotope profiles roughly f ollowing the approach described by Wassenaar and Hobson (1998) except that the Online Isotopes in Precipitation Calculator (OIPC) plug in for Google Earth (Bowen & Revenaugh 2003; Bowen et a l. 2005; Bowen & Wilkinson 2002) was used to evaluate !D values for precipitation during the growing season, following the recommendations of Meehan and colleagues (2004) Wing Size Analysis Forewing size was measured as a proxy for body size. Width was me asured from the white spot adjacent to the body to the point between the two white spots at the tip of the forewing (Figure 2 2). Although there were slight variations in wing patterns, these two markers were consistently used to ensure that homologous sec tions of the wing were measured. Measurements were recorded to the nearest millimeter. Differences in wing size by sex and origin were analyzed in R (R Development Core Team 2008). Because sample sizes were small (3 residents and 8 immigrants), I tested th e difference in size between resident and immigrant males by permuting the variable "origin" five hundred times and observing the distribution of F statistics from the one way ANOVA of each permuted dataset. The observed value was found less than 5% of the time (Figure 2 3).

PAGE 38

38 Larval Condition Analysis The ratio of 15 N relative to 14 N indicates whether or not an animal is nutritionally stressed and this ratio has been shown to hold across a wide range of taxa including insects (McCue & Pollock 2008). Unlike hydrogen and carbon, whose profile varies with geography, nitrogen varies according to the feeding status of the individual. Animals enduring starvation are enriched for 15 N relative to 14 N, such that the higher the relative amount of 15 N, the worse the an imal's condition (Gannes et al. 1998). Feeding experiments have not been conducted on monarch butterflies; however, experiments in another Lepidopteran, the Japanese oak silkworm ( Antheraea yamamai ), are consistent with enrichment of 15 N as a signal of res ponse to larval starvation (Mizota & Yamanaka 2011). Tibbets and colleagues (2008) also found a similar pattern of enrichment pattern between larvae and adults in insects that undergo complete metamorphosis. Because the keratin proteins that comprise wing tissue are fixed after metamorphosis, I assessed larval condition using the ratio of nitrogen isotopes in adult wings as a proxy for larval condition. Data were analyzed using R (R Development Core Team 2008). As for wing size, sample sizes were small and thus I tested the difference in 15 N between resident and immigrant males by permuting the variable "origin" five hundred times and observing the distribution of F statistics from the one way ANOVA of each permuted dataset. The observed value was found app roximately 10% of the time (Figure 2 4). Results Carbon and hydrogen stable isotope profiles for specimens collected in South Florida are presented in Table 2 2. These profiles indicate that eight specimens had a natal origin outside of Florida (Figure 2 5). Alt hough both male and female specimens had natal origins in Florida, there was a significant excess of males originating outside

PAGE 39

39 of Florida (Barnard's exact test, two sided: T = 2.063, N = 18, P = 0.047, see Table 2 3). Based on natal origin, migrant monarchs tha t entered the South Florida population originated from throughout the North American breeding territory, with each of the major migratory flyway routes roughly equally represented. No differences between the two South Florida collection sites were found, a nd the two sites were treated as a single population. Migrant male monarchs that entered the South Florida population were larger than resident males (Figure 2 6; P = 0.022, see Methods for details). However, the size of migrant males collected in South Florida was not significantly different from migrant males collected in Mexico (ANOVA: F 1,96 = 1.28, P = 0.26). Migrant females collected in Mexico were also signific antly larger than resident females from South Florida (ANOVA: F 1,96 = 44.2, P < 0.0001). There was no sexual dimorphism for body size in resident monarchs (Figure 2 6). Based on the 15 N ratio, t he immigrant males tended to be in worse condition than the r esident males (Figure 2 7), though the difference was not statistically significant ( P = 0.106, see Methods for details). In addition, immigrant males collected in Florida were in worse condition on average than males collected from Mexico (Figure 2 8; ANO VA, F 1,76 = 6.76, P = 0.011). Discussion My results suggest that migratory monarchs regularly enter the small, non migratory South Florida p opulation. It is easy to understand how m onarchs following the Atlantic flyway could end up in South Florida simply by continuing to follow the Atlantic coast, rather than heading west across Georgia or the Florida peninsula to continue along the Gulf coast. However, my stable isotope data suggest that monarchs in South

PAGE 40

40 Florida originated from locations throughout the s ummer range of the migratory population. All three proposed major migratory routes contribute butterflies to the South Florida population (Figure 2 9). Monarchs may arrive in South Florida by random processes such as displacement by weather, which serves a s a null hypothesis against which to test other explanations, or they may pursue an alternate migration strategy that leads to their presence in South Florida. The generally accepted paradigm is that monarchs emerging in the fall immediately enter reproduc tive diapause and remain in this state until the early spring (Urquhart 1960). A combination of temperature, day length, and light angle cues are believed to trigger both entry into and exit from diapause (Malcolm et al. 1987; Goehring & Oberhauser 2002; R eppert 2006). None of the known resident populations including South Florida, has the appropriate conditions to trigger diapause. I n fact, these same key environmental cues should trigger any diapausing animals entering a resident population to become rep roductively active. For example, Knight and Brower (2009) demonstrated that breeding occurs in the resident population in South Florida throughout the year. Consistent with their results, when I collected in late February and early March 2011, I observed m onarch adults in copula larvae at various instars, and eggs during my collection period in the early spring. Thus, environmental conditions during this time in South Florida were conducive to reproductive activity, rather than diapause. However, many of the animals I collected originated outside South Florida, indeed well north of the Gulf of Mexico. At this time, I do not know whether or not they were in diapause when they were collected. As monarchs that are reproductively active

PAGE 41

41 only live for 6 8 weeks (Herman & Tatar 2001), the immigrants I collected could not have become reproductively active any earlier than late December. However, by the end of October, most monarchs have disappeared from all but the gulf coast states (Journey North 2 013), so the im migrants I found whose isotopic profiles suggest they originated outside of the Gulf Coast states likely emerged no later than October. The conditions in South Florida should have triggered animals to exit diapause. Yet the animals cannot have exited diapause prior to December, or they would not have been alive when I was collecting in late February and early March. This leaves at least a two month period unaccounted for the monarchs could not be younger than October, ye t could not be older than December, if diapause were not maintained. Although detailed responses to triggers for exiting diapause have not been studied in monarchs, Zhu and colleagues (2009) demonstrated that monarchs in a diapause manipulation experiment would exit diapause within approximately two weeks when exposed to a combination of chemical and environmental cues. Dingle (1974) also showed that milkweed bugs ( Oncopeltus fasciatus) had fully exited diapause within a few weeks of being exposed to the en vironmental conditions triggering the return to reproductive competence. Therefore, either the immigrant monarchs I collected in South Florida stayed in diapause for several months after arriving in South Florida which seems unlikely given that other stud ies suggest that the switch should occur within a few weeks at the most or they spent at least a portion of the winter at an unknown location which had environmental conditions compatible with the maintenance of diapause prior to returning to South Florida

PAGE 42

42 Evaluating the diapause status of immigrants will involve further study. We can directly assess the reproductive condition of the animals at collection by dissecting the preserved gonadal tissue of the specimens I collected (Goehring & Oberhauser 2002). Regardless of the outcome, however, a key question that cannot be resolved from the set of animals I collected is how long do immigrants remain in diapause after encountering environmental conditions that trigger the termination of diapause, such as condit ions in South Florida. A controlled experiment using captive monarchs held under different temperature and light regimes is needed to provide data to answer this question. If the results of such study suggest that monarchs entering South Florida are unlike ly to remain in diapause long enough to explain the time delay I found, then they must have spent the winter elsewhere, and left this overwintering site well in advance of the bulk of the migratory population in Mexico Thus, investigation of where these i mmigrants spend the winter is warranted. Alternatively, males may employ condition dependent alternative migration strategies. Since environmental conditions in South Florida do not favor diapause, monarchs entering this population are likely to exit diapa use and begin reproducing without the need to overwinter for as long as their Mexican counterparts Thus, this alternative migration strategy may provide a reproductive advantage for males whose condition makes it unlikely they would survive the entire win ter in Mexico Physiological triggers for condition dependent responses have been shown to vary by sex for some Lepidoptera (Daniels 1995) Though I did not observe any female immigrants, barring sampling effort issues, it remains possible th a t females als o employ an alternative

PAGE 43

43 migration strategy, but with a threshold that was not triggered during the 2011 2012 winter season. Other possible explanations for recovering large numbers of migrants in South Florida include biases in my sampling protocol, or rel eases of farmed butterflies. However, the proportions of males and females were similar at both collection sites, and migrant butterflies were distributed across both collection sites, so my sampling protocol does not include obvious biases that would expl ain my findings. Another possibility is that the animals I captured were farm reared, and then released. However, the distribution of monarch farms (International Butterfly Breeders Association 2013) does not match the distribution of natal locations that I observed for migrants collected in South Florida. I thus plan to focus future efforts on testing the null hypothesis that immigrant monarchs arrive in South Florida due to random processes, the hypothesis that immigrant monarchs are entering South Florid a from an unknown overwintering destination, and the hypothesis that immigrants are following an alternative migration strategy. Wing size analyses showed there was no difference in South Florida migrant wing size relative to migratory males overwintering in Mexico On the contrary, my results indicate that migratory males are bigger than their resident counterparts. Because larger butterflies are stronger fliers (Dudley & Srygley 1994) this size advantage may translate into a reproductive advantage if these larger immigrants are better able to take down resident females. Therefore, migratory males might increase their reproductive success by switching to a resident life history strategy, migrating to a

PAGE 44

44 location where they would have the opportunity to mate immediately, rather than needing to survive until spring for mating opportunities at the overwintering colonies. There are two components of condition, n ot mutually exclusive, that might serve as triggers for selecting either a migratory or a resident life history strategy. Larval condition could determine the strategy that an adult male would pursue, for example, as determined by the quality of milkweed u pon which a larva feeds (Ladner & Altizer 2005; Atterholt & Solensky 2010). I used 15 N profiles as a proxy for larval condition. Interestingly, the immigrant males tended to be in poor condition relative to the resident males, though the difference was no t statistically significant. Of course, the most appropriate comparison to test for an alternative life history strategy would be between animals arriving in Florida and Mexico in the fall of the same year, rather than comparing animals collected after sur viving the winter; I do not have such data. When I compare my early spring collected Florida immigrant animals from 2011 to the early spring Mexican animals collected in 2012, i.e. same timeframe but in the following year, the Florida immigrants are more h ighly enriched for 15 N than the Mexican animals. Given the difference in collection years, these data must be interpreted with caution. For example, it is possible that all animals in 2012, including those arriving in Florida, were in better condition on average than all animals in 2011. However, my results are consistent with an alternative life history based on condition. The life history strategy could also be determined by the condition of the adult, and could actually change in the midst of migration if lipid stores fall below a certain threshold. The fitness of migratory monarchs overwintering in Mexico depends not only on traversing thousands of kilometers from their summer range to the overwintering

PAGE 45

45 colonies, but also on accumulating sufficient lipi d stores as adults along this journey to survive the reproductively inactive winter period (Alonso Mejia et al. 1997; Brower et al. 2006). Lipids could further be important via male investment in spermatophores, which is positively correlated with male mat ing success (Oberhauser 1988, 1989). Accordingly, males that are in poor condition may experience greater mating success if they pursue an alternative strategy by migrating a shorter distance, and/or to a location where they will be able to mate sooner Th e size advantage I observed might be particularly important if males were in poor condition with respect to lipid stores. Future investigation of lipid stores in migrants entering South Florida, as opposed to overwintering in Mexico is planned to directly assess adult condition of these butterflies. Much of what is known about monarch migration patterns comes from large scale, multi decadal mark recapture (Monarch Watch (2009)) or observational (Journey North (Howard & Davis 2004)) studies. These studies d emonstrated that there are corridors of high migration, or "flyways", from the northern summer breeding areas to the southern overwintering colonies. My results, combined with stable isotope data demonstrating trans Appalachian flight (Miller et al. 2011), and genetic data which failed to detect population structure within North America (Lyons et al. 2012), indicate that migration routes may be more complex than these broad studies indicate. It is also possible that changing climate or other unmeasured vari ables have altered migration patterns within the last decade. Additional studies using techniques such as stable isotope analyses and more variable and/or additional genetic markers may add further detail to the overall pattern of the major routes.

PAGE 46

46 Taborsk y and Brockmann (2010) outlined three criteria that condition dependent alternative strategies must meet to be evolutionarily stable: (1) environmental variation must be discrete, (2) environmental cues are reliable, (3) alternatives maximize fitness in di fferent environments. Because the environmental cues in South Florida do not favor continuation of diapause over the winter season, the migratory and resident alternative strategies fit the first two criteria, and my results indicate that these strategies may also meet the third criterion. Further investigation of my hypotheses on condition dependent alternate migration strategies is planned to evaluate how well these alternatives meet this third criterion, but the results of this study lead to an important overall conclusion: monarch migration behavior is more complex than implied by the major migratory routes. In particular, males and females do not seem to pursue the same strategies. The monarch migration is a unique phenomenon, and understanding sex spec ific behaviors and the role of alternative life history strategies in the migration is critical to understanding the evolutionary and conservation implications of these movements.

PAGE 47

47 Figure 2 1. Major migratory flyways of the fall monarch butterfly migratio n in eastern North America. Light orange arrows indicate the primary routes taken from summering locations throughout Canada and the United States to the overwintering colonies in central Mexico Width of the arrows is roughly correlated with number of mon archs migrating along a flyway. Gold arrow indicates hypothesized flyway connecting migratory population with resident South Florida population. Monarchs migrate in a general south or southwest orientation throughout the eastern range, so migrating monarch s are also found in areas not directly covered by the arrows. Red diamonds indicate sampling locations. Due to size constraints of the map, the location of the overwintering colonies is shown northeast of its actual location in central Mexico Based on the United States Forest Service ( 2008)

PAGE 48

48 Figure 2 2. Landmarks used to measure wing width. Width was measured in millimeters with a ruler aligned just below the white spot adjacent to the body, and the two white spots closest to the tip. The white line indicates the axis along which measurement was made. Photo taken by author Figure 2 3. Distribution of F statistics generated under the null hypothesis for permutation analysis of wing size. Vertical line indicates the F statistic value observed without permuting the data.

PAGE 49

49 Figure 2 4. Distribution of F statistics generated under the null hypothesis for permutation analysis of nitrogen isotope profiles. Vertical line indicates the F statistic value observed without permuting the data.

PAGE 50

50 Figure 2 5. Approximate natal origin s of specimens collected in South Florida, shown super imposed on the major migratory flyways. Specimens collected at the two sites in South Florida were assigned natal areas based on their carbon and hydrogen stable isotope ratios. Eight of the 18 specime ns (44%), all males, showed an origin outside of Florida. Dark red ellipses indicate the approximate origin of the each of these eight specimens.

PAGE 51

51 Figure 2 6. Wing size differences between monarchs collected in South Florida. Males shown in blue, females in red. Size and distribution of dots indicate individuals. Box limits show first and third quartiles, whiskers indicate 1.5 inter quartile range, and horizontal bar indicates median.

PAGE 52

52 Figure 2 7. Non resident males collected in Florida tend to be in worse condition than resident males, as inferred from enrichment in 15 N relative to 14 N. Box limits show first and third quartiles, whiskers indicate 1.5 inter quartile range, and horizontal bar indicates median.

PAGE 53

53 Figure 2 8. Non resident males collect ed in Florida in late February 2011 are in significantly worse condition than males collected at the Mexican overwintering colonies in early March 2012 as inferred from enrichment in 15 N relative to 14 N. Box limits show first and third quartiles, whiskers indicate 1.5 inter quartile range, and horizontal bar indicates median.

PAGE 54

54 Figure 2 9. Revised migratory flyway map illustrating the connectivity between each of the major migratory routes and the South Florida population. Light orange arrows indicate th e primary routes taken from summering locations throughout Canada and the United States to the overwintering colonies in central Mexico Width of the arrows is roughly correlated with number of monarchs migrating along a flyway. Migrating monarchs are also found in areas not directly covered by the arrows. Red diamonds indicate sampling locations. Due to size constraints of the map, the location of the overwintering colonies is shown northeast of its actual location in central Mexico

PAGE 55

55 Table 2 1. Collection site and specimen details Population Site name Location (Lat/Long) # Specimens Male Female South Florida a Homestead c N 25 27.996 W 80 26.178 6 4 South Florida Miami d N 25 41.466 W 80 17.682 5 3 Migratory b Cerro Pel—n e N 19 22.152 W 100 15.618 21 18 Migratory El Rosario e N 19 35.478 W 100 15.882 23 21 Migratory Gota de Agua e N 19 22.296 W 100 16.242 14 14 Migratory Piedra Herrada e N 19 06.450 W 99 50.802 21 21 Migratory Sierra Chincua e N 19 40.068 W 100 16.602 11 17 a Total 18 specimens (11 male, 7 female) b Total 196 specimens (97 male, 99 female) c Collected February 27 March 2, 2011 d Collected March 4, 21 and 22, 2011 e Collected March 7 9, 2012 Table 2 2. Stable isotope profile data Collection site Specimen Sex Hydrogen (!D) a Carbon ( 13 C) b Miami, FL MB8.1A Female 77.00 30.07 MB8.1B Female 96.22 30.07 MB8.1C Male 80.28 29.27 MB8.1D Male 81.02 27.71 MB8.1E Male 82.47 26.76 MB8.1F Male 86.77 28.48 MB8.1G Male 81.38 30.35 MB8.1H Female 91.92 31.50 Homestead, FL SFL4.1A Male 78.45 26.48 SFL4.1B Male 75.41 29.46 SFL4.1C Male 81.51 26.92 SFL4.1D Male 93.92 28.79 SFL4.1E Female 84.65 31.21 SFL4.1F Male 109.56 30.52 SFL4.1G Male 87.85 27.16 SFL4.1H Female 76.40 31.57 SFL4.1I Female 75.85 29.30 SFL4.1J Female 81.38 29.98 a Relative to Vienna Standard Mean Ocean Water (VSMOW) b Relative to Vienna Pee Dee Belemnite (VPDB)

PAGE 56

56 Table 2 3. Natal origin results Natal origin Male Female Florida 3 7 Not Florida 8 0

PAGE 57

57 CHAPTER 3 CANADIAN POPULATION The monarch breeding range is closely associated with the distribution of milkweeds, since these plants are the exclusive food source for monarch larvae (Urquhart 1960). During the spring and summer, milkweed g rows throughout much of North America, thus monarchs breed throughout the continent. The northern edge of the eastern portion of this breeding range was determined by Urquhart and Urquhart (1979) as running from the northern shore of Lake Superior to just south of Nova Scotia (Figure 3 2). Multiple milkweed species are native throughout southern Canada (Desmet & Brouillet 2013). The Journey North project (2013) has tracked observations of monarch butterflies since 1997, and milkweed plants since 2001. Eggs, larvae, and milkweeds have all been reported throughout New Brunswick and Nova Scotia north of the putative monarch breeding limit ( Table 3 1, Figure 3 2). Milkweeds have also been introduced on Labrador (Desmet & Brouillet 2013), although Journey North volunte ers have not reported any sightings that far north. Therefore, it is an open question whether monarchs are now breeding and successfully eclosing north of previously established breeding limit. To test the hypothesis that the monarch breeding limit has shi fted northward, I used stable isotope analysis to determine the natal origin of monarch butterflies collected both during their southward migration through Point Pelee National Park in Ontario, Canada, and from their overwintering colonies in Mexico Stabl e isotope ratios vary predictably with location, so these ratios can serve as intrinsic markers for tracing animal movement behavior (Hobson 1999). Because the carbon and hydrogen isotope ratios found in monarch wings are determined by the origin of the ho st plants on which

PAGE 58

58 animals fed as caterpillars and are stable following metamorphosis, measuring these ratios provides an estimate of the natal location (Wassenaar & Hobson 1998). I also used wing width measurements to evaluate whether monarchs originating from different geographic areas along the northern edge of the breeding range showed any obvious phenotypic differences from the rest of the migratory population. Methods I used stable isotope profiles to infer larval origins Stable isotope ratios vary predictably with location, so these ratios can serve as intrinsic markers for tracing animal movement behavior (Hobson 1999). The carbon and hydrogen isotope ratios found in monarch wings are determined by the origin of the hos t plants on which animals fed as caterpillars The keratin proteins that make up the wing are stable following metamorphosis, so measuring the stable isotope ratios of wing tissue provides an estimate of the natal location of the butterfly (Wassenaar & Hobson 1998). Specimen Collection The geography of southwestern Ontario serves as a funnel to Point Pelee, where monarchs are able to use the chain of small islands south of the Pelee peninsula as stepping stones to minimize the distance they must fly acr oss open water to reach the southern shores of the Great Lakes. Each year, monarchs consistently migrate through Point Pelee in late September (Parks Canada 2011), so this location is well suited to sampling monarchs migrating south from the northeastern s ummer range. Monarchs were collected near the tip of Point Pelee National Park in Ontario, Canada, and from five colonies spread across the overwintering region in Mexico ( Table 3 2). Live specimens were captured at Point Pelee, immediately placed in glassine en velopes, and stored at 20¡C for at least 12 hours prior to sex determination, dissection, and

PAGE 59

59 measurement. Although 196 butterflies were collected in Mexico only the 16 whose stable isotope profiles indicated an origin north of the published breeding lim it are included in this chapter (Chapter 4 discusses the analysis of the complete set of monarchs collected at the Mexican overwintering colonies). Due to permit restrictions in Mexico specimens were collected from the pool of dead butterflies littering t he forest floor at overwintering colony sites, and sorted into glassine envelopes at the end of the day. Sex was determined by the presence or absence of the androconium on the dorsal side of the hindwing before separating wings from bodies, and wings were returned to the glassine envelopes for later stable isotope analysis. Bodies were preserved for later DNA extraction in 2.0 mL tubes filled with 95% ethanol. Both wings and bodies were stored at 4¡C after specimens were brought to the lab. Natal Origin An alysis One hindwing from each South Florida specimen was washed in a 2:1 chloroform:methanol solution to remove surface impurities, lipids, and residual hemolymph then air dried. Small sections of hindwing (0.350 0.02 mg for hydrogen and 0.825 0.05 mg for carbon) were cut and placed into capsules (3x5mm silver capsules for hydrogen and 5x9mm tin capsules for carbon) for stable isotope analysis. To reduce small fluctuations in stable isotope profiles due to varying degrees of wing pigmentation (see Appe ndix), sections were consistently excised from the black portion of the wing margin. Carbon isotope profiles were measured at the Light Stable Isotope Mass Spec Lab at the University of Florida, and hydrogen isotope profiles were measured at the Central Ap palachians Stable Isotope Facility at the University of Maryland Center for Environmental Science Appalachian Laboratory. Approximate natal origins were assigned based on the combined isotope profiles following the approach

PAGE 60

60 described by Wassenaar and Hobso n (1998) and Miller and colleagues (2011), except that the Online Isotopes in Precipitation Calculator (OIPC) plug in for Google Earth (Bowen & Revenaugh 2003; Bowen et al. 2005; Bowen & Wilkinson 2002) was used to evaluate !D values for precipitation. Win g Size Analysis Forewing size was measured as a proxy for body size. Width was measured from the white spot adjacent to the body to the point between the two white spots at the tip of the forewing (Figure 3 1). Although there were slight variations in wing patterns, these two markers were consistently used to ensure that homologous sections of the wing were measured. Measurements were recorded to the nearest millimeter. Differences in wing size were analyzed in R (R Development Core Team 2008). Results Stab le isotope profiles indicate that 3 monarchs collected in Point Pelee, as well as 17 monarchs collected in Mexico had approximate natal origins north of the accepted breeding limit ( Table 3 3). Natal origins are shown in Figure 3 2, along with observations coll ected by the Journey North project for comparison. Origin did not significantly differ for males and females. The hydrogen stable isotope profile for the specimen (ER2.3AJ) with the northernmost origin (Figure 3 2) indicated an origin in northern Quebec. Although this origin is well above the published northern limit of the monarch range, citizen scientists have reported sightings at similar latitudes to the Journey North project (2013). All samples measured at CASIF were cali brated against USGS42, USGS43 standards, and a keratin quality control sample. Neither the readings of these standards nor the

PAGE 61

61 readings of samples measured before and after ER2.3AJ showed skew in their readings that would indicate a systematic error. Butte rflies collected in Canada ( N = 15) were significantly smaller than those collected in Mexico ( N = 17; Figure 3 3; ANOVA, F 1,28 = 10.086 P = 0.004 ). Specimens also showed sexual dimorphism, with females significantly larger than males (Figure 3 3; ANOVA, F 1,28 = 4.2186, P = 0.049). The interaction of collection country and sex was not significant (ANOVA, F 1,28 = 1.5514, P = 0.223). Discussion The Journey North project relies on volunteers to report sightings of monarch eggs, larvae, and adult butterflies. However, sightings of adult animal s could represent dispersal from animals eclosing within the breeding limit. Observations of eggs and caterpillars are strongly suggestive, but not definitive proof, that animals are successfully breeding. My results are c onsistent with the hypothesis that the monarch breeding limit has indeed moved northward S table isotope profiles show that some animals collected from both Point Pelee and Mexico had successfully eclosed north of the published breeding limit and these butterflies are arriving at the Mexican overwintering colonies Thus, m y data confirm the prediction from observational data that successful breeding is indeed happening north of the established limits. Stable isotope data complement observational, c itizen scientist collected data, which are concentrated in populated areas. Since much of northeastern North America is sparsely populated, this limits the areas in which observations are likely to occur Stable isotope profiles indicate larval origins, so my approach can detect butterflies that originated from both populated and unpopulated areas. Unsurprisingly, butterflies that were collected at Point Pelee originated from some of the same breeding areas

PAGE 62

62 previously reported by Journey North, but also fro m areas not included in the Journey North observations. Animals collected in Mexico in particular, show origins from wilderness regions of Quebec where citizen scientist observations are unlikely to occur. I found no evidence of phenotypic differences in monarchs originating from northern latitudes. Although size differences were detected between animals collected in Canada and animals collected in Mexico specimens north of the breeding limit did not drive these differences Using a captive population rea red in controlled laboratory conditions, Wensler (1977) demonstrated that monarchs reared in colder temperatures such as those found at northern latitudes followed Bergmann's rule (Van Voorhies 1996) and accordingly tend to be larger than monarchs reared in warmer temperatures. Moreo ver, I did not find a significant association between size and latitude (data not shown). Thus, my findings run counter to expectations. Butterflies were collected on the wing in Point Pelee. In Mexico however, monarchs were collected from natural mortality on the forest floor at the end of the overwintering season Thus, it is possible that the different collection methods used influenced the results; it is also possible that the difference between my two studies can be expla ined by natural selection. Accordingly, I have developed two complementary hypotheses: 1) I was not able to capture the full size range of animals in Canada. Since larger butterflies tend to be more vigorous flyers (Dudley & Srygley 1994), they may have ev aded capture or been roosting higher in the trees. If animals in Mexico died randomly with respect to size, then the Mexican samples represent a random sample of the size range of the population, while the P oint Pelee ones may not. 2) Alternatively, if sma ller animals are less likely to survive the migration then the population in the

PAGE 63

63 overwintering colonies is likely larger on average than animals beginning the journey. Again, I make the (untested) assumption that mortality in Mexico is random with respec t to size. Since the sample sizes collected at Point Pelee were relatively small future study will be necessary to evaluate these two hypotheses including capturing animals in P oint Pelee in a way that is less likely to result in size bias; and similarly collecting a live sample in Mexico in the same, unbiased way Moreover the apparent failure to follow Bergmann's rule in field caught animals is surprising, and bears further study. All the monarchs collected for this study were from the 2011 2012 southw ard migration and overwintering season, so the effects of annual variations in breeding season weather on the degree of monarch breeding in the far north are unknown. Environmental conditions conducive to monarch breeding are generally found during a fairl y brief period at these latitudes. Also, monarch larval success, including susceptibility to parasites, has been shown to vary with species of Asclepias on which larvae feed ( e.g., Bartholomew & Yeargan 2001; Calvert 1999; Cohen & Brower 1982, de Roode et al. 2008), and the set of Asclepias species found at northern latitudes differs from that found at southern latitudes (Desmet & Brouillet 2013). The combined effects of a short breeding season, cooler temperatures, and northern Asclepias species on monarch fitness are unknown, but a possible hypothesis is that monarch breeding in the far north produces fewer viable offspring for the same reproductive investment. In addition, wing size is a fairly limited proxy for phenotypic differences that may arise due t o northerly breeding. At a minimum, phenotypic characteristics that are correlated with migration and overwintering success, such as lipid content, are important to investigate to determine the degree to which breeding at northern latitudes

PAGE 64

64 affects these t raits, if at all. Solensky and Larkin (2003) also found that monarch larvae exhibit a degree of phenotypic plasticity in their coloration, with the degree of melanization inversely correlated with temperature. While darker larvae are better able to maintai n body temperatures suitable for development, they hypothesized that this was accompanied by a tradeoff in greater susceptibility to predation. Given that monarch larvae at northern latitudes are likely to experience colder temperatures than their southern counterparts, the degree to which melanization increases in northern latitudes, as well as the effects of this melanization on predation, are additional phenotypic traits that should be investigated to evaluate the overall effect of northern breeding on m onarch fitness. My data indicating that animals from beyond the previously established breeding limit indeed successfully mature and undertake the journey south are important additions to the growing literature on shifting population ranges, not only of mo narchs, but of many plant and animal species (reviewed in Parmesan & Yohe 2003) in the face of climate change. Ecological niche modeling has predicted monarch range shifts northward with changing climate (Batalden et al. 2007). The stable isotope data presented here confirm the model predictions, and underscore the utility of stable isotopes in studies of range expansion

PAGE 65

65 Figure 3 1. Landmarks used to measure wing width. Width was measured in millimeters with a ruler a ligned just below the white spot adjacent to the body, and the two white spots closest to the tip. The white line indicates the axis along which measurement was made. Photo taken by author

PAGE 66

66 Figure 3 2. Map of approximate origins, based on 13 C and !D, of monarch butterflies c ollected at Point Pelee National Park (red and blue circles). Specimens collected at five overwintering colonies in Mexico that originated north of the published breeding limit are also shown (red and blue triangles). Longitude for the Mexico specimens is especially approximate, since these specimens could also be placed in the Canadian Great Plains based on their carbon stable isotope profiles, although still at latitudes north of the published breeding limit. The brown dashed line indicates the previously published limit of the monarch breeding range. Purple dots mark sites where monarch eggs or larvae were reported north of the breeding limit by the Journey North project, 2006 2013.

PAGE 67

67 Figure 3 3. Size differences of specimens collected at Point Pelee and Mexico Size of points indicate number of individuals in a given size class; points for Canadian specimens with origins north of the published breeding limit are circled in gold (all Mexican individuals shown here originated north of the established limit ) Females shown in red; males in blue. Box limits show first and third quartiles, whiskers indicate 1.5 inter quartile range, and horizontal bar indicates median.

PAGE 68

68 3 1. Sightings, by year of observation, from the Journey North project suggesting northward expansion of the monarch breeding range. Year Milkweed a Eggs b Larvae b Adults 2013 5 0 0 2 2012 6 3 1 15 2011 1 1 0 11 2010 2 1 0 6 2009 2 0 0 7 2008 1 0 0 14 2007 2 1 0 28 2006 0 0 0 35 2005 2 1 1 6 2004 0 0 0 0 2003 0 0 0 0 2002 0 0 0 0 2001 0 0 0 1 2000 N/A 0 0 3 1999 N/A 0 0 8 1998 N/A 0 0 4 1997 N/A N/A N/A 1 a Milkweed sightings have only been tracked since 2001. b Egg and larvae sightings have only been tracked since 1998. 3 2. Collection site and specimen details Country Site name Location (Lat/Long) # Specimens Male Female Canada a Point Pelee N 41 54.706 W 82 30.530 7 8 Mexico b Cerro Pel—n N 19 22.152 W 100 15.618 1 0 El Rosario N 19 35.478 W 100 15.882 4 2 Gota de Agua N 19 22.296 W 100 16.242 1 2 Piedra Herrada N 19 06.450 W 99 50.802 2 1 Sierra Chincua N 19 40.068 W 100 16.602 4 0 a Collected September 12 14, 2011 b Collected March 7 9, 2012

PAGE 69

69 3 3. Stable isotope profile data Collection country Specimen Sex Hydrogen (!D) a Carbon ( 13 C) b Canada PP.2A M 100.1 28.29 PP.2B F 104.6 29.38 PP.2C F 114.4 28.08 PP.2D F 105.53 26.71 PP.2E M 109.95 28.29 PP.2F F 104.88 28.38 PP.2G M 102.41 28.19 PP.2H c F 119.17 26.57 PP.2I M 100.86 27.98 PP.2J M 101.13 28.99 PP.2K F 105.01 27.62 PP.2L c F 122.22 28.57 PP.2M M 110.94 30.19 PP.2N F 92.16 31.14 PP.2O c M 118.35 27.71 Mexico d CP.3J M 126.87 28.92 ER1.3N F 151.31 28.97 ER1.3O F 123.85 29.19 ER1.3AL M 138.77 28.20 ER2.3Q M 142.15 28.31 ER2.3AI M 132.13 29.28 ER2.3AJ M 175.68 28.87 GA.3A F 152.28 29.04 GA.3H F 132.76 28.17 GA.3AO M 122.75 29.99 PH.3AP F 136.54 29.49 PH.3AU M 125.91 28.99 PH.3BU M 127.59 29.65 SC.3AK M 120.73 29.81 SC.3AZ M 119.93 31.58 SC.3BD M 126.32 28.63 SC.3BJ M 136.66 29.99 a Relative to Vienna Standard Mean Ocean Water (VSMOW) b Relative to Vienna Pee Dee Belemnite (VPDB) c Canadian specimens with origins north of the published breeding limit d Only includes specimens collected in Mexico with origins north of the published breeding limit

PAGE 70

70 CHAPTER 4 MEXICAN OVERWINTERING COLONIES Since the ability to adapt to changing conditions is correlated with the amount of genetic variation in a species, a major concern in species conservation efforts is the preservation of adequate genetic variation both for the species as a whole, and in its distribution across the species' range. Thus, a key aspect of the planning process is assessment of the l evel of genetic variation and its distribution. For species such as the monarch butterfly ( Danaus plexippus ) that are threatened, but not yet endangered, investigating the distribution patterns of genetic variation, and the processes that lead to them pro vides critical data for the development of appropriate conservation policy. Bowlin and colleagues (2010) argued that migratory species offer unique natural systems in which to study natural variation in the responses to environmental change. Monarch butter flies in eastern North America undergo a seasonal migration from their summer range throughout the eastern United States and southern Canada to a cluster of overwintering colonies in central Mexico (Urquhart & Urquhart 1977). This migration is so unique that the area containing the overwintering colonies was named a UNESCO World Heritage site in 2008 (UNESCO 2008). Although monarch butterflies are not listed as an endangered species, the phenomenon of the monar ch migration has been a conservation concern for over three decades. The International Union for Conservation of Nature and Natural Resources (IUCN) added the monarch migration as a "threatened phenomenon" to its Red List in 1983 (Wells et al. 1983). Mexic o first began officially protecting the overwintering colony areas in the early 1980s; the Monarch Joint Venture was formed in the United States as a partnership between 13 federal, state, non governmental, and academic organizations in 2008; the Committee on the Status of

PAGE 71

71 Species at Risk in Ontario Canada listed monarchs as a "species of special concern" in 2009 (COSSARO 2009) The type of threats varies throughout the monarch range in eastern North America. Logging of the oyamel fir habitat has been a m ajor concern for the overwintering colonies in Mexico (e.g., Ramirez et al. 2003; Calder—n et al. 2004; Ramirez et al. 2007). In both the United States and Canada, possible negative effects of genetically modified crops on monarch reproduction and the loss of suitable breeding habitat have been focal areas of conservation research and policy (e.g., Zangerl et al. 2001; Macdonald & Yarrow 2003; Oberhauser & Rivers 2003; Anderson et al. 2005; Marvier et al. 2007; Prasifka et al. 2007). Because of this variati on in the degree and type of threat across the eastern North American monarch range, the distribution of genetic variation, and the degree of connectivity between monarchs found in geographically distant areas, are key for monarch conservation. Monarchs br eed throughout North America during the spring and summer months, with 4 6 generations produced during this period. Butterflies that emerge in the late summer or fall, however, enter reproductive diapause immediately after emergence, then undertake the ard uous migration to the overwintering grounds. In the eastern United States and southern Canada, these butterflies migrate during the fall along the major routes shown in Figure 4 1 to overwinter at a small cluster of colonies in central Mexico while monarc hs west of the Rocky Mountains migrate to a series of small colonies scattered along the California coast (Dingle et al. 2005; Leong 1990; Leong et al. 1995). These butterflies remain in diapause until the early spring, when they exit diapause and breed be fore leaving the overwintering colonies to begin the journey north. This overwintering generation of monarchs generally migrates far enough north to

PAGE 72

72 lay eggs in suitable milkweed habitat in southern Texas, and each subsequent generation continues the north ward migration to fan out across the full range of North America east of the Rockies. Because breeding occurs continuously during this migration, a possibility exists for an isolation by distance (IBD) pattern to form along the various migration routes. Th e overwintering colonies in Mexico however, may provide a means for monarchs from widely separated summering areas to interbreed, thus breaking down IBD patterns that may form during the spring and summer. A fter decades of intense effort, the precise loca tion of the overwintering area in Mexico was discovered (Brower et al. 1977; Urquhart & Urquhart 1977). Early efforts to study monarch movement and its effects on population structure relied on observational or mark recapture approaches. These studies were able to determine the major migratory flyways (Calvert & Brower 1986; Howard & Davis 2004; Howard & Davis 2009; Malcolm 1987; Schmidt Koenig 1985) Despite these major findings, however, both observational and mark recapture approaches are of limited util ity for investigating fine scale movements and population structure. In contrast, both genetic sequences and stable isotope profiles are intrinsic markers that do not rely on following the movements of individuals, nor on capturing the same individuals at multiple sites Thus, they are potentially more powerful, less prone to bias, and useful for investigations at a more fine scale. Prior to the widespread availability of gene sequencing, allozymes (naturally occurring variants of the same enzyme) were used to infer genetic differences. Eanes and Koehn (1978) used allozyme analysis to evaluate population structure in monarch butterflies, and found no evidence for geographic patterns of population structure. Because allozymes are post

PAGE 73

73 translational markers, t hey may mask genetic differences that are not carried forward to the protein such as synonymous and intronic mutations. Such mutations can be detected, however, if genomic DNA is sequenced. Thus, DNA may provide higher resolution and greater power to detec t subtle differences than allozymes. Brower and Jeansonne (2004) found a low degree of genetic variation in monarch butterflies using DNA sequences of gene fragments, and Lyons and colleagues (2012) used microsatellite markers to show that monarchs in Nort h America form a single panmictic population. While DNA markers are a means of directly assessing the distribution of genetic variation, movement patterns can only be inferred indirectly by analyzing the degree of differentiation between groups. In contras t, stable isotope analysis uses predictable geographic patterns in the distribution of different isotopes of key elements to measure animal movement between the location at which isotopes were incorporated into tissues, and the location at which specimens were collected. Because keratin in wing tissue is laid down at the time of metamorphosis, and does not change subsequently, isotopes of the carbon and hydrogen incorporated in these tissues can be used to assign natal origin to butterflies (Wassenaar & Hob son 1998). Thus, the ratio of light to heavy isotopes in an adult animal reflects the milkweed on which the larva fed. In monarch wings, hydrogen isotopes vary in conjunction with the isotopic signature of local meteoric water; carbon isotopes vary dependi ng on the milkweed species on which a larva fed prior to pupation (Hobson et al. 1999). As shown in Figure 4 2, hydrogen isotope ratios in eastern North America exhibit a roughly southeast northwest gradient, with increasing depletion of deuterium (D) to the northwest, whereas carbon

PAGE 74

74 isotope ratios show increasing depletion of 13 C moving from northeast to southwest. By determining stable isotope ratios for carbon and hydrogen, Wassenaar and Hobson (1998) found that there was no clear a ssociation between natal origin and overwintering colony site, and that a preponderance of monarchs originated from the Midwest or western Great Lakes. In contrast to previous studies, I integrated genetic and stable isotope analyses to test the hypothesis that monarchs at the Mexican overwintering colonies are part of a single panmictic population. To evaluate possible phenotypic differences in monarchs found at different colonies, I used wing size as a proxy for overall phenotype. Genetic sequences reflec t movements that persist over generations, wing size is a function of both genetics and of the condition of the larva, and stable isotopes reflect the conditions under which the larva fed. Thus, each line of evidence produces data that resolve patterns at different time scales, and integrating these three provides a more complete picture than any one alone. Methods I used stable isotope profiles to infer larval origins Stable isotope ratios vary predictably with location, so these ratios can serve as intri nsic markers for tracing animal movement behavior (Hobson 1999). The carbon and hydrogen isotope ratios found in monarch wings are determined by the origin of the host plants on which animals fed as caterpillars The keratin proteins that make up the wing are stable following metamorphosis, so measuring the stable isotope ratios of wing tissue provides an estimate of the natal location of the butterfly (Wassenaar & Hobson 1998).

PAGE 75

75 Specimen Collection Monarchs consistently congregate at a collection of sites scattered across the mountains of central Mexico northwest of Mexico City. Colonies to sample (Figure 4 3) were selected from throughout this geographic range based on accessibility, safety, and col lection permit restrictions. In March 2012, 181 specimens were collected from five overwintering colonies in Mexico ( Table 4 1). Due to permit restrictions, monarchs were collected from the pool of dead butterflies littering the forest floor at overwintering col ony sites, and sorted into glassine envelopes at the end of the day. Sex was determined by presence or absence of the androconium on the dorsal side of the hindwing before separating wings from bodies, and wings were returned to the glassine envelopes for later stable isotope analysis. Bodies were preserved for later DNA extraction in 2.0 mL tubes filled with 95% ethanol. Both wings and bodies were stored at 4¡C after specimens were brought to the lab. DNA Analysis DNA was extracted from head and thorax tis sue using the UltraClean Tissue & Cells DNA Isolation Kit (MO BIO Laboratories, Inc., Carlsbad, CA) according to the manufacturer's protocol for difficult tissue extractions. As summarized in Table 4 2, only a subset of the total number of specimens collected wa s used for analysis. Unused material was archived for analysis in future studies. To meet collection permit requirements, only 25 L aliquots of DNA extracts from Mexico were brought back to the University of Florida for PCR amplification and sequencing; t he remaining specimen material was archived (DNA @ 20¡C, bodies @ 4¡C) in the Laboratorio de GenŽtica de la Conservaci—n at UNAM Campus Morelia.

PAGE 76

76 Amplification of a hyper variable fragment of the nuclear Tektin gene was performed following the methods desc ribed by Whinnett and colleagues (2005) for the TektinA/Tektin3 primer pair. Each 25 L reaction volume contained 2 L (roughly 5 20 ng) genomic DNA extract, 12.5 L 2X PCR Master Mix (Promega, Madison, WI), 10 M each primer and ddH 2 O to bring the final v olume to 25 L. After optimizing the original PCR conditions, amplifications were performed with an initial denaturing of 2 minutes at 94¡C, followed by 35 cycles of 60 s at 94¡C, 60 s at 47.9¡C, 90 s at 72¡C, and a final extension of 10 minutes at 72¡C. P CR products were prepared for sequencing using either a QIAquick PCR Purification Kit (QIAGEN, Inc., Valencia, CA) or polyethylene glycol (PEG) and ethanol precipitation (Paithankar & Prasad 1991). All purified PCR products were sequenced at the DNA Analys is Facility on Science Hill at Yale University using the same primers as those used for PCR. Sequences were imported into Geneious (Drummond et al. 2012), forward and reverse reads were matched, and ends were trimmed to give a final product of 626 bp. Trim med contigs were aligned using the MUSCLE option (Edgar 2004) of Geneious with the default settings. Alignments were imported into Network (Fluxus Technology Ltd., Clare, Suffolk, UK) to generate median joining genetic network diagrams (Bandelt et al. 1999 ). Final diagrams illustrating the breakdown of genotypes by sex, and by overwintering colony were prepared in Adobe Illustrator (Adobe Systems Inc., San Jose, CA). Alignments were also used to generate basic neighbor joining tree with 10,000 bootstrap rep licates using built in Geneious tree builder with the default settings. Haplotypes were reconstructed using the PHASE 2.1 (Stephens et al. 2001, Stephens & Scheets 2005) option of DnaSP v5 (Librado & Rozas 2009) with the default

PAGE 77

77 parameters, except with 1,0 00 iterations instead of 100. As recommended in the PHASE 2.1 documentation, reconstruction was run five times with different starting seed values, with one run of 10,000 iterations. Results were compared for consistency (mean likelihood ranged from 932.5 27 to 932.014), and the results from the run with the highest likelihood were used for subsequent analyses. All subsequent analyses were performed in Arlequin 3.5 (Excoffier et al. 2007). To evaluate the degree to which genetic variation was unevenly dist ributed across the overwintering colony sites, measures of molecular diversity ( (Tajima 1983), (Tajima 1983), S (Watterson 1975)) were determined from these haplotypes for both the migratory population as a whole (all five colonies as a single popula tion), and for each colony separately. The degree of differentiation between the colonies was determined by calculating the pairwise genetic distances between the populations ( F ST ; Wright 1943), running an AMOVA with 1,000 permutations under the assumption that each colony represents a distinct population, and performing Raymond and Rousset's exact test of population differentiation (1995) with 100,000 Markov steps, which tests aga inst a null expectation of panmixia. Natal Origin Analysis One hindwing from each South Florida specimen was washed in a 2:1 chloroform:methanol solution to remove surface impurities, lipids, and residual hemolymph then air dried. Small sections of hindwi ng (0.350 0.02 mg for hydrogen and 0.825 0.05 mg for carbon) were excised from the wing and placed into capsules (3x5mm silver capsules for hydrogen and 5x9mm tin capsules for carbon) for stable isotope analysis. To reduce small fluctuations in stable isotope profiles due to varying degrees of wing pigmentation, sections were consistently excised from the black and

PAGE 78

78 white portion of the wing margin (see Appendix). Carbon isotope profiles were measured at the Light Stable Isotope Mass Spec Lab at the Univ ersity of Florida, and hydrogen isotope profiles were measured at the Central Appalachians Stable Isotope Facility at the University of Maryland Center for Environmental Science Appalachian Laboratory. Approximate natal origins were assigned based on the c ombined isotope profiles following the approach described by Wassenaar and Hobson (1998) and Miller and colleagues (2011), except that the Online Isotopes in Precipitation Calculator (OIPC) plug in for Google Earth (Bowen & Revenaugh 2003; Bowen et al. 200 5; Bowen & Wilkinson 2002) was used to evaluate !D values for precipitation. Origins were assigned to one of the three major migratory flyways (United States Forest Service 2008). In some cases, origins could not be assigned to a single flyway because of t he complex distribution of carbon and hydrogen isotopes in North America. Wing Size Analysis Forewing size was measured as a proxy for body size. Width was measured from the white spot adjacent to the body to the point between the two white spots at the ti p of the forewing (Figure 4 4). Although there are slight variations in wing patterns, these two markers were consistently present to ensure that analogous sections of the wing were measured. Measurements were recorded to the nearest millimeter. Difference s in wing size by sex and origin were analyzed in R (R Development Core Team 2008). Because sample sizes were small (13 females/colony, 11 16 males/colony), I tested the difference in size between overwintering colonies by permuting the variable "origin" f ive hundred times and observing the distribution of F statistics from the one way ANOVA of each permuted dataset.

PAGE 79

79 Results A 626 bp hypervariable region of the autosomal locus Tektin was analyzed to evaluate the possibility for genetic structure among the overwintering sites. There were 39 segregating sites (~6%), 21 of which were parsimony informative A ll but one site had two alleles ; the other had three Of 126 butterflies analyzed 92 (73%) were heterozygous for at least one site. Molecular diversity statistics are summarized in Table 4 3. Genotype networks show that haplotypes are homogenously distributed across the five sampled colonies ( Figure 4 5). Moreover, ha plotypes are distributed approximately equally between males and females ( Figure 4 6). Pairwise genetic distances were small, and permutation tests of these distances, as implemented in Arlequin showed that none were statistically significant ( Table 4 4; 1,023 permutations, 0.13 < P < 0.96). The distribution of genetic variation among the colonies was also not significant ( Table 4 5; AMOVA, P = 0.817 0.012). Finally, Raymond and Rousset's exact test of di fferentiation was unable to reject panmixia on either a global ( P = 0. 79 ) or pairwise basis ( ranging from 0.21 < P < 0.92). Phylogenetic analysis also supports a single homogenous population distributed across the overwintering colonies. T wo Queen butterfl y ( D. gilippus ) specimens collected in South Florida were also sequenced for the same region of Tektin to serve as an outgroup. Comparison between monarchs and this sister taxon shows 100% bootstrap support for separation between these taxa, but there was no support for differentiation within monarchs (Figure 4 7). I identified natal origin by using carbon and hydrogen stable isotopes to test the null hypothesis that animals congregate in particular overwintering sites without respect to their origin. I fou nd no significant relationship between origin (via !D and 13 C) and

PAGE 80

80 overwintering colonies with respect to north south ordering ( l inear regression, R 2 < 0.022, F < 0.53, P > 0.47). A lthough there is no clear association between specific flyways or summerin g areas and overwintering colony sites, sites do significantly differ with respect to origin ( Figure 4 8; ANOVA, main effect of site, F 4 = 3.35, P = 0.012). Sexes do not differ by origin (Figure 4 8; ANOVA, main effect of sex, F 4 = 1.00, P = 0.319), and site x sex is also not significant ( P = 0.068). The latter interaction seems to be driven primarily by exceptional females found in the Sierra Chincua colony (females at this site tend to be from the most southern locations). Wing s ize varied significantly between the overwintering colony sites ( Figure 4 9; ANOVA, F 4 = 4.18, P = 0.003). There were no significant differences in the size of male vs. female monarchs ( F 1 = 0.44, P = 0.507), nor was there a signific ant site x sex interaction ( F 4 = 0.670, P = 0.670). No significant differences were found between females from different colonies ( F 4 = 1.40, P = 0.24) M ales at the Sierra Chincua colony were significantly smaller than males at other colonies ( F 4 = 3.42, P = 0.012 ) but no significant differences were found between males from different colonies when Sierra Chincua was excluded from the analysis ( F 3 = 0.806, P = 0.494) In addition, I found no significant relationship between origin (via !D), size, and colony ( Figure 4 10, linear regression for each colony, R 2 < 0.064, F < 1.72, P > 0.20). Discussion Taken together, my genetic and stable isotope results support the hypothesis that monarchs at the Mexican overwinterin g colonies are part of a single panmictic population. The overall amount of genetic variation at the Tektin locus was low. Across all measures of diversity ( (Tajima 1983), (Tajima 1983), S (Watterson 1975) ), El Rosario consistently shows the highest, and Cerro Pel—n the lowest, amount of genetic

PAGE 81

81 variation. is a measure of the mean number of pairwise differences. Both and S are estimates of the population parameter which describes the amount of variation that can be expected under the neutral theory (Kimura & Ohta 1971 ) This parameter is defined as = 4 N e where N e is the effective population size, and is the mutation rate estimates based on (Tajima 1983 ); S is based on S the number of segregating (variable) sites (Watterson 1975). Since is based agreement between these two statistics is to be expected. Comparison between the colonies, however, shows that these differences in the distribution of genetic variation are not significant, since b oth the AMOVA which compares the amount of genetic variation among and within colonies and Raymond and Rousset's test which uses a Markov chain approach to test for panmixia found no supp ort for differentiation between the colonies. In addition, the Tektin genotype network indicate s that this genetic variation is distributed fairly uniformly across the five sampled overwintering colonies and fairly equally between the sexes. The degree of concordance between these different approaches to evaluating the distribution of genetic variation indicates that my finding of a single panmictic population is robust under different sets of assumptions This lack of genetic population structure is cons istent with previous work that has shown no genetic differentiation within the North American migratory monarch population (Eanes & Koehn 1978; Lyons et al. 2012). Stable isotope results complement the genetic data by providing evidence that monarchs sort into overwintering colonies randomly with respect to geographic origin. No relationship between geographic origin and colony was detected, but distribution of origins across the colonies was not uniform. One requirement of panmixia is that

PAGE 82

82 a nimals originat ing from different geographic origins must have the opportunity to mate with one an othe r Since monarchs begin breeding before leaving the overwintering colonies in the spring, the mix of geographic origins at the colonies fulfills this requirement, and is consistent with the hypothesis that any geographic structure that may develop during the spring and summer would be broken down at the overwintering colonies. Thus, while the genetic results support the hypothesis that the migratory population of monarch butterflies in eastern North America is genetically homogenous, it is the stable isotope data that strengthen the inference this homogeneity is due, at least in part, to panmixia. Although my findings are consistent with the hypothesis that monarchs at the Mexican overwintering colonies are part of a single panmictic population, my data also detected subtle differences between the overwintering colonies. These results suggest that, while monarchs from throughout the summering range mix, the colony sites are not equivalent. The ecological conditions are the colony sites I sampled vary widely in tree density, understory composition, exposure to weather, and colony size. Since all of these factors are important for monarch survival over the winter (Alonso Mejia et al. 1992; Alonso Mejia et al. 1998; James 1984), this ecological variation could be correlated with some of the differences I detected. For example, El Rosario is the largest colony, whereas Cerro Pel—n is one of the smallest. The colony site at El Ros ario is distributed along an altitudinal gradient with consistent cover of old growth oyamel firs, and butterflies in this colony tend to form massive clusters high in these trees. In contrast, the butterflies in Cerro Pel—n are found in smaller numbers an d lower densities. Large stands of old growth oyamel fir are also lacking at Cerro Pel—n, and the

PAGE 83

83 forest habitat is interspersed with open meadows. Thus, ecological conditions at the El Rosario site are better suited to overwintering survival than Cerro Pe l—n, even though monarchs consistently return to both sites each year (L. Brower, personal communication ). El Rosario displayed the largest variance in origins ; Cerro Pel—n displayed the largest variance in wing size Thus, a greater variety of origins wer e found at the overwintering colony with some of the best ecological conditions for overwintering survival, whereas some of the smallest animals were found at the colony with some of the worst ecological conditions. These differences in both origin and siz e indicate that different colonies attr act different types of monarchs, but the precise relationship between origin, size, and quality of ecological conditions at the overwintering colonies needs additional study. E cological differences in the colony site s may correlate with the patterns I observed, but detailed study of these potential correlations is needed. In particular, multi year observations are necessary to demonstrate whether or not between colony variation with respect to origin is a recurrent pa ttern, and the degree to which this pattern is consistent across years. My results lead to several questions that warrant future investigation. One key question is the persistence of the patterns I found across years. These specific patterns may consistent ly appear from year to year; differences between colonies may appear each year, but the particular pattern may differ; or colonies may show differences in some years, but not in others. Moreover, I collected towards the end of the overwintering period. Thu s, the profiles of animals arriving at the site with respect to

PAGE 84

84 genetics, natal origin, and size could be different from the animals found towards the spring, particularly if mortality is nonrandom with respect to any of the traits I studied. Finally, I co llected dead specimens, so the patterns for live butterflies may differ. For example, the majority of monarchs I collected originated in the areas that feed the Atlantic or Great Lakes flyways, but previous studies found a greater proportion of butterflies from the Midwest (Howard & Davis 2009; Wassenaar & Hobson 1998). Differences could be due to different methods used, random variation between years, or actual changes in monarch behavior patterns. However, although Howard and Davis used monarch sighting d ata collected via the Journey North project, Wassenaar and Hobson used a similar approach to ours (stable isotopes from dead animals), suggesting that different approaches are unlikely to explain my differing results. Determining if and how patterns vary a cross years is the first step in investigating the mechanism(s) that produce these patterns. Awareness of the fine scale differences between overwintering colonies can better inform conservation efforts focused on the overwintering colony sites. In additio n to the concerns raised over the years by various conservation organizations, the 2013 survey of the Mexican overwintering colonies by the World Wildlife Foundation Telcel Alliance, and Mexico 's National Commission of Protected Areas (CONAP) showed that t he number of monarchs congregating at these colonies was the lowest it has been in over two decades (World Wildlife Foundation 2013). Microclimate conditions at colony sites are key for monarch survival over the winter (Alonso Mejia 1996; Anderson & Brower 1996; Brower et al. 2009; Brower et al. 2008), but environmental conditions at different colony sites vary widely (Brower et al. 2002). My results show that the

PAGE 85

85 monarchs found at these sites vary as well with respect to genetic variation, origin and size, and that colony sites contain distinct combinations of animals, such that loss of any one of these sites would decrease the diversity of the species. The source of mortality is important in setting conservation priorities. For example, if mortality is ass ociated with particularly degraded sites, then restoring these sites may improve overwintering survival. If, however, mortality is due more to site extrinsic factors such as availability of suitable nectar sources and roosting sites along the migratory rou tes, then policies focused on the migration itself may be more important. Understanding the patterns and processes that lead to differences at the overwintering colony sites will provide a more complete picture of the potential impacts of different threats on the monarch migration.

PAGE 86

86 Figure 4 1. Major migratory flyways of the fall monarch butterfly migration in eastern North America. Light orange arrows indicate the primary routes taken from summering locations throughout Canada and the United States to the overwintering colonies in central Mexico Width of the arrows is roughly correlated with number of monarchs migrating along a flyway. Monarchs migrate in a general south or southwest orientation throughout the eastern range, so migrating monarchs are also found in areas not directly covered by the arrows. Due to size constraints of the map, the red diamond indicating sampling at the overwintering colonies is shown northeast of the actual location in central Mexico Based on the United States Forest Service ( 2008)

PAGE 87

87 Figure 4 2. Heavy isotope depletion gradients for carbon and hydrogen in monarch wing tissue. Arrows point in the direction of increasing depletion of heavy isotopes ( 13 C and D). Orange = carbon, blue = hydrogen. Based on data from Wassenaar an d Hobson (1998) ; hydrogen values modified according to Meehan and colleagues (2004)

PAGE 88

88 Figure 4 3. Map of overwintering colony sites. Cerro Pel—n and Gota de Agua are on opposite sites of the same mountain massif; all other sites are on separate massifs. Generated using Google Earth. Figure 4 4. Landmarks used to measure wing width. Width was measured in millimeters with a ruler aligned just below the white spot adjacent to the body, and the two white spots closest to the tip. The white line indicates t he axis along which measurement was made. Photo taken by author

PAGE 89

89 Figure 4 5. Network illustrating the distribution of genotypes among the overwintering colonies in Mexico Colonies are listed from north to south in legend. Node size is proportional to the number of samples sh aring a genotype, while each segment is proportional to the number of samples found at each overwintering colony. Lines connecting nodes represent a base pair difference. Small white circles represent unsampled intermediates.

PAGE 90

90 Figure 4 6. Network illustr ating the distribution of genotypes by sex. Node size is proportional to the number of samples sharing a genotype, while each segment is proportional to the number of samples of each sex. Lines connecting nodes represent a base pair difference.

PAGE 91

91 Figure 4 7. Neighbor joining tree illustrating the clear separation of monarchs ( D. plexippus orange nodes) from their sister species, the queen butterfly ( D. gilippus blue nodes), but no strong differentiation within monarchs. Branch labels are bootstrap suppor t based on 10,000 replicates. Although only the branch separating D. plexippus from D. gilippus is significant, other bootstrap values are shown to illustrate degree of support of branching within D. plexippus Colored circles at the tips indicate overwint ering colony of specimen. Overwintering colonies are listed from north to south in the legend.

PAGE 92

92 Figure 4 8. Origin differences at each overwintering colony site, as indicated by hydrogen isotope ratios (!D). Vertical axis reversed, so that more southerly origins (less negative !D values) are shown closer to the x axis. Colonies listed along x axis from south to north. Box limits show first and third quartiles, whiskers indicate 1.5 inter quartile range, and horizontal bar indicates median. Females shown in red; males shown in blue.

PAGE 93

93 Figure 4 9. Distribution of wing sizes at Mexican overwintering colonies. Colonies listed along x axis from south to north. Box limits show first and third quartiles, whiskers indicate 1.5 inter quartile range, and horizo ntal bar indicates median. Females shown in red; males shown in blue.

PAGE 94

94 Figure 4 10. Relationship between !D and wing size at the different overwintering colony sites. More northerly origins are indicated by more negative !D values. Colonies are listed fr om north to south in the legend. Colored lines show the linear regression for each colony, the black line is the overall regression, and the grey shadows show the confidence intervals for each regression. Circles indicate females; triangles indicate males.

PAGE 95

95 Table 4 1. Overwintering colonies at which specimens were collected. Colony a GPS Elevation Date Sierra Chincua N 19 40.068 W 100 16.602 3,300m 3/9/12 El Rosario N 19 35.478 W 100 15.882 3,200m 3/7/12 Cerro Pel—n N 19 22.152 W 100 15.618 2,650m 3/8/12 Gota de Agua N 19 22.296 W 100 16.242 2,900m 3/9/12 Piedra Herrada N 19 06.450 W 99 50.802 3,150m 3/8/12 a Colonies are listed north to south Table 4 2. Summary of specimens collected and analyzed. Colony a # Specimens Collected # Specimens Analyzed Male Female Male Female Sierra Chincua 34 34 16 13 El Rosario 38 34 15 13 Cerro Pel—n 27 24 14 13 Gota de Agua 23 21 11 13 Piedra Herrada 32 34 15 13 a Colonies are listed north to south Table 4 3. Summary of molecular diversity statistics Colony a n b S c c,d S c,d Sierra Chincua 21 13 0.0035 0.00034 2.20 0.382 3.02 0.326 El Rosario 27 23 0.0038 0.00032 2.37 0.303 5.05 0.354 Cerro Pel—n 27 12 0.0031 0.00027 1.96 0.361 2.63 0.294 Gota de Agua 27 16 0.0035 0.00029 2.17 0.339 3.51 0.317 Piedra Herrada 24 13 0.0037 0.00033 2.32 0.397 2.93 0.312 OVERALL 126 39 0.0035 0.00014 2.01 0.216 6.39 0.264 a Colonies are listed north to south b Number of individuals c Values are mean standard error d Per sequence Table 4 4. Pairwise F ST values for Mexican overwintering colonies Colony a Sierra Chincua El Rosario Cerro Pel—n Gota de Agua El Rosario 0.0177 Cerro Pel—n 0.0013 0.0027 Gota de Agua 0.0133 0.0082 0.0011 Piedra Herrada 0.0045 0.0022 0.0089 0.00274 a Colonies are listed north to south

PAGE 96

96 Table 4 5. AMOVA under the assumption that each overwintering colony is a separate population Source of Variation Degrees of freedom Sum of squares Variance component % of variation Among colonies 4 4.482 0.00041 0.04 Within colonies 247 271.728 1.10011 99.96 TOTAL 251 276.210 1.10052 Fixation Index ( F ST ) = 0.00037 1,023 permutations, P = 0.817 0.0119

PAGE 97

97 CHAPTER 5 SYNTHESIS My findings show that monarch butterflies have complex movement behaviors during and in addition to their mass migration. Therefore, the generally accepted paradigm that monarchs are distributed throughout North America during the spring and summer, but migrate to overwintering colonies masks fine scale comple xity that has implications for monarch conservation and evolution. T he monarch population lacks detectable genetic structure Therefore, it can be considered a single panmictic population However, my studies from three distinct geographic areas a putatively resident population in Miami, animals in Canada embarking upon the journey south and animals overwintering in Mexico all provided evidence of monarch movement behavior that is more complex than previously thought. The resident population in S outh Florida may serve as an alternative overwintering destination; viable adult butterflies are originating north of the published breeding limit; and the different Mexican overwintering colonies display subtle, but significant, differences in the geograp hic origins of their inhabitants All of these previously undocumented patterns raise interesting questions about monarch butterfly behavior and physiology that warrant further investigation. I discuss possible follow up studi es for each in turn. Resident Populations as Alternate Migratory Destinations By collecting monarchs throughout the year in South Florida, additional data on the degree of immigration and the reproductive status of immigrant butterflies can also be collected. Since juve nile hormone is a key signal in the maturation of monarch gonadal tissue, and its levels correlate with diapause status determining the titer of juvenile hormone in the hemolymph of freshly caught specimens provides a second

PAGE 98

98 means of assessing diapause status (Lessman et al. 1989). In addition to providing data about diapause status, t he nutritional condition of monarchs entering the South Florida population throughout the year c ould also be determined by measuring their lipid content (Gibo & McCurdy 1993) which would indicate whether immigrants show evidence of following a condition dependent alternate migration strategy If males immigrating to South Florida are following a condition dependent alternate migration strategy, then identifying the condition thresholds lea ding to migration to South Florida is critical to understanding the mechanism underlying this behavior. Another key question is whether these thresholds vary across the range, with time or with genetic variants. In addition, since I found that immigrant m ales were larger than resident butterflies, another unanswered question is whether this gives these males a reproductive advantage in South Florida that they would not experience in the Mexican overwintering colonies. I found no evidence for females using South Florida as an alternate migration destination I t is unclear whether females do not employ an alternate migration strategy, whether females have different condition thresholds that were not met during the 2010 2011 overwintering season, whether they use a different alternate destination, or whether as a result of my small sample size I simply failed to capture female immigrants. Across widely divergent taxa, alternative tactics appear more often in males than in females, most likely because of the gr eater energetic investment of females due to anisogamy (Taborsky & Brockmann 2010), so it is possible that monarch females do not employ an alternate migration strategy. Collecting additional specimens in South

PAGE 99

99 Florida, particularly throughout the overwint ering period, would indicate whether sample size was the reason I did not observe any immigrant females. In addition to more intensive sampling of the South Florida population, careful study of additional putatively resident populations in the United Stat es would be desirable to see whether or not males following alternative migration strategies are found at multiple sites as well as whether females are only found at some sites Other possibly resident populations have been reported in southern Texas and Arizona (Funk 1968; Journey North 2013). The locations and persistence of these populations may vary, and may shift with climate change or development pressure. The frequency with which immigrants from the main migratory population enter these additional r esident populations is also unknown. Evaluating the degree to which monarchs employ multiple alternate destinations is key to understanding the use of alternate migration strategies. Phenotypic Variation in the Mexican Overwintering Colonies Monarch phenot ypes vary across both their summer breeding range and the overwintering colonies in Mexico I found size differences at the overwintering colonies, which raises the question of whether there is additional phenotypic variation between the colonies For example, the shade s of orange wing pigments fall along a gradient from yellow orange (lighter color) to red orange (darker color) and these shades have been found to co vary with size larger monarchs have redder hues ( Davis 2009 ) Both adult wing and larval coloration vary with temperature, although these relationships appear to be different for males and females, and for monarchs originating from different parts of North America ( Davis et al. 2004; Davis et al. 2005). Insect size is also known to co v ary with temperature ( Dingle 1972 ; Wensler 1977 ) so it may be that the

PAGE 100

100 relationship between color and size is controlled by temperature, though this hypothesis needs investigation. Importantly, females have shown a significant, consistently strong prefer ence for males with less saturated orange shades (Davis et al. 2007) which should also have been smaller males, given the size color relationship described previously (Davis 2009) Perhaps surprisingly, Davis and colleagues found that females tended to pr efer larger males, although this preference was not significant across all experiments. Thus, investigating whether wing colors differ between overwintering colonies would provide insight into whether mating advantages vary across the colony sites for butt erflies with different wing colors. Only one forewing for each specimen I collected was used for stable isotope measurements so the digital image analysis used by Davis and colleagues could be applied to these specimens to evaluate wing color. My goal in such an analysis would be to evaluate whether butterflies at different overwintering colonies vary in their wing colors, whether color varies by natal origin as determined by hydrogen and carbon stable isotope profiles, if and how color co varies with size and whether color varies by larval condition, as indicated by nitrogen stable isotope profiles. Data from such a color analysis could also inform design of a controlled experiment to evaluate the interaction of temperature and larval nutritional condition on adult wing color and body size Differentiation of the Mexican Overwintering Colonies Wassenaar and Hobson (1998) found more animals from the western Great Lakes and Midwest regions than the Atlantic region Although my sampling regime was si milar, my study found a preponderance of butterflies from the eastern portion of their range. This discrepancy may be due to normal year to year variation, the period in the

PAGE 101

101 overwintering season in which the collections were made (their collections were in February, mine were in March ) or shifts in the distribution of the monarch population amongst the regions of their summer range Suitable breeding habitat varies widely throughout the summer range, so understanding how monarchs from different areas with in the range arrange themselves within the colonies provides information on the importance of particular parts of the summer range for the monarch population as a whole. I found that the monarchs from different summer ranges sorted into the colonies in dif ferent proportions. Since I analyzed dead monarchs, I have no data to indicate whether these proportions hold for monarchs that survive the overwintering season, or whether butterflies that originate from certain summering areas are more likely to perish d uring the winte r perhaps in a site specific manner due to microclimate differences among overwintering sites, and different sensitivity to condition based on natal origin In addition, mortality at the overwintering colonies has often been correlated with the severity of winter storms (Anderson & Brower 1996; Brower et al. 2004; Calvert et al. 1983). Gathering data that show how survival patterns differ at the various colonies could help the authorities of the monarch reserves in Mexico set colony specific policies for tourism, logging, and agriculture. My study at the Mexican overwintering colonies used dead monarchs collected from the forest floor near the end of the overwintering season. A logical follow up study would be to determine whether or not the patterns I observed are reproducible from one yea r to the next using the same sampling scheme. In addition, it would be useful to understand the association between patterns found in live butterflies and patterns found

PAGE 102

102 in dead ones An additional question of interest would be if and how differentiation between colonies by natal origin changes throughout the overwintering season. In addition, although colonies appear to be differentiated from one another with respect to natal origin, there is no obvious pattern in how they are differentiated I began the study with the null hypothesis that the overwintering colonies would not show differentiation with respect to natal origin. The obvious alternative hypothesis was that there would be a clear pattern perhaps from north to south in the range reflecting north to south natal origin; or perhaps clustering between colonies by origin, but without clear patterning with respect to the colonies themselves. Instead, b utterflies from areas covered by different flyways were present in different proportions at the differ ent colony sites, rather than there being any direct associations between flyways and specific sites Clear separation by natal origin would have been particularly interesting to observe, as such a pattern would have lent credence to the hypothesis that m onarchs mate non randomly with respect to natal origin. Of course, that animals are in close proximity does not mean there is mating going on. It remains unclear whether animals mate randomly with respect to geography; my finding of panmixia, along with th e findings of previous studies ( Lyons et al 2012; Wassenaar & Hobson 1998 ) suggests that they do However, Van Hook (1993) found that non random, dissortative mating based on size was occurring near the end of the overwintering season Unfortunately, she did not identify the natal origins of specimens in her study. Capturing animals in copula and analyzing their origins would be the definitive test of hypotheses regarding random mating with respect to geography.

PAGE 103

103 The overwintering colonies may be the b asis for panmixia. However, it is also possible that there is high connectivity across the summer range, a possibility that has not been rigorously explored. Applying my approach that integrates genetic and stable isotope analyses to butterflies collected at various sites during the summer breeding season would provide information on how areas within the summer breeding range are connected. Investigating butterflies in the summer range, particularly if studies were timed to sample each successive generation moving northward out of Mexico could provide detailed data on the connectivity between areas, and would complement the detailed data available for the southward migration (e.g., this study, Brindza et al. 2008; Monarch Watch 2009; Wassenaar & Hobson 1998 ). In addition, the stable isotope/natal origins approach could be particularly useful at the margins of the summer range, such as along the eastern edge of the Rocky Mountains the northwestern edge of the Canadian Great Plains and of course the northeas tern Canadian range into the Maritimes Given the evidence I found of range expansion along the latter northeast margin of the breeding range, it is possible that the summer range has extended through the Rocky Mountains to connect the eastern a nd western monarch populations. Future Directions Since I concentrated on regions in eastern North America, similar investigation of adjacent habitats, such as western North America, the Caribbean, Bermuda, and Central America, would indicate the effects of monarch movements in these areas. Substantial vicariant barriers (e.g., the Rocky Mountains and the Sonoran desert) exist that might impede gene flow between monarch populations in eastern and western North America. In addition, the generally accepted paradigm is that migrants from the west coast do not overwinter in the Mexican colonies (Urquhart & Urquhart 1977).

PAGE 104

104 However, no evidence of genetic varia tion has been found between these regions (Lyons et al. 2012). Using stable isotopes to determine the natal origins of butterflies found at the overwintering colonies on the California coast may provide some insights into the pathways that connect these po pulations. A map of carbon stable isotopes for milkweeds throughout the range of Danaus plexippus is an essential tool for continued use of stable isotopes to rigorously study monarch movements throughout their range The map developed by Hobs on, Wassenaar and Taylor (1999) is focused on the eastern breeding range, and was generated from stable isotope data collected for specimens from 33 sites distributed across the range. However, few sites were located either in the southeastern United States or along th e margins of the breeding range, and no sites outside the eastern breeding range were sampled. In addition, it would be useful to include data about variation in carbon isotopes for different milkweed species within the same geographic area. Expansion of t his map would be a prerequisite to applying my approach to investigate some of the key open questions suggested by my findings. Additional investigations looking at the fine scale processes at work throughout the monarch population will expand on the found ation laid by decades of investigation focused on the monarch migration. My findings show that the paradigm of a single migratory population in eastern North America needs to be revisited. The mass migration of the eastern population of D. plexippus overla ys fine scale patterns that indicate complexity in both migratory and breeding behavior. Rather than a monolithic entity, this panmictic monarch butterfly population exhibits subtle differences in space and time that have evolutionary and conservation impl ications for D. plexippus

PAGE 105

105 APPENDIX ISOTOPE FRACTIONATIO N IN DIFFERENT WING COLORS OF MONARCH BUTTERFLIES ( DANAUS PLEXIPPUS ) Introduction Stable isotopes have been used to as intrinsic markers to study wildlife migration. Because stable isotopes are incorporated into various tissues at different times, certain tissues carry isotopic signatures that reflect the geographic location in which the tissue was created. Bird feathers (Kelly et al. 2008), fish otoliths (Miller et al. 2010) and butterfly w ings (Hobson et al. 1999) are all tissues whose isotopic composition does not change after initial deposition. Thus, these tissues can be used to track movements between locations that exhibit different isotopic signatures (e.g., Bowen et al. 2005; Brattst ršm et al. 2008; Clegg et al. 2003; Miller et al. 2011; Szymanski et al. 2007). A key assumption underlying this technique is that samples from a particular type of tissue within a single individual are relatively homogenous, so that the isotopic composition measured in a small sample is representative of the tissue as a whole The unique seasonal migration of monarch butterflies in eastern North America has been investigated using analysis of stable carbon and hydrogen isotopes in wing tissue. Because monarch wings are composed chiefly of keratin proteins that are laid down du ring metamorphosis from larva to adult butterfly, wing tissue reflects the isotopic composition of the larval area ( Hobson et al. 1999 ). In monarch wings, hydrogen isotopes vary in conjunction with the isotopic signature of local meteoric water; carbon iso topes vary depending on the milkweed species on which a larva fed prior to pupation (Hobson et al. 1999). Hydrogen isotope ratios in eastern North America exhibit a roughly southeast northwest gradient, with increasing depletion of D

PAGE 106

106 to the northwest, whe reas carbon isotope ratios show increasing depletion of 13 C moving from northeast to southwest (Figure A 1). In contrast to the predictable geographic distribution of carbon and hydrogen isotopes, nitrogen isotopes in monarch wings are correlated with the nutritional condition of the larva. Animals enduring starvation are enriched for 15 N relative to 14 N, such that the higher the relative amount of 15 N, the worse the animal's condition (Gannes et al. 1998). Feeding experiments have not been conducted on mon arch butterflies; however, experiments in another Lepidopteran, the Japanese oak silkworm ( Antheraea yamamai ), are consistent with enrichment of 15 N as a signal of response to larval starvation (Mizota & Yamanaka 2011). Across animal taxa, nitrogen isotope enrichment consistently occurs between trophic levels, with higher trophic levels associated with greater enrichment in 15 N (Gannes et al 1998). Tibbets and colleagues (2008) also found a similar pattern of enrichment pattern between larvae and adults in insects that undergo complete metamorphosis. Thus, higher 15 N ratios are consistently found when nitrogenous compounds are recycled. A major assumption of previous studies using stable isotopes in monarch tissue (Hobson et al. 1999; Miller et al. 2011; Wassenaar & Hobson, 1998) is that homogenizing wing tissue is adequate to ensure consistent stable isotope compositions for all butterflies from a given area. Monarchs, however, exhibit some degree of natural variation in the intensity of their wing color s, from light to dark orange, and in the ratio of orange to black (Davis 2009; Davis et al. 2007). Fractionation of stable isotopes during synthesis and deposition of pigments involved in these wing colors could therefore

PAGE 107

107 affect stable isotope composition s of wings, but the degree to which this fractionation occurs has not been determined. The goal of this study was to test several hypotheses regarding isotope fractionation in orange and black monarch butterfly wing tissue. I hypothesized that both carbon and nitrogen would fractionate differently in different colors of monarch butterfly wing tissue. I also hypothesized that these fractionation differences would lead to different inferences about location based on stable carbon isotope ratios. Because stabl e nitrogen isotope ratios have no t shown significant differences in wing tissue between adults raised in similar conditions in previous studies (Tibbets et al. 2008), I further hypothesized that fractionation differences in nitrogen in orange and black segments would be too small to affect inferences about nutritional condition. Methods Specimen Selection To reduce the effects of possible confounding factors suc h as variability due to milkweed species or local variations in the composition of meteoric water, all specimens used for this analysis were collected from monarch butterflies raised in captivity at the Cockrell Butterfly Center at the Houston Museum of Na tural Science. All butterflies were raised in common cages with the same species of milkweed and the same source of local Houston water. In addition, all 13 specimens were from the same generation, so they were exposed to the same meteoric water. Finally, all wing sections were excised from homologous portions of the right hindwing of each specimen to minimize the effects of differences in pigment deposition across the wing ( e.g. wing sections adjacent to the body tend to be lighter)

PAGE 108

108 Stable Isotope Analysis All wing tissues were first washed in a 2:1 cholorform:methanol solution to remove surface contaminants and lipids. Specimens were air -dried; next, orange and black segments (~1.0 mg) were excised for analysis. Segments were loaded into tin capsules, and carbon and nitrogen isotope compositions were measured at the University of Florida Light Stable Isotope Mass Spec Lab. Isotope ratio measurements were corrected by the University of Florida Light Stable Isotope Mass Spec Lab using a calibrated lab standard measured before and after each wing color. !15N values are given relative to air, with a precision of 0.06; !13C values are given relative to VPDB, with a precision of 0.08. Isotope ratio data were grouped by wing color before analyzing in R ( R Development Core Team 2008). Measurements within each group were tested for normal distribution using the Shapiro -Wilk test (Shapiro & Wilk 1965). Variances were compared using F-tests, and Students two-sample t-tests were used to compare mean isotope ratios for each wing color. Results Corrected stable isotope ratio measurements for the 13 specimens included in this study are shown in Table A-1. Measurements within each color were normally distributed (Shapiro-Wilk, 0.931 W 0.963, 0.356 P 0.792), and variances for each color were equal for both carbon ( F12 = 1.067, P = 0.912) and nitrogen (F12 = 1.029, P = 0.961). Black wing segments were significantly more depleted in 13C than orange segments (Figure A-2, t24 = -3.039, P = 0.006), but no significant differences in !15N were found between wing colors (Figure A -3, t24 = 0.718, P = 0.479).

PAGE 109

109 Discussion My results are consistent with the hypothesis that carbon fractionates differently in different colors of monarch butterfly wing tissue. However, I found no support for fractionation differences with nitrogen. Although the difference in carbon stable isotope ratios in orange and black wing segments was statistically significant, this difference is generally too small to significantly affect the ide ntificati on of natal location. Although these differences mean that orange segments could nominally indicate more southerly origins than black segments, t he mapping of carbon stable isotope profiles is at a coarse scale (Figure A 1 bands are in 1 increments ), and the fractionation differences (mean standard error = 0.47 0.06) fall well below this scale. Thus, homogenized wing tissue is acceptable for identification of larval area, but using wing tissue homogenate may increase the amount of noise in the data. Since the intensity of orange pigments vary in monarch wings (Davis et al. 2007), any fractionation differences due to pigment density would further increase th e amount of noise. Black pigments, however, fluctuate little, so using black wing tissue would introduce as little noise as possible into stable isotope measurements The fractionation differences in the two colors present in monarch wing s raise questions about the kinetics involved in the physiological processes that create and deposit pigments in wing tissue. The wing is composed entirely of keratin proteins, but the pigments embedded in these keratins vary with color. Investigation of the molecular pathways involved in pigment deposition could illuminate the mechanisms underlying the differential fractionation observed in this study. In addition, specimens from wild populations should be analyzed to see if these results hold for wild populations. Finally, hydrogen isotopes should be analyzed in different wing colors to determine if hydrogen

PAGE 110

1 10 fracti onation also occurs in these tissues. If both hydrogen and carbon fractionate differently in wing tissue, then small errors in geographic assignment to larval area may be compounded and lead to misleading inferences. The results of this study indicate this is unlikely to be a severe problem, but repeating the approach to analyze hydrogen in each wing color would provide the data necessary to evaluate the validity of this conclusion.

PAGE 111

111 Figure A 1. Heavy isotope depletion gradients for carbon and hydrogen in monarch wing tissue. Arrows point in the direction of increasing depletion of heavy isotopes ( 13 C and D). Orange = carbon, blue = hydrogen. Red star indicates the sampling location at the Cockrell Butterfly Center in Houston, TX. Based on data from Wassen aar and Hobson (1998) ; hydrogen values modified according to Meehan and colleagues (2004)

PAGE 112

112 Figure A 2. Mean isotope fractionation in different colors of monarch butterfly wings is significantly different for carbon. Values are relative to VPDB. Box limi ts show first and third quartiles, whiskers indicate 1.5 inter quartile range, and horizontal bar indicates median.

PAGE 113

113 Figure A-3. Mean isotope fractionation in different colors of monarch butterfly wings is not significantly different for nitrogen. Values are relative to air. Box limits show first and third quartiles, whiskers indicate 1.5 inter -quartile range, and horizontal bar indicates median. Table A-1. Stable isotope data for different colored wing sections. Specimen ID "15N () "13C () Black Orange Differencea Black Orange Differencea CC.1A 4.13 3.89 0.24 -29.23 -29.34 0.11 CC.1B 1.57 1.29 0.28 -29.98 -29.51 -0.47 CC.1C 2.86 2.34 0.52 -29.60 -29.18 -0.42 CC.1D 2.46 1.95 0.51 -30.44 -30.04 -0.40 CC.1E 4.11 3.71 0.40 -29.71 -29.02 -0.71 CC.1F 3.86 3.47 0.39 -30.14 -29.40 -0.76 CC.1G 4.53 4.28 0.25 -30.13 -29.70 -0.43 CC.1H 4.03 3.63 0.40 -30.08 -29.53 -0.55 CC.1I 2.21 1.91 0.30 -29.63 -28.94 -0.69 CC.1J 0.75 0.65 0.10 -30.36 -29.78 -0.58 CC.1K 2.72 2.36 0.36 -30.05 -29.62 -0.43 CC.1L 1.03 0.61 0.42 -30.63 -30.11 -0.52 CC.1M 3.41 3.02 0.39 -30.46 -30.12 -0.34 aDifference between measurements for black and orange wing sections.

PAGE 114

114 LIST OF REFERENCES Alonso Mejia A (1996) Biology and conservation of overwintering monarch butterflies in Mexico. PhD Dissertation. University of Florida. Alonso Mejia A, Arellano Guillermo A, Brower LP (1992) Influence of temperature, surface body moisture and height above ground on survival of monarch butterflies overwintering in Mexico. Biotr opica 24 415 419. Alonso Mejia A, Glendinning JI, Brower LP (1993) The influence of temperature on crawling, shivering, and flying in overwintering monarch butterflies in Mexico. In: Biology and Conservation of the Monarch Butterfly (eds. Malcolm SB, Zalu cki MP), pp. 309 314. Natural History Museum of Los Angeles County, Los Angeles, CA. Alonso Mejia A, Montesinos Patino E, Rendon Salinas E, Brower LP, Oyama K (1998) Influence of forest canopy closure on rates of bird predation on overwintering Monarch but terflies Danaus plexippus L. Biological Conservation 85 151 159. Alonso Mejia A, Rendon Salinas E, Montesinos Patino E, Brower, LP. (1997) Use of lipid reserves by monarch butterflies overwintering in Mexico: implications for conservation. Ecological Appl ications 7 934 947. Anderson JB, Brower LP (1996) Freeze protection of overwintering monarch butterflies in Mexico: critical role of the forest as a blanket and an umbrella. Ecological Entomology 21 107 116. Anderson PL, Hellmich RL, Prasifka JR, Lewis LC (2005) Effects on fitness and behavior of monarch butterfly larvae exposed to a combination of Cry1Ab expressing corn anthers and pollen. Environmental Entomology 34 944 952. Atterholt AL, Solensky MJ (2010) Effects of larval rearing density and food availability on adult size and coloration in monarch b utterflies (Lepidoptera: Nymphalidae). Journal of Entomological Science 45 366 377. Bandelt H, Forster P, Rohl A (1999) Median joining networks for inferring intraspecific phylogenies. Molecular Biology and Evolution 16 37 48. Bartholomew CS, Yeargan KV (2001) Phenology of milkweed (Asclepiadaceae) growth and monarch (Lepidoptera: Nymphalidae) reproduction in Kentucky and ovipositional preference between common and honeyvine milkweed. Journal of the Kansas Entomological Society 74 211 220. Batalden RV, Oberhauser K, Peterson AT (2007) Ecological niches in sequential generations of eastern North American monarch butterflies (Lepidoptera : Danaidae): The ecology of migration and likely climate change implications. Environmental En tomology 36 1365 1373.

PAGE 115

115 Bojorquez Tapia LA, Brower LP, Castilleja G et al. (2003) Mapping expert knowledge: redesigning the Monarch Butterfly Biosphere Reserve. Conservation Biology 17 367 379. BopprŽ M (1993) The American monarch: Courtship and chemical communicaiton of a peculiar danaine butterfly. In: Biology and Conservation of the Monarch Butterfly (eds. Malcolm SB, Zalucki MP), pp. 29 41. Natural History Museum of Los Angeles County, Los Angeles, CA. Bowen GJ, Revenaugh J (2003) Interpolating the isotopic composition of modern meteoric precipitation. Water Resources Research 39 1299. Bowen GJ, Wassenaar LI, Hobson KA (2005) Global application of stable hydrogen and oxygen isotopes to wildlife for ensics. Oecologia 143 337 348. Bowen GJ, Wilkinson B (2002) Spatial distribution of 18 O in meteoric precipitation. Geology 30 315 318. Bowler DE, Benton TG (2005) Causes and consequences of animal dispersal strategies: relating individual behaviour to spatial dynamics. Biological Reviews 80 205 225. Brattstršm, O, Wassenaar, LI, Hobson, KA, kesson, S (2008) Placing butterflies on the ma p: Testing regional geographical resolution of three stable isotopes in Sweden using the monophagus peacock Inachis io Ecography 31 490 498. Brindza LJ, Brower LP, Davis AK, Van Hook T (2008) Comparative success of monarch butterfly migration to overwint ering sites in Mexico from inland and coastal sites in Virginia. Journal of the Lepidopterists' Society 62 189 200. Brower AVZ, Jeansonne MM (2004) Geographical populations and "subspecies" of new world monarch butterflies (Nymphalidae) share a recent ori gin and are not phylogenetically distinct. Annals of the Entomological Society of America 97 519 523. Brower LP (1961) Studies on the migration of the monarch butterfly I. Breeding population of Danaus plexippus and D. gilippus berenice in south central F lorida. Ecology 42 76 83. Brower LP (1985) New perspectives on the migration biology of the monarch butterfly, Danaus plexippus L. Contributions in Marine Science 27 748 785. Brower LP (1996) Monarch butterfly orientation: Missing pieces of a magnificent puzzle. Journal of Experimental Biology 199 93 103. Brower LP, Calvert WH, Hedrick LE, Christian J (1977) Biological observations on an overwintering colony of monarch butterflies ( Danaus plexippus Danaidae) in Mexico. Journal of the Lepidopterists' Society 31 232 242.

PAGE 116

116 Brower LP, Castilleja G, Peralta A, et al. (2002) Quantitative changes in forest quality in a principal overwintering area of the monarch butterfly in Mexico, 1971 1999. Conservation Biology 16 346 359. Brower LP, Fink LS, Walford P ( 2006) Fueling the fall migration of the Monarch butterfly. Integrative and Comparative Biology 46 1123 1142. Brower LP, Kust DR, Salinas ER, et al. (2004) Catastrophic winter storm mortality of monarch butterflies in Mexico during January 2002. In: Monarc h Butterfly Biology & Conservation (eds. Oberhauser KS, Solensky MJ). Cornell University Press, Ithaca, NY. Brower LP, Oberhauser KS, Boppre M, Brower AVZ, Vane Wright RI (2007) Monarch sex: ancient rites or recent wrongs? Antenna 31 12 18. Brower LP, Wil liams EH, Fink LS, Zubieta RR, Ramirez MI (2008) Monarch butterfly clusters provide microclimatic advantages during the overwintering season in Mexico. Journal of the Lepidopterists' Society 62 177 188. Brower LP, Williams EH, Slayback DA, et al. (2009) O yamel fir forest trunks provide thermal advantages for overwintering monarch butterflies in Mexico. Insect Conservation and Diversity 2 163 175. Bruce AB (1910) The Mendelian theory of heredity and the augmentation of vigor. Science 32 627 628. Calder—n RS, Alvarez BXM, Reyes JL, Garc’a Serrano E, Rodr’guez HS (2004) The Monarch Butterfly Biosphere Reserve, Michoac‡n, Mexico. In: Conserving Migratory Pollinators and Nectar Corridors in Western North America (ed. Nabhan GP), p. 190. University of Arizona Press, Tucson, AZ. Calvert WH (1999) Patterns in the spatial and temporal use of Texas milkweeds (Asclepiadaceae) by the monarch butterfly ( Danaus plexippus L.) during fall, 1996. Journal of the Lepidopterists' Society 53 37 44. Calvert WH, Brower LP (1986) The location of monarch butterfly ( Danaus plexippus L.) overwintering colonies in Mexico in relation to topography and climate. Journal of the Lepidopterists' Society 40 164 187. Calvert WH, Zuchowski W, Brower LP (1983) The effect of rain, sno w and freezing temperatures on overwintering butterflies in Mexico. Biotropica 15 42 47. Clegg SM, Kelly JF, Kimura M, Smith TB (2003) Combining genetic markers and stable isotopes to reveal population connectivity and migration patterns in a Neotropical migrant, Wilson's warbler ( Wilsonia pusilla ). Molecular Ecology 12 819 830. Cockerham CC, Weir BS (1993) Estimation of gene flow from F Statistics. Evolution 47 855 863.

PAGE 117

117 Cohen JA (1983) Chemical interactions among milkweed plants (Asclepiadaceae) and lepidopteran herbivores. PhD Dissertation. University of Florida. Cohen JA, Brower LP (1982) Oviposition and larval success of wild monarch butterflies (Lepidoptera: Danaidae) in relation to host plant size and cardenolide concentration. Journal of the Kan sas Entomological Society 55 343 348. Combreau O, Riou S, Judas J, Lawrence M, Launay F (2011) Migratory pathways and connectivity in Asian Houbara Bustards: Evidence from 15 years of satellite tracking. PLoS One 6 8. Committee on the Status of Species a t Risk in Ontario (COSSARO) (2009) COSSARO Candidate Species at Risk Evaluation Form for Monarch ( Danaus plexippus ): Assessed November 2009 by COSSARO as Special Concern. November 2009: Final. Ontario Ministry of Natural Resources, Peterborough, ON, Canada Cook BD, Bunn SE, Hughes JM (2007) Molecular genetic and stable isotope signatures reveal complementary patterns of population connectivity in the regionally vulnerable southern pygmy perch ( Nannoperca australis ). Biological Conservation 138 60 72. Copl en TB (1994) Reporting of stable hydrogen, carbon, and oxygen isotopic abundances. Pure & Applied Chemistry 66 273 276. Daniels JC (1995) Seasonal variation in the little sulphur butterfly, Eurema lisa lisa in central Florida: how it compares to other sympatric Eurema species (Lepidoptera: Pieridae). Holarctic Lepidoptera 2 59 65. Darwin C (1859) On the Origin of Species. In: From so Simple a Beginning: The Four Great Books of Charles Darwin (ed. Wilson EO ), pp. 441 760. W. W. Norton & Co., New York. Davis AK (2009) Wing color of monarch butterflies ( Danaus plexippus ) in eastern North America across life stages: migrants are "redder" than breeding and overwintering stages. Psyche 2009 1 5. Davis AK, Cope N Smith A, Solensky MJ (2007) Wing color predicts future mating success in male monarch butterflies. Annals of the Entomological Society of America 100 339 344. Davis AK, Farrey BD, Altizer S (2004) Quantifying monarch butterfly larval pigmentation using digital image analysis. Entomologia Experimentalis Et Applicata 113 145 147. Davis AK, Farrey BD, Altizer S (2005) Variation in thermally induced melanism in monarch butterflies (Lepidoptera : Nymphalidae) from three North American populations. Journal of Thermal Biology 30 410 421.

PAGE 118

118 de Roode JC, Pedersen AB, Hunter MD, Altizer S (2008) Host plant species affects virulence in monarch butterfly parasites. Journal of Animal Ecology 77 120 126. Desmet P, Brouillet L (2013) Database of Vascular Plants of Cana da (VASCAN): A community of contributed taxonomic checklist of all vascular plants of Canada, Saint Pierre and Miquelon, and Greenland. PhytoKeys 25 55 67. Dingle H (1972) Migration strategies of insects. Science 175 1327 1335. Dingle H, Zalucki MP, Rochester WA, Armijo Prewitt T (2005) Distribution of the monarch butterfly, Danaus plexippus (L.) (Lepidoptera: Nymphalidae), in western North America. Biological Journal of the Linnean Society 85 491 500. Dockx C (2002) Migration o f the North American monarch, Danaus plexippus to Cuba. PhD Dissertation. University of Florida. Dockx C (2007) Directional and stabilizing selection on wing size and shape in migrant and resident monarch butterflies, Danaus plexippus (L.), in Cuba. Biolo gical Journal of the Linnean Society 92 605 616. Dockx C, Brower LP, Wassenaar LI, Hobson KA (2004) Do North American monarch butterflies travel to Cuba? Stable isotope and chemical tracer techniques. Ecological Applications 14 1106 1114. Dudley R, Srygl ey R (1994) Flight physiology of neotropical butterflies: Allometry of airspeds during natural free flight. Journal of Experimental Biology 191 125 139. Drummond AJ, Ashton B, Buxton S, et al. (2012) Geneious v6.1.5. Available from http://www.geneious.com Eanes WF (1979) The monarch butterfly as a paradigm of genetic structure in a highly dispersive species. In: Movement of Highly Mobile Insects: Concepts and Methodology in Research (eds. Rabb RL, Kennedy GG), pp. 88 103. North Carolina State University P ress, Raleigh, NC. Eanes WF, Koehn RK (1978) Analysis of genetic structure in the monarch butterfly, Danaus plexippus L. Evolution 32 784 797. East EM (1910) A Mendelian interpretation of variation that is apparently continuous. American Naturalist 44 65 82. Edgar RC (2004) MUSCLE: multiple sequence alignment with high accuracy and high throughput. Nucleic Acids Research 32 1792 1797. Excoffier L, Laval G, Schneider S (2007) Arlequin (version 3.0): An integrated software package for population genetics data analysis. Evolutionary Bioinformatics 1 47 50.

PAGE 119

119 Fahrig L, Merriam G (1994) Conservation of fragmented populations. Conservation Biology 8 50 59. Felsenstein J (1974) The evolutionary advantage of recombination. Genetics 78 737 756. Foote AD, Newton J, Piertney SB, Willerslev E, Gilbert MTP (2009) Ecological, morphological and genetic divergence of sympatric North Atlantic killer whale populations. Molecular Ecology 18 5207 5217. Funk RS (1968) Overwintering of monarch butterflies as a breeding colony in southwestern Arizona. Journal of the Lepidopterists' Society 22 63 64. Frey D (1999) Resistance to mating by female monarch butterflies. In: 1997 North American Conference on the Monarch Butterfly (eds. Hoth J, Merino L, Oberhauser KS et al. ), pp. 79 87. Comisi—n Para La Cooperaci—n Ambiental/Commission for Environmental Cooperation, MontrŽal, Canada. Gannes LZ, del Rio CM, Koch P (1998) Natural abundance variations in stable isotopes and their potential uses in animal physiologica l ecology. Comparative Biochemistry and Physiology Part A: Molecular & Integrative Physiology 119 725 737. Garc’a Serrano E, Reyes JL, Alvarez BXM (2004) Locations and area occupied by monarch butterflies overwintering in Mexico from 1993 to 2002. In: Mon arch Butterfly Biology & Conservation (eds. Oberhauser KS, Solensky MJ), pp. 129 134. Cornell University Press, Ithaca, NY. Garland MS, Davis AK (2002) An examination of monarch butterfly ( Danaus plexippus ) autumn migration in coastal Virginia. American Mi dland Naturalist 147 170 174. Gibo DL, McCurdy JA (1993) Lipid accumulation by migrating monarch butterflies ( Danaus plexippus L.). Canadian Journal of Zoology 71 76 82. Goehring L, Oberhauser KS (2002) Effects of photoperiod, temperature, and host plant age on induction of reproductive diapause and development time in Danaus plexippus Ecological Entomology 27 674 685. GomŽz D’az E, Gonz‡lez Sol’s J (2007) Geographic assignment of seabirds to their origin: Combining morphologic, genetic, and biogeochemical analyses. Ecological Applications 17 1484 1498. Greenfield MD, Shelly TE (1985) Alternative mating strategies in a desert grasshopper: evidence of density dependence. Animal Behaviour 33 1192 1210. Gršnroos J, Green M, Alerstam T (2012) To fly or not to fly depending on winds: shorebird migration in different seasonal wind regimes. Animal Behaviour 83 1449 1457.

PAGE 120

120 Hans ki I (1998) Metapopulation dynamics. Nature 396 41 49. Hartl DL, Clark AG (2006) Principles of Population Genetics 4th edn. Sinauer Associates, Sunderland, MA. Heape W (1931) Emigration, Migration, and Nomadism W. Heffer and Sons, Ltd., Cambridge, UK. Herman WS, Tatar M (2001) Juvenile hormone regulation of longevity in the migratory monarch butterfly. Proceedings of the Royal Society of London Series B Biological Sciences 268 2509 2514. Hews DK, Knapp R, Moore MC (1994) Early exposure to androgens aff ects adult expression of alternative male types in tree lizards. Hormones and Behavior 28 96 115. Hilburn DJ (1989) A non migratory, non diapausing population of the monarch butterfly, Danaus plexippus (Lepidoptera: Danaidae), in Bermuda. Florida Entomologist 72 494 499. Hill Jr. HF, Wenner AM, Wells PH (1976) Reproductive behavior in an overwintering aggregation of monarch butterflies. American Midland Naturalist 95 10 19. Hobson KA (1999) Tracing origins and migration of wildlife using stable i sotopes: A review. Oecologia 120 314 326. Hobson KA, Wassenaar LI (1997) Linking brooding and wintering grounds of neotropical migrant songbirds using stable hydrogen isotopic analysis of feathers. Oecologia 109 142 148. Hobson KA, Wassenaar LI, Taylor O R (1999) Stable isotopes (delta D and delta C 13) are geographic indicators of natal origins of monarch butterflies in eastern North America. Oecologia 120 397 404. Hšglund J (2009) Evolutionary Conservation Genetics Oxford University Press, Oxford, UK. Houle D (1992) Comparing evolvability and variability of quantitative traits. Genetics 130 195 204. Houle D, Morikawa B, Lynch M (1996) Comparing mutational variabilities. Genetics 143 1467 1483. Howard E, Davis AK (2004) Documenting the spring movements of monarch butterflies with Journey North, a citizen science program. In: Monarch Butterfly Biology & Conservation (eds. Oberhauser KS, Solensky MJ), pp. 105 114. Cornell University Press, Ithaca, NY.

PAGE 121

121 Howard E, Davis AK (2009) The fall migration flyways of monarch butterflies in eastern North America revealed by citizen scientists. Journal of Insect Conservation 13 279 286. Hughes JM, Zalucki MP (1984) Genetic variation in a continuously breeding p opulation of Danaus plexippus L. (Lepidoptera: Nymphalidae). Heredity 52 1 7. Hughes RD (1979) Movement in population dynamics. In: Movement of Highly Mobile Insects: Concepts and Methodology in Research (eds. Rabb RL, Kennedy GG), pp. 14 34. North Caroli na State University Press, Raleigh, NC. International Butterfly Breeders Association (2013) Membership Distribution Maps. http://www.butterflybreeders.org/public/maps/membmaps.html Irwin DE, Irwin JH, Smith TB (2011) Genetic variation and seasonal migrato ry connectivity in Wilson's warblers ( Wilsonia pusilla ): species level differences in nuclear DNA between western and eastern populations. Molecular Ecology 20 3102 3115. Journey North (2013) Spring Monarch Butterfly Migration: Comparative Maps. http://ww w.learner.org/jnorth/tm/monarch/MigrationMaps Judge KA, Ting JJ, Gwynne DT (2008) Condition dependence of male life span and calling effort in a field cricket. Evolution 62 868 878. Kelly JF, Johnson MJ, Langridge S, Whitfield M (2008) Efficacy of stable isotope ratios in assigning endangered migrants to breeding and wintering sites. Ecological Applications 18 568 576. Kimura M (1983) The Neutral Theory of Molecular Evolution Cambridge University Press, Cambridge, UK. Kimura M, Ohta T (1971) Protein polymorphism as a phase of molecular evolution. Nature 229 467 469. Kirschner M, Gerhart J (1998) Evolvability. Proceedings of the National Academy of Sciences 95 8420 8427. Knight A (1998) A Population Study of Monarch Butterflies in North Central and South Florida. Master of Science Thesis. University of Florida. Knight A, Brower LP (2009) The influence of Eastern North American autumnal migrant monarch butterflies ( Danaus plexippus L .) on continuously breeding resident monarch popula tions in southern Florida. Journal of Chemical Ecology 35 816 823.

PAGE 122

122 Ladner DT, Altizer S (2005) Oviposition preference and larval performance of North American monarch butterflies on four Asclepias species. Entomologia Experimentalis Et Applicata 116 9 20. Leary CJ, Fox DJ, Shepard DB, Garcia AM (2005) Body size, age, growth and alternative mating tactics in toads: satellite males are smaller but not younger than calling males. Animal Behaviour 70 663 671. Leong KLH (1990) Microenvironmental factors a ssociated with the winter habitat of the monarch butterfly (Lepidoptera, Danaidae) in central California. Annals of the Entomological Society of America 83 906 910. Leong KLH, O'Brien E, Lowerisen K, Colleran M (1995) Mating activity and status of overwin tering monarch butterflies (Lepidoptera: Danaidae) in central California. Annals of the Entomological Society of America 88 45 50. Leong KLH, Sakai WH, Bremer W, Feuerstein D, Yoshimura G (2004) Analysis of the pattern of distribution and abundance of mon arch overwintering sites along the California coastline. In: Monarch Butterfly Biology & Conservation (eds. Oberhauser KS, Solensky MJ), pp. 177 185. Cornell University Press, Ithaca, NY. Lessman CA, Herman WS, Schooley DA et al. (1989) Detection of juven ile hormone I, II and III in adult monarch butterflies ( Danaus plexippus ). Insect Biochemistry 19 431 433. Lewis OT, Thomas CD, Hill JK, et al. (1997) Three ways of assessing metapopulation structure in the butterfly Plebejus argus. Ecological Entomology 22 283 293. Librado P, Rozas J (2009) DnaSP v5: a software for comprehensive analysis of DNA polymorphism data. Bioinformatics 25 1451 1452. Lushai G, Smith DAS, Goulson D, Allen JA, Maclean N (2003) Mitochondrial DNA clocks and the phylogeny of Danaus b utterflies. Insect Science and its Application 23 309 315. Lynch M, Conery J, Burger R (1995) Mutational meltdowns in sexual populations. Evolution 49 1067 1080. Lynch M, Gabriel W (1990) Mutation load and the survival of small populations. Evolution 44 1725 1737. Lyons JI, Pierce AA, Barribeau SM, et al. (2012) Lack of genetic differentiation between monarch butterflies with divergent migration destinations. Molecular Ecology 21 3433 3444. MacArthur RH, Wilson EO (1967) The theory of island biogeograph y Princeton University Press, Princeton, N.J.

PAGE 123

123 Macdonald P, Yarrow S (2003) Regulation of Bt crops in Canada. Journal of Invertebrate Pathology 83 93 99. Malcolm SB (1987) Monarch butterfly migration in North America: controversy and conservation. Trends i n Ecology & Evolution 2 135 138. Malcolm SB, Cockrell BJ, Brower LP (1987) Monarch butterfly voltinism: effects of temperature constraints at different latitudes. Oikos 49 77 82. Manel S, Schwartz MK, Luikart G, Taberlet P (2003) Landscape genetics : combining landscape ecology and population genetics. Trends in Ecology & Evolution 18 189 197. Marvier M, McCreedy C, Regetz J, Kareiva P (2007) A meta analysis of effects of Bt cotton and maize on nontarget invertebrates. Science 316 1475 1477. McCue MD, Pollock ED (2008) Stable isotopes may provide evidence for starvation in reptiles. Rapid Communications in Mass Spectrometry 22 2307 2314. Meehan TD, Giermakowski JT, Cryan PM (2004) GIS based model of stable hydrogen isotope ratios in North American growing season precipitation for use in animal movement studies. Isotopes in Environmental and Health Studies 40 291 300. Miller JA, Bellinger MR, Golden JT, Fujishin L, Banks MA (2010) Integration of natural and artificial markers in a mixed stock analys is of Chinook salmon ( Oncorhynchus tshawytscha ). Fisheries Research 102 152 159. Miller NG, Wassenaar LI, Hobson KA, Norris DR (2011) Monarch butterflies cross the Appalachians from the west to recolonize the east coast of North America. Biology Letters 7 43 46. Mizota C, Yamanaka T (2011) Diet consumer nitrogen isotope fractionation for prolonged fasting arthropods. Isotopes in Environmental and Health Studies 47 483 488. Monarch Watch. 2009. Monarch Watch: Dedicated to Education, Conservation & Researc h. Monarch Watch. The Kansas Biological Survey. University of Kansas. http://www.monarchwatch.org/index.html Muller HJ (1963) The need for recombination to prevent genetic deterioration. Genetics 48 903. Neves VC, Fraga JC, Schaefer H, et al. (2001) The occurence of the Monarch butterfly, Danaus plexippus L. in the Azores, with a brief review of its biology. Arquipelago Boletim da Universidade dos Acores Ciencias Biologicas e Marinhas 18A 17 24. Oberhauser KS (1988) Male monarch butterfly spermatophore m ass and mating strategies. Animal Behaviour 36 1384 1388.

PAGE 124

124 Oberhauser KS (1989) Effects of spermatophores on male and female monarch butterfly reproductive success. Behavioral Ecology and Sociobiology 25 237 246. Oberhauser KS, Frey D (1999) Coercive mating by overwintering male monarch butterflies. In: 1997 North American Conference on the Monarch Butterfly (eds. Hoth J, Merino L, Oberhauser KS et al. ), pp. 67 78. Comisi—n Para La Cooperaci—n Ambiental/Commission for Envir onmental Cooperation, MontrŽal, Canada. Oberhauser KS, Rivers ERL (2003) Monarch butterfly ( Danaus plexippus ) larvae and Bt maize pollen: a review of ecological risk assessment for non target species. AgBiotechNet 5 1 7. Paez DJ, Brisson Bonenfant C, Ross ignol O et al. (2011) Alternative developmental pathways and the propensity to migrate: a case study in the Atlantic salmon. Journal of Evolutionary Biology 24 245 255. Paithankar KR, Prasad KSN (1991) Precipitation of DNA by polyethylene glycol and etha nol. Nucleic Acids Research 19 1346. Parks Canada (2011) Point Pelee National Park of Canada: Monarch Migration. http://www.pc.gc.ca/eng/pn np/on/pelee/natcul/natcul5.aspx. Parmesan C, Yohe G (2003) A globally coherent fingerprint of climate change impacts across natural systems. Nature 421 37 42. Pawson S, Berndt L (2004) Observations of monarch butterfly overwintering behaviour in Christchurch. Weta 27 22 30. Pinnegar JK, Polunin NVC (2000) Contributions of stable isotope data to elucidating food webs of Mediterranean rocky littoral fishes. Oecologia 122 399 409. Pliske TE (1975) Courtship behavior of the monarch butterfly, Danaus plexippus L. Annals of the Entomological Society of America 68 143 151. Prasifka PL, Hellmich RL, Prasifka JR, Lewis LC (2007) Effects of Cry1Ab expressing corn anthers on the movement of monarch butterfly larvae. Environmental Entomology 36 228 233. Pulliam HR (1988) Sources, sinks, and popula tion regulation. The American Naturalist 132 652 661. Pyle, RM (1999) Chasing Monarchs: Migrating with the Butterflies of Passage Houghton Mifflin Co., Boston. R Development Core Team (2008) R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria.

PAGE 125

125 Rabb RL (1979) Regional Research on Insect Movement: Initial Considerations. In: Movement of Highly Mobile Insects: Concepts and Methodology in Research (eds. Rabb RL, Kennedy GG), pp. 2 12. North C arolina State University Press, Raleigh, NC. Ramirez MI, Azcarate JG, Luna L (2003) Effects of human activities on monarch butterfly habitat in protected mountain forests, Mexico. Forestry Chronicle 79 242 246. Ramirez MI, Miranda R, Zubieta R, Jimenez M (2007) Land Cover and Road Network Map for the Monarch Butterfly Biosphere Reserve in Mexico, 2003. Journal of Maps 181 190. Raymond M, Rousset F (1995) An exact test for population differentiation. Evolution 49 1280 1283. Reppert SM (2006) A colorful model of the circadian clock. Cell 124 233 236. Rothschild M (1978) Hell's angels. Antenna 2 38 39. Rubenstein DR, Hobson KA (2004) From birds to butterflies: animal movement patterns and stable isotopes. Trends in Ecology & Evolution 19 256 263. Schmidt Koenig K (1985) Migration strategies of monarch butterflies. Contributions in Marine Science 27 786 798. Shapiro SS, Wilk MB (1965) An analysis of variance test for normality (complete samples). Biometrika 52 591 611. Sharp Z (2007) Principles of Stable Isotope Geochemistry Pearson Prentice Hall, Upper Saddle River, NJ. Slayback DA, Brower LP (2007) Further aerial surveys confirm the extreme localization of overwintering monarch butterfly colonies in Mexico. American En tomologist 53 146 149. Slayback DA, Brower LP, RamÂ’rez MI, Fink LS (2007) Establishing the Presence and Absence of Overwintering Colonies of the Monarch Butterfly in Mexico by the Use of Small Aicraft. American Entomologist 53 28 40. Smith MD, Schrank HE Brockmann HJ (2013) Measuring the costs of alternative reproductive tactics in horseshoe crabs, Limulus polyphemus Animal Behaviour 85 165 173. Solensky MJ (2004) The effect of behavior and ecology on male mating success in overwintering monarch butter flies ( Danaus plexippus ). Journal of Insect Behavior 17 723 743.

PAGE 126

126 Solensky MJ, Larkin E (2003) Temperature induced variation in larval coloration in Danaus plexippus (Lepidoptera: Nymphalidae). Annals of the Entomological Society of America 96 211 216. Srygley RB, Lorch PD (2011) Weakness in the band: nutrient mediated trade offs between migration and immunity of Mormon crickets, Anabrus simplex Animal Behaviour 81 395 400. Stephens M, Scheet P (2005) Accounting for decay of linkage disequilibrium in h aplotype inference and missing data imputation. American Journal of Human Genetics 76 449 462. Stephens M, Smith NJ, Donnelly P (2001) A new statistical method for haplotype reconstruction from population data. American Journal of Human Genetics 68 978 989. Stimson J, Kasuya M (2000) Decline in the frequency of the white morph of the monarch butterfly ( Danaus plexippus plexippus L., Nymphalidae) on Oahu, Hawaii. Journal of the Lepidopterists' Society 54 29 32. Stimson J, Meyers L (1984) Inheritance and frequency of a color polymorphism in Danaus plexippus (Lepidoptera: Danaidae) on Oahu, Hawaii. Journal of Research on the Lepidoptera 23 153 160. Strecker U, Wilkens H (2000) Danaus plexippus (Linnaeus, 1758) new for Lanzarote (Canary Islands) (Lepidoptera, Nymphalidae). Atalanta 31 61 62. Szymanski ML, Afton AD, Hobson KA (2007) Use of stable isotope methodology to determine natal origins of mallards at a fine scale Midwest within the upper Midwest. Journal of Wildlife Management 71 1317 1324. Taborsky M, Brockmann, HJ (2010) Alternative reproductive tactics and life history phenotypes. In: Animal Behaviour: Evolution and Mechanisms (Ed. by Kappeler, P.), pp. 537 586. Springer Verlag, New Yo rk. Tajima F (1983) Evolutionary relationship of DNA sequences in finite populations. Genetics 105 437 460. Taylor LR, Taylor RAJ (1977) Aggregation, migration, and popultion mechanics. Nature 265 415 421. Tibbets TM, Wheeless LA, del Rio CM (2008) Isoto pic enrichment without change in diet: An ontogenetic shift in delta N 15 during insect metamorphosis. Functional Ecology 22 109 113. Turlure C, Baguette M, Stevens VM, Maes D (2011) Species and sex specific adjustments of movement behavior to landscape h eterogeneity in butterflies. Behavioral Ecology 22 967 975.

PAGE 127

127 United Nations Educational, Scientific and Cultural Organization (UNESCO) (2008) WHC 08/32.COM/8B.17: Examination of nomination of natural, mixed and cultural properties to the World Heritage Lis t Monarch Butterfly Biosphere Reserve (MEXICO). http://whc.unesco.org/en/decisions/1476. United States Forest Service (2008) Monarch Butterfly: North America's Migrating Insect US Dept. of Agriculture, Washington, DC. Urquhart FA (1960) The Monarch Butterfly University of Toronto Press, Toronto. Urquhart FA (1987) The Monarch Butterfly: International Traveler Nelson Hall, Chicago. Urquhart FA, Urquhart NR (1977) Overwintering areas and migratory routes of the monarch butterfly ( Danaus p. plexippus L epidoptera: Danaidae) in North America, with special reference to the western population. Canadian Entomologist 109 1583 1589. Urquhart FA, Urquhart NR (1978) Autumnal migration routes of the eastern population of the monarch butterfly ( Danaus p. plexippu s L.; Danaidae; Lepidoptera) in North America to the overwintering site in the Neovolcanic Plateau of Mexico. Canadian Journal of Zoology 56 1759 1764. Urquhart FA, Urquhart NR (1979a) Aberrant autumnal migration of the eastern population of the monarch b utterfly, Danaus plexippus plexippus (Lepidoptera: Danaidae) as it relates to the occurrence of strong westerly winds. Canadian Entomologist 111 1281 1286. Urquhart FA, Urquhart NR (1979b) Breeding areas and overnight roosting locations in the northern ra nge of the monarch butterfly ( Danaus plexippus plexippus ) with a summary of associated migratory routes. Canadian Field Naturalist 93 41 47. Van Voorhies WA (1996) Bergmann size clines: A simple explanation for their occurrence in ectotherms. Evolution 50 1259 1264. Van Hook TS (1993) Non random mating in monarch butterflies overwintering in Mexico. In: Biology and Conservation of the Monarch Butterfly (eds. Malcolm SB, Zalucki MP), pp. 49 60. Natural History Museum of Los Angeles County, Los Angeles. Velde LV, Van Dyck H (2013) Lipid economy, flight activity and reproductive behaviour in the speckled wood butterfly: on the energetic cost of territory holding. Oikos 122 555 562. Wassenaar LI, Hobson KA (1998) Natal origins of migratory monarch butterfl ies at wintering colonies in Mexico: New isotopic evidence. Proceedings of the National Academy of Sciences of the United States of America 95 15436 15439.

PAGE 128

128 Watterson GA (1975) On the number of segregating sites in genetical models without recombination. T heoretical Population Biology 7 256 276. Wells SM, Pyle RM, Collins NM (1983) Monarch butterfly: Threatened phenomenon: Mexican winter roosts. In: The IUCN Invertebrate Red Data Book (eds. Wells SM, Pyle RM, Collins NM), pp. 463 466. International Union f or Conservation of Nature and Natural Resources, Gland, Switzerland. Wensler RJ (1977) Ultrastructure of indirect flight muscles of monarch butterfly, Danaus plexippus (L.), with implications for fuel utilization. Acta Zoologica 58 157 167. Whinnett A, Br ower AVZ, Lee M M, Willmott KR, Mallet J (2005) Phylogenetic utility of Tektin a novel region for inferring systematic relationships among lepidoptera. Annals of the Entomological Society of America 98 873 886. Williams CB (1957) Insect migration. Annual Review of Entomology 2 163 180. Williams JJ, Stow DA, Brower LP (2007) The influence of forest fragmentation on the location of overwintering monarch butterflies in Central Mexico. Journal of the Lepidopterists' Society 61 90 104. World Wildlife Foundation (2013) Monarch Butterfly Survey Indicates Lowest Numbers in 20 Years, http://worldwildlife.org/stories/monarch butterfly survey indicates lowest numbers in 20 years Wright S (1938) Size of population and breeding structure in rel ation to evolution. Science 87 430 431. Wright S (1943) Isolation by distance. Genetics 28 114 138. Wright S (1946) Isolation by distance under diverse systems of mating. Genetics 31 39 59. Zalucki MP, Clarke AR (2004) Monarchs across the Pacific: The C olumbus hypothesis revisited. Biological Journal of the Linnean Society 82 111 121. Zalucki MP, Suzuki Y (1987) Milkweed patch quality, adult population structure, and egg laying in the monarch butterfly. Journal of the Lepidopterists' Society 41 13 22. Zangerl AR, McKenna D, Wraight CL, et al. (2001) Effects of exposure to event 176 Bacillus thuringiensis corn pollen on monarch and black swallowtail caterpillars under field conditions. Proceedings of the National Academy of Sciences of the United States of America 98 11908 11912. Zhu H, Gegear RJ, Casselman A, Kanginakudru S, Reppert SM (2009) Defining behavioral and molecular differences between summer and migratory monarch butterflies. BMC Biology 7 1 14.

PAGE 129

129 BIOGRAPHICAL SKETCH Carol Chaffee is broadly interested in how movement affects the distribution of genetic variation, and how this distribution affects evolution. Migration, dispersal, even daily movement around a local territory can all influence how genetic variation is distributed acro ss an organism's range. For her dissertation, she investigated how the unique migration behavior of the monarch butterfly ( Danaus plexippus ) affects the distribution of genetic variation in eastern North America. Carol followed a non traditional path towar d her doctorate. She graduated from the University of California, Berkeley, in 1989 with a Bachelor of Arts in social science, then went on to earn a Master of Arts in communications management at the University of Southern California. Following this, she pursued a career in software engineering with a focus on database design and information management. In 2002, Carol decided to switch career paths, and returned to San JosÂŽ State University to earn a Bachelor of Science in molecular biology with a minor in chemistry. She joined the Doctor of Philosophy program in zoology at the University of Florida in the fall of 2007. Carol plans to pursue a career focused on undergraduate biology education.